p3a dual mono?

[Big gain]

Greetings,

After going through 1325 different iterations of my P3a amplifier, I am about to throw in the towel and go to a true dual mono version of this beast. I am using a +/-42V supply with ON Semi's new MJE15034/5 drivers and MJL21193/4 outputs.

The problem is the two channels don't like to be run together at the same time off the same power supply. Operating each channel individually, they sound very good. Clean sound, very nice, Bravo! With no input, they are dead quiet. When I run them both together, the bias jumps on both channels from 94mA to about 300mA, and hum problems become quite loud. Moving wires helps only somewhat. I can hear the hum several feet from the speakers. Ground loop? I can't figure it out.

Sounds like a grounding problem to me but I have tried several different scenerios with no success. I have cut Rod's boards in half and each has it's own heatsink. Should each channel have it's own HQG point, or can they share one? I have tried both, no difference.

Has anyone had similar problems? Going dual mono looks like a simple solution, and I am willing to do it. Maybe a bit drastic, but I would like to get a final version of this amp up and going. Any ideas would be greatly appreciated.

Thanks much,

Big gain

[mhb]

Big Gain,

I've gone through many iterations as well, but never experienced grounding/hum issues.

E-mail me with specifics on your design, along with pictures you may have. I'll be happy to help you out.

Cheers,

M

[Big gain]

I don't have pics to send so I'll try to describe everything important. So the basic setup is like this. Rod's boards, cut in half. One power supply running +/- 42 V. 14,100uF per rail. Tranny is a 500VA 30-0-30 Avel Lindberg. Each channel has a separate heatsink. Grounding is as follows.

Thick wire from power supply common to a Star ground. To this star I have connected each channel's ground connection and input ground. Speaker ground has separate connection to power supply ground. A Velleman speaker protection unit also has a separate return to supply ground.

Source is a HK receiver using the preout connections.

If I decide to go the dual mono route, will the common ground for the pre-outs in the preamp cause problems with the separate power supplys?

This one has me stumped.

Thanks for any help,

Big gain

[sonnya]

It could very well be your ground connection to the input from the starground who is to thin. This way, a ground current loop will because of the high resistance turn in to voltage across the input.

[Andypairo]

Try making a drawing showing the layout (placement of boards/connectors/trafo/caps, etc), so that we can guess better

My P3A doesn't suffer from ground loops.

Cheers

Andrea

[mhb]

Big Gain,

The only way I have been able to make the P3a hum is to purposely route the input (+) and (-) over the toroid.

Make sure your input chassis wiring is shielded, and that your input jack is as close as possible to the PCB.

I use Avel toroids too, and they are quiet. There is more of a problem with hum running over the toroid than around, so if you are forced to run the inputs close to the tranny, run them low and not over the transformer top.

Your offset is too high, what are you using for a feedback capacitor? Offset with decently matched LTPs should be <30mV at your voltage, <20v with a 25-0-25.

I've attached a pic for a single channel P3a, ignore the tranny secondary wiring to the rectifier/PSU pcb, it's really routed nicely around the rectifiers.

Hope this helps,

MB

[sss]

i'm having almost the same problem

i had 2 p3a's running from single supply with everythin working perfect , i accidentally touched the heatsinks of the driver transistors (bd140/139) with a screw driver (oops ) and ........
so i changed the drivers and the second stage (gain stage) transistors to the new ones (mje15035/34)
now the same thing is happening to me ! when i power em up the bias jumps also !!
i can see that the supply voltage drops by ~3V

i think the problem is oscilation !!! not ground loop

i fixed it by increasing one of the stabilization caps (the one that connected to one of the drivrs)

BTW imo this happens because those are very fast transistors compared to my output transistors- mj15003/4

[mhb]

SSS,

I'm using MJE15032 and 3, never had oscillation problems.
Is your output high pass shunt in place? (10 ohm, .1uF cap)
And how about the input high pass? (220pF)

BTW, I'm using 100pF miller caps, no oscillation.

Rod's boards are 1oz copper, you have to re-solder with great care.

And I assumed you turned the MJE15034,5 transistors around, the pin config is a mirror image of the BD139/40 devices, so the heatsink side faces the output transistors.

Good luck,

M

[sss]

i'm not using Rods boards - thats one of the causes

about the output high pass shunt , is it important what type of cap i'm using ? i think its ceramic

[mhb]

sss,

I'd use a film cap, polyester would be fine, polycarbonate or pp are better. A 5mm lead spacing polyester film cap is <50 cents US.

You can use .1uF or .22uf, .22uF still gives you an F3 of about 75000hz. Make sure the 10 ohm resistor is rated at .5watts or better. Changing the resistor value will obviously necessitate a cap value change.

Leaving out this shunt circuit is not a good idea, IMO.

cheers,

m