Help with PS under load.

[JForged]

Hi peeps,


So I built a nice little headphone amp using a 12au7 and I have been running it off of my DC power supply for a bit when I decided I'd like to make this for my father-in-law but he would need it to be ac capable.

So I decide I am going to make these a two box setup mounted to one small board. A box for the PSU/fuse/switch and a box for the tube amp/ inputs/ volume.I put together a nice enclosure with my transformer and rectifier I test it and everything looks fine.



But after I connect a load to it. In this case I did the amp first with a horrible sound, then I tried a large wirewound resistor. This is what I get




I'm hoping someone here can help me sort this out as it has turned into a bit of a head scratcher for me.

[Mooly]

More info needed such as the circuit details of the PSU and the value of the load resistor you used. Also how much current does the headphone amp draw.

[krokkenoster]

Originally Posted by JForged

Hi peeps,


So I built a nice little headphone amp using a 12au7 and I have been running it off of my DC power supply for a bit when I decided I'd like to make this for my father-in-law but he would need it to be ac capable.

So I decide I am going to make these a two box setup mounted to one small board. A box for the PSU/fuse/switch and a box for the tube amp/ inputs/ volume.I put together a nice enclosure with my transformer and rectifier I test it and everything looks fine.



But after I connect a load to it. In this case I did the amp first with a horrible sound, then I tried a large wirewound resistor. This is what I get




I'm hoping someone here can help me sort this out as it has turned into a bit of a head scratcher for me.

It looks like if you have a bad rectifier it is difficult to judge as it is not clear what the oscilloscope are. From what I can see is that the rectifier is either shorted or built incorrectly

[DF96]

You did include a reservoir capacitor in the box with the transformer and rectifier?

[Robert Kesh]

More load needs bigger caps.

If you split the filament current from the rest you can use smaller caps on the low current part.

[krokkenoster]

Originally Posted by Robert Kesh

More load needs bigger caps.

If you split the filament current from the rest you can use smaller caps on the low current part.

Robert if I read the 'scope readings correctly, it is a sine wave, a reservoir cap too small will a give pulsating D.C. even a full wave will be pulsationd at double your supply frequency.

[Robert Kesh]

Originally Posted by krokkenoster

Robert if I read the 'scope readings correctly, it is a sine wave, a reservoir cap too small will a give pulsating D.C. even a full wave will be pulsationd at double your supply frequency.

The size of cap required to smooth "pulsating DC" is proportional to the current you are smoothing, ie proportional to load. A low value resistor in your tests should make more ripple than a high one given the same caps.

The largest current drawn by a 12au7 headphone amp will probably be going to the filament. The filament is much more tolerant of ripple (some people think AC sounds better than DC here), so you won't need big caps here.

So you can concentrate on properly smoothing the rest of the current, which will be smaller, and so require smaller caps to get the same level of ripple as if you smoothed all the supply in one go.

[krokkenoster]

Originally Posted by Robert Kesh

The size of cap required to smooth "pulsating DC" is proportional to the current you are smoothing.

The largest current drawn by a 12au7 headphone amp will probably be going to the filament. The filament is much more tolerant of ripple (some people think AC sounds better than DC here).

So you only have to smooth the rest of the current, which will be smaller, and so require smaller caps to get the same level of ripple as if you smoothed all the supply.

AGREED

[JForged]

OK everyone.

Thanks for the information.

Here is the current psu schematic.




and the schematic for the 12au7 headphone amp I pulled from this very site. I believe the credit goes to Roger gomez.




I'm not sure that the Tube amp is the problem though as any load on the PSU introduces that ripple effect.


From what I'm gathering I probably need to have a more detailed smoothing section after the rectifier.

Does this seem a correct assessment?

[Mooly]

Your bridge rectifier is wired incorrectly... maybe just a typo The two right hand diodes both need to be the other way round.

More smoothing yes The 47uf cap could usefully be a little bigger. Try at least 1000uf.

The centre tap from the tranny can be left floating. As it is you must ensure that it does not connect to the negative rail of the amp as if it does it will short out one winding of the tranny via the bridge rectifier.