EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: 001 on September 09, 2018, 08:09:06 am
-
Masteh HY1503D voltage rise in the "off" moment :wtf:
How to fix the problem?
The schematic is here:
-
A possible mechanism to cause a voltage rise on turning off the supply could be that the supply for the control circuit breaks down to fast. It would be especially a problem if the voltage over C2 drops much faster than at C1.
Adding an extra cap (e.g. 220-470 µF) in parallel to C2 might help - though it could cause an corresponding peak on turn on, because than the voltage in C2 rises slower than the positive side.
A better way would be enable the current through R107/DW2 only if the the negative voltage (C2) is sufficient. This could also help during turn on.
-
How to fix the problem?
A simple solution would be to fit an On/Off switch in the cable that connects to the device/circuit under test. Thereby, enabling the power supply output to be isolated before operating its mains switch.
-
The low-voltage shutoff is supposed to be done by zener DW3, which cuts off drive to the pass transistor when voltage there (+12V rail) goes below a threshold - the zener plus three transistors Vbe.
Think of it (R10/DW3) as a pullup resistor turning the pass transistor full-on, and the control op-amps pull down or shut off the pass transistor as required.
The problem is op-amp -6V rail falls before the +12V rail falls to a point where DW3 shuts it off.
So the op-amps are off and you get a surge in voltage.
Other makes of PSU take R10 to the raw DC (input) side of the 12V reg. and use a 12V zener for DW3.
I would upsize C2 from tiny 330uF to >1,000uF to increase hold time for the op-amps Vcc (-rail).
And check DW3's voltage rating and maybe move it.
-
May be I can build new circuit using this transformer and enclosure? Any recomendations?
-
I just got into the habit of pulling the banana plugs out before turning off power :palm:
But if there is a co-worker switch off or power failure, the voltage surge can blow things up.
You'd have to add an undervoltage lockout. Monitor the +12V reg input and shut off the pass transistor.
The circuit in the infamous chinese 0-30V kit shorts out the op-amp, as its UVLO approach.
https://www.eevblog.com/forum/projects/variable-linear-bench-power-supply-kit-0-30v-0-3a/msg1294394/#msg1294394 (https://www.eevblog.com/forum/projects/variable-linear-bench-power-supply-kit-0-30v-0-3a/msg1294394/#msg1294394)
Or add something with a TL431 that shuts off the pass transistor.
Everybody wants cheap and simple but something bullet proof is complicated.
-
Everybody wants cheap and simple but something bullet proof is complicated.
Yea!
There are many simple cheap and bullet proofs things around: rover bycicles, kalashnikow gun, chevrolet
what about "lab" power supply?
-
The basic circuit is not that bad. One point missing is to disable the output, when the supply for the control part is not there. This is relatively easy to add (e.g. a PNP, 2 resistors and ~15 V zener to replace DW3). This could check to have more than 15 V from the +12 V to the -6 V.
-
How to fix the problem?
A simple solution would be to fit an On/Off switch in the cable that connects to the device/circuit under test. Thereby, enabling the power supply output to be isolated before operating its mains switch.
Great idea. But it has a hole. What happens if the line power goes out while it’s connected to the DUT? DUT goes phut.
-
The basic circuit is not that bad. One point missing is to disable the output, when the supply for the control part is not there. This is relatively easy to add (e.g. a PNP, 2 resistors and ~15 V zener to replace DW3). This could check to have more than 15 V from the +12 V to the -6 V.
It is hard to understand what. Can You draw Yours idea?