I just did the maiden switch on of my pre-amp project that I've been working on for some time. I had it working fine on a board and a regulated supply.
Now its in chassis with dual mono choke loaded non-regulated supplies. I did all the modeling in PSU designer II prior to ordering the transformers. I included the model here along with the schematic.
One mistake I made in the model was just not showing the bleeder resistor.
Otherwise, I'm having 2 issues. Overall its down 29VDC. Normally, I use C1 to adjust voltage when I was using the old transformers. Now C1 makes no different whether its the value in the PSU model or all the way up to .1uF.
Second, the trans is supposed to be 200VAC output and its measuring 390VAC. Since its almost double then something is off. It was modeled as a full wave but it seems like something is wrong. I am going CT (5) to ground.
Now its in chassis with dual mono choke loaded non-regulated supplies. I did all the modeling in PSU designer II prior to ordering the transformers. I included the model here along with the schematic.
One mistake I made in the model was just not showing the bleeder resistor.
Otherwise, I'm having 2 issues. Overall its down 29VDC. Normally, I use C1 to adjust voltage when I was using the old transformers. Now C1 makes no different whether its the value in the PSU model or all the way up to .1uF.
Second, the trans is supposed to be 200VAC output and its measuring 390VAC. Since its almost double then something is off. It was modeled as a full wave but it seems like something is wrong. I am going CT (5) to ground.
Attachments
Hi there,
the transformer datasheet you show says 400V CT, so 390VAC sounds ok to me. Don't be confused by the 200VAC shown in the PSUD screenshot. For full-wave rectification, the value shown in PSUD is for one half of the winding - that means if PSUD says 200VAC in full-wave mode, the transformer needs to be 200-0-200 or, written differently, 400VAC CT.
Greetings,
Andreas
the transformer datasheet you show says 400V CT, so 390VAC sounds ok to me. Don't be confused by the 200VAC shown in the PSUD screenshot. For full-wave rectification, the value shown in PSUD is for one half of the winding - that means if PSUD says 200VAC in full-wave mode, the transformer needs to be 200-0-200 or, written differently, 400VAC CT.
Greetings,
Andreas
Thanks Andreas!
Well the good news is it makes music and although my PSU VAC is up significantly from when it was on a board (.045 now .145) output VAC is .009 whereas on the board I had it at .003.
But there doesn't appear to be excess destructive hum. I'm sure I can work on that but I need to try and get it up to the specs. Any ideas why changing C1 no longer affects voltage? I need to raise it 29VDC at the tube plate.
Well the good news is it makes music and although my PSU VAC is up significantly from when it was on a board (.045 now .145) output VAC is .009 whereas on the board I had it at .003.
But there doesn't appear to be excess destructive hum. I'm sure I can work on that but I need to try and get it up to the specs. Any ideas why changing C1 no longer affects voltage? I need to raise it 29VDC at the tube plate.
Hi,
the winding resistance of the trafo should equal the resistance of the secondary winding PLUS the reflected resistance of the primary winding. Did you take that into account?
Other things could be slightly off, too; for example the inductance of an iron-core choke is not constant but is a function of the average current in the choke...
Kenneth
the winding resistance of the trafo should equal the resistance of the secondary winding PLUS the reflected resistance of the primary winding. Did you take that into account?
Other things could be slightly off, too; for example the inductance of an iron-core choke is not constant but is a function of the average current in the choke...
Kenneth
There is something odd about your design. Are you attempting to make a choke input filter? 100nF is nothing when used as the first capacitor in a cap input filter. Might as well not be there. I just ran a quick and dirty SPICE simulation, and at that low value for the input cap, the overall output voltage behaves very much like a choke input filter. With a choke input filter you would also be fighting the need for the minimum, critical inductance of the choke, which might lead to further strangeness etc.
Try increasing the first cap to 10-30uF and check again.
Edit: If this is supposed to be a choke input filter, then the expected output voltage would be (transformer secondary voltage RMS times 0.9) minus (load current times total loss resistance (transformer, rectifier, choke)). With Vrms equal to 195V (390/2), then being 29V too low does sound quite plausible. 195 * 0.9 = 175V. The low secondary voltage could be because your mains voltage is a bit low.
- Frank.
Try increasing the first cap to 10-30uF and check again.
Edit: If this is supposed to be a choke input filter, then the expected output voltage would be (transformer secondary voltage RMS times 0.9) minus (load current times total loss resistance (transformer, rectifier, choke)). With Vrms equal to 195V (390/2), then being 29V too low does sound quite plausible. 195 * 0.9 = 175V. The low secondary voltage could be because your mains voltage is a bit low.
- Frank.
Last edited:
As Frank says in his post, but even 1uF should make a small, but discernible difference. His choke concerns more eloquently echo the concern I expressed in my last post.
Hi Frank,
I agree with all you said above, except
This will essentially turn the choke-input supply to a cap input supply and rise the output voltage significantly, approximately from 0.9 U_in to 1.4 U_in. This will probably be more than wished for...
Edit: The first post says it is going to be choke-loaded (see first diagram).
Greetings,
Andreas
I agree with all you said above, except
This will essentially turn the choke-input supply to a cap input supply and rise the output voltage significantly, approximately from 0.9 U_in to 1.4 U_in. This will probably be more than wished for...
Edit: The first post says it is going to be choke-loaded (see first diagram).
Greetings,
Andreas
Last edited:
You are right on all accounts. I was confused by the OP tweaking the output voltage via two caps of sizes 47 and 100nF. Kevin is probably more on the money with a few uF instead of the 10-30 I mention.
- Frank.
I think C1 at 0.1uF is way too low. This is the main reservoir capacitor and in circuits of this type is typically a few microFarads. A too small value would certainly lead to a low HT.
Cheers
Ian
I do want it to stay a choke loaded PSU. Its odd that in modeling it even the tiniest cap showed big voltage gains. That's why I settled on what is in the PSU model. Real world its not functioning that way.
I just spoke to Bud and he said as many here have that is to try 1.5uF and see what happens. I mentioned I didn't want to go to high and have it tilt to cap loaded.
Previously it was a hybrid choke loaded/regulated supply with an R-core transformer that I did while voicing. I am including it here.
The signal side has zero change. Its even the same parts.
The PSU has new transformers and chokes although the chokes are the same value as the previous one.
Apart from the voltage issue I am getting significantly more ripple. Previously on the output of the psu caps I measured .045vac. Now its .25vac which is identical on each channel (dual mono now). Output ripple has increased from .003 to .009vac.
I am including an earlier build shot and will shoot a new one later which all the wiring in place. The unit is playing now.
I just spoke to Bud and he said as many here have that is to try 1.5uF and see what happens. I mentioned I didn't want to go to high and have it tilt to cap loaded.
Previously it was a hybrid choke loaded/regulated supply with an R-core transformer that I did while voicing. I am including it here.
The signal side has zero change. Its even the same parts.
The PSU has new transformers and chokes although the chokes are the same value as the previous one.
Apart from the voltage issue I am getting significantly more ripple. Previously on the output of the psu caps I measured .045vac. Now its .25vac which is identical on each channel (dual mono now). Output ripple has increased from .003 to .009vac.
I am including an earlier build shot and will shoot a new one later which all the wiring in place. The unit is playing now.
Attachments
OK, did a 'proper' simulation. With 60Hz mains frequency (OP is in the US), the critical load current at Vout = 205V is about 6.3mA. So I did my simulation at 10mA load current.
With Rrect and Rtrafo each 50 ohm and Vtrafo,rms = 195V, I get that 330nF is too small, while 470nF is a bit too large to get 205 V out.
At these low load currents you really don't need more capacitance than this in order to 'tune' the output voltage. Exact values will of course depend on the actual operating conditions.
phrarod,
If you are looking for the voltage stability in a choke input filter, then it will be gone the moment you add enough capacitance on the input side to influence the output voltage. No real way around that, except to accept the lower output voltage without any input cap. There is unfortunately no way you will get 205V out for 195V in with a true choke input filter.
How high is your load current? Edit: Schematic, awesome.
20.5 mA.
- Frank.
With Rrect and Rtrafo each 50 ohm and Vtrafo,rms = 195V, I get that 330nF is too small, while 470nF is a bit too large to get 205 V out.
At these low load currents you really don't need more capacitance than this in order to 'tune' the output voltage. Exact values will of course depend on the actual operating conditions.
phrarod,
If you are looking for the voltage stability in a choke input filter, then it will be gone the moment you add enough capacitance on the input side to influence the output voltage. No real way around that, except to accept the lower output voltage without any input cap. There is unfortunately no way you will get 205V out for 195V in with a true choke input filter.
How high is your load current? Edit: Schematic, awesome.
20.5 mA.- Frank.
Frank - this makes sense as it reflects what PSU Designer showed. But in real world 470nF doesn't even bump the voltage more than maybe 1/2 volt. Ugh I spent a year and a half finding the sweet spot which is 200VDC on the plate with 12mA of bias. Now I'm at 171vdc at plate with 9mA of bias....
Well, that leaves two possibilities:
One is a wiring or component error, but that seems unlikely with dual mono construction.
Second is that your chokes aren't 29Hy at all. If they are too low by a fair margin, then you would get the behavior you describe. If I use a 5Hy choke in my simulation instead of 29, then I get an output voltage, which is nailed at about 180V almost irrespective of the size of the first capacitor (within limits, of course).
If the chokes are 29Hy, then they are plenty large enough at 10mA per channel to be a true choke input filter (with a bit of capacitive help if needed). They wouldn't even be close to go out of conduction.
Is there a chance they are not 29Hy after all, or do we need to look elsewhere?
- Frank.
One is a wiring or component error, but that seems unlikely with dual mono construction.
Second is that your chokes aren't 29Hy at all. If they are too low by a fair margin, then you would get the behavior you describe. If I use a 5Hy choke in my simulation instead of 29, then I get an output voltage, which is nailed at about 180V almost irrespective of the size of the first capacitor (within limits, of course).
If the chokes are 29Hy, then they are plenty large enough at 10mA per channel to be a true choke input filter (with a bit of capacitive help if needed). They wouldn't even be close to go out of conduction.
Is there a chance they are not 29Hy after all, or do we need to look elsewhere?
- Frank.
Hi there,
if chassis space ist not a problem:
One could probably find a transformer with two separate, sufficiently isolated 20-30V windings and connect one of them in series with each 200V winding of your power transformer. Mind correct phases/polarities.
If you really want a choke-input supply, there is no other way to raise the output voltage than to give it a higher AC input.
Greetings,
Andreas
if chassis space ist not a problem:
One could probably find a transformer with two separate, sufficiently isolated 20-30V windings and connect one of them in series with each 200V winding of your power transformer. Mind correct phases/polarities.
If you really want a choke-input supply, there is no other way to raise the output voltage than to give it a higher AC input.
Greetings,
Andreas
Frank was right per much. I was not thinking correctly. The circuit is very sensitive and reacts differently from the last psu. I found that .32uF - I paralleled a .1uF and a .22uF and I ended up with 201.9VDC
So real world is off just slightly on Franks modeling which is close enough to fiddle with. And here I ordered tons of caps that none are of the right value! I should've asked all of you first to confirm by modeling.
I also found that the cans that were not grounded (I have dampening material between then and the clamps) showed AC voltage on the cans. This might contribute to the hum. I'm grounding all of them now.
Time to source some .33uF film caps.
So real world is off just slightly on Franks modeling which is close enough to fiddle with. And here I ordered tons of caps that none are of the right value! I should've asked all of you first to confirm by modeling.
I also found that the cans that were not grounded (I have dampening material between then and the clamps) showed AC voltage on the cans. This might contribute to the hum. I'm grounding all of them now.
Time to source some .33uF film caps.
I found enough caps to parallel. Voltages are set. Now I have to deal with ripple. I don't know if it was an effect but now with the C1 in there @ .32uF I'm getting .3vac ripple whereas before on the board with the choke loaded regulated supply I had .045vac
Any tips on exploring possible causes is greatly appreciated.
Any tips on exploring possible causes is greatly appreciated.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.