My next project is upgrading the power supply on my k-12 Push Pull amp. I Know some people don't think that it is worth it, but I don't care. I am becoming addicted to the PP sound. I want to really clean up the power and find out what this amp can do. I am thinking of adding some very large caps and maybe tube rectification. Any ideas? Schematics? I am not very good at PS design. Thanks!
I filtered the B+ with a simple CRC filter. 100uF-100R-220uF instead of the stock 220uF.
Since I included a separate filament transformer the B+ was a bit higher anyway.
Higher voltage rating (350-450V) of these caps is not a bad idea since the B+ goes beyond 250V during warm-up (my stock caps were rated 250V).
Replace the bridge rectifier with four cheap UF4007 or other fast diodes or even Schottkys reduces switching noise (buzz). I put a 10nF 1000V film cap across the HT secondaries to reduce switching spikes.
This helps to get rid of most of the buzz.
Try to google VoltSeconds homepage, he has detailed descriptions of his mods to reduce buzz on the K-12.
There will still be 60Hz hum.
This is mostly coming from the AC filament heating. There is an easy trick to reduce that dramatically. Lift the ground reference of the filament winding (center tap) above the cathode voltage will prevent electrons to "escape" from the heater filaments to the cathode. This happens if the cathode is more positive than the filament.
I simply make a voltage divider with 1M and 150k resistors in series.
The ~30V between the resistors is stabilized by an electrolytic cap to ground (47uF will do). Now connect the center tap of the heater secondaries to this +30V point instead of connecting it to ground.
If you have decent wiring of your amp and keep signal wires away from any AC carrying wires or wires with high currents in them, your amp should be quite quiet.
My amp can be turned up to the max and there is no noise if I stick my ear in my relatively insensitive 90db speakers.
It revealed horrible signal-to-noise ratios of some of my sources, though. By the way, make sure there is no chance of any input signal while you do this "listening test". The amp can be LOUD and the speaker cone is not a safe place for your ear with the volume pot turned all the way up.
(Children don't do this at home...)
Sorry that I can not provide a schematic. I am on a conference in Montreal at the moment and my traveling notebook does not have any electronics-related software installed.
I hope this helps,
Martin
PS: I paralleled the 220uF B+ electrolytic with a 22uF Solen film cap. This made a surprisingly strong change of the overall sound. Highly recommended. I am sure a smaller cap would work also (or even better). This was just what I had laying around and I liked the sound
Since I included a separate filament transformer the B+ was a bit higher anyway.
Higher voltage rating (350-450V) of these caps is not a bad idea since the B+ goes beyond 250V during warm-up (my stock caps were rated 250V).
Replace the bridge rectifier with four cheap UF4007 or other fast diodes or even Schottkys reduces switching noise (buzz). I put a 10nF 1000V film cap across the HT secondaries to reduce switching spikes.
This helps to get rid of most of the buzz.
Try to google VoltSeconds homepage, he has detailed descriptions of his mods to reduce buzz on the K-12.
There will still be 60Hz hum.
This is mostly coming from the AC filament heating. There is an easy trick to reduce that dramatically. Lift the ground reference of the filament winding (center tap) above the cathode voltage will prevent electrons to "escape" from the heater filaments to the cathode. This happens if the cathode is more positive than the filament.
I simply make a voltage divider with 1M and 150k resistors in series.
The ~30V between the resistors is stabilized by an electrolytic cap to ground (47uF will do). Now connect the center tap of the heater secondaries to this +30V point instead of connecting it to ground.
If you have decent wiring of your amp and keep signal wires away from any AC carrying wires or wires with high currents in them, your amp should be quite quiet.
My amp can be turned up to the max and there is no noise if I stick my ear in my relatively insensitive 90db speakers.
It revealed horrible signal-to-noise ratios of some of my sources, though. By the way, make sure there is no chance of any input signal while you do this "listening test". The amp can be LOUD and the speaker cone is not a safe place for your ear with the volume pot turned all the way up.
(Children don't do this at home...)
Sorry that I can not provide a schematic. I am on a conference in Montreal at the moment and my traveling notebook does not have any electronics-related software installed.
I hope this helps,
Martin
PS: I paralleled the 220uF B+ electrolytic with a 22uF Solen film cap. This made a surprisingly strong change of the overall sound. Highly recommended. I am sure a smaller cap would work also (or even better). This was just what I had laying around and I liked the sound
bayermar said:
I found a cap across the secondary just caused ringing from the rectifier switching spikes. The transformer and cap are a simple oscilator !
I just used an ordinary 4007 so that could be why I had more trouble with the switching spikes.
In the end I put a 470nF across the rectified supply to ground and that helped a lot.
I also had trouble with the heater causing hum so I used a regulated DC supply in the end.
Link to Voltsecond's site:
http://www.siteswithstyle.com/VoltSecond/K-12M_AMP/K-12M_Push_Pull.html
Scott
http://www.siteswithstyle.com/VoltSecond/K-12M_AMP/K-12M_Push_Pull.html
Scott
Hi ambience,
I usually use resistors that are rated 3x over the actual power that is consumed by the resistor. This way the resistor stays reasonably cool (noise and reliability).
I don't think you can use a tube rectifier easily with the same transformer because of the higher voltage drop. You would definitely have to adjust the bias of the tubes. The input triode already runs on a quite low plate voltage in the stock configuration.
Martin
I usually use resistors that are rated 3x over the actual power that is consumed by the resistor. This way the resistor stays reasonably cool (noise and reliability).
I don't think you can use a tube rectifier easily with the same transformer because of the higher voltage drop. You would definitely have to adjust the bias of the tubes. The input triode already runs on a quite low plate voltage in the stock configuration.
Martin
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.