Does any one know what Class A3 is ? Is it a modified "linear Class C" operation, or something crazy like that ?
I have been into electronics for decades and never heard of this Class, despite being an avid researcher on Audio design. I stumbled across it when looking at Electraprint output transformers as they make single ended Class A3 output transformers. The circuits Electraprint presents as A3, to me look like standard Class A1 or A2 designs - nothing special.
What advantages does it have over classical SE Class A designs ? How does it sound to the ear ? Does it have lower distortion, more power, better bandwidth; etc ? What drive requirements does A3 need ?
If my suspicions are right - this looks like one of those "new American inventions", where they change a part or two on on old well known design ( ie: re-invent the wheel ) and then claim they have invented something new; then suddenly rush out before anyone knows about it and get a patent put on it with the rights to sell it for a licence fee?
I say this because Electraprint won't allow you to commercially use their Class A3 tube designs without getting a licence from them. So my apologies for sounding sarcastic on these so called new-fangled inventions - as the electronics industry wreeks of such.
I also did a Google search and nothing came up that even discussed its operation.
I have been into electronics for decades and never heard of this Class, despite being an avid researcher on Audio design. I stumbled across it when looking at Electraprint output transformers as they make single ended Class A3 output transformers. The circuits Electraprint presents as A3, to me look like standard Class A1 or A2 designs - nothing special.
What advantages does it have over classical SE Class A designs ? How does it sound to the ear ? Does it have lower distortion, more power, better bandwidth; etc ? What drive requirements does A3 need ?
If my suspicions are right - this looks like one of those "new American inventions", where they change a part or two on on old well known design ( ie: re-invent the wheel ) and then claim they have invented something new; then suddenly rush out before anyone knows about it and get a patent put on it with the rights to sell it for a licence fee?
I say this because Electraprint won't allow you to commercially use their Class A3 tube designs without getting a licence from them. So my apologies for sounding sarcastic on these so called new-fangled inventions - as the electronics industry wreeks of such.
I also did a Google search and nothing came up that even discussed its operation.
Because there is no "Class A3", it is just a marketing term dreamt up by E-P to sell their hi-end transformers. To use the A3 transformers, you need Sovtek 300B tubes that can withstand the higher B+, so "Class A3" operation seems to imply the following: higher plate voltage -> higher primary impedance -> lower idle current -> lower THD, but that's just speculation on my part.😀
Last edited:
"Class A3" is simply Class A2 where the Vgk never drops below 0V, and grid current flows through 360deg of the input cycle.
"What advantages does it have over classical SE Class A designs ? How does it sound to the ear ? Does it have lower distortion, more power, better bandwidth; etc ? What drive requirements does A3 need ?"
The main advantage is that it avoids the wild swing in grid impedance from a zero grid current condition to that sudden drop when grid current is drawn. This can reduce distortion since there is always some voltage dropped across the Zo of the grid driver. This should be of no consequence with a very Lo-Z grid driver, such as a source follower.
As for drive requirements, they're the same as Class A2: a very Lo-Z output, and adequate current sourcing capability, both easily obtained from a source follower.
"What advantages does it have over classical SE Class A designs ? How does it sound to the ear ? Does it have lower distortion, more power, better bandwidth; etc ? What drive requirements does A3 need ?"
The main advantage is that it avoids the wild swing in grid impedance from a zero grid current condition to that sudden drop when grid current is drawn. This can reduce distortion since there is always some voltage dropped across the Zo of the grid driver. This should be of no consequence with a very Lo-Z grid driver, such as a source follower.
As for drive requirements, they're the same as Class A2: a very Lo-Z output, and adequate current sourcing capability, both easily obtained from a source follower.
Yeah, in Class A1 grid current never flows since Vgk never goes above 0V, and in Class A3, grid current always flows. They're kinda opposites.
Interesting, learned something new. Are there some readily available examples of Class A3 designs? Or just use an A2 drive circuit and make sure Vgk never drops below 0V.
I had never heard of Class A3 before this thread, but it seems remarkably similar to screen grid drive.
I pulled the term A3 out of a warm dark place in my very first web page in 2003. At the time I was experimenting with the 811A and other tubes that worked with positive grid voltages.
The issue here is that the grid impedance changes abruptly during the transition from negative bias to positive bias causing the driver to distort unless it can source the required grid current. At the time I had not discovered mosfets or PowerDrive.
I had discovered that several tubes like the 811A and the 6HV5 can be operated on low enough plate voltage that the grid voltage remains in the positive grid region during the entire audio cycle. This eliminates the transition from high to low grid impedance. You can in fact drive the grid of a high Mu triode from the 16 ohm tap of the OPT in a conventional tube audio amp.
I got Jack to built the OPT's in my first Tubelab TSE amp. He seemed put off by the fact that I was using mosfets to drive the grid of a 45 or a 300B. Now he builds a special transformer to drive a DHT from a chip amp, and requires a license to build it for commercial use?????
A copy of my first web page used to be on my web site, but the link is broken now. I will be working on the site in the coming weeks, I can but it back. It may still be available on the way back machine but it was 11 years ago.
The issue here is that the grid impedance changes abruptly during the transition from negative bias to positive bias causing the driver to distort unless it can source the required grid current. At the time I had not discovered mosfets or PowerDrive.
I had discovered that several tubes like the 811A and the 6HV5 can be operated on low enough plate voltage that the grid voltage remains in the positive grid region during the entire audio cycle. This eliminates the transition from high to low grid impedance. You can in fact drive the grid of a high Mu triode from the 16 ohm tap of the OPT in a conventional tube audio amp.
I got Jack to built the OPT's in my first Tubelab TSE amp. He seemed put off by the fact that I was using mosfets to drive the grid of a 45 or a 300B. Now he builds a special transformer to drive a DHT from a chip amp, and requires a license to build it for commercial use?????
A copy of my first web page used to be on my web site, but the link is broken now. I will be working on the site in the coming weeks, I can but it back. It may still be available on the way back machine but it was 11 years ago.
Member
Joined 2009
Paid Member
Thanks everyone for your informative input.
It seems like Class A3, is as one person put it, is more a marketing term.
As for it having higher plate primary Z, such would effect high frequency response with its higher interwinding capacitance and dampening factor AND by also having a low Z grid drive would open it up for potential non-linear input situations; hence higher distortion. On that basis, I can't see that it really holds any great benefits.
It possibly would suit zero bias tubes ( ie 809 triodes ) should one decide to go down that track, but for what performance advantage ?
As for Electraprint aiming to use a WE300B in Class A3 - it seems to undermine all the advantages of having a 300 B in the first place as a reference amplifier.
It seems like Class A3, is as one person put it, is more a marketing term.
As for it having higher plate primary Z, such would effect high frequency response with its higher interwinding capacitance and dampening factor AND by also having a low Z grid drive would open it up for potential non-linear input situations; hence higher distortion. On that basis, I can't see that it really holds any great benefits.
It possibly would suit zero bias tubes ( ie 809 triodes ) should one decide to go down that track, but for what performance advantage ?
As for Electraprint aiming to use a WE300B in Class A3 - it seems to undermine all the advantages of having a 300 B in the first place as a reference amplifier.
If the "A3" condition only occurs at Vgk=0V where Rin stays high, it is only meaningful at the full power output using sinusoidal signal, is it really applicable to musical signal?
Anyone looked at the A3 schematics on the Electraprint site? I don't see anything special there. http://www.electra-print.com/docs/a3circuits.pdf
What has the waveform got to do with it? Provided Vgk>0V throughout the cycle it should not matter.jazbo8 said:
How do you ensure that Vgk>0V throughout the whole cycle with the way the tube is biased in the E-P example?
Which E-P example? There are four examples in the link shown in post 12, some of which can't be A3. You can't do A3 with cathode resistor bias!!! That makes me suspicious that none of them may be A3.
If A3 means grid positive throughout the cycle, then provided you achieve that then the signal does not matter. How you achieve that is a separate issue.
If A3 means grid positive throughout the cycle, then provided you achieve that then the signal does not matter. How you achieve that is a separate issue.
All of them... I understand what you are saying, as they say the devil is in the details, A3 depends on how to go about achieving it, anyway, I think we have spend enough time on this gimmick...
I don't understand the EP example. A 300B with 500 volts on it's plate will need a whole bunch of NEGATIVE bias to reach his specified plate current. This looks like an A1 design, maybe A2 if driven hard enough, but A2 doesn't work with a coupling cap on the grid. It can be made to work with a GOOD driver transformer and no cap. Jack does make good transformers.
I don't know too many 300B's that will live at 500 volts either. Sovteks and Shuguangs will runaway at anything over 400 volts. I have no experience with high dollar tubes, but I wouldn't feed an expensive tube 500 volts unless the manufacturer specifies it for such use.
operating a tube in continuous grid current does itself not cause a high plate resistance. The type of tubes that work in this region (continuous grid current) tend to have high plate resistance because of their design. They are typically high Mu triodes.
The transition from no grid current (Vg less than 0) to grid current (Vg positive) will cause a typical cap coupled driver to distort as the grid draws current regardless of the reason the grid was driven positive. It usually causes the bias to shift as the coupling cap discharged into the grid creating a recovery time after an overload. This is a common problem in guitar amps.
A3 was my first attempt to eliminate this "blocking distortion" or "farting out" as the guitar guys call it. That was 2003. Today we have mosfet drivers and augmented cathode followers that eliminate the problem by eliminating the coupling cap.
I don't know too many 300B's that will live at 500 volts either. Sovteks and Shuguangs will runaway at anything over 400 volts. I have no experience with high dollar tubes, but I wouldn't feed an expensive tube 500 volts unless the manufacturer specifies it for such use.
operating a tube in continuous grid current does itself not cause a high plate resistance. The type of tubes that work in this region (continuous grid current) tend to have high plate resistance because of their design. They are typically high Mu triodes.
The transition from no grid current (Vg less than 0) to grid current (Vg positive) will cause a typical cap coupled driver to distort as the grid draws current regardless of the reason the grid was driven positive. It usually causes the bias to shift as the coupling cap discharged into the grid creating a recovery time after an overload. This is a common problem in guitar amps.
A3 was my first attempt to eliminate this "blocking distortion" or "farting out" as the guitar guys call it. That was 2003. Today we have mosfet drivers and augmented cathode followers that eliminate the problem by eliminating the coupling cap.
This is what I noted, when I first saw it.
And without changing the subject I might also mention >
There was an good short article on page 46 of the Oct 1960 - Audio - Magazine on patents and inventions.
The gist of the article was stating that "many inventions don't count as true inventions but are actually no different to already existing designs as they don't actually do anything different to pre-existing circuits" ( the article words it better than I have expressed ). As a result, such Inventors trying to patent these regurgitated ideas are effectively "damaging the community" by destroying the progress of society. The statment was implying the idea was driven by "greed and corruption and not innovation".
Maybe Hafler and Keroes fall into this category, as evidence of Ultra-liner was found as far back as 1937 but under other names; so it was not a really a new idea when Hafler & Keroes were working with it in 1950 - they just repackaged it and submitted it as a "new idea" - probably because no one had placed a patent on it beforehand - as it was not considered a NEW Invention, even in 1937. In fact one website says "Ultra-linear" was used in 1935 or 36 by an Australian School Teacher for school training of his students; but due to a school fire all records of it's existence were lost.
The US Patents Office has lot to answer for, by issuing patents for these types of ideas.
I dont l know if Electra-print actually have a patent on Class A3 - but they seem to have licence rights. Licence rights for something that to me is not new at all.
Hi guys,
The Electra Print transformer set is still available at Ebay. I am very curious to find out how the EL84 single-ended power amplifier could have 12 watt per channel output. Wondering if any can share the schematic?? I understand that the circuit use extreme high voltage which exceeds the max plate volt as per the EL84 datasheet.
Johnny
The Electra Print transformer set is still available at Ebay. I am very curious to find out how the EL84 single-ended power amplifier could have 12 watt per channel output. Wondering if any can share the schematic?? I understand that the circuit use extreme high voltage which exceeds the max plate volt as per the EL84 datasheet.
Johnny