Dennis Had Inspire Amps

[Rosco65]

Originally Posted by CoGT3

What do think about adding this power circuit to the 45 SET? Seems like you would get the same benefit as with the lower powered SE pentode amps, cleaner break up when pushed? Also BearCity mentioned Dennis will be wiring his PSE up with this circuit. With the large output of the PSE relative to the 45 SET, would think it be as beneficial if your are not pushing the PSE hard? (Using 96dB speakers)?

Dennis wrote to me: "In regards to the sound, this “floating” voltage transfer system I called a infinite impedance circuit does in fact show up in signal purity at normal listen levels as well". He invited me to return my new amp to him to be upgraded to the new circuit. My experience with him is that he is not one who unnecessarily likes to spend his customer's money.

I think that we can safely assume that a 20 watt amplifier (PSE) driving a 96 dB speaker will clip even if momentarily when playing dynamically demanding material. I would be willing to wager that this clipping happens more frequently than we realize.

[CoGT3]

Sounds like:
A) The 45 SET may be taking a return trip to Cary (if the PS works on his 45 circuit)
B) The EML 5Z3 won't be needed for much longer
C) Time to find some more 5V4Gs to play with

I also agree, Dennis does not like to gobble up monies unnecessarily. Real pleasure to deal with for sure.

[Musica Amantem]

Originally Posted by CoGT3

Sounds like:
A) The 45 SET may be taking a return trip to Cary (if the PS works on his 45 circuit)
B) The EML 5Z3 won't be needed for much longer
C) Time to find some more 5V4Gs to play with

I also agree, Dennis does not like to gobble up monies unnecessarily. Real pleasure to deal with for sure.

Any impressions you may want to share on the SET 45 from Dennis? Thanks!

[Bombadil]

Originally Posted by Rosco65

I think that we can safely assume that a 20 watt amplifier (PSE) driving a 96 dB speaker will clip even if momentarily when playing dynamically demanding material. I would be willing to wager that this clipping happens more frequently than we realize.

Not unless one is playing at fairly loud levels. Such a setup should be able to play peaks of over 105 dB. I almost never play that loud. My typical average volume is in the mid to low 80s. I almost never have peaks of over 100.

[Rosco65]

Loud is relative to the listener, the room and to the music being played. If we posit a 96dB speaker fed by a 16 wpc amp, we can expect 108dB at 1 meter. Two speakers brings that to 111dB, and if our listening distance is 2 meters (fairly near field) we are back to 105dB. If we listen even further away, say 4 meters, we are now down to 99dB.

If we assume dynamic peaks of 15dB in music (not uncommon), then in our setup we a fine as long as average listening levels are less than 90dB. In your case Bombadil you're right in the zone, and would rarely be clipping. However, if we enjoy more dynamically demanding material that has 25dB peaks, we will almost certainly clip from time to time.

[Analog Addict]

Originally Posted by CoGT3

What do think about adding this power circuit to the 45 SET? Seems like you would get the same benefit as with the lower powered SE pentode amps, cleaner break up when pushed? Also BearCity mentioned Dennis will be wiring his PSE up with this circuit. With the large output of the PSE relative to the 45 SET, would think it be as beneficial if your are not pushing the PSE hard? (Using 96dB speakers)?

Well, that is a question above my tech grade, but I do know that one of the beneficiaries of the new PS is the screen grid voltage. The screen grid’s purpose is to reduce capacitance that arises between the control grid and the plate. Such parasitic capacitance can cause the tube’s circuit to become self-resonant at some frequencies, and it will reduce the tube’s achievable gain at higher frequencies. This problem is called the Miller Effect, and the screen grid helps to resolve it. A triode amplifier will typically require some type of “neutralization” circuit outside of the vacuum tube to avoid the detriments of the Miller Effect. ((G6B10) Vacuum Tubes - Ham Radio School.com)
Any variance in the voltage at the screen grid will cause fluctuations in the electron flow, introducing unwanted variance in the signal.

From Dennis' comments...."What I discovered was when an output device, in my case a vacuum tube, became non-linear and produced various spurious distortion products back fed to the power supply. In the case of a basic power supply supplying both the output devices along with the driving front gain sections I was able to detect these modulated parasitic products in the driver stage compounding the problem of output distortion of the final product.

So what he is saying (I think) is that the clipping/distortion produced by a transient dynamic peak is compounded by bleeding into the (assumed) regulated power supply feeding the driver tubes and the output tube screen grids, causing even greater distortion. Call it unintentional negative feedback of the worst type.

So, in a flash of Inspire-ation he came up with the idea of using the elements of the rectifier tube to isolate the power supply from any distortion or variance produced by the output tubes. Think of it kinda like a isolation or interstage transformer in some ways. There is no direct contact linking the two ends of the circuit. Instead, Dennis is using the vacuum of the rectifier tube to isolate the regulated power supply from the rest of the circuit, hence the "infinite impedence" part of the circuit. The novel part of it is that he is running direct current through the rectifier, and not AC as would normally be applied to a rectifier.

Now with respect to the 45 amp, remember that there's no screen grid in the 45, so you would lose the benefit of this effect. However, any amp will use driver tubes, so they would benefit from this application. My undereducated guess is that the amp would benefit, but not so much as a tetrode amp.

Originally Posted by Rosco65

Dennis wrote to me: "In regards to the sound, this “floating” voltage transfer system I called a infinite impedance circuit does in fact show up in signal purity at normal listen levels as well". He invited me to return my new amp to him to be upgraded to the new circuit. My experience with him is that he is not one who unnecessarily likes to spend his customer's money.

The reason for increased signal purity is explained above. Realize that any improvement in even microvolt fluctuations in the power supply will show up further down the signal chain as amplification is applied.

Originally Posted by Rosco65

I think that we can safely assume that a 20 watt amplifier (PSE) driving a 96 dB speaker will clip even if momentarily when playing dynamically demanding material. I would be willing to wager that this clipping happens more frequently than we realize.

One of the things we did was run the Hot Rod with the 6550s on the bench. Dennis was able to absolutely drive the $hit out of the amp, clipping like hell up to 20 WPC, and the 2nd order distortion was still the predominant wave with the clipping symmetrical on both sides of the waveform. So yes, undoubtedly people are driving his amps into clipping, but many may not even notice it if they are not listening carefully.



One thing that some of you may be interested are some of the recordings we use to test the changes and modifications. The "Honey dripping from the speakers" quote is a direct reference to one of the test albums, Honey in the Horn by Al Hirt, specifically I Can't Get Started, the first track

[ame="https://www.youtube.com/watch?v=ctNU-Wm51EA"]https://www.youtube.com/watch?v=ctNU-Wm51EA[/ame]

This album hit #3 in 1963, so there are a lot of copies of it floating around the thrift stores for a buck. I've found two or three copies, and a copy is always at the shop.



Another one we've been using a lot is Stardust by Willie Nelson, specifically Georgia on My Mind...

[ame="https://www.youtube.com/watch?v=FNlw0IDFKDA"][ame="https://www.youtube.com/watch?v=FNlw0IDFKDA"]https://www.youtube.com/watch?v=FNlw0IDFKDA[/ame][/URL]

Willie got a Grammy for this song, and it was on the charts for 10 years, also making it ubiquitous at the thrifts....



Lastly, Gold Dust Woman by Fleetwood Mac...

[ame="https://www.youtube.com/watch?v=cMp57bUzOB8"]https://www.youtube.com/watch?v=cMp57bUzOB8[/ame]

[pstrisik]

I look forward to impressions from the brave early upgraders! It sounds like an upgrade worth considering, particularly for the earlier SEPs, if I understand your description.

[Rosco65]

Originally Posted by pstrisik

I look forward to impressions from the brave early upgraders! It sounds like an upgrade worth considering, particularly for the earlier SEPs, if I understand your description.

I suppose that I could pay for the upgrade to my SEP by selling my very expensive U52 rectifier.

[Bombadil]

20 wpc for me, using a true 96dB speaker, would be way more than enough. When I use my Inspire SE EL84, which makes maybe 5 wpc, on my 88dB speakers with a powered subwoofer handling 100Hz on down, I have plenty of volume. Probably clips a bit now and then. In my bedroom system I've been running a 30 wpc amp into 85dB speakers for over 10 years now and I haven't once heard noticeable clipping.

Finding a true 96dB speaker can be a hunt. Stereophile measured a DeVore spec'ed at 96 and found it was 90.5. (or was that 90.7?). Two Zu speakers spec'ed at 97, measured 91.5 and 92.5. Someone owning a speaker which they think is a 97, but is really a 91.5, would need nearly 4X the power to reach a specific volume level.

I'd like to find a bookshelf with a nice clean sound with a real efficiency of around 92. Haven't heard one yet which I liked.

[Rosco65]

Originally Posted by Bombadil

20 wpc for me, using a true 96dB speaker, would be way more than enough. When I use my Inspire SE EL84, which makes maybe 5 wpc, on my 88dB speakers with a powered subwoofer handling 100Hz on down, I have plenty of volume. Probably clips a bit now and then. In my bedroom system I've been running a 30 wpc amp into 85dB speakers for over 10 years now and I haven't once heard noticeable clipping.

Finding a true 96dB speaker can be a hunt. Stereophile measured a DeVore spec'ed at 96 and found it was 90.5. (or was that 90.7?). Two Zu speakers spec'ed at 97, measured 91.5 and 92.5. Someone owning a speaker which they think is a 97, but is really a 91.5, would need nearly 4X the power to reach a specific volume level.

I'd like to find a bookshelf with a nice clean sound with a real efficiency of around 92. Haven't heard one yet which I liked.

Troels Gravesen has interesting opinions about high efficiency speakers and how manufacturers frequently overstate the sensitivity of their speakers.

Bookshelf speakers generally sacrifice low end to get reasonable efficiency and/or power handling. Since you seem amenable to running a subwoofer, that wouldn't appear to be an issue. I have had great success with the SUPER 3i from Omega Loudspeakers. I can attest that they are a genuine 94dB and sound wonderful and much bigger than they actually are. If you were planning on putting them on speaker stands, you might be better off with the SUPER 3XRS, which offers a bit more bass extension and does not need a stand.

A lot of people have had great success with these running them with Decware Super Zen Triode amps - 2wpc triode-connected SE EL84 amps.