Re: [time-nuts] Fast risetime pulse generator

[Morris Odell]

Hi all,

Brooke Clark wrote:

> In the HP 5100 Frequency Synthesizer they use a heater on the SRD in order to 
> increase it's 
> lifetime so that the 3 MHz drive will generate the needed harmonics. 
> http://www.prc68.com/I/HP5100.shtml These 
> assemblies are no longer available and getting an SRD with a lifetime long 
> enough to work with a 3 MHz input is a common 
> as chicken teeth.  A solution is to get a large stud mounted Varactor and 
> these do have long lifetimes.

I was so pleased to see this. It mirrored my experience exactly. A few years 
ago I restored a 5100A and its 5110A exciter. The SRD heater in the 5110A had 
developed a failed thermostat and was a charred blob. I made up a test jig and 
tested lots and lots of different diodes and as it turned out, the best 
candidate was a stud mounted varactor out of a junked transmitter from the old 
VHF cellphone system. It's worked like a charm ever since. This post (and 
webpage) has explained it beautifully.

Morris



___
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

Re: [time-nuts] Fast risetime pulse generator

[paul swed]

Bill
Now this discussion gets really interesting. You are doing something and
will report results. You have my attention.
Regards
Paul
WB8TSL

On Fri, Apr 15, 2016 at 12:27 PM, BIll Ezell  wrote:

> I'm amazed at how much traffic this has generated.
> Thanks Brooke, you're exactly on-topic as usual.
>
> As I said originally, I wanted a quick-and-dirty-and-cheap, and I'm using
> an SMMD-835 SRD. Making up the test layout tonight.
> I'm starting with a 100 uA bias on the diode, and a 20 pF diode-to-load
> coupling cap, driving the SRD with a 5V 1Mhz square-wave.
>
> I'm also going to try the 2N3904 Vce-reverse-bias avalanche idea just for
> fun.
>
> I'll post again after I have some results.
>
> PS - as I've mentioned before, I trim these things. I hate the
> endlessly-requoted replies, impossible to parse what's actually been said.
>
> --- Brooke post, trimmed --
> Hi:
>
> The idea with a step recovery diode is that during forward bias the diode
> stores charge.  When the polarity reverses the diode continues to conduct
> until the charge has been depleted.  At that time the diode stops
> conducting and opens.
> Suppose this was done using a square wave. The length of the positive
> (charging) half of waveform needs to be long enough so that the diode if
> fully charged (a function of the diode lifetime and charging current).
> Then on the negative half (discharging) when all the charge has been
> extracted the diode turns of making a sharp edge.  So the diode lifetime is
> important in relation to input drive frequency.  For a sinewave drive the
> lifetime should be about 1/4 the period at the recommended drive level.
>
> --
> Bill Ezell
> --
> The day Microsoft makes something that doesn't suck
> will be the day they make vacuum cleaners.
> Or maybe Windows 10.
>
> ___
> time-nuts mailing list -- time-nuts@febo.com
> To unsubscribe, go to
> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.
>
___
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

Re: [time-nuts] Fast risetime pulse generator

[BIll Ezell]

I'm amazed at how much traffic this has generated.
Thanks Brooke, you're exactly on-topic as usual.

As I said originally, I wanted a quick-and-dirty-and-cheap, and I'm 
using an SMMD-835 SRD. Making up the test layout tonight.
I'm starting with a 100 uA bias on the diode, and a 20 pF diode-to-load 
coupling cap, driving the SRD with a 5V 1Mhz square-wave.


I'm also going to try the 2N3904 Vce-reverse-bias avalanche idea just 
for fun.


I'll post again after I have some results.

PS - as I've mentioned before, I trim these things. I hate the 
endlessly-requoted replies, impossible to parse what's actually been said.


--- Brooke post, trimmed --
Hi:

The idea with a step recovery diode is that during forward bias the 
diode stores charge.  When the polarity reverses the diode continues to 
conduct until the charge has been depleted.  At that time the diode 
stops conducting and opens.
Suppose this was done using a square wave. The length of the positive 
(charging) half of waveform needs to be long enough so that the diode if 
fully charged (a function of the diode lifetime and charging current).  
Then on the negative half (discharging) when all the charge has been 
extracted the diode turns of making a sharp edge.  So the diode lifetime 
is important in relation to input drive frequency.  For a sinewave drive 
the lifetime should be about 1/4 the period at the recommended drive level.


--
Bill Ezell
--
The day Microsoft makes something that doesn't suck
will be the day they make vacuum cleaners.
Or maybe Windows 10.

___
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

Re: [time-nuts] Fast risetime pulse generator

[David]

On Wed, 13 Apr 2016 22:52:28 -0500, you wrote:

>On Wed, April 13, 2016 5:32 pm, David wrote:
>> Step recovery diodes are very similar in construction to varactor
>> diodes so the later can be used in step recovery applications with
>> qualification.
>
>Agilent app note AN1054 discusses using PIN diodes as frequency
>multipliers.  I believe that the PIN diodes were used in the manner of a
>step recovery diode, as a less expensive and more easily attainable
>replacement.

PIN diodes will certainly work as well and maybe even better since
they will have less capacitance.
___
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

Re: [time-nuts] Fast risetime pulse generator

[Brooke Clarke]

Hi:

The idea with a step recovery diode is that during forward bias the diode stores charge.  When the polarity reverses the 
diode continues to conduct until the charge has been depleted.  At that time the diode stops conducting and opens.
Suppose this was done using a square wave. The length of the positive (charging) half of waveform needs to be long 
enough so that the diode if fully charged (a function of the diode lifetime and charging current).  Then on the negative 
half (discharging) when all the charge has been extracted the diode turns of making a sharp edge.  So the diode lifetime 
is important in relation to input drive frequency.  For a sinewave drive the lifetime should be about 1/4 the period at 
the recommended drive level.

http://www.prc68.com/I/Diodes.html
Diodes that store charge have different names and that's related to how they are characterized.  So PIN, Varactor, SRD 
and Noise are all diodes that store charge, and can all be used in any of those application if you have good 
characterization data.  In the HP 5100 Frequency Synthesizer they use a heater on the SRD in order to increase it's 
lifetime so that the 3 MHz drive will generate the needed harmonics. http://www.prc68.com/I/HP5100.shtml These 
assemblies are no longer available and getting an SRD with a lifetime long enough to work with a 3 MHz input is a common 
as chicken teeth.  A solution is to get a large stud mounted Varactor and these do have long lifetimes.


Note a "Comb Generator" is an SRD with an input matching circuit (typically 100 MHz) specified in the frequency domain, 
but can also be thought of as a fast pulse generator.  I designed and built a hermetically sealed cylindrical module 
that was 1 100 MHz input comb generator at Aertech. http://www.prc68.com/I/Aertech.shtml#Comb 


The HP 8406 is a lower frequency bench top instrument.
http://www.prc68.com/I/HP8406.shtml

PS Talking about impedance with a non linear device like a diode can lead to 
unexpected results.

--
Have Fun,

Brooke Clarke
http://www.PRC68.com
http://www.end2partygovernment.com/2012Issues.html
The lesser of evils is still evil.

 Original Message 

Agilent app note AN1054


___
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

Re: [time-nuts] Fast risetime pulse generator

[jimlux]

On 4/13/16 7:50 PM, bownes wrote:


Maybe a krytron? If you are able to get one anyway. ;)


that was my first thought.. or some form of triggered sparkgap
Went and looked up the data sheet for a KN6

Their rise time isn't that fast, what sets them apart from other high 
current switching devices is that they are low jitter, while switching 
high currents.  Driving them is a pain (they need moderate high voltage, 
etc.).  Switching hundreds of amps with tens of ns jitter is useful in 
some applications, but in time nuts terms, not so wonderful.








On Apr 13, 2016, at 18:32, David  wrote:


On Wed, 13 Apr 2016 00:03:03 -0500, you wrote:

...

If you are building up something, you probably want a SRD (step
recovery diode) or tunnel diode. But both may be hard to find and
expensive these days.


Step recovery diodes are very similar in construction to varactor
diodes so the later can be used in step recovery applications with
qualification.

I have also seen the 1N4148 used as a step recovery diode when driven
by an avalanche pulser but I do not know if it was qualified.
___
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

___
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.



___
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

Re: [time-nuts] Fast risetime pulse generator

[Chris Caudle]

On Wed, April 13, 2016 5:32 pm, David wrote:
> Step recovery diodes are very similar in construction to varactor
> diodes so the later can be used in step recovery applications with
> qualification.

Agilent app note AN1054 discusses using PIN diodes as frequency
multipliers.  I believe that the PIN diodes were used in the manner of a
step recovery diode, as a less expensive and more easily attainable
replacement.

-- 
Chris Caudle




___
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

Re: [time-nuts] Fast risetime pulse generator

[bownes]

Maybe a krytron? If you are able to get one anyway. ;)

> On Apr 13, 2016, at 18:32, David  wrote:
> 
>> On Wed, 13 Apr 2016 00:03:03 -0500, you wrote:
>> 
>> ...
>> 
>> If you are building up something, you probably want a SRD (step
>> recovery diode) or tunnel diode. But both may be hard to find and
>> expensive these days.
> 
> Step recovery diodes are very similar in construction to varactor
> diodes so the later can be used in step recovery applications with
> qualification.
> 
> I have also seen the 1N4148 used as a step recovery diode when driven
> by an avalanche pulser but I do not know if it was qualified.
> ___
> time-nuts mailing list -- time-nuts@febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.
___
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

Re: [time-nuts] Fast risetime pulse generator

[Magnus Danielson]

On 04/14/2016 12:32 AM, David wrote:

On Wed, 13 Apr 2016 00:03:03 -0500, you wrote:


...

If you are building up something, you probably want a SRD (step
recovery diode) or tunnel diode. But both may be hard to find and
expensive these days.


Step recovery diodes are very similar in construction to varactor
diodes so the later can be used in step recovery applications with
qualification.

I have also seen the 1N4148 used as a step recovery diode when driven
by an avalanche pulser but I do not know if it was qualified.


Check Fig.7 in the datasheet:
http://www.nxp.com/documents/data_sheet/1N4148_1N4448.pdf

A SRD source I just found is:
http://www.adsemi.com/diodes/step_recovery.shtml

Cheers,
Magnus
___
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

Re: [time-nuts] Fast risetime pulse generator

[David]

On Wed, 13 Apr 2016 00:03:03 -0500, you wrote:

>...
> 
>If you are building up something, you probably want a SRD (step
>recovery diode) or tunnel diode. But both may be hard to find and
>expensive these days.

Step recovery diodes are very similar in construction to varactor
diodes so the later can be used in step recovery applications with
qualification.

I have also seen the 1N4148 used as a step recovery diode when driven
by an avalanche pulser but I do not know if it was qualified.
___
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

Re: [time-nuts] Fast risetime pulse generator

[Gerhard Hoffmann]

Am 13.04.2016 um 07:03 schrieb Bill Byrom:

I agree with earlier comments that the Analog Devices SiGe voltage
comparators appear to be a good choice, with 37 ps typical rise/fall
(20%/80%):
http://www.analog.com/media/en/technical-documentation/data-sheets/ADCMP580_581_582.pdf
The evaluation board single unit price is about $299:
http://www.analog.com/media/en/technical-documentation/evaluation-documentation/310187823ADCMP572_3_80_1_2EB_0.pdf
  


If that is too much, you can do it in the style of my home-etched quickie:
18 * 14 mm, 0.5mm thick FR4. The trick is to get from chip output into
the semi-rigid coax as quickly as possible. The CML outputs have an
integrated 50 Ohms back termination. SMA launchers will probably
complicate things.

This is what has generated the waveforms in my previous post. Note that
the 20 GHz sampler is already close to its own risetime limit, so what
you see is probably worse than reality. I should have used the 50 GHz
plug in with its 9 ps risetime but then I usually avoid that because
I don't want to put it in jeopardy.

If one wants more voltage, an alternative might be a driver for 10 GBPS
DFB laser diodes. 5V in 35 ps seems possible.  ADN2525 or so.

Hittite is now AD. That is a good thing. Mouser or Digikey and no longer
prices and old data sheets under NDA.

If you think at Potato Semi chips, watch the loading conditions. Buffers
that can barely drive a 0.6 pF FET probe are a bad joke.

regards, Gerhard


___
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

Re: [time-nuts] Fast risetime pulse generator

[Bill Byrom]

The Tektronix 067-0681-01 was widely used in the 1970's and 1980's for
oscilloscope calibration. It produces a <125 ps risetime edge using a
tunnel diode. But it was designed for use with a large driving voltage
slow pulse (60 to 100 V p-p):
http://w140.com/tek_067-0681-01.pdf
http://w140.com/tekwiki/wiki/067-0681-01
 
The Tektronix SD-24 TDR sampling head (which was used in the 1180x and
CSA80x mainframes, such as the 11801B and CSA803A) generated a 27 ps
risetime rising or falling 250 mV step from one or two channels. But
you would need a working sampling scope mainframe to operate the
sampling head.
 
There are many other products (past and present) which can generate <150
ps risetime pulses. For example, the Tektronix AWG70002A arbitrary
waveform generator has <27 ps risetime (10 bit D/A analog output, 500
mVp-p single-ended or 1 Vp-p differential) and <35 ps marker (digital 1
V p-p) outputs. This product uses SiGe technology. But I'm guessing that
you are looking for something inexpensive that you can lash up rather
than a commercial generator.
 
If you are building up something, you probably want a SRD (step
recovery diode) or tunnel diode. But both may be hard to find and
expensive these days.
 
I agree with earlier comments that the Analog Devices SiGe voltage
comparators appear to be a good choice, with 37 ps typical rise/fall
(20%/80%):
http://www.analog.com/media/en/technical-documentation/data-sheets/ADCMP580_581_582.pdf
The evaluation board single unit price is about $299:
http://www.analog.com/media/en/technical-documentation/evaluation-documentation/310187823ADCMP572_3_80_1_2EB_0.pdf
 
--
Bill Byrom N5BB
 
 
 
On Tue, Apr 12, 2016, at 09:51 PM, Rob Sherwood. wrote:
> I have both HP and Tektronix pulse generators.  Neither are
> expensive and
> both have been very reliable.  Since I need some really slow rep rates
> for receiver testing, I use my Tek 115 triggered with an HP 8904A
> synthesizer.
> Rob, NC0B
>
> Sent from my iPad
>
>> On Apr 12, 2016, at 5:05 PM, "Logan Cummings"
>>  wrote:
>>
>>
>>
>> Hi Bill,
>>
>> Lots of resources out there on EEVblog forums and elsewhere. One nice
>> alternative to the avalanche pulser is a very fast comparator
>> like the
>> ADCMP580 from Analog Devices. Generate a pulse of whatever width
>> you want
>> with slower gear, and have the comparator sharpen the edges.
>>
>> See here:
>> http://www.eevblog.com/forum/testgear/inexpensive-diy-50ps-ultra-fast-pulse-generator/
>> and
>> http://www.starlino.com/build-a-really-fast-pulse-generator-50ps-rise-time-using-an-ultra-fast-sige-comparator.html
>>
>> For a couple of examples. The EEVblog thread also has a link to
>> info on an
>> avalanche pulser using a transistor that avalanches at 30V rather
>> than 70+.
>>
>> -Logan
>>
>> On Tue, Apr 12, 2016 at 2:28 PM, Alan Melia
>> 
>> wrote:
>>
>>> Bill, Avalanche pulse gens only require high voltage because of
>>> the high
>>> VBRcbo and the gain of normal NPN transistors. I cant find the
>>> reference
>>> now it might have been a 1970s Ham Radio but if you use the same
>>> circuit as
>>> Jim but put an NPN "upside down" that is emitter where the
>>> collector is in
>>> Jim's circuit you can fire off fast pulses from a 12v supply,
>>> instead of
>>> requiring 70 to 100v. I do wish I could locate the source as I
>>> have had
>>> several arguments about it :-))in the nicest possible way of
>>> course.
>>>
>>> Alan
>>> G3NYK
>>>
>>> - Original Message - From: "BIll Ezell" 
>>> To: 
>>> Sent: Tuesday, April 12, 2016 9:30 PM
>>> Subject: [time-nuts] Fast risetime pulse generator
>>>
>>>
>>>
>>> (cross-posted to volt-nuts)
 After paying only limited attention to this topic, I suddenly have
 a need
 for a pulse generator that has <150 ps risetime and a pulse width
 of at
 least 2 ns. 100mv amplitude or more is fine. I've looked at the
 classic Jim
 Williams avalanche generator, but I don't want to have to deal
 with the
 (relatively) high voltage source needed.

 I've done microwave design using Gunn diodes, so I'm drawn to
 using a
 step-recovery diode. The topology seems very straightforward, and
 I can
 build it right onto a BNC connector, no PCB.

 I'm thinking using an SMD835 diode, biased at ~1ma. The (sketchy)
 datasheet claims a T of 20 nsecs and a Tr of 85 ps, Cj of 0.4 to
 0.8 pf.

 Questions:

 The obvious, is it reasonable?

 Is the bias current reasonable? I'm assuming the bias current is
 actually
 dependent on the repetition rate, you need enough current to
 replenish the
 charge within one pulse cycle. I suppose I could compute it from
 the stated
 junction capacitance, but I'm not sure that's the only factor.

 Will the stored charge actually give me the desired transition
 rate into
 50 ohms? Hmm, again I should be 

Re: [time-nuts] Fast risetime pulse generator

[Rob Sherwood .]

Oops, sorry I didn't realize you needed 150 ps rise time. Rob

Sent from my iPad

> On Apr 12, 2016, at 6:02 PM, "jimlux"  wrote:
> 
> 
> 
>> On 4/12/16 1:30 PM, BIll Ezell wrote:
>> (cross-posted to volt-nuts)
>> After paying only limited attention to this topic, I suddenly have a
>> need for a pulse generator that has <150 ps risetime and a pulse width
>> of at least 2 ns. 100mv amplitude or more is fine. I've looked at the
>> classic Jim Williams avalanche generator, but I don't want to have to
>> deal with the (relatively) high voltage source needed.
>> 
>> I've done microwave design using Gunn diodes, so I'm drawn to using a
>> step-recovery diode. The topology seems very straightforward, and I can
>> build it right onto a BNC connector, no PCB.
>> 
>> I'm thinking using an SMD835 diode, biased at ~1ma. The (sketchy)
>> datasheet claims a T of 20 nsecs and a Tr of 85 ps, Cj of 0.4 to 0.8 pf.
>> 
> 
> how about a fast digital buffer?
> I don't think Potatochip semi is quite in the 150ps range (although you could 
> look at their website and see).
> 
> But we use microwave dividers and such all the time at 8 GHz.  Check places 
> like Hittite.
> 
> 
> 
> 
>> Questions:
>> 
>> The obvious, is it reasonable?
>> 
>> Is the bias current reasonable? I'm assuming the bias current is
>> actually dependent on the repetition rate, you need enough current to
>> replenish the charge within one pulse cycle. I suppose I could compute
>> it from the stated junction capacitance, but I'm not sure that's the
>> only factor.
>> 
>> Will the stored charge actually give me the desired transition rate into
>> 50 ohms? Hmm, again I should be able to compute this, but any other
>> factors ignoring the non-diode ones like cap inductance?
>> 
>> How should I compute the coupling cap from the diode to the load? Use
>> the impedance at the pulse rep rate? Seems reasonable. BTW, I don't care
>> about droop in the  pulse, just the risetime.  (measuring overshoot in
>> an HF amp). Again, just want to verify that the obvious answer is the
>> correct one. I clearly need to be very careful about the inductance.
>> 
>> Thanks, Bill
>> 
> 
> ___
> time-nuts mailing list -- time-nuts@febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.
> 
> 
> 
> --
> If this email is spam, report it to
> https://support.onlymyemail.com/view/report_spam/ODExMjI6MTg3MTg2MDE0ODpyb2JAbmMwYi5jb206ZGVsaXZlcmVk
> 
___
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

Re: [time-nuts] Fast risetime pulse generator

[Rob Sherwood .]

I have both HP and Tektronix pulse generators.  Neither are expensive and both 
have been very reliable.  Since I need some really slow rep rates for receiver 
testing, I use my Tek 115 triggered with an HP 8904A synthesizer.  
Rob, NC0B

Sent from my iPad

> On Apr 12, 2016, at 5:05 PM, "Logan Cummings"  
> wrote:
> 
> 
> 
> Hi Bill,
> 
>Lots of resources out there on EEVblog forums and elsewhere. One nice
> alternative to the avalanche pulser is a very fast comparator like the
> ADCMP580 from Analog Devices. Generate a pulse of whatever width you want
> with slower gear, and have the comparator sharpen the edges.
> 
> See here:
> http://www.eevblog.com/forum/testgear/inexpensive-diy-50ps-ultra-fast-pulse-generator/
> and
> http://www.starlino.com/build-a-really-fast-pulse-generator-50ps-rise-time-using-an-ultra-fast-sige-comparator.html
> 
> For a couple of examples. The EEVblog thread also has a link to info on an
> avalanche pulser using a transistor that avalanches at 30V rather than 70+.
> 
> -Logan
> 
> On Tue, Apr 12, 2016 at 2:28 PM, Alan Melia 
> wrote:
> 
>> Bill, Avalanche pulse gens only require high voltage because of the high
>> VBRcbo and the gain of normal NPN transistors. I cant find the reference
>> now it might have been a 1970s Ham Radio but if you use the same circuit as
>> Jim but put an NPN "upside down" that is emitter where the collector is in
>> Jim's circuit you can fire off fast pulses from a 12v supply, instead of
>> requiring 70 to 100v. I do wish I could locate the source as I have had
>> several arguments about it :-))in the nicest possible way of course.
>> 
>> Alan
>> G3NYK
>> 
>> - Original Message - From: "BIll Ezell" 
>> To: 
>> Sent: Tuesday, April 12, 2016 9:30 PM
>> Subject: [time-nuts] Fast risetime pulse generator
>> 
>> 
>> 
>> (cross-posted to volt-nuts)
>>> After paying only limited attention to this topic, I suddenly have a need
>>> for a pulse generator that has <150 ps risetime and a pulse width of at
>>> least 2 ns. 100mv amplitude or more is fine. I've looked at the classic Jim
>>> Williams avalanche generator, but I don't want to have to deal with the
>>> (relatively) high voltage source needed.
>>> 
>>> I've done microwave design using Gunn diodes, so I'm drawn to using a
>>> step-recovery diode. The topology seems very straightforward, and I can
>>> build it right onto a BNC connector, no PCB.
>>> 
>>> I'm thinking using an SMD835 diode, biased at ~1ma. The (sketchy)
>>> datasheet claims a T of 20 nsecs and a Tr of 85 ps, Cj of 0.4 to 0.8 pf.
>>> 
>>> Questions:
>>> 
>>> The obvious, is it reasonable?
>>> 
>>> Is the bias current reasonable? I'm assuming the bias current is actually
>>> dependent on the repetition rate, you need enough current to replenish the
>>> charge within one pulse cycle. I suppose I could compute it from the stated
>>> junction capacitance, but I'm not sure that's the only factor.
>>> 
>>> Will the stored charge actually give me the desired transition rate into
>>> 50 ohms? Hmm, again I should be able to compute this, but any other factors
>>> ignoring the non-diode ones like cap inductance?
>>> 
>>> How should I compute the coupling cap from the diode to the load? Use the
>>> impedance at the pulse rep rate? Seems reasonable. BTW, I don't care about
>>> droop in the  pulse, just the risetime.  (measuring overshoot in an HF
>>> amp). Again, just want to verify that the obvious answer is the correct
>>> one. I clearly need to be very careful about the inductance.
>>> 
>>> Thanks, Bill
>>> 
>>> --
>>> Bill Ezell
>>> --
>>> The day Microsoft makes something that doesn't suck
>>> will be the day they make vacuum cleaners.
>>> Or maybe Windows 10.
>>> 
>>> ___
>>> time-nuts mailing list -- time-nuts@febo.com
>>> To unsubscribe, go to
>>> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>>> and follow the instructions there.
>> 
>> ___
>> time-nuts mailing list -- time-nuts@febo.com
>> To unsubscribe, go to
>> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>> and follow the instructions there.
> ___
> time-nuts mailing list -- time-nuts@febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.
> 
> 
> 
> --
> If this email is spam, report it to
> https://support.onlymyemail.com/view/report_spam/ODExMjI6MTg3MTg0ODczMjpyb2JAbmMwYi5jb206ZGVsaXZlcmVk
> 
___
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

Re: [time-nuts] Fast risetime pulse generator

[Bruce Griffiths]

On Wednesday, April 13, 2016 12:36:54 AM Gerhard Hoffmann wrote:
> Am 12.04.2016 um 22:30 schrieb BIll Ezell:
> > (cross-posted to volt-nuts)
> > After paying only limited attention to this topic, I suddenly have a
> > need for a pulse generator that has <150 ps risetime and a pulse width
> > of at least 2 ns. 100mv amplitude or more is fine. I've looked at the
> > classic Jim Williams avalanche generator, but I don't want to have to
> > deal with the (relatively) high voltage source needed.
> 
> This is what an Analog Devices ADCMP580 comparator produces. Available
> from Digi-Key.
> 
> regards, Gerhard
The LTC6954 programmable divider evaluation board has ECL outputs with 110ps 
transition times. Accepts input up to 1.4GHz.
I have cascaded a couple of these to quickly produce a low jitter output 
divided down from 100MHz.

Bruce
___
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

Re: [time-nuts] Fast risetime pulse generator

[Bruce Griffiths]

http://www.its.caltech.edu/~hajimiri/pdf/non-linear.transmission.pdf
The above indicates how cascaded NLTLs using CMOS processes can be used to 
construct NLTLs that sharpen both edges of a pulse. 2.5ps fwhm impulses can 
also be produced.
Bruce

On Tuesday, April 12, 2016 04:42:06 PM jimlux wrote:
> On 4/12/16 1:30 PM, BIll Ezell wrote:
> > (cross-posted to volt-nuts)
> > After paying only limited attention to this topic, I suddenly have a
> > need for a pulse generator that has <150 ps risetime and a pulse width
> > of at least 2 ns. 100mv amplitude or more is fine. I've looked at the
> > classic Jim Williams avalanche generator, but I don't want to have to
> > deal with the (relatively) high voltage source needed.
> > 
> > I've done microwave design using Gunn diodes, so I'm drawn to using a
> > step-recovery diode. The topology seems very straightforward, and I can
> > build it right onto a BNC connector, no PCB.
> > 
> > I'm thinking using an SMD835 diode, biased at ~1ma. The (sketchy)
> > datasheet claims a T of 20 nsecs and a Tr of 85 ps, Cj of 0.4 to 0.8 pf.
> 
> how about a fast digital buffer?
> I don't think Potatochip semi is quite in the 150ps range (although you
> could look at their website and see).
> 
> But we use microwave dividers and such all the time at 8 GHz.  Check
> places like Hittite.
> 
> > Questions:
> > 
> > The obvious, is it reasonable?
> > 
> > Is the bias current reasonable? I'm assuming the bias current is
> > actually dependent on the repetition rate, you need enough current to
> > replenish the charge within one pulse cycle. I suppose I could compute
> > it from the stated junction capacitance, but I'm not sure that's the
> > only factor.
> > 
> > Will the stored charge actually give me the desired transition rate into
> > 50 ohms? Hmm, again I should be able to compute this, but any other
> > factors ignoring the non-diode ones like cap inductance?
> > 
> > How should I compute the coupling cap from the diode to the load? Use
> > the impedance at the pulse rep rate? Seems reasonable. BTW, I don't care
> > about droop in the  pulse, just the risetime.  (measuring overshoot in
> > an HF amp). Again, just want to verify that the obvious answer is the
> > correct one. I clearly need to be very careful about the inductance.
> > 
> > Thanks, Bill
> 
> ___
> time-nuts mailing list -- time-nuts@febo.com
> To unsubscribe, go to
> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the
> instructions there.

___
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

Re: [time-nuts] Fast risetime pulse generator

[Gerhard Hoffmann]

Am 12.04.2016 um 22:30 schrieb BIll Ezell:

(cross-posted to volt-nuts)
After paying only limited attention to this topic, I suddenly have a 
need for a pulse generator that has <150 ps risetime and a pulse width 
of at least 2 ns. 100mv amplitude or more is fine. I've looked at the 
classic Jim Williams avalanche generator, but I don't want to have to 
deal with the (relatively) high voltage source needed.
This is what an Analog Devices ADCMP580 comparator produces. Available 
from Digi-Key.


regards, Gerhard
___
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

Re: [time-nuts] Fast risetime pulse generator

[jimlux]

On 4/12/16 1:30 PM, BIll Ezell wrote:

(cross-posted to volt-nuts)
After paying only limited attention to this topic, I suddenly have a
need for a pulse generator that has <150 ps risetime and a pulse width
of at least 2 ns. 100mv amplitude or more is fine. I've looked at the
classic Jim Williams avalanche generator, but I don't want to have to
deal with the (relatively) high voltage source needed.

I've done microwave design using Gunn diodes, so I'm drawn to using a
step-recovery diode. The topology seems very straightforward, and I can
build it right onto a BNC connector, no PCB.

I'm thinking using an SMD835 diode, biased at ~1ma. The (sketchy)
datasheet claims a T of 20 nsecs and a Tr of 85 ps, Cj of 0.4 to 0.8 pf.



how about a fast digital buffer?
I don't think Potatochip semi is quite in the 150ps range (although you 
could look at their website and see).


But we use microwave dividers and such all the time at 8 GHz.  Check 
places like Hittite.






Questions:

The obvious, is it reasonable?

Is the bias current reasonable? I'm assuming the bias current is
actually dependent on the repetition rate, you need enough current to
replenish the charge within one pulse cycle. I suppose I could compute
it from the stated junction capacitance, but I'm not sure that's the
only factor.

Will the stored charge actually give me the desired transition rate into
50 ohms? Hmm, again I should be able to compute this, but any other
factors ignoring the non-diode ones like cap inductance?

How should I compute the coupling cap from the diode to the load? Use
the impedance at the pulse rep rate? Seems reasonable. BTW, I don't care
about droop in the  pulse, just the risetime.  (measuring overshoot in
an HF amp). Again, just want to verify that the obvious answer is the
correct one. I clearly need to be very careful about the inductance.

Thanks, Bill



___
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

Re: [time-nuts] Fast risetime pulse generator

[Logan Cummings]

Hi Bill,

Lots of resources out there on EEVblog forums and elsewhere. One nice
alternative to the avalanche pulser is a very fast comparator like the
ADCMP580 from Analog Devices. Generate a pulse of whatever width you want
with slower gear, and have the comparator sharpen the edges.

See here:
http://www.eevblog.com/forum/testgear/inexpensive-diy-50ps-ultra-fast-pulse-generator/
and
http://www.starlino.com/build-a-really-fast-pulse-generator-50ps-rise-time-using-an-ultra-fast-sige-comparator.html

For a couple of examples. The EEVblog thread also has a link to info on an
avalanche pulser using a transistor that avalanches at 30V rather than 70+.

-Logan

On Tue, Apr 12, 2016 at 2:28 PM, Alan Melia 
wrote:

> Bill, Avalanche pulse gens only require high voltage because of the high
> VBRcbo and the gain of normal NPN transistors. I cant find the reference
> now it might have been a 1970s Ham Radio but if you use the same circuit as
> Jim but put an NPN "upside down" that is emitter where the collector is in
> Jim's circuit you can fire off fast pulses from a 12v supply, instead of
> requiring 70 to 100v. I do wish I could locate the source as I have had
> several arguments about it :-))in the nicest possible way of course.
>
> Alan
> G3NYK
>
> - Original Message - From: "BIll Ezell" 
> To: 
> Sent: Tuesday, April 12, 2016 9:30 PM
> Subject: [time-nuts] Fast risetime pulse generator
>
>
>
> (cross-posted to volt-nuts)
>> After paying only limited attention to this topic, I suddenly have a need
>> for a pulse generator that has <150 ps risetime and a pulse width of at
>> least 2 ns. 100mv amplitude or more is fine. I've looked at the classic Jim
>> Williams avalanche generator, but I don't want to have to deal with the
>> (relatively) high voltage source needed.
>>
>> I've done microwave design using Gunn diodes, so I'm drawn to using a
>> step-recovery diode. The topology seems very straightforward, and I can
>> build it right onto a BNC connector, no PCB.
>>
>> I'm thinking using an SMD835 diode, biased at ~1ma. The (sketchy)
>> datasheet claims a T of 20 nsecs and a Tr of 85 ps, Cj of 0.4 to 0.8 pf.
>>
>> Questions:
>>
>> The obvious, is it reasonable?
>>
>> Is the bias current reasonable? I'm assuming the bias current is actually
>> dependent on the repetition rate, you need enough current to replenish the
>> charge within one pulse cycle. I suppose I could compute it from the stated
>> junction capacitance, but I'm not sure that's the only factor.
>>
>> Will the stored charge actually give me the desired transition rate into
>> 50 ohms? Hmm, again I should be able to compute this, but any other factors
>> ignoring the non-diode ones like cap inductance?
>>
>> How should I compute the coupling cap from the diode to the load? Use the
>> impedance at the pulse rep rate? Seems reasonable. BTW, I don't care about
>> droop in the  pulse, just the risetime.  (measuring overshoot in an HF
>> amp). Again, just want to verify that the obvious answer is the correct
>> one. I clearly need to be very careful about the inductance.
>>
>> Thanks, Bill
>>
>> --
>> Bill Ezell
>> --
>> The day Microsoft makes something that doesn't suck
>> will be the day they make vacuum cleaners.
>> Or maybe Windows 10.
>>
>> ___
>> time-nuts mailing list -- time-nuts@febo.com
>> To unsubscribe, go to
>> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>> and follow the instructions there.
>>
>
> ___
> time-nuts mailing list -- time-nuts@febo.com
> To unsubscribe, go to
> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.
>
___
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

Re: [time-nuts] Fast risetime pulse generator

[Alexander Pummer]

Hi Bill
check out the classical fast rise time pulse generator the mercury 
whetted relay, with a good lay-out you could get easily 50psec rise 
time, but rather use an SMA than BNC.

You could look for an old Tektronix pulse generator used that principle
73
KJ6UHN
Alex

On 4/12/2016 1:30 PM, BIll Ezell wrote:

(cross-posted to volt-nuts)
After paying only limited attention to this topic, I suddenly have a 
need for a pulse generator that has <150 ps risetime and a pulse width 
of at least 2 ns. 100mv amplitude or more is fine. I've looked at the 
classic Jim Williams avalanche generator, but I don't want to have to 
deal with the (relatively) high voltage source needed.


I've done microwave design using Gunn diodes, so I'm drawn to using a 
step-recovery diode. The topology seems very straightforward, and I 
can build it right onto a BNC connector, no PCB.


I'm thinking using an SMD835 diode, biased at ~1ma. The (sketchy) 
datasheet claims a T of 20 nsecs and a Tr of 85 ps, Cj of 0.4 to 0.8 pf.


Questions:

The obvious, is it reasonable?

Is the bias current reasonable? I'm assuming the bias current is 
actually dependent on the repetition rate, you need enough current to 
replenish the charge within one pulse cycle. I suppose I could compute 
it from the stated junction capacitance, but I'm not sure that's the 
only factor.


Will the stored charge actually give me the desired transition rate 
into 50 ohms? Hmm, again I should be able to compute this, but any 
other factors ignoring the non-diode ones like cap inductance?


How should I compute the coupling cap from the diode to the load? Use 
the impedance at the pulse rep rate? Seems reasonable. BTW, I don't 
care about droop in the  pulse, just the risetime. (measuring 
overshoot in an HF amp). Again, just want to verify that the obvious 
answer is the correct one. I clearly need to be very careful about the 
inductance.


Thanks, Bill



___
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

Re: [time-nuts] Fast risetime pulse generator

[Alan Melia]

Bill, Avalanche pulse gens only require high voltage because of the high 
VBRcbo and the gain of normal NPN transistors. I cant find the reference now 
it might have been a 1970s Ham Radio but if you use the same circuit as Jim 
but put an NPN "upside down" that is emitter where the collector is in Jim's 
circuit you can fire off fast pulses from a 12v supply, instead of requiring 
70 to 100v. I do wish I could locate the source as I have had several 
arguments about it :-))in the nicest possible way of course.


Alan
G3NYK

- Original Message - 
From: "BIll Ezell" 

To: 
Sent: Tuesday, April 12, 2016 9:30 PM
Subject: [time-nuts] Fast risetime pulse generator



(cross-posted to volt-nuts)
After paying only limited attention to this topic, I suddenly have a need 
for a pulse generator that has <150 ps risetime and a pulse width of at 
least 2 ns. 100mv amplitude or more is fine. I've looked at the classic 
Jim Williams avalanche generator, but I don't want to have to deal with 
the (relatively) high voltage source needed.


I've done microwave design using Gunn diodes, so I'm drawn to using a 
step-recovery diode. The topology seems very straightforward, and I can 
build it right onto a BNC connector, no PCB.


I'm thinking using an SMD835 diode, biased at ~1ma. The (sketchy) 
datasheet claims a T of 20 nsecs and a Tr of 85 ps, Cj of 0.4 to 0.8 pf.


Questions:

The obvious, is it reasonable?

Is the bias current reasonable? I'm assuming the bias current is actually 
dependent on the repetition rate, you need enough current to replenish the 
charge within one pulse cycle. I suppose I could compute it from the 
stated junction capacitance, but I'm not sure that's the only factor.


Will the stored charge actually give me the desired transition rate into 
50 ohms? Hmm, again I should be able to compute this, but any other 
factors ignoring the non-diode ones like cap inductance?


How should I compute the coupling cap from the diode to the load? Use the 
impedance at the pulse rep rate? Seems reasonable. BTW, I don't care about 
droop in the  pulse, just the risetime.  (measuring overshoot in an HF 
amp). Again, just want to verify that the obvious answer is the correct 
one. I clearly need to be very careful about the inductance.


Thanks, Bill

--
Bill Ezell
--
The day Microsoft makes something that doesn't suck
will be the day they make vacuum cleaners.
Or maybe Windows 10.

___
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to 
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there. 


___
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.