DCAudioDIY.com

This is anecdotal evidence, but...

At CAF 2023, Joseph Audio demonstrated its Pearl Graphene speakers, which have a sensitivity of 86 dB. The sound was remarkably immersive, with an enormous soundstage. Beyond that, it was difficult to single out any one or two sonic characteristics in which they particularly excelled. Rather, everything came together seamlessly to create a captivating, effortless listening experience.

The speakers were driven by 1 kW monoblock amplifiers. Jeff mentioned that high-powered amplifiers were essential to achieving that sense of effortlessness.

Of course, we know that 86 dB sensitivity speakers don't necessarily require 1 kW of power. Perhaps the substantial amplifier headroom, along with the ability to deliver high current when demanded, contributed to precise control of the cone movements.

Whatever the explanation, the result of high-powered amp is impressive.

_________________
Walt

Walt,

The point of my post isn’t about how loudly the amp can drive the speakers. It’s about how precisely the amplifier can control the cone’s movement—how quickly it can stop the cone’s excursion without overshoot, reverse direction, and accelerate it in the opposite direction. And do it thousands of times every second.

Walt,

The point of my post isn’t about how loudly the amp can drive the speakers. It’s about how precisely the amplifier can control the cone’s movement—how quickly it can stop the cone’s excursion without overshoot, reverse direction, and accelerate it in the opposite direction. And do it thousands of times every second.[/quote]

Here we go down the "Rabbit Hole." Based on publications from decades ago, and my own experience -- amplifiers with very low output-impedance (aka. so-called damping factor) actually encourage ringing and are not good at arresting unsolicited voice-coil movement. Series resistance with the voice coil, especially with woofers and/or whole multi-way systems using passive crossovers provides electronic restriction. I found that about two ohms gets the job done with resistor power ratings suitable to the speakers power rating. Although the bigger the 2-ohms in terms of power handling is better at limiting thermal hysteresis. You can play with values. Anywhere from 1-ohm to 5-ohms.

I read a review of a high-end power amp. (the brand? I don't remember) that offers that feature as a variable adjustment.

And yes -- I just discredited everything you read on the subject. Ain't the internet fun?

_________________
Walt

Once the amplifier damping factor gets above about 10, there is no substantive gain. The DCR of the voice coil limits how much current the amplifier with very low output impedance can sink to damp cone movements. On the other hand, a servo can supply reverse voltages to the voice coil which will very much add to the damping, but that is another issue. The Crown Studio Reference amp I have for my woofers has a damping factor of about 20K or about 2 millohms output impedance which means diddly squat as far as driver damping. It is however a bullet proof amplifier that does not care what the load impedance is.

Here we go down the "Rabbit Hole." Based on publications from decades ago, and my own experience -- amplifiers with very low output-impedance (aka. so-called damping factor) actually encourage ringing (a well designed amplifier with high damping factor will not encourage ringing. A poorly designed amp with excessive negative feedback that has poor gain/bandwidth product can not only ring but can try to correct a problem after it is over. Very bad.) and are not good at arresting unsolicited voice-coil movement. Series resistance with the voice coil, especially with woofers and/or whole multi-way systems using passive crossovers provides electronic restriction. I found that about two ohms gets the job done with resistor power ratings suitable to the speakers power rating. Although the bigger the 2-ohms in terms of power handling is better at limiting thermal hysteresis. You can play with values. Anywhere from 1-ohm to 5-ohms.

I read a review of a high-end power amp. (the brand? I don't remember) that offers that feature as a variable adjustment.

And yes -- I just discredited everything you read on the subject. Ain't the internet fun?[/quote]

I have a nice stereo 845 SE amp, but the B+ is only 500V. 20 years ago, when I was younger and poorer, I bought about 2-dozen GM70’s when they were dirt cheap, with the plan of building by unicorn amp: monoblocks with a B+ of no more than 800v. I probably won’t start them until I retire in 6 years, but I’d like to design and price them out now. I regularly work with 600v DC, and know the dangers that start around 1000v. Do I need them? Hell no. But it’s the principle of not getting rid of all those GM70s I own, and creating my crowning achievement. Also, I am just obsessed with the concept. Thanks for the advice. Seems like neither SE or class A push-pull has a great sonic advantage, so that frees up my options. And I already have a good schematic for a push-pull class A 845, which looks to have a good driver.

Well, if you have so many GM70 tubes, it would be shame not to put them to good use. They are more powerful than an 845, takes less voltage to drive, you can get some good A2 performance out of them, and SE output transformers are easily available. I built a pair of GM70 SE amps, transformer coupled, that really were a lot of fun, and never seemed shy of power or LF performance (they actually got a positive mention in a show report on our DIY room at CAF several years ago). Replaced them with 211 Class A2 sets (the GM70 amps were too big for me to move around, real boat anchors, and they were disassembled for parts). I have a few GM70s waiting for an amp, and that might be back in the cards.

1000V power supply just takes care, thinking about layout, insulation, component voltage ratings, turn-on sequencing, cap bleed, etc. so that the supply can be energized and de-energized safely with no residual charge if you need to work on it. Since 1000V exceed the rating of most multimeters, good to plan for any check points be in a divider network (or use a high-voltage probe). You should go with the voltage needed to get the most out of the tube rather than hamstring it.

Driving in A2 takes more than just a transformer. You likely will want to use a direct coupled driver, a low impedance cathode/pentode/MOSFET follower, that can easily source the current the grid draws in A2. My 211 amps are only 12W operating in class A, but double that power in A2. You need to factor in the additional voltage swing from the preceding stage(s). Since a 845 takes more swing than an GM70 anyway, you might find that using that driver should put you in the ballpark. You want to pick your idle operating point to equalize the output voltage swing across the transformer at the A2 positive grid current limit without running out of swing on the other end. Play around with a loadline calculator to get ideas, I found it to be a great tool. A higher primary impedance transformer allows you to push the operating point to the right, getting more swing and power output. I think with an optimized design, you can get 50W from a single GM70.

David

FWIW, I was not referring to woofers or bass frequencies. Speakers in question are Pearl Graphene.
Lets get out of the way. We have hijacked this thread enough.

Here's an interesting project over at Audiokarma. I like this fellow, he's very sensible and always shows his work. If you scroll down a bit you can see the results--31 wpc at 5% THD. He was pleasantly surprised at the quality of the Hammond interstages. The OPTs were Magnequest FS-100s. He has another thread using an even simpler circuit with Hammond OPTs, I'll see if I can find it.

You could easily adapt this for GM70s.

https://audiokarma.org/forums/threads/t ... t-17747461