Re: [time-nuts] TTimelab question

[Christopher Hoover]

One way to do cheap, short-run enclosure costs is to use extruded aluminum
cases of this sort
.
They extrusion only needs to be cut to length.   It is completed with a
front and back panel that is just a simple stamping or lasering of sheet
metal.   You get a lot of pieces out of one sheet.   Add optional silk
screening.

For full stamped and braked enclosures, I've gotten quite decent prices
from United Sheet Metal for short runs (100's).   They often send the job
to Taiwan.

Protocase is good for 1-10 pieces.

-ch




On Wed, Feb 22, 2017 at 6:13 AM, John Ackermann N8UR  wrote:

> I really like the setup that Mark is describing.  As to TAPR's plans,
> we've found that enclosures are a challenge -- metalwork is pretty
> expensive unless you get significant volume, and in our niche market,
> that's hard to do.
>
> But I am hoping to find an inexpensive clamshell-type enclosure with flat
> front and rear panels, and then do up designs (perhaps with Front Panel
> Express) for those panels.  That can be done at a reasonable cost, and at a
> minimum we can make design files available so people can order their own
> panels.
>
> For my own use, I'm also going to do a couple of 2U rack enclosures -- one
> to hold two TICCs operating independently, and another for the "megaTICC"
> -- four units slaved together to make an 8 channel counter, with a
> Raspberry Pi controller along the line of what Mark described. (In
> multi-board mode, each TICC outputs on its own USB line, so the RPi's main
> purpose is to deal with the 8 channels of data from 4 USB connections.)
>
> I'll make the design files for those enclosures available as well, but it
> may be a while as my entire lab is now packed up as we are in the final
> stages of moving from Atlanta back to Dayton.
>
> Also, in a day or three I'll be announcing a simple project that sprung
> out of the TICC assembly and testing process that some of you might find
> useful.  We're still finalizing details on that.
>
> John
> 
>
> On 02/21/2017 11:45 PM, Mark Sims wrote:
>
>> I doubt that it is something TAPR would do.   Building complete systems
>> gets into all sorts of issues (mainly regulatory).   But it is easy enough
>> to build.   They sell a nice case that the RPI3 and touchscreen mounts in.
>> The PI+touchscreen+case sells for around $110.   The TICC(s) connect to it
>> via USB.
>>
>> There are also some Win10 tablets with 1024x600 touch screens that sell
>> for around $60 (apparently Microsoft doesn't charge manufacturers for Win10
>> on tablets with small/low res screens).
>>
>> I  am thinking about laying out a front-end board for the TICC.  It would
>> have some switchable (relay?) 50 ohm input terminators,  switchable PICDIV
>> dividers for PPS/1MHz/5MHz/10MHz/15MHz (or 2.5 MHz)  inputs,  footprints
>> for a decent reference oscillator (MV89/8663/DIP/etc), and a 12V to 5V
>> (3A?) power converter for the  TICC and PI... most of the better surplus
>> oscillators run off of 12V.  Also maybe add a data multiplexer for
>> combining the outputs of two TICC boards into one data stream (but Heather
>> could do that in software).  John has some ideas for a similar board.
>>
>> -
>>
>>   Wow!  If you can persuade John and TAPR to produce that, I would be
>>> there
>>>
>> with my chequebook before the ink had dried on the web-page! :-)
>> ___
>> time-nuts mailing list -- time-nuts@febo.com
>> To unsubscribe, go to https://www.febo.com/cgi-bin/m
>> ailman/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/m
> ailman/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] TTimelab question

[Bob Bownes]

We have a local metalworking place that will laser cut steel and Al to 3/8". 
For small runs (like 1), they are very competitive to simply buying the 
materials alone. I had them make me 3 1U panels from 3/16" steel with 12 D 
holes to build N patch panels. $15 ea. Powder coated in my color of choice. 
Blanks were about 6 iirc. Well worth not punching all those holes!

Take a look in your search engine of choice. If you can't find one, our guy 
will ship. ;)

> On Feb 22, 2017, at 16:44, Bob Stewart  wrote:
> 
> Hi John,
> 
> I have to agree with you about enclosures.  They make the difference between 
> a DIY unit and a nice commercial unit, but they are very expensive.  The 
> enclosures I use are about $20 each from Mouser.  But the end panels are 
> another $20 each: milled front and back.  So, that's $60 per unit.  One 
> suggestion I might make is to check into what it would cost to make the end 
> panels from circuit board.  At volume, it might be significantly cheaper that 
> $20 per panel.  Learning KICAD well enough to make an end panel isn't that 
> difficult.
> 
> OSHPark has better prices on larger volume orders.  Off the cuff without 
> looking, I think I pay about $19 per mainboard for an order of 10, so that 
> would probably work out to $15 for a pair of PCB end panels.  That's still 
> $35 per enclosure, though.  You'd probably need to order 15 of each to meet 
> their minimums.  There are other suppliers that have better prices, but I 
> have no experience with them or their costs.  It wouldn't be as nice as 
> milled aluminum, but it should work just as well.
> Bob 
> -
> AE6RV.com
> 
> GFS GPSDO list:
> groups.yahoo.com/neo/groups/GFS-GPSDOs/info
> 
>  From: John Ackermann N8UR 
> To: Discussion of precise time and frequency measurement  
> Sent: Wednesday, February 22, 2017 8:13 AM
> Subject: Re: [time-nuts] TTimelab question
> 
> I really like the setup that Mark is describing.  As to TAPR's plans, 
> we've found that enclosures are a challenge -- metalwork is pretty 
> expensive unless you get significant volume, and in our niche market, 
> that's hard to do.
> 
> But I am hoping to find an inexpensive clamshell-type enclosure with 
> flat front and rear panels, and then do up designs (perhaps with Front 
> Panel Express) for those panels.  That can be done at a reasonable cost, 
> and at a minimum we can make design files available so people can order 
> their own panels.
> 
> For my own use, I'm also going to do a couple of 2U rack enclosures -- 
> one to hold two TICCs operating independently, and another for the 
> "megaTICC" -- four units slaved together to make an 8 channel counter, 
> with a Raspberry Pi controller along the line of what Mark described. 
> (In multi-board mode, each TICC outputs on its own USB line, so the 
> RPi's main purpose is to deal with the 8 channels of data from 4 USB 
> connections.)
> 
> I'll make the design files for those enclosures available as well, but 
> it may be a while as my entire lab is now packed up as we are in the 
> final stages of moving from Atlanta back to Dayton.
> 
> Also, in a day or three I'll be announcing a simple project that sprung 
> out of the TICC assembly and testing process that some of you might find 
> useful.  We're still finalizing details on that.
> 
> John
> 
> 
> ___
> 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] TTimelab question

[Bob Stewart]

Hi John,

I have to agree with you about enclosures.  They make the difference between a 
DIY unit and a nice commercial unit, but they are very expensive.  The 
enclosures I use are about $20 each from Mouser.  But the end panels are 
another $20 each: milled front and back.  So, that's $60 per unit.  One 
suggestion I might make is to check into what it would cost to make the end 
panels from circuit board.  At volume, it might be significantly cheaper that 
$20 per panel.  Learning KICAD well enough to make an end panel isn't that 
difficult.

OSHPark has better prices on larger volume orders.  Off the cuff without 
looking, I think I pay about $19 per mainboard for an order of 10, so that 
would probably work out to $15 for a pair of PCB end panels.  That's still $35 
per enclosure, though.  You'd probably need to order 15 of each to meet their 
minimums.  There are other suppliers that have better prices, but I have no 
experience with them or their costs.  It wouldn't be as nice as milled 
aluminum, but it should work just as well.
Bob 
-
AE6RV.com

GFS GPSDO list:
groups.yahoo.com/neo/groups/GFS-GPSDOs/info

  From: John Ackermann N8UR 
 To: Discussion of precise time and frequency measurement  
 Sent: Wednesday, February 22, 2017 8:13 AM
 Subject: Re: [time-nuts] TTimelab question
   
I really like the setup that Mark is describing.  As to TAPR's plans, 
we've found that enclosures are a challenge -- metalwork is pretty 
expensive unless you get significant volume, and in our niche market, 
that's hard to do.

But I am hoping to find an inexpensive clamshell-type enclosure with 
flat front and rear panels, and then do up designs (perhaps with Front 
Panel Express) for those panels.  That can be done at a reasonable cost, 
and at a minimum we can make design files available so people can order 
their own panels.

For my own use, I'm also going to do a couple of 2U rack enclosures -- 
one to hold two TICCs operating independently, and another for the 
"megaTICC" -- four units slaved together to make an 8 channel counter, 
with a Raspberry Pi controller along the line of what Mark described. 
(In multi-board mode, each TICC outputs on its own USB line, so the 
RPi's main purpose is to deal with the 8 channels of data from 4 USB 
connections.)

I'll make the design files for those enclosures available as well, but 
it may be a while as my entire lab is now packed up as we are in the 
final stages of moving from Atlanta back to Dayton.

Also, in a day or three I'll be announcing a simple project that sprung 
out of the TICC assembly and testing process that some of you might find 
useful.  We're still finalizing details on that.

John

   
___
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] TTimelab question

[Bob Camp]

Hi


> On Feb 21, 2017, at 11:45 PM, Mark Sims  wrote:
> 
> I doubt that it is something TAPR would do.   Building complete systems gets 
> into all sorts of issues (mainly regulatory).

There’s also the issue of cost. If you need to sell maybe 200 gizmos at $100 to 
make things work, The same process is a bit more 
exciting when it’s 200 gizmos at $800. In the first case, somebody needs to 
front the money for a $20,000 project. In the second 
case they may need to take out a second mortgage. If 20 of the $100 devices 
don’t sell, they have $2K in inventory to write off.
In the second case they have $16K. As you can quickly see on their web site, 
some of their projects hang around in inventory 
for a *long* time.  Yes, that’s all in sales dollars to make the math easy. 
Somehow I doubt that TAPR is a high margin / high profit
organization :)

Bob

>   But it is easy enough to build.   They sell a nice case that the RPI3 and 
> touchscreen mounts in.  The PI+touchscreen+case sells for around $110.   The 
> TICC(s) connect to it via USB.  
> 
> There are also some Win10 tablets with 1024x600 touch screens that sell for 
> around $60 (apparently Microsoft doesn't charge manufacturers for Win10 on 
> tablets with small/low res screens).
> 
> I  am thinking about laying out a front-end board for the TICC.  It would 
> have some switchable (relay?) 50 ohm input terminators,  switchable PICDIV 
> dividers for PPS/1MHz/5MHz/10MHz/15MHz (or 2.5 MHz)  inputs,  footprints for 
> a decent reference oscillator (MV89/8663/DIP/etc), and a 12V to 5V (3A?) 
> power converter for the  TICC and PI... most of the better surplus 
> oscillators run off of 12V.  Also maybe add a data multiplexer for combining 
> the outputs of two TICC boards into one data stream (but Heather could do 
> that in software).  John has some ideas for a similar board.  
> 
> -
> 
>> Wow!  If you can persuade John and TAPR to produce that, I would be there
> with my chequebook before the ink had dried on the web-page! :-)
> ___
> 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] TTimelab question

[John Ackermann N8UR]

I really like the setup that Mark is describing.  As to TAPR's plans, 
we've found that enclosures are a challenge -- metalwork is pretty 
expensive unless you get significant volume, and in our niche market, 
that's hard to do.


But I am hoping to find an inexpensive clamshell-type enclosure with 
flat front and rear panels, and then do up designs (perhaps with Front 
Panel Express) for those panels.  That can be done at a reasonable cost, 
and at a minimum we can make design files available so people can order 
their own panels.


For my own use, I'm also going to do a couple of 2U rack enclosures -- 
one to hold two TICCs operating independently, and another for the 
"megaTICC" -- four units slaved together to make an 8 channel counter, 
with a Raspberry Pi controller along the line of what Mark described. 
(In multi-board mode, each TICC outputs on its own USB line, so the 
RPi's main purpose is to deal with the 8 channels of data from 4 USB 
connections.)


I'll make the design files for those enclosures available as well, but 
it may be a while as my entire lab is now packed up as we are in the 
final stages of moving from Atlanta back to Dayton.


Also, in a day or three I'll be announcing a simple project that sprung 
out of the TICC assembly and testing process that some of you might find 
useful.  We're still finalizing details on that.


John

On 02/21/2017 11:45 PM, Mark Sims wrote:

I doubt that it is something TAPR would do.   Building complete systems gets 
into all sorts of issues (mainly regulatory).   But it is easy enough to build. 
  They sell a nice case that the RPI3 and touchscreen mounts in.  The 
PI+touchscreen+case sells for around $110.   The TICC(s) connect to it via USB.

There are also some Win10 tablets with 1024x600 touch screens that sell for 
around $60 (apparently Microsoft doesn't charge manufacturers for Win10 on 
tablets with small/low res screens).

I  am thinking about laying out a front-end board for the TICC.  It would have 
some switchable (relay?) 50 ohm input terminators,  switchable PICDIV dividers 
for PPS/1MHz/5MHz/10MHz/15MHz (or 2.5 MHz)  inputs,  footprints for a decent 
reference oscillator (MV89/8663/DIP/etc), and a 12V to 5V (3A?) power converter 
for the  TICC and PI... most of the better surplus oscillators run off of 12V.  
Also maybe add a data multiplexer for combining the outputs of two TICC boards 
into one data stream (but Heather could do that in software).  John has some 
ideas for a similar board.

-


  Wow!  If you can persuade John and TAPR to produce that, I would be there

with my chequebook before the ink had dried on the web-page! :-)
___
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] TTimelab question

[Mark Sims]

I doubt that it is something TAPR would do.   Building complete systems gets 
into all sorts of issues (mainly regulatory).   But it is easy enough to build. 
  They sell a nice case that the RPI3 and touchscreen mounts in.  The 
PI+touchscreen+case sells for around $110.   The TICC(s) connect to it via USB. 
 

There are also some Win10 tablets with 1024x600 touch screens that sell for 
around $60 (apparently Microsoft doesn't charge manufacturers for Win10 on 
tablets with small/low res screens).

I  am thinking about laying out a front-end board for the TICC.  It would have 
some switchable (relay?) 50 ohm input terminators,  switchable PICDIV dividers 
for PPS/1MHz/5MHz/10MHz/15MHz (or 2.5 MHz)  inputs,  footprints for a decent 
reference oscillator (MV89/8663/DIP/etc), and a 12V to 5V (3A?) power converter 
for the  TICC and PI... most of the better surplus oscillators run off of 12V.  
Also maybe add a data multiplexer for combining the outputs of two TICC boards 
into one data stream (but Heather could do that in software).  John has some 
ideas for a similar board.  

-

>  Wow!  If you can persuade John and TAPR to produce that, I would be there
with my chequebook before the ink had dried on the web-page! :-)
___
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.