Just my WAG, but try placing a small (10 to 20pF) capacitor from pin 5 to GND on both CD3 and CD4 ic's, if you have any. Let us know the effect, please.

 

Folks, I think the time has come to put this one to bed.

The breadboard worked, the PCB doesn't. That suggests the design is good. There is a problem in the implementation.

This device uses components that I don't understand well. The currents are at low levels. I don't have the knowledge or equipment to properly troubleshoot this device.

Adding an oscilloscope to the mix might help but it will be a long road to success.

I think my best bet is to do another breadboard from scratch. Then I could swap sections which might ultimately get me a PCB that works.

I really appreciate the suggestions from everybody in this thread.

 

Did you try my suggestion?

 

Did you try my suggestion?

No, sorry. I am pretty sure the last two counters are working fine. All four are probably good. Why do you think capacitors there would help? Not too keen on following WAGs because soldering can make a mess.

 

My thinking is that the breadboard adds extra capacitance to the reset pulse that isn't there with on the PCB, so the counters reset erratically.

 

My thinking is that the breadboard adds extra capacitance to the reset pulse that isn't there with on the PCB, so the counters reset erratically.

OK, effects like that might be happening, but probably not at those counters. I get 36 flashes of LED1 for every actuation of the second LED, which is as it should be. It indicates that the last two counters are working correctly. Most likely the problem is upstream of the counters.

 

Folks, I think the time has come to put this one to bed.

The breadboard worked, the PCB doesn't. That suggests the design is good. There is a problem in the implementation.

This device uses components that I don't understand well. The currents are at low levels. I don't have the knowledge or equipment to properly troubleshoot this device.

Adding an oscilloscope to the mix might help but it will be a long road to success.

I think my best bet is to do another breadboard from scratch. Then I could swap sections which might ultimately get me a PCB that works.

I really appreciate the suggestions from everybody in this thread.

No, this suggests that the design is marginal.
The hallmark of a good clean design is that it works- every time.

I have designed MANY circuits that depend on clean, perfect AC line frequency/phase detection to work correctly, I can attest that it's not trivial to achieve this.
Your zero crosser is almost guaranteed to generate extra pulses that will drive you bonkers until you fix this.
Counting fast/erratic behavior is a clear symptom of this problem.

 

No, this suggests that the design is marginal.
The hallmark of a good clean design is that it works- every time.

I have designed MANY circuits that depend on clean, perfect AC line frequency/phase detection to work correctly, I can attest that it's not trivial to achieve this.
Your zero crosser is almost guaranteed to generate extra pulses that will drive you bonkers until you fix this.
Counting fast/erratic behavior is a clear symptom of this problem.

Yes, OK, I agree with you.

But I'm beyond my limits, so I can't proceed for now.

Maybe I will get a scope. Maybe I'll discover interesting things.

 

What happened to the PCB pictures?
Did you post them?
What sort of doctor are you?
Good luck.............

 

What happened to the PCB pictures?
Did you post them?
What sort of doctor are you?
Good luck.............

The PCB pics are in post 16.

Anybody can use Doctor in their handle if they want - but I'm actually a real one, in chemistry.

Chemists have solutions - sometimes.

 

I don't know if you are aware, but almost the whole circuit could be replaced with a single ATtiny85 8-pin MCU.
It could accurately output a separate pulse every 1sec, 1min, and 1hour (on separate pins) and even be pre-synced to a real time clock.

 

I don't know if you are aware, but almost the whole circuit could be replaced with a single ATtiny85 8-pin MCU.
It could accurately output a separate pulse every 1sec, 1min, and 1hour (on separate pins) and even be pre-synced to a real time clock.

Wow. I better read up on it.

Procrastination pays off again! (Haven't bought the scope yet.)

 

Better yet, I could use the electronics out of a Walmart clock. It takes 1.5VDC into the big near-square surfaces on the right and delivers 1 sec pulses to the contacts on the left.

 

There used to be an MM5369 IC that generated 60 Hz from a colour burst crystal.
Now you can program a PIC12C508 to do the same thing. (Note that the title says 60 MHz in error.)
As a matter of fact, you can generate a 1-Hz signal from just about any MCU.

https://ww1.microchip.com/downloads/en/appnotes/4_008.pdf

 

Do you need any more help? I have designed for commercial sale line synchronised devices. Use a simple moc 28 optoisolator connected direct to the line ( with dropping resistors and diode reverse protection) far simpler design than you have.

 

My 10c for Doctor_Ed. I have read this whole thread with interest, and I would just like to add that you cannot really hope to debug problematic circuits without the basic tools, foremost of these being a scope. It could even be argued that you can get by without a multimeter – but not a scope. Thanks to Far East manufacturing, these are FAR cheaper than they once were.

 

@kaindub and @Baker Steve
I have ordered a scope from Amazon. It's just a simple hand-held unit. But I think it will do the trick.

I was hesitant because devices tend to be more complicated than necessary, have a steep learning curve, and have very bad manuals. And a scope might be a very frustrating device for me. But after watching some utubes, I think it will be OK. It might be a lot of fun.

Trying a few things at random has not solved this circuit. Indeed, I need the right tools.

Thanks for your support and interest!

 

@kaindub and @Baker Steve
I have ordered a scope from Amazon. It's just a simple hand-held unit. But I think it will do the trick.

I don't think you will regret it. And should you get to the stage where you want to go further, it could be argued that Digilent have the most cost-effective tool set.

 

@kaindub and @Baker Steve
I have ordered a scope from Amazon. It's just a simple hand-held unit. But I think it will do the trick.

I was hesitant because devices tend to be more complicated than necessary, have a steep learning curve, and have very bad manuals. And a scope might be a very frustrating device for me. But after watching some utubes, I think it will be OK. It might be a lot of fun.

Trying a few things at random has not solved this circuit. Indeed, I need the right tools.

Thanks for your support and interest!

How do you know they are more complicated, steep, bad manuals?
2nd hand opinions from people you do not know?
Come on!!

 

I would suggest putting a small ceramic capacitor (maybe .05uF) across the right-hand diode feeding the optocoupler. That will filter out most high frequency noise coming from the power line, but will still allow 60Hz.