I have built a full wave center tapped power supply and despite the fact that nothing is wrong, it does not work properly. All voltages given are rounded to a whole volt.
I wanted a power supply around 32 to 38vdc to power a TDA7498 amplifier. I found a couple of 60W 24 VAC transformers in my stockpile and decided to use them in series to give me 24vac 0vac 24vac. With the full wave center tap and a capacitor output I should get 24 VAC x 1.414 = 34 VDC. Much to my surprise I only got around 20 VDC. Since then I moved the primary taps to the low voltage condition and I was able to get 29 VAC from the transformer and 24 VDC out and have left the connections there for the rest of my readings.
With 29 VAC I should get 29 VAC x 1.414 = 41 VDC but I only get 24 VDC. The pictures show my physical wiring except for one hidden wire. The cathode of the dual rectifier is connected to the TO220 case. The wire that connects to this point is under the heat sink going to a ring terminal on the screw holding the TO220 rectifier to the heat sink. I have two red wires in the ring terminal for higher current operation. The other end of this wire goes to the + connection.
Other than that there is a resistor to bleed the capacitors at power off and two .1uF caps across the large electrolytics to reduce voltage transits.
Each transformer measures 29 VAC.
Across both transformers I measure 58 VAC.
I put a scope on the output to make sure it was full wave and not half wave and it had a very nice waveform that agreed with the voltmeter.
I also put a voltmeter from each anode to the output + and ground to check for symmetry. I checked these connections with both DC and AC settings on the DVM to look for any differences, but none were found.
In short nothing is wrong, it just does not work!
I wanted a power supply around 32 to 38vdc to power a TDA7498 amplifier. I found a couple of 60W 24 VAC transformers in my stockpile and decided to use them in series to give me 24vac 0vac 24vac. With the full wave center tap and a capacitor output I should get 24 VAC x 1.414 = 34 VDC. Much to my surprise I only got around 20 VDC. Since then I moved the primary taps to the low voltage condition and I was able to get 29 VAC from the transformer and 24 VDC out and have left the connections there for the rest of my readings.
With 29 VAC I should get 29 VAC x 1.414 = 41 VDC but I only get 24 VDC. The pictures show my physical wiring except for one hidden wire. The cathode of the dual rectifier is connected to the TO220 case. The wire that connects to this point is under the heat sink going to a ring terminal on the screw holding the TO220 rectifier to the heat sink. I have two red wires in the ring terminal for higher current operation. The other end of this wire goes to the + connection.
Other than that there is a resistor to bleed the capacitors at power off and two .1uF caps across the large electrolytics to reduce voltage transits.
Each transformer measures 29 VAC.
Across both transformers I measure 58 VAC.
I put a scope on the output to make sure it was full wave and not half wave and it had a very nice waveform that agreed with the voltmeter.
I also put a voltmeter from each anode to the output + and ground to check for symmetry. I checked these connections with both DC and AC settings on the DVM to look for any differences, but none were found.
In short nothing is wrong, it just does not work!
Attachments
With these types of power supplies there are a few things to consider....
The Vac RMS x 1.414 is only valid for NO-LOAD voltage or a very tiny load current....
Once you place a load on the output filter cap, you then have a time constant that discharges the filter cap "after" the crest of the sine wave....until it intersects the next waveform cycle...then you have a conduction angle, time period to charge the filter cap back up...
Are the filter caps of correct size???
Are you accounting for voltage drop on primary and secondary winding due to load current ?
correct secondary phasing ???
The Vac RMS x 1.414 is only valid for NO-LOAD voltage or a very tiny load current....
Once you place a load on the output filter cap, you then have a time constant that discharges the filter cap "after" the crest of the sine wave....until it intersects the next waveform cycle...then you have a conduction angle, time period to charge the filter cap back up...
Are the filter caps of correct size???
Are you accounting for voltage drop on primary and secondary winding due to load current ?
correct secondary phasing ???
Cerrem, all the measurements i posted are no load.
If the secondaries were not phased properly i would not get 58 VAC across them.
If the secondaries were not phased properly i would not get 58 VAC across them.
The setup is a bit of a mess, it is difficult to follow exactly the wiring, but I see a number of issues:
Your transformers secondaries have three connections, not two. What exactly are you doing with this third wire?
If it goes to another supply sharing a ground or anything else with this one, you are in trouble.
Even when eventually everything works as intended, it won't be usable: you cannot use two different transformers for a center-tap winding: each will work in halfwave, and this will saturate the core for currents much lower than the nominal value.
When the windings are on the same core, they compensate each other and there is no net magnetization of the core.
You have to parallel the windings and use a bridge as rectifier.
Your transformers secondaries have three connections, not two. What exactly are you doing with this third wire?
If it goes to another supply sharing a ground or anything else with this one, you are in trouble.
Even when eventually everything works as intended, it won't be usable: you cannot use two different transformers for a center-tap winding: each will work in halfwave, and this will saturate the core for currents much lower than the nominal value.
When the windings are on the same core, they compensate each other and there is no net magnetization of the core.
You have to parallel the windings and use a bridge as rectifier.
Elvee, I did not want to complicate things talking about the third wire but I needed a +- 15VDC supply at 50 mA so that is what they are for. This is a regulated supply with 7915 and 7815. And you are right sharing a ground with the other supply was a problem.
Elvee said
“Even when eventually everything works as intended, it won't be usable: you cannot use two different transformers for a center-tap winding: each will work in halfwave, and this will saturate the core for currents much lower than the nominal value.”
I have never heard of a transformer being sold with a warning that it cannot be used for a half wave rectifier.
Elvee said
“Even when eventually everything works as intended, it won't be usable: you cannot use two different transformers for a center-tap winding: each will work in halfwave, and this will saturate the core for currents much lower than the nominal value.”
I have never heard of a transformer being sold with a warning that it cannot be used for a half wave rectifier.
Last edited:
1. Change the scope.
2. Change the rectifier with other type, verified.
3.Change the caps.
4.If after (1&2&3) it`s the same, something it`s wrong with wires or transformers.
Good lock!
It may work better if you connect the transformers in the same way (mirror). 4-4-GND with taps 8 (or 7) going to the PSU. You will need to change the polarity of the primary of one of the transformers.
power supplies obey the laws of physics, yours is no exception, you did something wrong that is why it does not work...
Of course you can: it is just that it will saturate at low currents, be noisy and overheat. If you are prepared to tolerate that, it's perfectly OK
Now where have I heard that before? Take it from an oldie: if it doesn't work properly then something is wrong. In most cases this type of problem turns out to be eitherpowerbob said:
1. a faulty component
2. a faulty build being compared with a faulty understanding, and hence being considered correct
Draw us a circuit diagram of exactly what you think you have built. Show every connection you have made - including the ones which you think are of no consequence. Let's not get sidetracked into an argument about half-wave rectification and what it allegedly does to transformers.
Disconnect the amp and measure the PSU on its own.
I have cleaned up the wiring to make it easier to verify, and also completely disconnected the amplifier. I also took a scope pic of both secondary’s with the ground lead on the center tap.
No load output, 31.2 vdc, wires hanging in the air.
Voltage of each secondary measured at DC ground at the anodes of the rectifier, is 27.8 vac and 28 vac.
Voltage across the two secondary’s measured at the anodes 54.2 vac. Notice it does not properly add up the two separate voltages 27.8 + 28 = 55.8 vac. This may be a clue, I measured this twice to verify.
This is with the primary’s in series. In the next post I reverse one primary.
No load output, 31.2 vdc, wires hanging in the air.
Voltage of each secondary measured at DC ground at the anodes of the rectifier, is 27.8 vac and 28 vac.
Voltage across the two secondary’s measured at the anodes 54.2 vac. Notice it does not properly add up the two separate voltages 27.8 + 28 = 55.8 vac. This may be a clue, I measured this twice to verify.
This is with the primary’s in series. In the next post I reverse one primary.
Attachments
You've linked a text book drawing again. I'll bet that your implementation is not that. Can you draw out your circuit instead?
This waveform is not normal:
http://www.diyaudio.com/forums/atta...es-laws-physics-scope-pic-both-secondarys.jpg
I suspect your diodes have an insufficient voltage rating and are breaking down.
Maybe they are schottky's?
http://www.diyaudio.com/forums/atta...es-laws-physics-scope-pic-both-secondarys.jpg
I suspect your diodes have an insufficient voltage rating and are breaking down.
Maybe they are schottky's?
How you take rectified voltage from the rectifier catods? I don`t see any solid contact there.
The problem could be that the primaries are in series and not parallel. The voltage is not split equaly under load.
Last edited:
In the first post i posted the pdf for the diode i used. I bought it new from digikey along with the capacitors.
Primaries
Primaries must be in parallel, and must be in phase (same on each one)
Then connect the secondaries in series and test voltages under load.
The place where the secondaries are connected would be your center tap.
Makes it like one transformer with two primaries, with each transformer being one side of the center tap.
Hope this helps!
Primaries must be in parallel, and must be in phase (same on each one)
Then connect the secondaries in series and test voltages under load.
The place where the secondaries are connected would be your center tap.
Makes it like one transformer with two primaries, with each transformer being one side of the center tap.
Hope this helps!
Attachments
Elvee, i agree the waveform on the secondary's looks distorted. If you can think of other points to scope out i will post the pictures. I have an isolation transformer on the scope so i can put a probe anywhere in the circuit. And as i said before the output looks like a straight line on DC coupling. With AC coupling from memory i went down to 2 mV a division and there was maybe 6 to 10 mV of ripple.
- Status
- Not open for further replies.