IN-18 Modular Arduino clock layout updated

[Terry S]

Ok, John shamed me into tweaking the layout. Here's the latest revs. I did address the lines crossing the hole keepouts, as well as rationalizing out a whole bunch of vias. My criteria for a clean design includes as few as possible jogs in the traces, minimal number of vias within reason, and an adherence to x on one layer and y on the other.... until the routing is done, then around the margins, it becomes OK to violate that rule. Symmetrical layouts of repeated circuitry. Also only 45 degree bends, (90 or 45 at vias OK) clean pad entry, (no acid traps), trace widths no narrower than the design justifies... very conservative rules. Makes for high yields and reliable boards. Anyway, now I'm pretty happy with both of these boards, and after some careful design validation, I'll be ready to order boards. So your feedback, please. Thanks, Terry

[Jeff Walton]

While I know that you don't care for LED backlighting, many of the designs I've seen use (2) per IN-18 with one on each side of the tube seal for best effect.  You appear to have more than enough room for that.

 

Very nice boards!

 

Jeff

--
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+...@googlegroups.com.
To post to this group, send email to neoni...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/44df85be-01c4-4df7-bed6-1959abe369de%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Nicholas Stock]

The PV electronics IN18 clocks (old style now) use 3x1206(?) type LED's under each tube. Much as some people don't like them, I find a nice orange LED complements the neon glow quite well...they can also make great night lights when the tubes are shut down at night...;-)

Nick

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/54ee8d52.88196b0a.5a63.ffffdd20%40mx.google.com.

[Terry S]

You guys persuaded me to add a 2nd LED under each tube. Done. 3 - not going to happen.

I also liked Pete's cascode anode driver idea so much I've adopted that as well.

Interesting side effect of doing that -- no need for a ground connection to the 2 digit extension boards. That seems strange even to me... but the nixies don't need it, the cathode drivers are all on the main board, and the LED and colon drivers don't require a ground connection either, as they are low side drivers on the main board.

This frees up two pins on the 20 pin inter-board connectors. Hmmm.... Thinking out loud, I'm wondering if I can use one of the pins to identify to the Arduino just how many extension boards there are -- therefore allowing the software to auto-configure based on the hardware. A pull-up on the extension board and a pull-down on the main board would form a voltage divider I could feed into an analog input. 2 extension boards would change the voltage level. The code can then make the mux rate and colons behave as required, based on the number of tubes present.

Getting close to ordering boards,,,,, if I can just put a stop to the creeping elegance.

Terry

[Mich...@aol.com]

:)

Still interested.   How many boards are you doing?

 

Also, thoughts of a 7971 driver board?  =:D

 

Michail

 

 

--
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+...@googlegroups.com.
To post to this group, send email to neoni...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/deb17d20-5ec4-49c5-a8b4-8f3e8ff9e17e%40googlegroups.com.

[Terry S]

I'll probably go with OSHPark, the minimum order for the first round, which means 3 main and probably 6 extension boards. No price advantage in building more, so that'll be it until I'm confident in the design. Once it's proven out, I'll order more if there is still interest. Or I'll set it up on OSHPark so you can order your own.

If I can eventually put together an order for about 20 sets, the price drops dramatically.

I'm not personally interested in doing a B7971 board, as I've used up or sold off all my tubes. I do have a plethora of IN-12's, so that would be my logical next board set. I also want to do a numitron clock.

Terry

[John Rehwinkel]

> Ok, John shamed me into tweaking the layout.

That wasn't really my goal, but you did ask for comments.

> Here's the latest revs. I did address the lines crossing the hole keepouts, as well as rationalizing out a whole bunch of vias. My criteria for a clean design includes as few as possible jogs in the traces, minimal number of vias within reason, and an adherence to x on one layer and y on the other.... until the routing is done, then around the margins, it becomes OK to violate that rule. Symmetrical layouts of repeated circuitry. Also only 45 degree bends, (90 or 45 at vias OK) clean pad entry, (no acid traps), trace widths no narrower than the design justifies... very conservative rules. Makes for high yields and reliable boards. Anyway, now I'm pretty happy with both of these boards, and after some careful design validation, I'll be ready to order boards. So your feedback, please.

I really like this version! Those are really satisfying layouts!

- John

[John Rehwinkel]

> I also liked Pete's cascode anode driver idea so much I've adopted that as well.
>
> Interesting side effect of doing that -- no need for a ground connection to the 2 digit extension boards. That seems strange even to me... but the nixies don't need it, the cathode drivers are all on the main board, and the LED and colon drivers don't require a ground connection either, as they are low side drivers on the main board.

My digit boards ended up the same way. No ground, no ground plane. Just anodes and cathodes.

I peered at the engineering drawings of various right-angle male and female headers to see if I could figure out what their mated width was, so I could put the holes in the right place for
my boards to plug together neatly. I finally gave up, ordered a few from various vendors, assembled them, and measured the result. Headers are cheap, having boards made is ... less cheap.

How did you come up with the spacing for yours? The silkscreens look really nice.

> This frees up two pins on the 20 pin inter-board connectors. Hmmm.... Thinking out loud, I'm wondering if I can use one of the pins to identify to the Arduino just how many extension boards there are -- therefore allowing the software to auto-configure based on the hardware. A pull-up on the extension board and a pull-down on the main board would form a voltage divider I could feed into an analog input. 2 extension boards would change the voltage level. The code can then make the mux rate and colons behave as required, based on the number of tubes present.

Slick! I guess you'd use one pin to power the pull-up, and the other pin for the sense lead?

> Getting close to ordering boards,,,,, if I can just put a stop to the creeping elegance.

Believe me, I know the feeling! I just tweaked my silkscreens to make pin 1 clearer and show the transformer on the Tayloredge board to make the orientation clear, added a voltage source jumper, rerouted the power to avoid some weird crossovers, and moved a few parts to make the trace runs nicer. It's so tempting to order a set, the renders on the board house site look really nice to me.

- John

[Terry S]

I did ask for comments -- and I appreciate the feedback and all the helpful suggestions.

Regarding the connector spacing -- I still need to verify that. The parts come from the Eagle library. From a quick look just now at the part drawings, they look reasonably accurate. I need the board edges to touch to get the tube and colon spacing right, although even if they don't the colon can be adjusted somewhat. I've used similar headers before and they typically mate fully.

The new ID circuit only required one additional pin -- the 5 volts for the pull-up was already on the extension board. So I still have one open pin....

Terry