Your title, originally confused me, as to what you were trying to achieve. I thought from the title, you wanted to learn about diodes. I did not realise it was a delay for PSU circuit, you wanted).
The title is still correct, just the thread has become double laned by now. Probably because I talk to much in trying to explain what and why I did things. Nevertheless I am really, really thankful for the help for the root cause of my diode issues, be it the ujt or the lm431. Or the more grass roots approach of David, which I have to look more into it. At the end it will be a delay, but I would like to understand what I am doing and so far I've failed at a rather basic stage, the diodes. So this is not really over.
To your suggestion: According to the first look I would not mind the the lm431, as it does not seem to need a power supply. I am not against ICs, in this case I just have no stable supply to reliably power them for time they are needed. The supplyvoltage would slowly rise as the smoothing capacitors are being charged.
For a PUT as for the lm431, as long as the reference is raising faster than the voltage over the capacitor, I should be safe. In my theory.
Anyway, sorry for the confusion. I've replaced the lm7805 with a lm317 and will do more testing once the fuse arrives. It is still about diodes and why they do what they do and hopefilly I will then be able to ask more specific questions about their behaviour.
I understand. Threads, do sometimes go in odd directions and can be hard to accurately name, initially.
The LM431, because it only seems to need a few microamps (2uA typical, 4uA guaranteed) to run, plus any current the rest of the circuitry needs, may allow your longer time constants, such as 10 seconds, and yet have a capacitor, which is not ridiculously big. Anyway, they are fairly cheap (probably a lot less than a dollar, each), so could make an interesting experiment, for more messing/learning about diodes.
The LM431 seems to have been designed, to eliminate or reduce, some of the issues, people can have with normal zener diodes. Such as they (zeners) are nonadjustable, somewhat/partly inaccurate (i.e. about +/- 5% or so) and the voltage is not that precisely regulated, especially at lower currents, etc.
In theory, you could get similar results, by using multi transistor circuits, and conventional diodes or zeners, but these days it is often not worth the effort, when there are small, low cost ICs, which will do the job.
Other solutions in this thread should also work. But some of them would have limitations E.g. Neons need high voltages (around 90V plus) and some of the alternative transistor types mentioned, can be harder to obtain and/or a bit pricey.
But I suspect it could be done, with a pile of discrete components (including transistors). Doing it that way, could be more educational, than using ICs. But these days, using ICs is the more common way, fairly often.
tl;dr
LM431 is one way which might work. But there are probably millions of other ways of doing it. Just like most things in electronics.
Home
Help
Search
Login
Register
