20 Oct
2025
20 Oct
'25
7:01 p.m.
-----Original Message-----
From: Peter Coghlan via cctalk <cctalk(a)classiccmp.org>
Sent: 19 October 2025 18:32
To: General Discussion: On-Topic and Off-Topic Posts <cctalk(a)classiccmp.org>
Cc: Peter Coghlan <cctalk(a)beyondthepale.ie>
Subject: [cctalk] Re: Rainbow H7842 PSU
suspicious.
I measured Vz at 4.3V. In my notes I have previously measured 5V or 5.4V,
possibly on different H7842 PSUs (my notes are not clear). I have found the
Zener diode, it is hard to make out the markings without lifting it, but I
*think* the markings are
00
426
Not sure if this means the Zener voltage is supposed to be 6V? I will lift
it when I get a bit of time, probably in a couple of days, so I can look at
it properly and check it out of circuit with my DMM.
-----Original Message-----
From: Rob Jarratt <robert.jarratt(a)ntlworld.com>
Sent: 19 October 2025 14:04
To: rob(a)jarratt.me.uk; 'General Discussion: On-Topic and Off-Topic Posts'
<cctalk(a)classiccmp.org>
Subject: RE: [cctalk] Re: Rainbow H7842 PSU
> -----Original Message-----
> From: Rob Jarratt via cctalk <cctalk(a)classiccmp.org>
> Sent: 14 October 2025 21:42
> To: 'General Discussion: On-Topic and Off-Topic Posts'
> <cctalk(a)classiccmp.org>
> Cc: Rob Jarratt <robert.jarratt(a)ntlworld.com>
> Subject: [cctalk] Re: Rainbow H7842 PSU
>
>
>
> Oh my! I have clearly got my understanding the wrong way around,
> not sure how I did that because I read the datasheet carefully.
> Somehow, I got confused. I re-measured and found 1IN+=5.5V,
> 1IN-=9.4V, Power OK=6.7V, but GND (on the 393) is -13V. So as you
> say Power OK should be -13V and AC OK H would be asserted. I guess
> this must mean that the comparator itself is faulty. I have some
> 393s, so I will replace it and see what
happens.
>
Today I replaced the LM393P with a new LM393N. The AC OK H signal is
now working correctly. However, the outputs are a bit on the low
side. The +5V output is 4.37V, the 6% tolerance quoted in the
technical manual would allow a minimum of 4.7V. The +12V output is
11.3V, 6% tolerance would allow a minimum of 11.3, so only just OK.
Any suggestions on what I could look at here, maybe the output
capacitors? I checked the two capacitors on the +5V output and the
ESR seems OK when measured in circuit, but perhaps the big one before
the smoothing choke should be replaced? NB Raw 5V is about 4.2V
Thinking about this a bit more, I am thinking the 5V output capacitors
because the 5V output is proportionally lower than the 12V output. But
they are both too low and I guess that implicates the drive control. I
seem to remember replacing the PWM fairly recently so I would hope its
not that, but maybe one of its inputs. Vstart is only 11V, which seems a bit Either the 7812 is bad or something is pulling it
down. But it looks
like I also replaced the 7812. I might try using a bench PSU to
provide power to the 7812 to see how it behaves.
I still think the first thing to be done is to figure out whether Vz is correct or
not, ideally by reading the markings on the zener diode or otherwise figuring
out what voltage is supposed to be across it and then measuring the voltage
directly across it to confirm or deny that it is correct.
Thanks
Rob
>
> The most likely causes of only 11V coming out of the 7812 are either the input
> to it is low (it needs to be at least several volts higher than 12V) or the current
> drawn from it is excessive, in which case it will be hot or it is
> faulty. There is also a remote possibility that is oscillating due to a bad
> decoupling capacitor.
>
> (It is possible that nothing cares that Vstart is a bit low, some of the
> comparator inputs come from it via voltage dividers but the more critical ones
> seem to be referenced to Vz.)
>
> There is also the odd 0.7V difference between two ground connections in
> different parts of the power supply that needs to be accounted for or fixed if
> it turns out to be a fault.
>
> Once these items are dealt with, the low outputs can be addressed if they are
> still low after dealing with the other issues.
>
> Regards,
> Peter Coghlan.