Moving On

[kt88pppamp]

My PPP units are complete. They hopefully remain stable when I go out auditioning for a new pair of speakers. They put out 130 watts continuous with no visible distortion on the SA at around 5K and below. The upper frequencies report some distortion, mostly 2nd order. Up there, those harmonics are pretty much inaudible anyways. I need a distortion analyzer to see some real numbers. The project was fun.

Frankly, I had more fun researching designing and building them than listening to them. I bet you all know how that is. I ate an irrinductable speghetti dinner to celebrate.



What do you all think of the construction and the circuit?

[LinuxGuru]

Very nice build :-)

For distortion analysis you can use high-quality sound card with input protection, and ARTA software.

[TomMcNally]

Very nice monster amps ! Time to disturb the neighbors.

[kt88pppamp]

I attend RIT and they will be on display at a festival they have there called Imagine RIT. Anyone can attend if they are in the Rochester, NY area. Carlp, you up to coming? It is Saturday, May 5th. Mark your calenders.

http://www.rit.edu/imagine/

I will be located on the ground floor of the College of Applied Sciences and Technologies Building (Building 82) in a conference room.

My speakers, both amps, a preamp, my phono stage, my turntable, and record collection will be there. Yes, I do take requests for music.

[kt88pppamp]

I am finishing the testing of Rev. 2 of my monster amps. At 1k, the distortion appears to be 0.25% at 130ish watts RMS with the new driver.

The previous driver was unstable so I was forced to use a local feedback network to stop oscillation problems. The plate-to-plate feedback put the clipping power at around 80 watts.

The new driver, even without an input appears to be stable. I tried it without a source connected and a few speakers. There was no whistling. I cannot hear any hum and very little hissing.

I will post the new schematic and some scope pictures tomorrow.

[DeathRex]

I am finishing the testing of Rev. 2 of my monster amps. At 1k, the distortion appears to be 0.25% at 130ish watts RMS with the new driver.

WOW.

Gotta see the new driver.

[Blair]

Yes, please!

What iron did you use?

Thanks,

Blair

[kt88pppamp]

Here is a preliminary schematic (see attached).

I had Heybor Transformers in Grand Haven, MI wind me the power and output iron two years ago. The circuit really is nothing too special. A 12AV6 can be used in place of the 12AX7 if some wiring changes are made. The only problem is they are no longer made.

About 15dB of feedback is used to set the sensitivity to around 1.5V. Lowering the FB to 10dB will raise the distortion to 0.3%ish and the sensitivity will drop to 0.8ish volts.

Open loop gain is 40ish dB. With feedback it is 25-30dB. Without FB the distortion probably is more than tolerable for some folks. Only problem is the amp is sensitive at 0.35 volts for full power.

[DeathRex]

The circuit really is nothing too special. A 12AV6 can be used in place of the 12AX7 if some wiring changes are made.

I guess it looks alot like the Eiclone. Didn't know you could get that little of distotion.
I tried 6AV6s and kept having problems with hum from the heater. It varied greatly from tube to tube, but there was so much more than from 6AB4s, or 6C4s. I don't remember if I raised the DC on the heater or not.

I've had nothing but good times with the LM334Z, like in the Posiden. Just that with only a LM334Z you have to add an additional capacitor.

[kt88pppamp]

This morning, I tested the amp and here are the results:

- The square wave at 10k has practically zero ringing. There is no phase compensation except for an 82pF cap across the feedback resistor.

- The test load was an 8 ohm 225 watt resistor.

- Input sensitivity is 2VPk (1.414 volts RMS) for 138 watts.

- THD is around 0.25% at 138 watts out, at most frequencies every octave.

- Attached are my published test results.

I never even looked at the eiclone circuit or heard much about it for that matter. I guess me and Shannon happened across very similar circuitry. After all, tubes circuits are very simple. The LM334 is a semiconductor and has popped on me under certain conditions. That is why I shy away from it. I thought I would have to sacrifice some performance for a resistor instead of an LM334. But no, a phase compensation pot does the same job. Why use semiconductors where they do not belong? Furthermore, adding a cap places another pole in the system, rendering it coherently less stable. Why would you make a circuit less stable for a CCS (especially for PPP circuits)?

Questions:

- I was blown away by the numbers with the amount of feedback I am using. Did anyone else think without an LM334 low distortion was possible? Has anyone done this with such high power, or at all, without a CCS?

- The distortion increases exponentially and becomes really apparent after 10k. This happens in all amps I have worked on, even solid state. Why is this? Harmonics > 10k are inaudible anyways, so I guess they are moot.

- With a simpler circuit, I am producing comparable numbers to the figures McIntosh and Audio Research produce. Why do they use more complicated circuits and carefully selected tubes, where with run of the mill parts, I can produce similar numbers? What is with their high prices for what is really nothing special, or in some cases, inferior?

- I asked Heyboer for a 130 watt power handling transformer. Why is it still producing low numbers at 138 watts?

- What is up with the square wave? Is it low leakage inductance?



- What do you think of the wiring job (see above)?

[Blair]

Looks good! The freq response Heyboer provided has a "minimum" frequency response +/- a FB or do. The rolloff is typically more incremental with better iron. Look at the original A-470 specs. They'll do 60W, but not quite the same response as their claimed 30W response.

What primary impedance did you request for your outputs?

[kt88pppamp]

The OPT primary has a 1.9k primary and a 4, 8, common secondary.