Re: [neonixie-l] Re: First Nixie Project Schematic Review

[Jan Rychter]

While we're on the subject of MOSFETs, I thought I'd share my finds, even 
though they might not be directly applicable to the original question.

For reaching >180V directly, I've had very good experiences with IRFH5025PbF. 
Vds up to 250V, RdsON 100mΩ. I found that the low-profile package really helps 
with electromagnetic noise — those leads on your IRF644's TO-220 package act as 
antennas once you start pumping amps of current through them!

For lower voltages, there is the IRFH5207PbF: Vds 75V, RdsON of just 9.6 mΩ.

Now, none of them is really operable from 3.3V, but I think with these HV boost 
converters one should really think about having a charge pump or a gate driver 
and drive the gates at 8 or 10V. It will help in a number of ways and if you 
use an integrated driver it might not even be complicated.

Also, be wary of the large gate charge: 37nC and 39nC for these two guys. 
That's a lot. You do not want to switch them too fast. I run mine at around 
50kHz.

--Jan Rychter
  PartsBox
  https://partsbox.io/

> On 11 Jun 2015, at 17:05, gregebert  wrote:
> 
> If you need a good MOSFET with a low Vgs(on), consider the DMN6040SVT. It has 
> a low Rds(on), around 60 milliohms. My wristwatch uses this for the DC-DC 
> converter for these reasons, and also because it has very low leakage. I 
> drive the gate at 3.2V .
> 
> BUT.if you use this device, you must use a transformer (basically 2 
> coupled inductors), to reduce the "kickback" spike that occurs while the HV 
> cap is charging. The DMN6040SVT will only tolerate 60V, so if you want to 
> generate 180V, you will want a transformer that is at least a 4:1 turns 
> ratio. Mine is 10:1 .
> 
> Finding low Vgs(on) devices with a high Vds is difficult; there are not very 
> many and this was the best one I could find for my application.
> 
> You may want to look into circuit changes that will give you a higher Vgs.
> 
> -- 
> You received this message because you are subscribed to the Google Groups 
> "neonixie-l" group.
> To unsubscribe from this group and stop receiving emails from it, send an 
> email to neonixie-l+unsubscr...@googlegroups.com.
> To post to this group, send email to neonixie-l@googlegroups.com.
> To view this discussion on the web, visit 
> https://groups.google.com/d/msgid/neonixie-l/075bfa73-0dee-43fe-bd62-138960054193%40googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/9CB61943-6C8D-4294-8B82-4041FBC034E1%40gmail.com.
For more options, visit https://groups.google.com/d/optout.

Re: [neonixie-l] Re: IN-1 reliability

[Jan Rychter]

Just took a quick look at my stash, and I also have a bunch of IN-1s coded 
9205. I'm not sure if it's worth it to design something around them, then — if 
they fail so quickly and easily.

--J.


> On 17 May 2015, at 07:27, gregebert  wrote:
> 
> My driver circuit is a direct-drive constant-current source, and it was set 
> to the IN-1 spec current of 3.0mA. I was only using one numeral in each tube, 
> so all undriven cathodes were floating. The clock uses 15 tubes to display 1 
> thru 12 on the clock face (it's a large neon-equivalent of a regular 
> mechanical clock) so the numeral is displayed constantly.
> 
> What I found is that after a few days there would be 2 glowing cathodes: The 
> desired numeral, and the adjacent cathode closest to the front of the tube. I 
> confirmed a low-resistance short of a few ohms was present. I kept a lot of 
> notes, but I would need to dig around for them. I also found that the short 
> was caused by a tiny filament that formed between the cathodes by passing 
> about 100mA which caused it to glow. It usually took 300 to 500mA to zap the 
> short; it acted like a fuse. Afterwards the tube functioned as expected.
> 
> As I ran my clock longer, I noticed more tubes failing in the same manner, 
> and each time I was able to zap them back to life. It got ridiculous having 
> to zap another tube almost every day, so I replaced them with Burroughs tubes 
> and never had any trouble.
> 
> I suspect the filament was growing in the direction of the electric field 
> between the cathode and anode, and grew until it shorted the adjacent 
> cathode. I dont have the equipment to analyze the chemical makeup of the 
> cathodes or the filament that formed. I do know it was very small diameter; 
> not visible until it was made to glow.
> 
> NASA has done considerable research on tin whiskers, which is probably a 
> similar mechanism to what was failing in my IN-1 tubes. They form in electric 
> fields.
> 
> I've heard that IN-1 tubes do not contain mercury, whereas Burroughs (and 
> many other brands) do contain mercury. Perhaps there is some odd role mercury 
> plays in preventing the formation of these filaments.
> 
> I suspect that after the USSR broke into separate nations that different 
> materials or processes were used in IN-1 manufacturing; there may have been 
> relaxed quality standards as well. My tubes are all date-coded 9205.
> 
> I have an A101 dekatron spinning 24/7 on the same clock, and it's been 
> running flawlessly for more than 2 years. Despite a much higher operating 
> voltage (hence higher electric field), it's fine.
> 
> -- 
> You received this message because you are subscribed to the Google Groups 
> "neonixie-l" group.
> To unsubscribe from this group and stop receiving emails from it, send an 
> email to neonixie-l+unsubscr...@googlegroups.com.
> To post to this group, send email to neonixie-l@googlegroups.com.
> To view this discussion on the web, visit 
> https://groups.google.com/d/msgid/neonixie-l/53216fa6-0cd4-4b97-a0e1-7b0daccf73e6%40googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/C899E94F-B0FA-40A9-BD62-95C04A8593F1%40gmail.com.
For more options, visit https://groups.google.com/d/optout.

Re: [neonixie-l] Steampunk

[Jan Rychter]

I will second what others have already said here: I have a clock in a wooden 
(laser-cut plywood) enclosure, but: 

* my HV PSU 
(http://jan.rychter.com/high-voltage-power-supply-for-nixie-tube-projects) has 
current limiting, so most faults in the HV section are handled gracefully,
* the whole unit is powered from a 12V/500mA brick.

I considered the safety issues, but then determined that it is *really* 
difficult to start a fire with 6W of power. And with most supplies having 
overcurrent protection you’d have to draw exactly 6W, a short would just 
trigger the protection and all you’d get would be the power supply turning on 
and off periodically.

So, assuming that your wall brick is reasonably safe (hint: if it’s very cheap, 
it likely isn’t safe), a wooden enclosure should be perfectly fine.

—J.

> On 22 Feb 2015, at 06:03, Matt  wrote:
> 
> True.  A *good* (and I stress good) switched mode power supply will become a 
> constant current power supply when you exceed its capacity. So if you short a 
> 12 volt, 2 amp power supply, you will get no more than 2ish amps.  If the UL 
> or CE listing is accurate, the power supply may get warm, but that should not 
> be a problem.  So to be safe, buy a switched mode power supply that can 
> deliver no more than the amperage that you need. That way, if something 
> shorts in the clock, that is all the power it is going to get, which is 
> hopefully not enough to start a fire. To be sure, you may want to test this.
> 
> Unfortunately, there are many power supplies on the market that are unsafe 
> for various reasons.  Make sure to get one from a reputable source, or have 
> the power supply properly examined. Here are some articles and videos about 
> unsafe power supplies:
> 
> http://www.righto.com/2012/10/a-dozen-usb-chargers-in-lab-apple-is.html
> 
> https://www.youtube.com/watch?v=wi-b9k-0KfE
> 
> https://www.youtube.com/watch?v=_exfmbsPqEI
> 
> http://hackaday.com/2012/10/10/raspberry-pi-foundation-looks-a-counterfeit-apple-power-supplies/
> 
> https://www.youtube.com/watch?v=q9wKjZXDPWI&t=9m39s (USB power supply begins 
> 9:39 in, ends at 17:52)
> 
> Also, poorly designed power supplies can add noise to your A/C line or RF 
> interference for nearby devices.  Here is one power supply that mostly works, 
> but poor power filtering prevents a specific device from working (while other 
> devices work fine):
> https://www.youtube.com/watch?v=8wkoxZw53Sk
> 
> This video explains how a simple switched mode power supply works:
> https://www.youtube.com/watch?v=bmtW_oA1fRo
> 
> On 02/21/2015 09:54 PM, Jeff Walton wrote:
>> Regarding your concern about the wooden housing:
>> 
>> I had an old nixie design that I did in college in the 1970 timeframe.  It 
>> used voltage doubler circuitry for the nixies and housed a filament 
>> transformer for the main power supply, which were all mounted inside the 
>> housing.  My old design with point to point wiring and discrete devices is 
>> dark ages compared to current designs.  That clock ran perfectly for over 
>> thirty years with a wooden and plexiglass case and was no problem - until 
>> 2005 when the voltage doubler shorted and caught fire.  It had far more 
>> power available inside the housing when things went wrong.
>> 
>> I think that the designs from present day with the external power supply 
>> (and limited energy) are quite safe to run with wooden enclosures. The power 
>> supplies are quite safe and if there was a problem with something, it would 
>> most likely not be with the clock itself.
>> 
>> On Saturday, January 25, 2014 at 12:10:21 AM UTC-6, Michel wrote:
>> 
>>I would feel a bit anxious to leave a wooden clock switched on for say 24 
>> hrs/day, especially if it is not my own design. Just
>>worried it would catch fire at a moment I am asleep or not there. Is that 
>> just me or do other people have that same worry?
>> 
>>I quite like the design though, not an immediate "wow" factor as some 
>> other clocks, but I wouldn't mind having a clock like
>>this one.
>>Michel
>> 
> 
> -- 
> You received this message because you are subscribed to the Google Groups 
> "neonixie-l" group.
> To unsubscribe from this group and stop receiving emails from it, send an 
> email to neonixie-l+unsubscr...@googlegroups.com.
> To post to this group, send an email to neonixie-l@googlegroups.com.
> To view this discussion on the web, visit 
> https://groups.google.com/d/msgid/neonixie-l/54E9632D.8040400%40matthewc.net.
> For more options, visit https://groups.google.com/d/optout.

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/CBFFA6C8-9417-4E1C-93D6-682FEFF103B0%40gmail.com

Re: [neonixie-l] Little machining Job - slightly OT

[Jan Rychter]

I’ve been looking at Chinese machining services for a while now, wondering why 
everybody seems to want to buy a CNC router from China instead of buying 
services. If you search the web you’ll find lots of people buying CNC machines 
or building their own. At the same time, a quick look at alibaba.com shows 
hundreds of companies seemingly willing to do prototyping work even in 
single-unit quantities. Just search alibaba.com for something like “cnc 
machining prototyping” or similar and you’ll see what I mean.

I intend to try some of those out, but perhaps someone can do that first :-)

I’ve had very good experiences with both buying from aliexpress.com and with 
ordering PCBs from iteadstudio.com — inexpensive, with quality meeting 
expectations every time. I don’t know why CNC machining would be any different.

To keep this message on subject, I’ve been thinking about a nixie clock with an 
enclosure made from CNC-machined brass, requiring a 5-axis machine.

So, you might want to look at some of those companies and try them out. It 
shouldn’t cost much. And please let us know if you do!

—J.


> On 04 Feb 2015, at 17:30, Paul Parry  wrote:
> 
> Hi Guys,
> 
> Anyone suggest where I can get some little spacers made ( brass preferably ) 
> I need 15mm O/D with a 9.5mm I/D hole and 7mm in height. They are to allow me 
> to fit some little glass tubes that hold little Neon colon indicators into 
> some regular 15mm copper pipe fittings.
> It would take someone with a lathe about half an hour, but it isn't something 
> I have. I would need 20 of them and clean out of ideas. One engineering 
> company quoted me a 3 figure price just to set a lathe up :(
> 
> Kind regards,
> Paul
> 
> 
> -- 
> You received this message because you are subscribed to the Google Groups 
> "neonixie-l" group.
> To unsubscribe from this group and stop receiving emails from it, send an 
> email to neonixie-l+unsubscr...@googlegroups.com.
> To post to this group, send email to neonixie-l@googlegroups.com.
> To view this discussion on the web, visit 
> https://groups.google.com/d/msgid/neonixie-l/bd5cee25-efef-402b-9bca-501b2c00a102%40googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/AD0F28E8-62A3-4BE6-8D70-CC0C8CF291DA%40gmail.com.
For more options, visit https://groups.google.com/d/optout.

Re: [neonixie-l] Nixie tube sockets - 3D printer!

[Jan Rychter]

> For A-101's I use 0.093" crimp-style connector socket pins (they cost about 
> $0.10 US from digikey, part #A14097-ND). After soldering wire to the 
> crimp-pin, I cover with heat-shrink tubing. It's very solid. Also works for 
> IN-1's, though I advise using a better tube.

On a related note, does anybody have part numbers for pins that can be used to 
mount nixies like IN-12 and IN-18? I mean the through-hole kind, so if you 
solder them into your PCB, the other ends form a kind of a socket.

These sometimes appear on eBay, but in small quantities and with uncomfortably 
high prices. I tried looking around, but navigating the world of through-hole 
pins is an excercise in frustration.

--J.

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/18DE897B-D8CD-43AA-9900-1C85FC65DF36%40gmail.com.
For more options, visit https://groups.google.com/d/optout.

Re: [neonixie-l] HV PSU design based on the TPS40210

[Jan Rychter]

> By any chance, did you price any options for assembly of the smd components?
> What method do you use to bake your own?

I just assemble the boards myself, on the cheap, so I don't have any pricing 
for assembly. What I do is:

* order the boards from iteadstudio,
* clean the board with isopropyl alcohol,
* apply solder paste from a syringe, using a needle,
* place components using tweezers,
* dump the whole thing into a cheap mini-oven,
* press the button on my MSP430 Launchpad.

The Launchpad board controls an SSR which PWMs power to the oven (just 4 times 
a second). It also has a tiny board connected to it, hosting a MAX31855 
thermocouple interface. The thermocouple gets inserted into the oven, and my 
software does the PID control. It was just a quick hack.

This contraption reflows boards surprisingly well (see here for the curve I 
normally get:http://jan.rychter.com/enblog/home-made-reflow-oven-for-smd). The 
only issue is with solder paste application: it takes time and effort, and I 
usually end up applying too much. I recently ordered some stencils from OSH 
Stencils, we'll see how that goes.

The photos show varied results, you might notice some components are slightly 
shifted, and there is some flux residue. This is mostly because of the "too 
much paste" problem: I ended up reworking the main chips using a hot air 
station.

Still, overall, the process works so well that I have no fear of SMD and in 
fact I prefer it to pulling wires through holes.

--J.

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/21D38F37-861F-4FEA-A612-A72D5933B22B%40gmail.com.
For more options, visit https://groups.google.com/d/optout.

Re: [neonixie-l] HV PSU design based on the TPS40210

[Jan Rychter]

On 5 maj 2014, at 19:27, John Rehwinkel  wrote:

>> I finished my HV PSU design and decided to share it as open-source hardware, 
>> for anyone to use 
>> freely:http://jan.rychter.com/high-voltage-power-supply-for-nixie-tube-projects
> 
> That's a nice bit of work, thank you for sharing it!  I'll just offer one 
> tip: when supporting multiple package
> variants and not wanting your schematic to look wonky, make a custom Eagle 
> part that has a footprint
> with the package variants, and use that.  A side advantage is that you don't 
> have to try to overlap parts
> to get the arrangement you want, avoiding design rule check errors.

Yes, that's definitely the better approach. But I'm wary of investing too much 
time into Eagle, as I think it is on its way out (looking forward to switching 
to KiCad), so I don't want to spend too much time tweaking libraries.

>> The supply generates up to 220V from a 12V input. In addition to that, it 
>> also provides 2*Vout (so, up to 440V, for dekatrons), and two outputs for 
>> powering digital logic: 5V and 3.3V. The primary HV boost circuit reaches 
>> 88% efficiency when going from 12V to 185V at 55mA, with a 3% output ripple.
> 
> Those are some useful parameters.  That would also make a nice tube/CRT 
> supply.  The 5V output could be
> adjusted to 6.3V for heaters, and the doubler could be extended to a tripler 
> (or more) for CRTs that want more voltage. The regulation, efficiency, and 
> available current would all be less, but CRTs don't need much current.

And this is why making a design public makes sense -- I would never have known!

> I really appreciate details like mounting holes!

Yes, I'm tired of devices that don't have them, and as a result are only useful 
as a bench toy. I don't know how people mount the other PSUs I've seen -- cast 
them in resin? Make rails? Use clips?

> [...]
> I'd be intereseted if you care to share more details on loop stability or the 
> parasitic ringing on the switching node and
> how to tame it with a snubber.

I'd love to, but unfortunately there isn't much to share. The loop stability 
calculations were done using SwitcherPro. It did require some tweaking, as it 
wasn't really meant to design a supply like that. I managed to get a decent 
margin, but I can't be certain of the actual behavior, because:

* the output cap plays a role, but isn't easily controllable, especially its 
ESR,
* the design is intended to be used with various inductors and caps.

And since I don't have access to a network analyzer, I can't really verify the 
calculations. All I can see is that I get no oscillations or unstable behavior, 
and that load transient behavior is nice and clean, in the circuits I've built.

As to the snubber, I just went with TI recommendations again -- I don't think I 
can properly design a snubber with my 100MHz scope.

TI has good literature on the topic:
http://focus.ti.com/general/docs/litabsmultiplefilelist.tsp?literatureNumber=slup100
http://focus.ti.com/general/docs/litabsmultiplefilelist.tsp?literatureNumber=slva255
 

The test points placed on the board should be enough for someone to connect a 
network analyzer, so perhaps some day this will happen :-)

--J.

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/CB663D74-6DD5-48E9-B4AA-BD7B3E35154E%40gmail.com.
For more options, visit https://groups.google.com/d/optout.

[neonixie-l] HV PSU design based on the TPS40210

[Jan Rychter]

Hi,

I finished my HV PSU design and decided to share it as open-source hardware, 
for anyone to use freely: 
http://jan.rychter.com/high-voltage-power-supply-for-nixie-tube-projects

The supply generates up to 220V from a 12V input. In addition to that, it also 
provides 2*Vout (so, up to 440V, for dekatrons), and two outputs for powering 
digital logic: 5V and 3.3V. The primary HV boost circuit reaches 88% efficiency 
when going from 12V to 185V at 55mA, with a 3% output ripple.

I designed it because I couldn't find anything that would make sense for my 
Nixie projects. There are plenty of tiny power supply modules available on 
eBay, but most of them end up being impractical: no 3.3V (for my 
microcontroller) and 5V (for my 74141 nixie drivers), no mounting holes, no 
>400V output for powering dekatrons. Some supplies make a token gesture towards 
practicality by sticking a 7805 on the same board, but you quickly find out 
that the current draw of 6x74141 is enough to require a large heat sink on a 
12V-powered 7805 (one 74141 consumes 12.5mA!). This means that instead of a 
single-board power supply you end up routing your input power all over the 
place, implementing your power supply in several places.

The version I'm posting online is not perfect, but works quite well in a number 
of my projects. I decided I'd rather publish it as it is now rather than keep 
it locked forever.

The design is based on the TPS40210 from Texas Instruments, as I've grown tired 
of the MAX1771. I just couldn't get it to work reliably, didn't like the 
pricing, didn't like the availability.

Hope it helps someone, at least to understand boost converters better (I 
certainly learned a lot while building this!).

--J.

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/57C4EEA4-A8CA-4FEF-9F69-877A7B0739B4%40gmail.com.
For more options, visit https://groups.google.com/d/optout.

Re: [neonixie-l] Favourite chips...

[Jan Rychter]

On 4 maj 2014, at 18:10, gregeb...@hotmail.com wrote:

> Firstly, I only do direct-drive.
> 
> I dont use driver chips. Instead, I use either NPN or NMOS discrete devices, 
> because I can ensure constant-current thru the cathode.

Which transistors do you use? I'm especially interested in tiny SMD ones. 
Dropping some hints would be very helpful -- it is difficult to just start 
searching for semiconductors out there.

--J.

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/B05ABAFE-440C-4BB7-B331-611BBAB93199%40gmail.com.
For more options, visit https://groups.google.com/d/optout.

Re: [neonixie-l] Need some help wrapping my mind around driving IN-9's with A TLC-5940

[Jan Rychter]

Ok, I realize I'm really late to this thread, but I've just only found it going 
through the archives and I thought I'd pitch in, because it might help someone 
in the future: I spent quite a bit of time looking for inexpensive 
multi-channel DACs.

I have a "bargraph" project on the back burner and I wanted to drive 8-12 IN-13 
tubes. The best solution I could find was a ROHM BU2501FV DAC. It's a 
12-channel DAC with rail-to-rail output capability, driven via a serial 
shift-register-like interface. Seems perfect for this kind of application. I 
was planning to use it with transistors to drive the tubes. I got the chips 
from RS Components and they were not overly expensive. The BU2500FV is 5V, 
BU2501FV is 3V.

The nice thing is that it also has a VrefL pin that allows you to supply a 
lower reference voltage for the DAC. I was planning to supply it with a diode 
drop, to offset the B-E voltage of the driver transistors, which would give me 
a lot of room to work with.

You might also want to look at BH2221FV.

--J.

On 4 gru 2013, at 00:04, Jon  wrote:

> 
> On Tuesday, December 3, 2013 2:24:03 AM UTC, Chill4844 wrote:
> I am attempting to use an Arduino in conjunction with a TLC5940 to drive an 
> array of IN-9's (And not kill myself or blow anything up) the problem being 
> if i understand it right (which I probably don't) is that instead of a 
> positive PWM out the arduino the TLC5940 works by sinking the current on the 
> cathode side. I don't think i can use this with my standard mpsa42 (and I'm 
> not going to attempt anything until i get some advice) I do have some 
> mpsa92's available.
>  
> I don't think you want the PWM capability of this chip for IN9. For what 
> you're trying to do, take a look at the TLC5628 which is an 8 channel DAC 
> with a simple 4 wire interface to the microcontroller. Hook the output of 
> each DAC channel up to the simple MPSA42 current sinks per Jeff's paper, and 
> that's all you need. You can manage around the Vbe offset in software to keep 
> the hardware simple. That's the technology which sits behind this: 
> http://youtu.be/mQ1567EFCY0
>  
> Grahame rolled his own version and documented it here 
> (http://www.sgitheach.org.uk/ss.html) including a full schematic.
>  
> BTW, IN9 are very variable and suffer a lot from cathode poisoning. To get an 
> array of them to perform consistently well and in a uniform way is going to 
> take quite some tube selection and conditioning. IN13 may be a better bet, 
> though they are rather dim to my eye.
>  
> Cheers,
>  
> Jon.
> 
> -- 
> You received this message because you are subscribed to the Google Groups 
> "neonixie-l" group.
> To unsubscribe from this group and stop receiving emails from it, send an 
> email to neonixie-l+unsubscr...@googlegroups.com.
> To post to this group, send an email to neonixie-l@googlegroups.com.
> To view this discussion on the web, visit 
> https://groups.google.com/d/msgid/neonixie-l/ec596498-2b0c-4a67-99f8-836eb94ef88a%40googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/A0393D25-CE1C-412B-9661-DC97FD40D52D%40gmail.com.
For more options, visit https://groups.google.com/groups/opt_out.

Re: [neonixie-l] Anyone used Iteadstudio for PCBs?

[Jan Rychter]

On Tuesday, December 17, 2013 4:11:06 PM UTC+1, Nick wrote:

> On Monday, 16 December 2013 14:51:14 UTC, Nigel Walker wrote:
>>
>>  Just got my Itead boards. That's about two and a half weeks from order 
>> to delivery to the UK.
>>
>
> Quality? Photos? What's the solder mask/silk screen registration like? All 
> holes in the middle of pads? etc...
>

I can show you what some of my boards look like. The fragment shows a 
QFN-24 chip, and some 0603 caps, the traces are 10mil. Looks uglier than in 
reality because of reflections (that's not smeared silkscreen paint you see 
on the traces, just reflections).

I've never seen problems with drill alignment, but silkscreen can be a 
little off, note the difference between the pictures (another board from 
the same batch). Never found it to be a problem in practice. Overall, I'd 
say I got exactly what I expected. Don't expect miracles, but the quality 
is fairly good, and price/performance is excellent.

--J.





-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/6508a371-47fc-4ec3-8c44-3e1c17b10216%40googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

Re: [neonixie-l] OT: Sensor networks... (motes)

[Jan Rychter]

On 11 gru 2013, at 14:41, Nick  wrote:

>> On Wednesday, 11 December 2013 12:38:52 UTC, Jan Rychter wrote:
>> If you don't specifically need 802.15.4 for compatibility reasons and if you 
>> aren't extremely size-constrained, you might find that it is much cheaper to 
>> go with an MSP430 and a nRF24L01+ module. This is what I'm doing in a 
>> project right now: an MSP430G2412IRSA16T (about $1) and a Chinese nRF24L01+ 
>> module (complete module, with a PCB trace antenna, for about $1.15). The 
>> Nordic chips work fairly well and are suitable for many applications. And 
>> getting a complete radio-networked microcontroller solution for $2.15 is 
>> really hard to beat. 
>> 
>> To keep this on topic, I've been considering using those radio modules in a 
>> home automation system, to light up clocks whenever someone is present in 
>> the room :-) 
>> 
> Oddly, I've been looking at exactly those modules - the  nRF24L01+ modules 
> are astonishingly good value for money - I just bought a bunch from 
> http://imall.iteadstudio.com/catalogsearch/result/?q=NRF24L01+ - 2 of each 
> type - my application may need the longer range of the LNA/PA version, but 
> for that sort of money I just bought a pair to try out.

If you decide they are right for you, I'd look for larger quantities on 
aliexpress.com — this is where I get my modules at $1.15 - $1.30 a piece.

> The Nordic stuff is excellent and longer range generally than the 
> 802.15.4/ZigBee stuff, though I;ve noticed that TI do SoCs with inbuilt 
> wireless but normally with an ARM core - the MSP430 variants are <1GHz.
> 
> So many extraordinarily cheap neat devices out there...

You might want to keep an eye on the upcoming Freescale Kinetis W series, which 
is comparable, but with a 32-bit ARM core: 
http://www.freescale.com/webapp/sps/site/taxonomy.jsp?code=KINETIS_W_SERIES

However, I decided these kinds of devices are not for me. First, you need to 
know at least a little bit about RF design, which at 2.4GHz isn't obvious. 
Second, last I checked the components were not easily to obtain: neither the 
wireless chips themselves, nor the surrounding elements (such as tiny inductors 
that you need for the balun). Given the price of complete working assembled 
modules it just didn't seem worth the effort.

--J.

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/290803A4-5528-4219-9DC7-707A61F01F67%40gmail.com.
For more options, visit https://groups.google.com/groups/opt_out.

Re: [neonixie-l] OT: Sensor networks... (motes)

[Jan Rychter]

On 11 gru 2013, at 12:52, Jan Rychter  wrote:

> On 11 gru 2013, at 05:31, Nick  wrote:
> 
>> I've ben thinking about the MSP430 series and their SoC/ wireless chips 
>> (mostly 802.15.4 based) - very nice suite of devices
>> Trouble is I've always been an Atmel guy, so it's a big switch though I 
>> suspect that the learning curve is not too bad...
>> 
>> Perhaps its time to change. ..

[replying to myself, as there is one more thing I forgot to mention]

If you don't specifically need 802.15.4 for compatibility reasons and if you 
aren't extremely size-constrained, you might find that it is much cheaper to go 
with an MSP430 and a nRF24L01+ module. This is what I'm doing in a project 
right now: an MSP430G2412IRSA16T (about $1) and a Chinese nRF24L01+ module 
(complete module, with a PCB trace antenna, for about $1.15). The Nordic chips 
work fairly well and are suitable for many applications. And getting a complete 
radio-networked microcontroller solution for $2.15 is really hard to beat.

To keep this on topic, I've been considering using those radio modules in a 
home automation system, to light up clocks whenever someone is present in the 
room :-)

--J.

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/ED48D452-ED7D-4443-A088-9EC1D1EF7B75%40gmail.com.
For more options, visit https://groups.google.com/groups/opt_out.

Re: [neonixie-l] OT: Sensor networks... (motes)

[Jan Rychter]

On 11 gru 2013, at 05:31, Nick  wrote:

> I've ben thinking about the MSP430 series and their SoC/ wireless chips 
> (mostly 802.15.4 based) - very nice suite of devices
> Trouble is I've always been an Atmel guy, so it's a big switch though I 
> suspect that the learning curve is not too bad...
> 
> Perhaps its time to change. ..

I'd really recommend taking a look at the MSP430 devices. To keep this on 
topic, I use the MSP430 in my nixie clocks. It's straightforward to program, 
*very* well documented, inexpensive, and runs for weeks on a single oily 
electron. The low-power modes are really easy to use. Average current 
consumption for my CPU module when running on battery backup is down to single 
microamps.

My clocks usually use the MSP430G2553, mostly because I'm lazy — I could fit 
into a smaller part. For designs that are very price- or size- sensitive, and 
where I don't need the ADC, I use the MSP430G2412 in a QFN-16 package. It's 
tiny, it's really cheap, and needs only two external components (a pullup on 
the reset pin and a decoupling cap). Oh, all MSP430 chips are 16-bit. And it's 
easy to get started with the launchpad boards.

I recently started looking at Freescale Kinetis devices, and I will be 
migrating to those in most of my designs. I'd recommend taking a look — pricing 
for the low-end KL05 chips is comparable to the MSP430, and the architecture is 
nicer (32-bit ARM Cortex M0+ core). For bigger devices I'll be using the KL25 
and K20.

One thing I noticed already is that Texas Instruments really has top-notch 
documentation, which means the learning curve isn't bad at all. Freescale's is 
harder to understand, isn't always complete, I have found mistakes, and is 
overall harder to use.

As an aside, I don't really understand why Atmel microcontrollers are so 
popular in hobbyist circles.

--J.

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/5D9198CA-9555-4BC2-9EE7-2187EA20B7C9%40gmail.com.
For more options, visit https://groups.google.com/groups/opt_out.

Re: [neonixie-l] Anyone used Iteadstudio for PCBs?

[Jan Rychter]

I use them regularly and so far I'm very happy with the results.

Note that you should not expect miracles: those are boards produced in China at 
rock-bottom prices, so mistakes will sometimes be made. One of my orders got 
lost once and I had to ask for it to be produced and sent again. But 
considering my 15+ orders so far, I think this is a very good result.

As far as quality is concerned, I had no issues. The boards were always 
perfect. Also, they officially send you 10 boards, but I noticed that I 
actually often get 11 or 12.

They also recently lowered prices for the larger sizes 5x10cm and 10x10cm.

One thing that many people miss is that milling is included in the price. I 
often have fancy board outlines, rounded corners, or mounting screw cutouts 
(these take up less space than mounting holes). They are always perfect. I was 
actually worried recently, because I had a 20mm x 20mm board with milled 
cutouts on all four sides, but it was manufactured perfectly.

The biggest problem with their business is the incredibly slow Hong Kong Mail, 
and even more disastrous China Post. In my case (shipping to Poland), it takes 
2+ weeks for the package to make it out of Hong Kong, and then several days for 
it to arrive at my doorstep (it usually spends nly 
1-2 days in Poland). Recently they shipped one package using China Post and 
that one took 5 weeks to get here. On the average, I've come to expect 3-4 
weeks from clicking "order" to receiving the package from the mailman.

Gerbers are not a problem in my case, they provide a job file for EAGLE, and it 
takes all of 30 seconds to produce them, zip them up and send as an E-mail 
attachment.

All in all, a great deal.

--J.

On 9 gru 2013, at 15:16, Nick  wrote:

> http://imall.iteadstudio.com/open-pcb/pcb-prototyping.html (note that their 
> prices are for 10 boards)
> 
> I was considering using them but was wondering if anyone here had any 
> experiences to share, or another alternative that's better?
> 
> Thanks
> 
> Nick
> 
> -- 
> You received this message because you are subscribed to the Google Groups 
> "neonixie-l" group.
> To unsubscribe from this group and stop receiving emails from it, send an 
> email to neonixie-l+unsubscr...@googlegroups.com.
> To post to this group, send an email to neonixie-l@googlegroups.com.
> To view this discussion on the web, visit 
> https://groups.google.com/d/msgid/neonixie-l/a5c5ff03-e2bc-49b2-98c7-3b2339a96bb1%40googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/73ABFBEF-9EFB-4CB7-BFA0-EF302AA10D3E%40gmail.com.
For more options, visit https://groups.google.com/groups/opt_out.

Re: [neonixie-l] Arduino and MAX6921

[Jan Rychter]

On 21 lis 2013, at 23:47, Adam Jacobs  wrote:

> I used to think it was strange too and always included the clock 
> functionality in the microcontroller. However, I was always fine-tuning the 
> accuracy. I believe that someone was posting to this list a little while back 
> about his own Atmel microcontroller clock-code accuracy issues.

I had clock accuracy issues with my MSP430-based clock until I remembered to 
properly set the load capacitance for the 32.768kHz crystal. After changing it 
to 12pF the clock runs incredibly well.

As for battery backup, the chip detects external power failing and quickly 
switches into a power-saving mode, where it does pretty much nothing except 
wake up every second to increase the counter (I use the watchdog timer for 
that). It also checks the voltage. In this mode, the power consumption is down 
to single microamps, so a single CR2032 can last for years. It worked better 
than I expected it to.

I was never fond of external RTC chips because of their prices. I try to design 
cheaply, I think this approach forces you to design better. And I try to never 
design in a part that is not available (samples don't cut it). I'll use it only 
if Farnell or RS stock it. This rules out most Maxim parts, unfortunately, 
because while often the best fit, they are usually unobtainium.

Exceptions can be made for chips like 74141 or the MAX6921, which can be gotten 
from eBay.

Going back to the RTC, I think implementing the loops that count seconds, 
minutes and hours is a really fun experience, the "going to the basics" kind 
you very rarely get.

--J.

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/E5579BE3-89A4-4E86-9C6D-56AC30DEF42D%40gmail.com.
For more options, visit https://groups.google.com/groups/opt_out.

Re: [neonixie-l] New power supply design

[Jan Rychter]

Those are impressive numbers.

As I'm designing my own supply (getting about 90% efficiency for a boost from 
12V to 180V at up to 80mA, will share the design once done), I'd be interested 
in knowing more about the design? Are you willing to share some information?

The biggest design problem I encountered was that the Nixie load can be highly 
variable (think dimming, multiplexing, etc), so with a booster design you can't 
just operate in Continuous Conduction Mode (CCM) all the time, which is what 
most high-power designs do. What's worse, the transition between CCM and DCM 
(Discontinuous Conduction Mode) is poorly defined. This is why I decided to 
finally use a smaller inductance and stay in DCM all the time, even with full 
load.

Actually, is your design a boost at all? Or do you use a flyback transformer?

When you measure output power, do you just stick a meter and measure average 
current/voltage, or do you do something fancy to take ripple into account?

--J.


On 12 sie 2013, at 06:45, taylorjpt  wrote:

> 
> 77W test data:  http://ppl.ug/A2EYKy2BwmY/
> 
> 383mA/201Vout at 36V input (48V abs max);  Had to pre-load my load bank to 
> get over the 250mA limit.
> 
> jt
>  
> 
> 
> -- 
> You received this message because you are subscribed to the Google Groups 
> "neonixie-l" group.
> To unsubscribe from this group and stop receiving emails from it, send an 
> email to neonixie-l+unsubscr...@googlegroups.com.
> To post to this group, send an email to neonixie-l@googlegroups.com.
> To view this discussion on the web, visit 
> https://groups.google.com/d/msgid/neonixie-l/9944e216-7d04-4c1e-8a5f-8e37f5c4f687%40googlegroups.com?hl=en-GB.
> For more options, visit https://groups.google.com/groups/opt_out.

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to neonixie-l+unsubscr...@googlegroups.com.
To post to this group, send an email to neonixie-l@googlegroups.com.
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/954448DA-D5B3-4F75-85A3-430D5B82EAF3%40gmail.com?hl=en-GB.
For more options, visit https://groups.google.com/groups/opt_out.

Re: [neonixie-l] Re: Why is multiplexing nixie's bad

[Jan Rychter]

On 20 wrz 2012, at 19:50, Adam Jacobs  wrote:

[about multiplexing]

> You forgot the other pro's:
> 
> - reduced power consumption
> - dramatically increased tube life at equal cathode current.

Ok, so has that actually been proven? I've seen opinions stating that tube life 
is dramatically increased and others saying that it is dramatically reduced. 
Some people say it has no influence at all. I'd really like to see a definitive 
(e.g. fact- or reasoning- based) statement on the matter.

Anyone?

--J.

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To post to this group, send an email to neonixie-l@googlegroups.com.
To unsubscribe from this group, send email to 
neonixie-l+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

Re: [neonixie-l] where can I get IN-18 tubes at a reasonable price?

[Jan Rychter]

On 2 lip 2012, at 19:33, Terry S wrote:

>> I'll wager that Dieter also has a growing bin, laden with stinkers ;)
>> 
> Dieter is growing a Bin Laden?

Ok, so now this group is right in the sights of the NSA, CIA, Echelon and 
whatever other agencies are monitoring internet communications in the U.S. — 
with this combination of key words it simply has to be a terrorist threat!

Especially with all these people discussing high voltages, gas mixtures, 
clocks, and worst of all — *cathode poisoning*!

--J.

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To post to this group, send an email to neonixie-l@googlegroups.com.
To unsubscribe from this group, send email to 
neonixie-l+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/neonixie-l?hl=en-GB.

Re: [neonixie-l] MAX1771 PSU instability (solved)

[Jan Rychter]

Thanks for the excellent advice. I will try a toroidal inductor, I don't know 
why I never thought of that. I am also building an experimental design based on 
the TPS40210 — that part is much cheaper despite requiring some more passive 
components, and it has an automotive variant, which means it will be around 
pretty much forever. We'll see how it behaves, the devil is in the details. 
Also, I will use an IRFH5025 FET for switching: it's SMD, the specs look nice, 
and the price is great.

I have one more question -- did you use a potentiometer in any of your designs? 
This *really* makes a huge difference in my tests.

As for the drain being the source of noise -- right, this is the pulsed 
inductor output, with monstrous voltage peaks and high currents. But there is a 
limit to how small one can make it and how far away from the feedback network 
it can be. You might be right about an SMD FET being better here, though.

Great tip about several resistors in parallel for RSENSE, too :-)

thanks,
--J.

On 24 sty 2012, at 03:16, misty01a wrote:

> Hello.
> 
> In my "high current" MAX1771 PSU design I made several years ago, the
> best results I got when I used a toroidal inductor with a current
> rating of 10.3 A (Bourns 2200LL-470-RC), and a RSENSE peak current
> setting of 5.33 A.  I tried other inductors, including a physically
> smaller 2100LL series inductor (with a smaller current limit), but the
> efficiency was not good as the 2200LL.  I don't know why.
> 
> Several notes that may help (based on my personal experience only):
> 
> - The FET drain tab is the source of a great amount of noise.  The FET
> drain should be placed as far away from the FB pin trace as reasonably
> possible.  An SMD (D2PAK) FET may be better than a TO-220 one because
> of the horizontal placement on the PCB (also for better cooling of the
> FET).
> - I have never observed any interference from the inductor, but then I
> use toroidal inductors which are supposed to be self-shielding.  I
> once did get good results with an unshielded Bourns 1110 series
> inductor, however.
> - For RSENSE I use several 0.075 Ohm 1/4 W resistors in parallel.
> They are cheaper than a single larger resistor, and it's easier to
> adjust the peak current value.
> - A diode between the REF and FB pins may help if the MAX1771 has
> trouble starting up.  The MAX1771 can enter 12 V fixed output mode if
> the supply voltage is larger than 12 volts and the voltage at the FB
> pin is too low (because of the high input:output ratio of the voltage
> divider).  On the other hand, many other people's designs seem to be
> working fine without this diode.
> 
> My present design uses a Bourns 2200LL-470-V-RC inductor and an
> FDB28N30TM FET (0.129 Ohm ON resistance).  The maximum power output in
> a testbench situation was 24.9 W at 85 % efficiency with 14 V input
> and 220 V output.  Board size is 50 mm x 40 mm.  I use three of these
> units in parallel to provide 40 watts at 250 V in a (non-Nixie)
> project.
> 
> -- 
> You received this message because you are subscribed to the Google Groups 
> "neonixie-l" group.
> To post to this group, send an email to neonixie-l@googlegroups.com.
> To unsubscribe from this group, send email to 
> neonixie-l+unsubscr...@googlegroups.com.
> For more options, visit this group at 
> http://groups.google.com/group/neonixie-l?hl=en-GB.
> 

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To post to this group, send an email to neonixie-l@googlegroups.com.
To unsubscribe from this group, send email to 
neonixie-l+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/neonixie-l?hl=en-GB.

Re: [neonixie-l] MAX1771 PSU instability (solved)

[Jan Rychter]

On 23 sty 2012, at 19:46, John Rehwinkel wrote:

>> A 0R025 RSENSE resistor means that we're ramping the current up to 4A before 
>> releasing the energy stored in the inductor. 4A is a lot of current. The 
>> magnetic field produced by the inductor is so strong that it affects 
>> everything around it, even very carefully routed traces
> 
>> I did some experimenting and:
>> 
>> -- moving the inductor away is a solution, but it needs to be *far* away, 
>> 8cm or so,
>> 
>> -- shielding the inductor, while improving the situation, increases the 
>> current consumption 2x, produces audible noise and heats up the shield 
>> considerably, so it is not a solution.
>> 
>> All in all, I don't think a circuit with an RSENSE of 0R025 is feasible, at 
>> least not with a 5cm x 5cm board. I'll be going back to 0R050 and 0R100 and 
>> smaller inductors.
> 
> Have you tried a toroidal inductor?  They're better behaved in this respect, 
> even though they're physically large, especially for the high current one 
> you'd need.

Hmm, I haven't considered that. They might indeed behave better. The inductor I 
use is from a Polish manufacturer: http://www.feryster.pl/polski/dsz14x15v.php 
(except I used a 5A version).

But overall I think I'll try the smaller RSENSE values first, they should be 
enough to provide the output currents I need.

--J.

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To post to this group, send an email to neonixie-l@googlegroups.com.
To unsubscribe from this group, send email to 
neonixie-l+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/neonixie-l?hl=en-GB.

Re: [neonixie-l] MAX1771 PSU instability (solved)

[Jan Rychter]

A while ago I posted about my problems with a MAX1771-based PSU that I was 
designing and building, asking for advice. I did get some advice, which was 
right -- but only partly so. Read on if you're interested in the details.

My board was based on Nick de Smith's design. What I wanted, though, is higher 
current capability: I wanted to be able to draw 80mA and possibly more. I also 
wanted to have a second HV output through a voltage doubler (to get ~400V for 
dekatrons) and some additional stabilizers for digital circuitry.

I followed the advice given by Maxim in the datasheet and by Nick on his site 
-- reduced RSENSE to 0R025 (25mOhm) and used a large inductor (rated for 5A). 
The inductor is heavier than all the rest of the board combined :-) However, 
while the voltmeter happily indicated 180V at the output, the circuit was not 
stable: the MAX1771 was going into shutdown and there was a huge (30V) 
low-frequency ripple on the output. Ugly.

The advice I got from this group was that this was likely caused by noise on 
the feedback pin, and that my FB trace routing was no good.

Several board prototypes later (thank heavens for toner transfer, you can't 
beat the 30 minute turnaround time)...

As it turns out, the advice about the feedback trace was right, but only partly 
so. My problems were due to noise on the FB pin allright, but no amount of 
careful routing would have solved them. One way to get the circuit to behave in 
a stable way was to increase the current through the voltage divider. Even with 
0805 resistors there is room to spare and going from 1.5MOhm total to about 
300kOhm total meant that at least this part of the circuit was less susceptible 
to interference. However, I couldn't find a way to make the thing work with a 
(vertical) potentiometer and I don't think there is a way.

A 0R025 RSENSE resistor means that we're ramping the current up to 4A before 
releasing the energy stored in the inductor. 4A is a lot of current. The 
magnetic field produced by the inductor is so strong that it affects everything 
around it, even very carefully routed traces. A scope probe held in the air 
near the pot picks up almost .6V of interference! This only goes up as one 
approaches the inductor. Depending on how the inductor is mounted (which side 
of the board) you can get voltage increases or voltage drops on your FB pin.

I did some experimenting and:

-- moving the inductor away is a solution, but it needs to be *far* away, 8cm 
or so,

-- shielding the inductor, while improving the situation, increases the current 
consumption 2x, produces audible noise and heats up the shield considerably, so 
it is not a solution.

All in all, I don't think a circuit with an RSENSE of 0R025 is feasible, at 
least not with a 5cm x 5cm board. I'll be going back to 0R050 and 0R100 and 
smaller inductors.

I can post some scope traces if people are interested. I have learned a lot in 
the process, and as a nice side effect I have a pretty clean PCB layout that is 
realizable with a single-sided board (requires one jumper) -- this might be 
useful for people who make their own PCBs. Once I get it verified and tested I 
plan to post it for people to use.

--J.

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To post to this group, send an email to neonixie-l@googlegroups.com.
To unsubscribe from this group, send email to 
neonixie-l+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/neonixie-l?hl=en-GB.

Re: [neonixie-l] MAX1771 PSU instability

[Jan Rychter]

On 15 sty 2012, at 00:37, Jon wrote:
> On Jan 14, 10:24 pm, Jan Rychter  wrote:
>> On 14 sty 2012, at 18:37, marta_kson wrote:
>> 
>> I think I'll need to redesign the board without the potentiometer, thus 
>> bringing the FB resistors much closer to the chip.
>> 
> 
> I can't speak to the arcane world of the 1771 circuit, never having
> used it. But if it's important to you to have an easily variable
> output voltage and it's the pot which is killing you, then there is a
> completely different strategy which works well. You can make very
> satisfactory nixie and dekatron power supplies using microcontrollers
> with on-board ADC and PWM modules.The PWM pin drives the switching
> element in a standard boost SMPS circuit, and the HV feedback line
> comes from a fixed potential divider into the ADC pin. You then
> process the ADC output in firmware to determine the PWM duty cycle,
> which gives you a continuously variable, programmable HV supply which
> is great for playing around with things. Want to test the nixie at
> 160V or 200V? Just tell the microcontroller what you want the output
> to be, and let the magic of negative feedback sort it out.
> 
> The downside is that it costs you some time overhead in the
> microcontroller to keep the HV going, but with the right chip you can
> do this, run a USB link and still have plenty of cycles left for a
> funky user program. You probably don't get as good efficiency as
> Nick's beautifully optimised 1771 circuit, but that is rarely a
> terminal design issue in my experience. The efficiency drop has never
> required me to heatsink the FET or anything like that.
> 
> I've run this strategy up to about 12W output - no doubt less than you
> could get out of a thoroughly tweaked 1771 or John Taylor's awesome
> modules - but plenty enough for most purposes.

I develop software for a living, and I know how difficult it is to make 
reliable software. In the solution above I'd worry what would happen if my 
software went astray and the PWM controller happily continued to pump the 
voltage without any control whatsoever. I'd much rather use solutions which are 
self-balancing.

As for other suggested solutions (555 timers, etc) -- I learned already that a 
switching PSU is a complex beast, even though it might seem superficially 
simple. There are many corner cases that I'd rather have covered, such as: how 
does the circuit behave when the output gets shorted? When the short gets 
released? Is there overcurrent/overvoltage protection? Does it start reliably? 
Does it start when input voltage ramps up slowly? How does it react to sudden 
load changes? There are many things which one can get wrong, so I prefer an 
IC-based solution, where someone thought about all the corner cases.

--J.

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To post to this group, send an email to neonixie-l@googlegroups.com.
To unsubscribe from this group, send email to 
neonixie-l+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/neonixie-l?hl=en-GB.

Re: [neonixie-l] New to the group.

[Jan Rychter]

On 23 sty 2012, at 04:10, threeneurons wrote:

>> Although, if you are using a voltage doubler to get your HV supply, you
>> will find that the supply sags as current increases
>> 
>> -Adam W7Q
>> 
> 
> Yep, the ripple voltage increases dramatically, per multiplier stage,
> given the same current draw. But the nixies are the devices that draw
> the most current, with the dekatron drawing not even a milliamp. So
> make the nixie voltage straight out, without the multiplier, then
> double or triple the voltage for the dekatron.  ie, 180V nixie & 3x ->
> 540V dekatron, or 220V nixie & 2x -> 440V dekatron.

I built a PSU with a voltage doubler (200V doubled to get 400V). With 1mA of 
output current on the doubled 400V output the ripple isn't a problem at all. I 
think it was around 10V P-P.

--J.

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To post to this group, send an email to neonixie-l@googlegroups.com.
To unsubscribe from this group, send email to 
neonixie-l+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/neonixie-l?hl=en-GB.

Re: [neonixie-l] New to the group.

[Jan Rychter]

On 18 sty 2012, at 07:08, dr pepper wrote:

> Hi guys I'm new to the group, I have built a couple of clocks before,
> I'm working on a IN1 nixie clock with an OG4 dekatron and a IN9 neon
> bargraph (seconds) at the moment, it uses a software module for the
> pic micro to decode the msf time signal.
> All the electronics (well nearly all) and software is my own.
> The latest development is a switching power supply that uses a
> cockroft walton voltage multiplier so that I can have just one power
> supply inverter with 2 outputs 200 and 400v for the nixies and
> dekatron respectively.

Ahh, interesting, that is very similar to what I'm planning. My PSU produces 
two voltages as well (I'll post a writeup soon about the stability problems I 
had, I have them partially solved now).

I'll be using a combination of IN-1 and A-101 dekatrons and I was planning on 
pushing the nixies slightly to 200-210V, so that the dekatron can get its 
400-420V. I was worried about damaging the IN-1s, though -- not sure if they'll 
take to 210V kindly.

--J.

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To post to this group, send an email to neonixie-l@googlegroups.com.
To unsubscribe from this group, send email to 
neonixie-l+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/neonixie-l?hl=en-GB.

Re: [neonixie-l] MAX1771 PSU instability

[Jan Rychter]

On 14 sty 2012, at 18:37, marta_kson wrote:
> The PCB is very hard to follow with those dreadful colors, Eagle?

Yes, sorry about that. But Eagle is the only low cost (or free) option for 
hobbyists and it isn't bad for simple designs.

> The FB trace is definitely no good. It's imperative to keep it just as
> a minimum pad right at the chip with just the two resistors connected.

Hmm. To make the low voltage part of the FB trace any shorter, I'd need to get 
rid of the potentiometer. Otherwise it can't be made any shorter, the resistors 
can't be physically closer to the chip.

[time passes]

Well, after several experiments it turned out that after removing the pot and 
shorting it with a horizontally-laid piece of wire everything works just fine. 
So it seems that the pot (a Bourns 3296-style multi-turn potentiometer, about 
1cm of height, as in http://www.bourns.com/pdfs/3296.pdf) was picking up 
interference. Could that be because of its vertical design?

But I'd like to better understand the comments (yours and Nick's) about the FB 
trace being too long. I thought that the critical part of that trace is the 
low-voltage part, from the resistor divider towards the FB pin. In my design 
the "long" part is actually just HV. I did not consider that to be very 
sensitive to EMI, as any voltage picked up here gets attenuated 20x or so by 
the resistor divider. Am I right?

The length of the HV trace to the resistor divider isn't different in my design 
and in Nick's design. It's just that I placed the output connector at the top 
of my board, while Nick has it on the bottom. But it's the same HV and it still 
needs to go around the whole chip and RSENSE resistor in particular. I could 
eliminate one row of vias next to RSENSE, which would make that trace 1mm 
shorter, but is it really worth the effort?

I think I'll need to redesign the board without the potentiometer, thus 
bringing the FB resistors much closer to the chip.

--J.

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To post to this group, send an email to neonixie-l@googlegroups.com.
To unsubscribe from this group, send email to 
neonixie-l+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/neonixie-l?hl=en-GB.

Re: [neonixie-l] MAX1771 PSU instability

[Jan Rychter]

Thanks for your suggestions!

On 13 sty 2012, at 17:17, John Rehwinkel wrote:
>> I have a supply that kind of works, but I get a large sawtooth-like ripple 
>> on the output. The ripple is 17V peak-to-peak at 170V and the frequency is 
>> 23-80Hz (depending on the load).
> 
> There are four things I would look at: current limiting, feedback 
> oscillation, power supply problems, and thermal issues.
> 
> The MAX1771 has built-in current limiting, which could cause the whole 
> shebang to cycle on and off, which could appear as a sawtooth after the 
> output filtering.  Make sure you have solid, low impedance connections to the 
> current sense resistor and back to the CS input.

The board layout is right here, in case you'd care to take a look: 
https://skitch.com/jrychter/g2tq6/max1771-psu-1.0

This is heavily based on Nick de Smith's layout, although I might have made it 
worse. Of course I can share the full design if I manage to work out the bugs.

> Power supply problems are the easiest - see if there's any similar 
> oscillation at the power supply input pin.  Nick's diagram shows separate 
> power inputs for the inductor and the MAX1771 - are you powering them 
> separately?  Either way, look at the input power, if it's a lab supply that's 
> going into current limit and retrying, you could get behaviour like this.  If 
> you're not powering them separately, it might be worth a try.

Hmm. I'm powering the board from a (cheap) lab power supply integrated into my 
soldering station. It is supposedly capable of providing 1A @ 12V and I'm only 
drawing between 50mA and 200mA, but…

I looked at the supply with a scope and there are dips of around .75V that 
correspond to my ripple frequency. I thought those are to be expected. Nothing 
more serious, though. Would 1771 be that sensitive?

I have a single supply trace. There is a 100uF capacitor, then the inductor, 
and then the 10uF tantalum cap right next to the 1771. I never thought about 
separate traces for power.

> For feedback oscillation, you'd have a phase shift happening somewhere in the 
> voltage regulation loop - or something that pretends to be phase shift.  I'd 
> look closely at the feedback resistors and make sure there aren't any 
> parallel capacitors inadvertently hooked to them.  I'd also carefully check 
> the output capacitors - both the electrolytic and the high frequency one.  If 
> they're not doing their job of absorbing rapid spikes and hash, the 
> controller chip can get confused about what's really going on.  Scoping those 
> points is tricky, due to all the high voltage, high current stuff going on 
> nearby, which will tend to couple into the probe lead, upsetting the 
> regulation further and confusing the scope display.

I thought feedback oscillation would result in higher frequency ripples. I 
don't see any parallel capacitors, just the ground plane below. My feedback 
path is substantially longer than in Nick's design, but it originates straight 
at the 4.7uF output reservoir capacitor and is relatively far away from 
anything else (everything around it is ground), so I thought I would be fine 
here.

The other modifications I made to the feedback path are: 1) I used two 
resistors in series because of high voltage and 2) my divider is really 1360k - 
8k2 (+5k pot) instead of the more usual 1M5 - 10k (+5k pot), because of the 
resistors I had. I did not expect this to make a difference.

Oh, also, the final 100nF capacitor is not mounted yet (I don't have one). I 
assumed this would only make a difference for higher frequency switching noise.

> The last thing that occurs to me is a thermal issue.  The MAX1771 will 
> throttle down its on-time when it gets hot, which could lead to thermal 
> cycling. 

This is definitely not the case, nothing gets hot except for the load 
resistors, I haven't gotten anywhere near peak currents this should be capable 
of.

I will keep looking based on your suggestions.

--J.

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To post to this group, send an email to neonixie-l@googlegroups.com.
To unsubscribe from this group, send email to 
neonixie-l+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/neonixie-l?hl=en-GB.

[neonixie-l] MAX1771 PSU instability

[Jan Rychter]

Nixie Veterans,

I'm looking for advice: I thought some of you here might have dealt with 
MAX1771 and know its quirks.

I've built a power supply based on Nick de Smith's excellent web page. I added 
a voltage doubler (for dekatrons), and a couple of linear stabilizers for other 
circuitry. I also built a larger current version (large inductor, RSENSE of 
25mOhm). I tried to follow the PCB design advice very carefully.

I have a supply that kind of works, but I get a large sawtooth-like ripple on 
the output. The ripple is 17V peak-to-peak at 170V and the frequency is 23-80Hz 
(depending on the load). I'm loading the PSU with 25kOhms of resistance, which 
should pull 6.8mA -- so this is not an unloaded run. The ripple frequency seems 
to change with the load, the amplitude remains the same, though. Smaller loads 
seem to decrease ripple frequency, larger loads increase it.

I expected to see a ripple of 1-2%, but not 10%. Something seems to be wrong.

Any ideas or suggestions on where to look now?

--J.

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To post to this group, send an email to neonixie-l@googlegroups.com.
To unsubscribe from this group, send email to 
neonixie-l+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/neonixie-l?hl=en-GB.

Re: [neonixie-l] Tube voltage

[Jan Rychter]

On 9 gru 2011, at 04:53, threeneurons wrote:

>>> "If I'm not mistaken the tube sets its voltage. "
>>> If I say 'zener diode', does that help a bit ? Or forward bias voltage
>>> of an 'ordinary' diode.
>>> "the lamp will "see" the full 200V (or more). Doesn't that damage the
>>> lamps?"
> No damage. Don't worry. Its current, not voltage that can kill a
> nixie. And then you need a lot of it. See my nixie page, here:
> 
> http://threeneurons.wordpress.com/nixie-power-supply/
> 
> It has a drawing showing what has been already discussed.

Thanks to everyone! What I was missing is that Nixies are actually "active" 
devices: they set their own voltage, so the sustain voltage is a *given*. A 
Nixie is not a resistor.

Now that that is clear I was able to calculate my anode resistors and light up 
my nixies properly. My IN-12As have a sustain voltage of 136V (which surprised 
me), so I get a 54V drop across the anode resistor. And they glow even at 1.5mA 
(which also surprised me).

Special thanks to threeneurons for his excellent pages and making the whole 
circuit collection public. It's a great learning resource!

Also, a related question: if I want bo build a mixed Nixie/Dekatron circuit 
that needs 180V and 400V, do you think it's better to:

a) build a PSU that produces 200V with a voltage doubler to 400V,
b) build a PSU that produces 400V right away and then put a voltage divider 
after it?

I looked at component availability and both approaches seem feasible.

--J.

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To post to this group, send an email to neonixie-l@googlegroups.com.
To unsubscribe from this group, send email to 
neonixie-l+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/neonixie-l?hl=en-GB.

[neonixie-l] Tube voltage

[Jan Rychter]

This might sound like a basic question, but I've done quite a bit of reading 
and can't quite understand the voltage requirements. I'm hoping for advice from 
more senior nixie hackers.

Let's look at an IN-12 lamp (which is what I have in hand). The datasheet says 
that normal anode voltage should not exceed 170V. Further down it also mentions 
200V as the *minimal* voltage for lighting up the tube, and 2.5mA as the 
minimal current.

Assuming I have a 180V supply, I need to have 160V (so as not to exceed the 
lamps specs?), so I need to "drop" 20V on the resistor at 2.5mA, which gives 
8kOhms. Does that sound right? But what about the start-up voltage?

What also worries me is -- let's assume I have a 200V (or higher) power supply. 
No matter what resistor I put in, as long as there is no current flowing (e.g. 
before the nixie lights up), the lamp will "see" the full 200V (or more). 
Doesn't that damage the lamps?

I know there are calculators out there, but I really need to understand this 
stuff.

thanks,
--J.

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To post to this group, send an email to neonixie-l@googlegroups.com.
To unsubscribe from this group, send email to 
neonixie-l+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/neonixie-l?hl=en-GB.

Re: [neonixie-l] Yet more 7441, 74141 questions

[Jan Rychter]

On 23 lis 2011, at 21:27, chuck richards wrote:
> Indeed, *is* there any modern low-current bcd-to-decimal
> high voltage ic that provides a one-package 16 pin DIP
> solution to this requirement?   Is there?  Please let me
> know if there is.

This is not exactly an answer to your question, but -- I thought I'd share m 
findings.

I was looking for something that would let me drive nixies through a serial 
interface.

The best I could find was the MC33999. It won't replace the 74141s, because of 
50V max output drain voltage, but it might save some circuitry if somebody 
wants to direct-drive nixies through a serial/SPI connection. While searching, 
I also found the MCP23S18 (SPI) and MCP23018 (I2C) I/O expanders, which could 
also be useful. You still need a driving transistor + resistor per cathode.

In my current design I decided to go with 74HC595s and 74141s; with 6 nixies 
you still need two 16-bit chips anyway, so three 595s are not a big deal.

--J.

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To post to this group, send an email to neonixie-l@googlegroups.com.
To unsubscribe from this group, send email to 
neonixie-l+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/neonixie-l?hl=en-GB.

Re: [neonixie-l] switching off -- do I need an anode driver at all?

[Jan Rychter]

On 23 lis 2011, at 18:26, Adam Jacobs wrote:

> I built a direct-drive clock with Moses' chip years ago with IN-8's. It turns 
> off for about 12 hours out of each day using only the blanking feature on the 
> 74141's. I don't notice any glow, the tubes seem to turn very cleanly "off".
> It might be because I used the soviet 74141's, though.. or maybe it only 
> works with specific tubes.
> I agree with David, the cleanest solution would be to simply turn off your HV 
> supply.
> 

I do have the soviet 74141s — could it be that they are different (don't have 
the zener diodes) from the 74141s made elsewhere?

My HV supply will be on a separate board, so turning it off isn't that obvious, 
but perhaps I'll redesign it to incorporate an "HV enable" pin. The simplest 
solution seems to be to implement a common driver for all the anodes with a 
MPSA92/MPSA42 pair. A single MPSA92 should have no problem driving 6 nixies.

--J.


> On 11/23/2011 8:22 AM, Jan Rychter wrote:
>> I'd like to be able to periodically switch off my nixies in order to extend 
>> their lifespan. I have a non-multiplexed direct-drive circuit with 74141s 
>> driving the nixies. I've been wondering - is it enough to keep the 74141s 
>> switched off, or do I need to build in a circuit that will remove the anode 
>> voltage?
>> 
>> Normally I would just leave the anodes powered and there should be no 
>> current flowing, but nixies and high voltages are not something I understand 
>> very well.
>> 
>> --J.
>> 
> 
> -- 
> You received this message because you are subscribed to the Google Groups 
> "neonixie-l" group.
> To post to this group, send an email to neonixie-l@googlegroups.com.
> To unsubscribe from this group, send email to 
> neonixie-l+unsubscr...@googlegroups.com.
> For more options, visit this group at 
> http://groups.google.com/group/neonixie-l?hl=en-GB.
> 

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To post to this group, send an email to neonixie-l@googlegroups.com.
To unsubscribe from this group, send email to 
neonixie-l+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/neonixie-l?hl=en-GB.

[neonixie-l] switching off -- do I need an anode driver at all?

[Jan Rychter]

I'd like to be able to periodically switch off my nixies in order to extend 
their lifespan. I have a non-multiplexed direct-drive circuit with 74141s 
driving the nixies. I've been wondering - is it enough to keep the 74141s 
switched off, or do I need to build in a circuit that will remove the anode 
voltage?

Normally I would just leave the anodes powered and there should be no current 
flowing, but nixies and high voltages are not something I understand very well.

--J.

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To post to this group, send an email to neonixie-l@googlegroups.com.
To unsubscribe from this group, send email to 
neonixie-l+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/neonixie-l?hl=en-GB.