Acrosound PP 6550: Any Advice?

[zman3]

Hello all,

I'm considering building a pair of 6550-based 60W monoblocks around the Acrosound schematics shown on page 13 of this .pdf (High Power Ultra-Linear Williamson):

http://www.pmillett.com/file_downloads/acrosound.pdf

Does anybody have any experience with this circuit? What, if anything, is there to look out for? Are there any glaring mistakes in the schematic?

I've considered several designs by Joseph Norwood Still, but his designs seem to push several components - namely tubes and power transformers - to their limits (I'm not a fan of forced-air cooling). I have a pair of Hammond 278CX transformers just waiting to be used, and I think they'd be quite well suited for this project (if not a bit underrated!).

Thanks in advance for any wisdom anyone could share on this subject.

[boywonder]

What I found most interesting about your attached PDF is that it cost 2 cents postage to mail it back in the day! A very fun read...

Sorry I am too much of a newbie to critique the circuits.

[HollowState]

Originally posted by zman3

Does anybody have any experience with this circuit? What, if anything, is there to look out for? Are there any glaring mistakes in the schematic?

I see no glaring mistakes in the design. It''s a pretty standard straight forward circuit, but there are a few things I would change if I were to build it.

1) Use a full wave bridge in the bias supply.

2) Use at least a 100ufd cap after the choke.

3) Replace the single 10K bias/balance pot with two seperate ones. One for each tube. Also loose the 25K/5W pot to a fixed resistor.

4) Put 10 ohm resistors in the output cathodes in order to measure current with a small onboard meter and switch. And loose the fuses.

5) (Extreme and basically designers choice) Rebias the 6SN7 driver stage to allow direct coupling the input thus eliminating two coupling caps. Go all out with a CCS in the cathodes. This may or may not make it sound any better.

BTW, forget about the 100W amp on page 11. The input can't work right as shown. A 6SN7 cannot work into a 2700 ohm load. But it would work with a (transistor) current source in each cathode.

Victor

[zman3]

Thanks for the input, guys. I have absolutely no desire to build a 100-watt monster amp with 6146's, even thought I have scads of those tubes (they're currently set aside for my HW-101 ham transceiver).

Another question does bug me though. The acrosound schematic calls for the TO-330 output tranny, which has a primary impedance of 3.3k. The closest thing I can get in reasonably-priced iron (e.g. Edcor or Hammond) is around 4.3k. I have a pair of Hammond 1650R's, but they're 5k. Any suggestions?

Thanks again!

[SY]

Yes. Look at the impedance curve of any speaker, think about the transformation ratio and what the plate load will REALLY be, then stop stressing about it.

[HollowState]

Originally posted by zman3
The acrosound schematic calls for the TO-330 output tranny, which has a primary impedance of 3.3k.

Actually, according to my hard copy of the Acrosound catalog, the TO-330 has a primary impedance of 3800 ohms. What SY is saying is that the actual load presented by any speaker system is variable over a range at different frequencies. The numbers given for impedance are general guides only. A transformer rated at 4300 primary ohms will work good for 6550s. A rating of 5000 ohms is on the high side for these tubes, but would still work albeiit with a little more distortion and greater output. 3300 ohms is on the low side of what 6550s want. So don't be overly concerned with strict tight numbers. I think the Edcor's would be a good choice, as personally, I'm not a big Hammond fan.

For whatever transformer you use other then a TO-330, some slight adjustment of the feedback 47pfd capacitor will probably be necessary when observing a square wave. But this is tweeking after you get it built and running.

Victor

[zman3]

Thanks, Sy and Hollowstate.

Yeah, the "3300" was definitely an oversight on my part. You guys are right about the mismatch. A friend of mine who repairs MRI machines says it really shouldn't be a huge deal. I really don't feel so bad about it now.

I agree with you on most of the circuit changes. The full-wave bridge is definitely an improvement over the simple diode. I was also planning to use improved filtering in the power supply, and to implement the metering scheme you suggested.

The 25k pot in the bias supply will probably remain, as the possibility of using EL34's intrigues me. The 5W rating does bother me. Why so high? This is a bias supply running less than 50V at what I would suspect to be very little current.

As for the re-biasing of the driver stage... well let's just say I'm not a 'newbie" as such, but I'm not quite ready for prime-time either!

I've read a few good things about Edcor iron. Do you have any personal experience with it?

Thanks again!

[weinstro]

For an interesting tutorial on using load lines to determine the OPT, do a search for Patrick Turner Audio.

[Tubes4e4]

Hi hollowstate,

BTW, forget about the 100W amp on page 11. The input can't work right as shown. A 6SN7 cannot work into a 2700 ohm load.

Of course it can, and perfectly will. That´s just the so called Van Scoyoc (or "cross coupled") inverter. If you are still in doubt, just look it up in RDH4, or even more convincing: Breadboard and measure through it.

Regards,

Tom Schlangen