The general rule with circuits is that if it doesn't behave as it should then it isn't the circuit you think it is. Either a component fault or a wiring error. Note that some circuits can cause trouble because a misunderstanding which results in a design fault is then propagated into testing or debugging. Someone can check twelve times that it is wired correctly but each check uses the same false information as the original build so the check says it is fine but reality disagrees.
Nevertheless, the circuit will misbehave if the diode at the negative end of the bridge gets reverse biased because too much current is being drawn from the negative supply.
Try simulating it.
Nevertheless, the circuit will misbehave if the diode at the negative end of the bridge gets reverse biased because too much current is being drawn from the negative supply.
Try simulating it.
DF96 is right about this circuit. Can't expect it to behave the way you would like.
Maybe think about it like this: When the single diode connected to the 20v tap conducts, the current loop to charge the low voltage supply capacitor through the transformer is flowing clockwise. Think about what that implies for how the bridge diode are biased at the same time. There is no clockwise path to charge the capacitor through the lower bridge diodes.
Maybe think about it like this: When the single diode connected to the 20v tap conducts, the current loop to charge the low voltage supply capacitor through the transformer is flowing clockwise. Think about what that implies for how the bridge diode are biased at the same time. There is no clockwise path to charge the capacitor through the lower bridge diodes.
Last edited:
It looks like a second bridge to charge the lower capacitor, but what about it's load? If the load is floating and connected across the capacitor, fine. But if one end of the low voltage load is connected to ground, then the capacitor isn't filtering it.
I think the bottom line is the basic scheme you are trying for isn't going to work.
I think the bottom line is the basic scheme you are trying for isn't going to work.
Guys, I designed the military things in the past that still fly )) the wiring "error" is out of question. I actually re-done it from scratch this time, but noticed very same slow -U rising behavior same as a month ago.
All diodes and caps have been changed. The trafo is same, but, as stated, I test-connected the bridge to 180, to 20, and to 200 AC and each time the output is good for 30 mA with insignificant drop, so the trafo/winding is fine.
Loads - I told you what they are: simple plain resistors, measured with multimeter of course. The load onto particularly -U is 820 ohms, how can be it "too high"?
Simulation, kiss my **** )) I now understand why such an "obvious" solution just does not exist in Internet. Heavily searched for couple hours - found any another (im)possible variation except the very this one. That's why I asked for your input, hoped if there is some trick to make it working.
FMB, I see, could come handy indeed in another build, but I need the + tied to GND. Well, I think my best/easiest solution is an additional bifilar 76-turn winding, which gives me 22VAC with a CT, will do, not a big deal though some elegance will be lost.
All diodes and caps have been changed. The trafo is same, but, as stated, I test-connected the bridge to 180, to 20, and to 200 AC and each time the output is good for 30 mA with insignificant drop, so the trafo/winding is fine.
Loads - I told you what they are: simple plain resistors, measured with multimeter of course. The load onto particularly -U is 820 ohms, how can be it "too high"?
Simulation, kiss my **** )) I now understand why such an "obvious" solution just does not exist in Internet. Heavily searched for couple hours - found any another (im)possible variation except the very this one. That's why I asked for your input, hoped if there is some trick to make it working.
FMB, I see, could come handy indeed in another build, but I need the + tied to GND. Well, I think my best/easiest solution is an additional bifilar 76-turn winding, which gives me 22VAC with a CT, will do, not a big deal though some elegance will be lost.
As I said more than once, you should not expect to be able to draw more current from the negative supply than is being supplied by the positive supply. This ensures that the diode at the grounded end of the bridge conducts on AC peaks, which in turn ensures that the junction of the caps is connected to the secondary.Shef said:
Humour me. Plot a graph of -ve voltage versus -ve current. Then draw some current from the +ve rail and repeat the exercise. If I am right you should see that you can now draw more current from the -ve before the -ve voltage droops. If you don't want to simulate then careful measurement is your only option to aid understanding of the circuit.
The reason that the voltage only slowly recovers is that it is leakage current through the caps which is returning things to normal behaviour.
I note your confidence. I sometimes get confident too.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.