Something is but I don't know what!

Sorry, thanks for your patience.

I've triple checked and all connections are secure.

A stab in the dark here but could the 47uF decoupling cap for the 4069 be too large?

 

If it is across Vcc and Gnd, no. It sould be OK even if it were a 4700 uF. Are you sure your bench supply is solid and not dropping voltage as the relays pick up?

 

It's not regulated, its basically just a 12v SLA battery. It's good though.

Further digging reveals the problem may lie with the inverter. Probing the input shows it HIGH after connecting to the power supply when it should be LOW before connecting it to 12v. Can't work out why.

EDIT: Now works, but only with the voltmeter connected. Probably because the trigger wire is not grounded whereas it would be in situ.

2nd EDIT: To test the theory of the grounded trigger I placed a pull-down resistor between the input of the inverter and ground. The circuit now operates as desired. If I'm right, I won't need this resistor in the final circuit as the input with tied to ground anyway when it is inactive.

 

And yet again, something goes wrong.

I've now connected the other side of the 556 in exactly the same way. While the first side works as it should, the second side doesn't. The inverter output is LOW all the time, even though everything is wired exactly the same as the other side.

 

Try putting a 1K resistor from the input of the inverters to ground. An open input will assume high in most cases. From previous posts, my impression was that the input to the inverter would either be +12 or Ground, but never an open. Do you have two DPST or DPDT switches? If so, tie +12 to one side, Gnd to the other side and the common to the input of the inverter. Then you will better duplicate what the Lock/Unlock lines do in the car.

 

In situ, you are correct, it will never be open. I'm just trying to simulate the conditions on a breadboard to make sure it all works and seemingly failing.

The side that works now has a 10k pull-up resistor between the trigger and 12v and a 10k resistor between the inverter input and ground. The other side has the same layout but refuses to work.

 

I really don't have any advise other than to keep checking voltages at key points. Timing cap should charge to 2/3 Vcc then discharge, Inverter output should always be opposite input, etc.

 

As mentioned Bill, your continued help and patience really is appreciated.

Now, I've managed to relieve a switch from an old PC PSU to simulate the conditions as they will occur in use. I now have one side of the 556 working as desired, with no need for pull-up or pull-down resistors, as per my schematic (remembering the 10k resistors are part of the lock out circuit and not pull-up)

The problem I now have is the other side of the 556 refuses to work correctly. I've narrowed the culprit down to the fact the output is remaining high once triggered. The capacitor charges to 2/3 Vcc on triggering but when it gets there it stays there, the output doesn't go low.

 

That sounds like either the wire going to the discharge pin not being connected properly OR a bad 556

 

Can anything else cause it?

The discharge and threshold pins are tied together and then connected between the resistor and capacitor. Wired up properly and securely.

As for a bad 556, chances of all 3 of my ICs being bad?

 

If the cap it tied to the discharge line and is reaching 2/3 Vcc, the internal comparator should trigger and toggle the output line AND turn on the internal transistor which takes the discharge line effectively to ground, thus discharging the capacitor. Check voltage at the discharge PIN of the IC as well as the voltage at the threshold pin. (RIGHT ON THE IC PIN, not a wire that is supposed to be connected to it..) There has to be a bad connection somewhere or all of your 556 ICs could have been blown if placed into an improper connection. Is the cap voltage continuing to climb beyond 2/3 Vcc? I'm sure that when you do find the problem it will turn out to be something very simple. Sometimes they are the hardest to find. I had a problem one time that a pin on the IC folded in underneath the body instead of going into the socket. Checking at the pins themselves was the only way I could find the problem.

 

They have all been connected up correctly at all times. I've even grounded myself when handling them to prevent ESD

Maybe I have just been unlucky with 3 bad ICs. I'll get some more in the next few days.

 

If you're actually using a TS556CN, that's a CMOS 556 timer.

If your timing capacitor has too high of a leakage rate, it won't charge high enough to trip the comparator.

I hate to bring this up, but the Terms of Service of the Boards has changed, and all "automotive modifications" are now on the "Restricted Topics" list. See section 6 of the TOS, near the bottom. Links to the TOS are at the bottom of every page on the Forum.

 

Yes, it is a TS556CN.

I think the capacitor is fine. As I say, one side of the 556 works perfectly but the other doesn't. The capacitor is charging properly but the output stays HIGH. I'm not sure what a high leakage rate is determined to be or what the leakage rate of my capacitors are, I'd need to find out.

Yes, I've seen. I can understand the reasoning for the change of ToS, if not the implementation. I fully suspect this thread to be closed soon, even though it isn't as if there are safety concerns with this circuit in my opinion. Lighting and ECUs I can understand however, which I suspect is why the change came about in the first place. However, that is the decision that has been made and I will respect that.

Before it is closed I would like to say thank you to all who have helped, especially Bill. Many thanks.

 

OK, well if you are sourcing current to the cap via a 100k resistor, and the cap has less than 200k internal resistance, it'll never get to the 2/3 threshold.

One "work-around" you might try is to connect a 200k resistor from the CTRL input to GND; that will change the upper threshold to 1/2 Vcc, as the 200k will be in parallel with the lower 200k of the internal trigger/threshold voltage divider. That will also make the lower threshold 1/4 Vcc.

You don't want to try to measure the resistance of the cap directly, as you might damage your meter.

If your supply is actually 12v, then when the cap is charged to 2/3 Vcc, the resistor will have 12/3/100k = 40uA current through it. If the capacitor's leakage rate exceeds 40uA with 8v across it, it will never charge to 2/3 Vcc.

Now for the clincher: an acceptable leakage rate for a 47uF electrolytic cap rated for 25wv is 0.19mA, or 190uA! If your cap is not less than 1/5 of that leakage, it will never be charged high enough to trip the timer.

Your 556's weren't bad, and neither is the cap. It simply has too high of a leakage rate for the application as-is.

Another work-around is change the 100k resistor to 1 MEG, the 47uF cap to 4.7uF, but a different type of cap - metal poly film that will have a much lower leakage rate. Electrolytics are popular because they are relatively compact for their uF ratings, but they generally have a higher leakage rate than other types of capacitors.

But, adding a 200k resistor from CTRL to GND is the "quick and dirty" fix.

 

Thanks SgtWookie.

I've just looked at the data for the caps I bought and it states a leakage current of 3uA though as with most things, I expect the stated specifications aren't realistic for everyday use? They are rated 16v.

Supply is nearer to 12.3v. (An old spare SLA battery)

Can't get my head around why one side always works, but the other won't.

I will look into your suggestions though. Thanks again.

I will look into some of your suggestions.

 

At this point, it is kind of awkward - but the ToS says we must close this thread.