you can use a dual secondary transformer and convert that to centre tapped by connecting two of the secondary taps.
Or you can use a centre tapped transformer.
The converted dual, or the centre tapped, will both perform as well as each other.
The smoothing capacitors try to store the peak voltage from the AC mains waveform.
For a sinewave which is what the mains is supposed to be, the peak is defined as sqrt(2) * Vac.
Thus the peak of 50Vac is 70.71Vpk
But that is not the output of your transformer.
Take a 300VA 230:50-0-50Vac 5% regulation transformer fed with variable mains voltage.
The maximum output is:
Outputmax = Vmains/rated mains * rated secondary * (1+regulation)
For a mains voltage tolerance that allows upto 253Vac, the output max is 253/230*(50+50)*(1+0.05)*sqrt(2) = 163.3Vpk
Subtract the ~1V for the two diode drops inside a bridge rectifier passing near zero current. That leaves 162.3V across the series connected capacitors. The voltage across each capacitor is ~81.2Vdc
Too high for 80V capacitors.
When you set up your amplifier plus PSU, the supply rail voltage will be lower (due to quiescent current draw), possibly as much as 3V lower, (in the 50Vac example that would get you down to ~±78Vdc, when mains is at 253Vac) and more typically around ±75Vdc when mains is more normal.
35Vac, 40Vac, or 45Vac would be more suitable depending on what you require from the amplifier and how you build it.
Or you can use a centre tapped transformer.
The converted dual, or the centre tapped, will both perform as well as each other.
The smoothing capacitors try to store the peak voltage from the AC mains waveform.
For a sinewave which is what the mains is supposed to be, the peak is defined as sqrt(2) * Vac.
Thus the peak of 50Vac is 70.71Vpk
But that is not the output of your transformer.
Take a 300VA 230:50-0-50Vac 5% regulation transformer fed with variable mains voltage.
The maximum output is:
Outputmax = Vmains/rated mains * rated secondary * (1+regulation)
For a mains voltage tolerance that allows upto 253Vac, the output max is 253/230*(50+50)*(1+0.05)*sqrt(2) = 163.3Vpk
Subtract the ~1V for the two diode drops inside a bridge rectifier passing near zero current. That leaves 162.3V across the series connected capacitors. The voltage across each capacitor is ~81.2Vdc
Too high for 80V capacitors.
When you set up your amplifier plus PSU, the supply rail voltage will be lower (due to quiescent current draw), possibly as much as 3V lower, (in the 50Vac example that would get you down to ~±78Vdc, when mains is at 253Vac) and more typically around ±75Vdc when mains is more normal.
35Vac, 40Vac, or 45Vac would be more suitable depending on what you require from the amplifier and how you build it.
Last edited:
Hello,
I have a question regarding grounding. I must have made a mistake, any thoughts are welcome.
When I link the grounds from both sides of the PSU board my dim bulb tester shows bright. This is linking gnd 2.2 and 3.3. When I cut the link all is well. The voltages are corrects at about 24v both sides.
What have I done?
These show the link I soldered and had to cut. If I connect both side of the ground to my SL-60 and on to chassis ground it also shows short. If I connect either side by it's self all is ok.
I have a question regarding grounding. I must have made a mistake, any thoughts are welcome.
When I link the grounds from both sides of the PSU board my dim bulb tester shows bright. This is linking gnd 2.2 and 3.3. When I cut the link all is well. The voltages are corrects at about 24v both sides.
What have I done?
These show the link I soldered and had to cut. If I connect both side of the ground to my SL-60 and on to chassis ground it also shows short. If I connect either side by it's self all is ok.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
Thank you guys,
Here is a pic of the power coming into the PSU. not the best lighting but I think it shows how I have it setup.
I wanted to send this prior to trying any changes, I hope (and know) I just made a rookie mistake.
An externally hosted image should be here but it was not working when we last tested it.
Very large version here
Thanks 6L6, yes, you are correct that when both grounds are linked to the chassis though a stm-cl60 at the same time the bulb is bright as can be. If either side is grounded it acts as expected.
Note on the photo, I have not cleaned up the board yet or fixed a few small soldering bits and have not added the LED's, was waiting for my new shipment.
An externally hosted image should be here but it was not working when we last tested it.
sml.JPG){kind=link}
large version here
Thank you all,
brian
Connected to the middle of the chassis, you can see it in the pic's, it's got a yellow spring clip on it right now. I just tried using two separate CL-60's, one for each ground and the light bulb dimmed then slowly went bright and the CL-s got up to 100 degrees so I turned it off. with one side connected by it's self the CL stays room temp. I wonder if I have a bad part someplace or most likely I have made a silly mistake I can't see
Thanks,
brian
Thanks,
brian
Is anything touching the bottom of the PCB when you test? You've got it resting on cardboard or plastic or something and it's not touching the chassis, yes?
If the CL is heating up that means there's a bunch of current flowing, so you've got a short somewhere... that CL is there to add it's 10ohm cold resistance and elevate PSU ground off the chassis so you don't get ground loops through the AC safety earth. It should be stone cold in normal operation.
If the CL is heating up that means there's a bunch of current flowing, so you've got a short somewhere... that CL is there to add it's 10ohm cold resistance and elevate PSU ground off the chassis so you don't get ground loops through the AC safety earth. It should be stone cold in normal operation.
Just a warning.
Some of the pcb tracks on the rectifier boards come too close to the mounting holes so that if you are using washers, the washers will be covering the pcb tracks.
The only thing preventing a short is the thin varnish.
Always remember Murphy's Law reigns supreme.
To prevent issues it's probably best not to use washers here.
Some of the pcb tracks on the rectifier boards come too close to the mounting holes so that if you are using washers, the washers will be covering the pcb tracks.
The only thing preventing a short is the thin varnish.
Always remember Murphy's Law reigns supreme.
To prevent issues it's probably best not to use washers here.
Last edited:
Another question on the power supply for stereo F5T. Would I be safe using 40volt caps or is that pushing it?
I can get Mundorf 33000mf 40 volt caps that will fit or I can get Mundorf 22000mf 63volt caps. Roughly same price but big difference in capacitance. One supply per channel. 176000mf vs 264000mf.
I can get Mundorf 33000mf 40 volt caps that will fit or I can get Mundorf 22000mf 63volt caps. Roughly same price but big difference in capacitance. One supply per channel. 176000mf vs 264000mf.
a dual 25Vac transformer will generate ~37Vdc on a lower bias ClassAB amplifier.
Expect a couple of volts less into a Full ClassA amplifier.
That gets your caps down to ~+35Vdc.
Then add on 10% for max mains tolerance and your 40Vdc is still not exceeded.
But what if the amplifier is disconnected? How high will your 25Vac transformer take the PSU when loading is open circuit?
Expect a couple of volts less into a Full ClassA amplifier.
That gets your caps down to ~+35Vdc.
Then add on 10% for max mains tolerance and your 40Vdc is still not exceeded.
But what if the amplifier is disconnected? How high will your 25Vac transformer take the PSU when loading is open circuit?