Tubes go dark after reset calibration

More
3 months 3 weeks ago #10268 by w8an
I hooked up the bench supply at 12 volts then re-ran the high voltage calibration. I no longer get the pulsing like before, but still get no visible digits (after the calibration), but instead get a faint plasma glow in each tube.

I've hooked the scope to the gate of the MOSFET.

What I'm seeing is a very unstable waveform consisting of sets of 8 pulses, 19uS apart, and occurring every 1.5mS. I'm assuming that I should be getting a nice stable and constant pulse train.

Scope video: https://youtu.be/mDcvqbyyvI4

Is it possible that the power supply that was putting out over 16 volts might have damaged something? This board worked perfectly after the build and then for a few weeks, before it was shelved in the closet for a couple of months.

Please Log in or Create an account to join the conversation.

More
3 months 3 weeks ago #10269 by Ty_Eeberfest
That gate pulse does not look good at all. Before we go any further into the HV generator parts I think you should try to measure the frequency present at the crystal pins (9 & 10) of the Atmega chip. Steady 16MHz is expected. My thinking is this: the characteristics of the gate signal are controlled by code - no physical components should be able to affect it like we are seeing. Code does not suddenly break, but if the CPU clock is wandering all over the place it can certainly act funny. We have seen several cases in the fairly recent past where crystals have mysteriously failed resulting in all sorts of bizarre random behaviour of the CPU.

Other concerns:

You said the 5V power is in spec. This makes me think the 16V input didn't hurt anything because the only things that see the full input voltage are the 5V regulator (still working) and the HV MOSFET (which I think is okay since the gate signal seems to be the problem).

To go back to the LDR subject, I just wanted to mention that the LDR does not affect the high voltage. Dimming is done by varying digit on-times in the multiplexing code, not by varying the HV voltage.

Look into it later when the dust is clearing off the crater.

Please Log in or Create an account to join the conversation.

More
3 months 3 weeks ago #10270 by Ty_Eeberfest
Also - in the event that checking the crystal doesn't show us the problem, the next thing I'd like to see is a video clip of the HV Calibration routine running.

Look into it later when the dust is clearing off the crater.

Please Log in or Create an account to join the conversation.

More
3 months 3 weeks ago #10271 by w8an
I wouldn't have expected the kind of operational behavior that this clock exhibited after finding what the problem was. I just started looking at signals with my scope and saw a strange waveform (choppy and drop outs) on the minutes tube high voltage pin. Turns out the opto-isolator on that tube had failed. I replaced it and the thing is working great now.

Thanks for all the suggestions and troubleshooting aid.

Please Log in or Create an account to join the conversation.

More
3 months 2 weeks ago #10272 by Ty_Eeberfest
Wow. I too would never have expected a failed opto to produce the symptoms you had. Glad you found it, otherwise we'd probably be at this for days. Now I'm trying to understand how the opto could mess up the HV. I'm gonna guess it was somehow shorting HV momentarily to ground (whenever the opto was fired maybe?) but my problem is that the only ground present on that opto is on the LED side - not the switching transistor side - of the isolation barrier. Weird...

Thanks for putting up with the false starts and bling leads.....

Look into it later when the dust is clearing off the crater.

Please Log in or Create an account to join the conversation.

More
3 months 2 weeks ago #10273 by w8an

Ty_Eeberfest wrote: Thanks for putting up with the false starts and bling leads.....


You gotta start somewhere and work through the problem. The end result is all that matters. Thanks again!

--Steve
The following user(s) said Thank You: Ty_Eeberfest

Please Log in or Create an account to join the conversation.

Moderators: AccutronTy_EeberfestIan
Time to create page: 0.290 seconds

Search

Tube Suppliers

Go to top
JSN Boot template designed by JoomlaShine.com