It's been a while since the messages above... Priorities changed after our second son was born 🙂. I was only now able to get back to my F5 and I'd like to thank you both for your suggestions. My F5 is now assembled, biased and even playing some music through test speakers. Had to do a few iterations to reduce hum, but now it's around 0.1mV on the DMM, looks ok. It will soon meet my Alpairs 7.3...
I am having a mechanical hum problem with my F5 amplifier. At first I thought the problem was my 300va transformer being too small, so I replaced it with a 600va. Well the 600va trasformer is humming too. I now am wondering if there is a problem with my power supply. I am using this board with 30,000uf 35V caps. The amp sounds wonderful but what is odd is that I am getting ~22VDC on the positive rail and ~22.7VDC on the negative. Any ideas on what might be going on here? I'm not getting any major hum on the outputs and my only idea is that maybe something is wrong with one of my rectifiers or power supply caps. Any ideas on where I should start?
DC on the mains? Look for DC blocking such as this: Mains DC and Transformers I think it has solved other people's hum problems.
Transformer hum problem.
Does the transformer still make a mechanical hum when the F5 is disconnected from the power supply voltage rails?
Does the transformer still make a mechanical hum when the F5 is disconnected from the power supply voltage rails?
Good question.
Build and test in stages.
Add modules after the previous have passed all tests.
Secondaries of a transformer may have a differing Vac value.
But with a bridge rectifier, each secondary winding charges both rails.
For the bridge rectifier, two diodes are strictly for the positive rail, the other two for the negative supply line.
Power diodes have a (small) forward voltage spread, you are not using an integrated diode bridge, but individual diodes.
Worst case, rail difference would be twice the max Vf spread.
Electrolytics have a much larger tolerance spread, in the order of +/-20 percent.
Worst case, smoothing capacitance value on one rail would be 1.5 times higher than the other phase [(100 + 20)/(100-20)]
You spoke of rail voltage, but supplied (little) intel about the transformer(s)
The transformer Vac should be within the tolerance of the manufacturer.
I would expect no worse than 1% match between same winding voltages.
and if bifillar wound then that reduces to 0% difference.
Once you connect that through a 4diode bridge and smooth it, the bridge and capacitor tolerances have an insignificant effect on the smoothed voltage difference.
I would expect no more than 0.1Vdc difference and usually much less, if your voltameter is capable of reading to the required resolution.
Now add on the same load to both polarities. Now the transformer and rectifier and smoothing tolerances will give more potential for voltage differences. But still less than 0.1Vdc
Change to unequal loading on the two polarities and the voltages will go awry. They can vary enormous amounts.
your gross difference of 0.7Vdc indicates a wiring error, or a different current loading.
I would expect no worse than 1% match between same winding voltages.
and if bifillar wound then that reduces to 0% difference.
Once you connect that through a 4diode bridge and smooth it, the bridge and capacitor tolerances have an insignificant effect on the smoothed voltage difference.
I would expect no more than 0.1Vdc difference and usually much less, if your voltameter is capable of reading to the required resolution.
Now add on the same load to both polarities. Now the transformer and rectifier and smoothing tolerances will give more potential for voltage differences. But still less than 0.1Vdc
Change to unequal loading on the two polarities and the voltages will go awry. They can vary enormous amounts.
your gross difference of 0.7Vdc indicates a wiring error, or a different current loading.
Last edited:
Ok, here is what I have tested and found. If I disconnect the amplifier boards from the PS the mechanical hum disappears. I can only hear it if I press a screwdriver to the transformer and press my ear to the screwdriver (old mechanics trick). When I remove the amp boards my positive rail is 26.44V and my negative rail is 26.42V, I have a good meter so I trust my measurements. If I load the power supply with just the left channel board I get positive~23.8 and negative~24.4 and a mechanical hum If I load the power supply with just the right board I get positive ~ 23.9 and negative ~24.4 and a mechanical hum. Unless there is an error on the cviller v1.0 boards that causes uneven current draw on the positive rail I believe the problem must lie with the PS or how I have wired in the transformer. I do not believe that there is problem with the transformer itself at this point. What should I test next?
jacco vermeulen both transformers are antek transformers with 18v secondaries. The first was a 300VA, the second a 600va.
Last edited:
one more question. Is there anyway that this problem could possibly be caused by having one of my secondaries reversed to my power supply board?
From your #14994 post, your meter is okay and the offset is consistent with both channels and the components on your amp boards are okay as the amp's functioning okay with no/little dc bias on the o/p, about the only thing left to check is the MUR860 diodes from Jim's Audio that might also be producing the transformer hum, as per your query on post $14982 - check for temp &/or simply replace them for new ones
Incidentally, if replacing the MUR860's, I'd suggest the Soft Recovery MSR ones (ie. MSR1560, etc)
I doubt that the caps would produce this difference but, if so, one or more caps will be hotter than the others
A bit of a puzzle ...
Incidentally, if replacing the MUR860's, I'd suggest the Soft Recovery MSR ones (ie. MSR1560, etc)
I doubt that the caps would produce this difference but, if so, one or more caps will be hotter than the others
A bit of a puzzle ...
Insert a 1r0 resistor into each PSU output lead.
Do not connect speaker.
Measure the voltage drop after powering ON.
Do not connect speaker.
Measure the voltage drop after powering ON.
We have reached post 15000
