Esp P3a

[audioPT]

Guys,

I'm working two ESP's P3A in bridge mode with a load of 4Ohm.

Rod wrote that this must not be done.

My question is:
There is any problem with my amp, since the PSU is made with a 18x0x18V 380VA xformer?

[paulb]

You're effectively driving a 2-ohm load from each amplifier. This will be very efficient and you may find the output transistors get really hot with only moderate power going to the speaker. You may want to double up all the output transistors; see the original P3 article (I think) for details.

[audioPT]

Yep, I know that I'm driving 2 Ohm...

I also forgot to write that I'm using 2 pairs of MJL in each amp.
Also read Rod's text about bridging the amps and if I'm using 25-0-25V, I think the amps are working in class A (as explained in P3Ab project).

But now I have a doubt:
If this amps are working in class A, why they don't get hot?

They are driving an isobaric subwoofer, and after "The day after tomorrow" movie, the heatsink still cold... I mean, not very hot.

My LM12 'gainclone' gets more hot...

I've measured the output and got this at 100Hz / 4Ohm (bridge):
Umax=19,7V
Imax=4,9A

Does it mean a power of 96,53W?

[Mark25]

the p3a is not class A, IIRC p3b is. some poeple increase the quiecent current of the p3a to 300ma, etc to get part of the output in class A.

what quiecent current are you running? you can go quite low if you are only driving a sub.

[audioPT]

I can't get more than 230mA...

Only know that my neighbours are not happy

I've said Class A because I've a 18x0x18V xformer

[Mark25]

The fundamental difference between Rod's Class A (P3B) and his class B (P3A) is the quiecent current. The P3A runs at 40 - 100mA, the P3B runs at 1.5 amps, nessesitating a BIG heatsink and different power supply topology.

If i were using a p3a for a sub, i'd start with the quiecent current at 10 mA, increasing it to max 40mA, ONLY if it made an audable difference, i suspect it wouldn't.

when measuring the quiecent current, it can be useful to measure the actual emmiter resistance value. The actual value of those high power 0.33 ohm resistors can vary from 0.33 quite a bit in my experience.

[audioPT]

Wanna know something?

I've changed some things in both amps (of course, they are bridged and must be equal):

Output resistors -> from 1 x (0,33 Ohm 5% 5W) to 3 x (1Ohm 2% 2W);
Q9 (bias) from BC546 to BD139 and attached at the heatsink;

The surprise:
Iq down to 28mA. And after one hour of true "earkilling", visceral shaking and other tremours with "Chemical Brothers" at full power, the Iq just stop at 41mA.

By the way, the heatsink was not very warm...

This word is for Rod Elliott: Very nice amp, apart from being undestructable!

obs: if any word wasn't well written, I'm just sorry


Regards!

[richie00boy]

You should not attach Q9 to the heatsink because the CFP output stage configuration requires you to sense the temperature of the driver transistors.

[Mark25]

Originally posted by audioPT
The surprise:
Iq down to 28mA. And after one hour of true "earkilling", visceral shaking and other tremours with "Chemical Brothers" at full power, the Iq just stop at 41mA.

out of interest, how low could you go before it made an audible difference?

[audioPT]

It's only for low consuption in idle.
In low frequencies, I think nobody can hear the difference.

You can call me anything you want, but in these frequencies, Class A or Class B or Class D sound the same at the same power output, and the power consumming is very different, right?

regards