Examples of Class A push/pull

[Retrovert]

increased performance, as in more power, less distortion. I'm a car guy, "performance" is understood in the context of "more", which only occasionally crosses paths with the question of how easy it is to live with. Cathode bias has its good points, but its got problems too. So does fixed bias for that matter. You don't buy a Ferrari to tow a boat, and you don't enter a Mack truck in a high speed road coarse.

Ahh, but that is a jargon definition which is not applicable across all domains. From the OED:

https://www.lexico.com/en/definition/performance
2 mass noun The action or process of performing a task or function.
2.1 A task or operation seen in terms of how successfully it is performed. ‘pay increases are now being linked more closely to performance’
count noun ‘it was a tremendous all-round performance by Wigan’
2.2 The capabilities of a machine, product, or vehicle. ‘the hardware is put through tests which assess the performance of the processor’
as modifier ‘a performance car’​

The key is the process by which the task or function is accomplished, success in terms of meeting goals, or capabilities, not merely a single facet, i.e. power.

I am reminded of a story I heard about a hedge-fund manager with far too much money who purchased a racing rear-end ratio for his sports car and then unintentionally spun the wheels at every light. The performance of that car in the city environment was obviously quite poor, but on a race track would be quite good. The concept of "performance" generally embodies suitability for a particular task and set of goals. He also had a giant belly which did not easily fit behind the wheel.

In the 1990s I saw Malcolm Forbes and his boyfriend cram themselves into Forbes' Lamborghini Diablo, mere inches from the ground. It was like watching a plumber with severe middle-aged spread, hip arthritis, a bad back, and casts on all limbs forcibly pack himself underneath a kitchen sink whilst attempting to replace a water feed to the faucet. I watched with the sort of amazement as one watching a python swallow a goat. This was not a cat settling into a cardboard box three sizes to small. The boytoy, being young, cut, and flexible had no such trouble.

Race cars, to use your analogy, offer very poor performance when it comes to comfort, such as cabin noise, handling, braking, the radio, or the windows and doors. Plus the trunk is really small.

Tying it back to the amplifier, the goals must be clearly set forth. Is it a single factor, such as output power, THD, IMD, transient response, tube longevity, lack of adjustment, stability, etc., or is it a combination of factors? Cathode bias offers better or worse performance depending upon the design goals.

 

[Pio1980]

Ahh, but that is a jargon definition which is not applicable across all domains. From the OED:

https://www.lexico.com/en/definition/performance
2 mass noun The action or process of performing a task or function.
2.1 A task or operation seen in terms of how successfully it is performed. ‘pay increases are now being linked more closely to performance’
count noun ‘it was a tremendous all-round performance by Wigan’
2.2 The capabilities of a machine, product, or vehicle. ‘the hardware is put through tests which assess the performance of the processor’
as modifier ‘a performance car’​

The key is the process by which the task or function is accomplished, success in terms of meeting goals, or capabilities, not merely a single facet, i.e. power.

I am reminded of a story I heard about a hedge-fund manager with far too much money who purchased a racing rear-end ratio for his sports car and then unintentionally spun the wheels at every light. The performance of that car in the city environment was obviously quite poor, but on a race track would be quite good. The concept of "performance" generally embodies suitability for a particular task and set of goals. He also had a giant belly which did not easily fit behind the wheel.

In the 1990s I saw Malcolm Forbes and his boyfriend cram themselves into Forbes' Lamborghini Diablo, mere inches from the ground. It was like watching a plumber with severe middle-aged spread, hip arthritis, a bad back, and casts on all limbs forcibly pack himself underneath a kitchen sink whilst attempting to replace a water feed to the faucet. I watched with the sort of amazement as one watching a python swallow a goat. This was not a cat settling into a cardboard box three sizes to small. The boytoy, being young, cut, and flexible had no such trouble.

Race cars, to use your analogy, offer very poor performance when it comes to comfort, such as cabin noise, handling, braking, the radio, or the windows and doors. Plus the trunk is really small.

Tying it back to the amplifier, the goals must be clearly set forth. Is it a single factor, such as output power, THD, IMD, transient response, tube longevity, lack of adjustment, stability, etc., or is it a combination of factors? Cathode bias offers better or worse performance depending upon the design goals.

Horses for courses. Some of the tube console amps are quite elegant for a sinple topology and service free use, limited only by the service life of the tubes. Ten WPC used to be the low-midpower standard and perfectly adequate when matched to the speakers and reasonable expectation.

 

[arts]

In practice, a power supply should not fail and changes in tubes with age are minor. But power supplies built using electrolytics, inherently unreliable, do fail and the the smallest unit of time perceivable by a human is the amount of time between "WTF, the sound just stopped" and "WTF, my output pair is magma, the output transformer is in flames, the rise in voltage chernobyled the other channel and all the tubes, my rectifier given to me by Schade himself is slag, my power transformer winding just went to Valhalla, and I'm only halfway through Inna-Gadda-Da-Vida!".[1]

I'll second that,but in my case it was In the Court of the Crimson King.

 

[crooner]

How about the Marantz 9 monoblock running in triode mode? I would venture this is Class A or very close to it. Certainly sounds like it. 40 watts using 4 EL34's in Push Pull - Parallel and oversized output transformer. This was designed by Sid Smith who also engineered the Radio Craftsmen 500. Perhaps Dave G would want to elaborate further? Thanks!

 

[dcgillespie]

Crooner -- From the schematic of the 9, the quiescent operating current of each output tube is about 55 mA, with a plate voltage of about 415 vdc. The missing piece of information to make an absolute determination as to triode mode class of operation is the reflected load of the OPT. However, with a power output, B+, and quiescent current level all remarkably similar to those of the Dynaco ST-70 (if the channels were paralleled), it wouldn't be unrealistic to use 1/2 of the Dynaco A-470 transformer's reflected impedance for the purposes of estimating the primary impedance of the 9's output transformer, or 2150Ω.

From the EL34 triode mode family of plate curves, using a full power plate voltage of 400 vdc, Class A operation of a single tube is achieved with a quiescent current of (about) 62 mA, and requires a plate load impedance of 4230Ω. After output transformer losses, this results in about 7.25 watts of real world RMS power output. Converting these operating parameters into a two tube Class A push-pull configuration, a reflected plate-to-plate load impedance of about 8460Ω would be required, producing about 14.5 watts RMS power output.

In the push-pull-parallel configuration of the Marantz 9 then, the first step to achieving true Class A output stage operation in triode mode would be to bump the output tube quiescent current up slightly from 55 mA to 62 mA. But with four tubes, now the required reflected plate-to-plate load is 4230Ω, against an estimated 2150Ω offered by the 9's output transformer. Therefore, the second step would be to move the speaker connection down to the next lower impedance tap (operating 8Ω speakers on the 4Ω tap, 16Ω speakers on the 8Ω tap, etc) relative to the nominal impedance of the speaker used, which will then double the estimated reflected load to the tubes to 4300Ω -- or very close to the required 4230Ω calculated previously. With these two moves, the Marantz 9, operating in triode mode, will produce a maximum power output of about 29-30 watts RMS as a true Class A amplifier, with each tube dissipating about 26 watts under quiescent conditions, which is an entirely safe operating condition for EL34 tubes when configured for triode operation as discussed.

If the 9 is simply switched to triode mode operation without making these adjustments, then it must still be considered to be operating as a Class AB1 amplifier.

Dave

 

[crooner]

Thank you Dave for the excellent technical explanation.
I had long suspected the Marantz 9 design had the potential for achieving true Class A operation.

I am currently using the 16 ohm tap to utilize the entirety of the output transformer secondary winding. But I may switch to the 8 ohm tap and increase the quiescent current to 62 mA as you have suggested. There may be a slight power loss, but with the benefit of true class A operation.

Now, the Marantz 9's meter is marked for 55 mA. I may have to measure it with a external meter so that I don't exceed 62 mA. I hope the Mullard XF2's I'm using don't develop hot spots on the plate at that dissipation level. Although they are very tough tubes, some do have the tendency to develop hot spots, in my experience.

Crooner -- From the schematic of the 9, the quiescent operating current of each output tube is about 55 mA, with a plate voltage of about 415 vdc. The missing piece of information to make an absolute determination as to triode mode class of operation is the reflected load of the OPT. However, with a power output, B+, and quiescent current level all remarkably similar to those of the Dynaco ST-70 (if the channels were paralleled), it wouldn't be unrealistic to use 1/2 of the Dynaco A-470 transformer's reflected impedance for the purposes of estimating the primary impedance of the 9's output transformer, or 2150Ω.

From the EL34 triode mode family of plate curves, using a full power plate voltage of 400 vdc, Class A operation of a single tube is achieved with a quiescent current of (about) 62 mA, and requires a plate load impedance of 4230Ω. After output transformer losses, this results in about 7.25 watts of real world RMS power output. Converting these operating parameters into a two tube Class A push-pull configuration, a reflected plate-to-plate load impedance of about 8460Ω would be required, producing about 14.5 watts RMS power output.

In the push-pull-parallel configuration of the Marantz 9 then, the first step to achieving true Class A output stage operation in triode mode would be to bump the output tube quiescent current up slightly from 55 mA to 62 mA. But with four tubes, now the required reflected plate-to-plate load is 4230Ω, against an estimated 2150Ω offered by the 9's output transformer. Therefore, the second step would be to move the speaker connection down to the next lower impedance tap (operating 8Ω speakers on the 4Ω tap, 16Ω speakers on the 8Ω tap, etc) relative to the nominal impedance of the speaker used, which will then double the estimated reflected load to the tubes to 4300Ω -- or very close to the required 4230Ω calculated previously. With these two moves, the Marantz 9, operating in triode mode, will produce a maximum power output of about 29-30 watts RMS as a true Class A amplifier, with each tube dissipating about 26 watts under quiescent conditions, which is an entirely safe operating condition for EL34 tubes when configured for triode operation as discussed.

If the 9 is simply switched to triode mode operation without making these adjustments, then it must still be considered to be operating as a Class AB1 amplifier.

Dave

 

[dcgillespie]

The tubes should be fine at 62 mA. The plate of an EL34 is rated for 25 watts of dissipation (using a rating system similar to the old conservative Design Center rating system used in America), and because the tube is a true pentode with non-aligned grids, the screen is rated for 6 watts (12 watts on dynamic peaks) so as to be able to handle the significant amount of current flow it can be exposed to. In triode mode, these ratings can be summed for a conservative 31 watt rating, since the plate and screen grid elements are tied together. If the operating B+ in your amplifiers is reasonably close to that as shown on the schematic, then operation at 62 mA has the tubes operating at 84% of the conservative rating for the tubes as determined above, which is very conservative operation in deed for the power tubes in any true Class A amplifier.

Dave

 

[Pio1980]

Some have advocated operating the 8/8B in higher current bias, probably in triode for best sound.
I personally stay on the factory recommended mark in tapped screen mode for my own use, so cannot comment.

 

[crooner]

How about the 8B? is it possible to run it in Class A triode as well? The power supply is not as robust as the Model 9, though. The nominal quiescent current is 50 mA per EL34, a tad lower than the 9.

By the way, I checked the Model 9 service manual and the B+ for the output stage is listed at 427V. This is for the design center voltage of 117V AC.

 

[crooner]

For tonight's summer's solstice listening session, I switched to the 8 ohm tap and ran the Marantz 9's in triode mode. I didn't measure the quiescent current on each tube to be exactly 62 mA, but I tried to approach it by observing the total current draw on each monoblock. When running at the bias mark, they draw 1.3 A each from its bucking transformer between 110 to 112V AC. By a simple rule of three, I determined 62 mA on each EL34 would increase the total current draw of each amplifier to around 1.465 A

I therefore adjusted the bias on the pots until the total current draw was 1.4 A on the external meters.

The first thing that struck me was the naturalness of the human voice. However, other things also became apparent. Switching to the 8 ohm tap decreased loudness significantly. Perhaps as much as 3 dB, or a halving of actual power delivery. So the amplifiers must have been delivering 20 watts each instead of close to 40 watts as rated for triode mode.

The sound also became softer, too smooth actually. With some high frequency attenuation.

I ended up switching back to my preferred 16 ohm tap. I think the magic resides here. No doubt about it. Specially with my LS3/5A's. The slam and dynamics are just unbelievable along with that glorious triode midrange and luscious extra detail.

I do believe these classic amplifiers were optimized for 16 ohms, as most high quality speakers of the time were so rated, including the KLH Nine electrostatics, often used with Marantz 9's.

I also adjusted the current back to the rated 55 mA per tube. The amplifiers sound amazing this way in triode. But it was a fun experiment for sure!

My quad of EL34's show some hot spots on the plate, and at 62 mA they seem to be more visible, no doubt due to the increased dissipation. I think it's normal for Mullard double halo XF2's to show these from time to time. Not harmful to them at all, in my experience.

 

[dcgillespie]

So may I assume that your speakers are nominally rated at 16Ω?

Because of the characteristics of vacuum tubes in general (but particularly with triode configured tubes), they do not follow the hard and fast rule of SS equipment, where raising the load impedance by a factor of two results in a 50% loss of power. With such a scenario using vacuum tubes, less peak current drawn by the load means that the plate voltage can pull lower than it does with normal loading, which results in an element of compensation. The opposite is also true, where if the load on the amplifiier is increased (lower numerical impedance), then the plate voltage is not allowed to drop as low as it normally would, which allows the tube to pull more current, again providing an element of compensation to deal with the increased load demand. Actually, this element of compensation is really the famous low output impedance that triodes display, which causes them to be less sensitive to the ever changing load that a loudspeaker presents to an amplifier. For PPP EL34s operating with the 9's B+ levels using the quiescent current and loading conditions suggested, Class A operation will be achieved with the load line indicating that a maximum power output of 29-30 watts RMS will be produced, which is supported in practice as well.

All of this of course is on the technical side of the equation. If you prefer the 9's original operating conditions, then there is certainly no harm no foul in doing that, either! Thanks for sharing your experience with the alternate mode of operation!

Dave

 

[crooner]

Hi Dave,

Yes, the speakers are Chartwell LS3/5A BBC monitors rated at 15 ohms nominal impedance.

They are known to work best on the 16 ohm tap when tube amplification is used.

In practice, I have found this to be true and yesterday's experiment confirms it.

I think the loss in power I experienced was more of a load mismatch than anything. I noticed I had to advance the volume on the preamp to a higher setting to achieve the same loudness as before.

It would be interesting to try Class A again with a different set of speakers. Now, that I think about it, perhaps the reason double pairs of KLH Nine's were often used with the Model 9's. The nominal impedance when connected in parallel halves to 8 ohms. I'd love to hear the amps running in triode Class A with them.

So may I assume that your speakers are nominally rated at 16Ω?

Because of the characteristics of vacuum tubes in general (but particularly with triode configured tubes), they do not follow the hard and fast rule of SS equipment, where raising the load impedance by a factor of two results in a 50% loss of power. With such a scenario using vacuum tubes, less peak current drawn by the load means that the plate voltage can pull lower than it does with normal loading, which results in an element of compensation. The opposite is also true, where if the load on the amplifiier is increased (lower numerical impedance), then the plate voltage is not allowed to drop as low as it normally would, which allows the tube to pull more current, again providing an element of compensation to deal with the increased load demand. Actually, this element of compensation is really the famous low output impedance that triodes display, which causes them to be less sensitive to the ever changing load that a loudspeaker presents to an amplifier. For PPP EL34s operating with the 9's B+ levels using the quiescent current and loading conditions suggested, Class A operation will be achieved with the load line indicating that a maximum power output of 29-30 watts RMS will be produced, which is supported in practice as well.

All of this of course is on the technical side of the equation. If you prefer the 9's original operating conditions, then there is certainly no harm no foul in doing that, either! Thanks for sharing your experience with the alternate mode of operation!

Dave

 

[Retrovert]

Now, that I think about it, perhaps the reason double pairs of KLH Nine's were often used with the Model 9's. The nominal impedance when connected in parallel halves to 8 ohms. I'd love to hear the amps running in triode Class A with them.

In my experience the double Model Nine configuration is purely about improved dispersion, as the Model Nine tends to beam the sound. Great if one is sitting right in the sweet spot, otherwise, well, less great.

I have a double pair of Model Nines in storage, temporarily unused because (a) (i) the power supplies are failing and need a total teardown and rebuild with modern diodes and new capacitors, and (ii) I need to find some high-voltage friends to work on these with me, and borrow a portable defibrillator (7,0000 VDC will readily kill one very dead) (b) the second set came to me in rough condition and need work, including new stands, and (c) my living room is unfortunately too small to properly set up one pair, let alone the double configuration. The sound is, of course, truly amazing, and the angled dispersion is a vastly superior sound, completely filling the room. I always say that I don't know what music will be played at the end of the Universe, but that I know it will be played on a KLH Model Nine.

The difficulty in driving the Model Nine is that it presents a highly capacitive load to the amplifier, which can easily cause a marginally stable amplifier to descend into oscillation, and move an otherwise stable and not troublesome amplifier into oscillation. Doubling up on that capacitance — the parallel capacitance increases even though the impedance decreases — only makes the problem worse. One of the tests of amplifier stability is placing a capacitor across the speaker terminals. The KLH Model Nine is that very test. So it's not an impedance issue

I would not try to drive the Model Nine from a low-power Class A tube amplifier because it would likely eat the amplifier and the speakers. Not enough power output. The KLH Model Nine specification stated that each speaker requires about 35 Watts to properly drive. That's per channel. Not very efficient. Only a solid-state amplifier able to drive a capacitive load at reasonable power can properly drive them. Yes, sigh, high-power tube amplifiers did exist. But these were uncommon and generally not Class A. Solid-state amplifiers generally do better driving a higher-impedance loads, so that's generally not an issue. Also, too much voltage will readily fry the panels. That's always a warning to be seriously taken. Given that they ain't makin' 'em no more, best to be careful.

 

[crooner]

[QUOTE="Retrovert, post: 12856402, member: 129008]
I would not try to drive the Model Nine from a low-power Class A tube amplifier because it would likely eat the amplifier and the speakers. Not enough power output. The KLH Model Nine specification stated that each speaker requires about 35 Watts to properly drive. That's per channel. Not very efficient. Only a solid-state amplifier able to drive a capacitive load at reasonable power can properly drive them. Yes, sigh, high-power tube amplifiers did exist. But these were uncommon and generally not Class A. Solid-state amplifiers generally do better driving a higher-impedance loads, so that's generally not an issue. Also, too much voltage will readily fry the panels. That's always a warning to be seriously taken. Given that they ain't makin' 'em no more, best to be careful.[/QUOTE]

The Marantz Model 9 running in triode is capable of 40 watts per monoblock. And it can be made to operate in Class A by increasing the quiescent current and using the 8 ohm tap as explained by Dave in an earlier post. This means double KLH Nines would be ideal for it and the speakers were actually demonstrated this way back in the day...

 

[Retrovert]

The Marantz Model 9 running in triode is capable of 40 watts per monoblock. And it can be made to operate in Class A by increasing the quiescent current and using the 8 ohm tap as explained by Dave in an earlier post. This means double KLH Nines would be ideal for it and the speakers were actually demonstrated this way back in the day...

You are quibbling and putting words in my mouth. Don't do that.

I clearly stated, "I would not try to drive the Model Nine from a low-power Class A tube amplifier", which is accurate. Again, "LOW POWER". The specifications from KLH call for 40 to 75 WPC. (I'd have to OCR the text to drop in the full text, and I doubt anyone cares that much.) The Model Nine is about as efficient as an AR-3. So that's not usually appropriate for the typical Class A tube amplifier of the day, generally under 10 to 15 WPC, using apples-to-apples watts before the FCC change in terminology. (My dad drove his AR-3's using a Dynaco ST-70, which was among the best setups in 1963. That was one of the few amplifiers able to properly drive those speakers.)

Talk to anyone who has a pair of Model Nines. (Again, I have TWO pairs.) Driving from a low-power amplifier isn't optimal, and Class A tends to be lower power. Very few Class A amplifiers exist in the necessary wattage. Many owners uses solid-state; the previous (original) owner of my first pair did exactly that and for good reason.

The Marantz Model 9 delivers 40 WPC in Class A? Wow. I thought it was Class AB push-pull, either ultralinear or triode. Oh, riiiiiiight. You're talking about a MODIFIED amplifier. Sure, when MODIFIED it can run Class A. Well, that's nice, but that MODIFIED amplifier was not the factory-shipped configuration so arguing that some super-fancy hot-rodded MODIFIED Class A amplifier could drive the Model Nine does not mean that it would be generally true to say that any amplifier could drive it, particularly when most were CLASS AB, not CLASS A and the Marantz was not Class A FROM THE FACTORY.

And my point was about LOW-POWER CLASS A. Not that no tube amplifier could drive it. My ST-70 could drive it as well. BUT THE ST-70 IS NOT A LOW-POWER CLASS A AMPLIFIER! IT IS A HIGH-POWER CLASS AB! (NB: high-power for its day.)

The stability of the Marantz Model 9 was legendary so I do not doubt it could drive a capacitive load. Many amplifiers cannot. I suggest reading about what the Quad ESL-57 or 63 does to marginally stable amplifiers or ones that can't drive low-impedance loads; it isn't pretty. The impedance of an electrostatic generally halves above a few kHz, so the parallel load of 8 ohms is going to drop to about 4. Some amplifiers don't care, others do.

My statement that the KLH Model Nine can cause instability in many amplifiers remains true, as does my statement about why a low-power Class A amplifier isn't suitable:

I would not try to drive the Model Nine from a low-power Class A tube amplifier because it would likely eat the amplifier and the speakers. Not enough power output. The KLH Model Nine specification stated that each speaker requires about 35 Watts to properly drive. That's per channel. Not very efficient. Only a solid-state amplifier able to drive a capacitive load at reasonable power can properly drive them. Yes, sigh, high-power tube amplifiers did exist. But these were uncommon and generally not Class A.​

Again, my statement was about low-power Class A tube amplifiers.