This is my circuit.

When the side lights are off. So are the Orange LED markers.
When the side lights are on. So are the Orange LED markers.
When the side lights are off, and then the indicator is on, then the appropriate side Orange LED marker flashes in time with the Indicator until indicating is cancelled and then returns to the off state.
When the side lights are on, and then the indicator is on too, then the side light request to the appropriate Orange LED is overridden by the indicator request, causing it to flash in time with the Indicator until indicating is cancelled, and then returns to the on state.

The pulse converters seem to be the only parts of the circuitry which are working. The 4011 element of the circuit(s) only seem to work with this one rogue chip, as described above....!? why? I need help to fix this. Please.

One thing you might look at is that the CD logic draws almost no current on the inputs and the regulators have big caps on them with nothing to bleed them off. You could add a resistor to ground to each regulator to discharge them. Maybe 100 ohms or so should do.
Edit:
Sorry, wrong circuit. Add a 1K resistor from pin 1 to ground so it will be ground when the transistor is off.

 

I doubt you want to change direction now, but you can make this much simpler.

 

OK, now that we have established that this is a 'real' but a very badly designed circuit. Here may be the problem: CMOS latch up.
This can occur when the input voltage is higher than the power supply voltage. I suspect that the 0.1 caps having nothing but something shy of 100 megohms to discharge thru, are holding the input voltage at least 2 volts higher than the supply voltage when power is re-applied. If a reasonable discharge path was provided, I think the problem would probably go away.

I think you have come up with an unintentional crow bar circuit and the 12volt battery is winning every time at the expense of the chips. They probably do not even get a chance to get warm.

 

When the side lights are off. So are the Orange LED markers.
When the side lights are on. So are the Orange LED markers.
When the side lights are off, and then the indicator is on, then the appropriate side Orange LED marker flashes in time with the Indicator until indicating is cancelled and then returns to the off state.
When the side lights are on, and then the indicator is on too, then the side light request to the appropriate Orange LED is overridden by the indicator request, causing it to flash in time with the Indicator until indicating is cancelled, and then returns to the on state.

Could you summarize the functionality you want in a truth table? Your description is difficult for me to follow and your schematic is even harder to follow.

The bottom line is that you've made some questionable choices. Your use of the voltage regulators is completely inappropriate. CMOS logic can operate up to 15V so you don't need to operate it at 9V and level shift any inputs. If you're worried about voltage spikes above 15V, you can clamp the appropriate inputs.

 

No it wouldn't. I need to regulate all the inputs to the logic to the same level as VDD. So, side lights and indicator (pulse and constant separately)

The reason I picked 9v is because an automotive environment can swing between 10.5v - 14v depending on whether the engine is running or not etc.

So if you regulate the power supply to the chips, the logic voltages will be regulated. You can do that with one regulator. After that, you just need to "condition" the external inputs signals and provide appropriate outputs circuit to drive the LEDS. If you describe what the actual external input signals are, we can probably help. But you really need to begin with a truth table to describe how the logic should function.

 

If you're willing to have the LED blink the opposite of the turn indicator when both the side marker and indicator are on, all you need is a single XOR for each side.

You can operate the circuit from the battery voltage and you won't need any level shifting. If you needed level shifting, there are better ways to do it. If you're worried about voltage spikes, you can clamp the inputs.

EDIT: Corrected OFF/OFF truth table entry.

 

OK, now that we have established that this is a 'real' but a very badly designed circuit. Here may be the problem: CMOS latch up.
This can occur when the input voltage is higher than the power supply voltage. I suspect that the 0.1 caps having nothing but something shy of 100 megohms to discharge thru, are holding the input voltage at least 2 volts higher than the supply voltage when power is re-applied. If a reasonable discharge path was provided, I think the problem would probably go away.

Yes, I agree to this. So can this be prevented by putting a 10K resistor across the output of the LM7809 to ground?

I think you have come up with an unintentional crow bar circuit and the 12volt battery is winning every time at the expense of the chips. They probably do not even get a chance to get warm.

Can you explain a little bit more on this crow bar circuit? Where was it formed?

Thanks.

Allen

 

Fixed it!

I was at the gym thinking about this, and as soon as I got back, I tried it, and it works! My "poorly designed" circuit is now working!

All i had to do, is tie the collector of the BC558C PNP transistors down to 0V via a 470K resistor, to give it a ground path for holding those dependent inputs down when there is no active 9V coming through the transistor. First time i've had the whole thing working at once! I'm very pleased.

Now i'm gonna go back to the drawing board and redesign the whole thing using a completely different approach. I already got ideas for streamlining the circuit, and simplifying it using different components.

Thank you for everyone's input. I know i can be a difficult person, but if you knew some of the stuff i've been through in my past, you'd have a better understanding of me. I'd still be interested in having input from people, maybe collect a few different good ideas on how else this circuits functionality can be achieved, but in a simpler, more robust way.

 

Can you explain a little bit more on this crow bar circuit? Where was it formed?

https://en.wikipedia.org/wiki/Crowbar_(circuit)

Also, see this: read note 1 at bottom.

http://www.ece.drexel.edu/courses/ECE-E431/latch-up/latch-up.html

Load on 9vREGS should provide minimum recommended load current. (~ 10mA). See data sheet.

Also, I really hope the CMOS output to transistor interface is not part of the 'real' circuit as well. If it is, this would be a major problem for the 4011's survival..

 

https://en.wikipedia.org/wiki/Crowbar_(circuit)

Also, see this: read note 1 at bottom.

http://www.ece.drexel.edu/courses/ECE-E431/latch-up/latch-up.html

Load on 9vREGS should provide minimum recommended load current. (~ 10mA). See data sheet.

Also, I really hope the CMOS output to transistor interface is not part of the 'real' circuit as well. If it is, this would be a major problem for the 4011's survival..

I think the 7809 has enough load internal. It's not like the 317 in that way.
Are you worried about the 9 ma from the 4011?

 

I think the 7809 has enough load internal. It's not like the 317 in that way.
Are you worried about the 9 ma from the 4011?

We learned from a recent post for a 7912 regulation problem that all of the LM78xx/LM79xx regulators require a minimum 5mA load. The datasheets don't mention it explicitly, but the specs show a minimum current of 5mA.

Regulation without the minimum load is possible as many have probably used LM78xx without one and didn't see problems.

I recently had an LM317T that wouldn't regulate with a 10mA load, but I've had others that worked with a 5mA load.

 

We learned from a recent post for a 7912 regulation problem that all of the LM78xx/LM79xx regulators require a minimum 5mA load. The datasheets don't mention it explicitly, but the specs show a minimum current of 5mA.

Regulation without the minimum load is possible as many have probably used LM78xx without one and didn't see problems.

I recently had an LM317T that wouldn't regulate with a 10mA load, but I've had others that worked with a 5mA load.

I see in the data sheets the spec for load regulation, but nothing for minimum load.
Not something I would think you would leave out as they are so clear on the 317.
Did the one here have the old type (L78XX)? I think maybe they did need a load.
Anyway sounds like rissy is okay now.
If you can find that old thread let me know, I'm curious now.

 

If you can find that old thread let me know, I'm curious now.

It was this thread. Solution was in post #8. The minimum load for the low power versions is 1mA.

 

It was this thread. Solution was in post #8. The minimum load for the low power versions is 1mA.

Very interesting.

Thanks!

 

My PSU built around several LM 3-terminal fixed regulators never had any permanent load, just caps at the output. They show the same stable voltage as when a load is applied. 5, 9, 12, 15 & 18V positive/negative.

BTW what I recall from reading the datasheet for the LM117, 237 and 337 is that the first requires just 5mA to work while the others require a minimum of 10mA.

 

My PSU built around several LM 3-terminal fixed regulators never had any permanent load, just caps at the output. They show the same stable voltage as when a load is applied. 5, 9, 12, 15 & 18V positive/negative.

According to the datasheet, the behavior you've been seeing isn't guaranteed.

Designers have sometimes operated a device outside of the guaranteed specs and saw some behavior they wanted (e.g. exceeding common mode input voltage range). But you need to be aware that manufacturers could change the behavior at any time and designs using those features that weren't guaranteed could find themselves scrambling to find another source with the desired behavior.