ECC83 Power Amp

[Steve Conner]

Teemu: My investigation of OPT saturation in PP amps never took into account DC imbalance caused by asymmetrical clipping. I've seen scope shots posted by Loudthud of the output of a cranked Plexi, that suggest there could be a fair bit of DC. This would then make the tone more middle-y like my cheap SE OPT does when biased heavily.

Tilman: I've built it! It looks like a cheesy 3 watt practice amp, but will blow your head off on the high power setting. But I doubt it would catch on for the reasons explained by MWJB etc. That is to say, it would be built in small quantities, so the actual electronic guts would only account for a fraction of the price, so you might as well make it all tube: more people would buy it.

I fully agree with his other comments.

Next I intend to build one with a switchmode power supply, that will piss off the cork sniffing brigade even more.

[R.G.]

The tricky thing is IMHO to capture the damping factor of the tube amp and amplify it so, that the ear thinks the SS amp has a non HiFi damping factor too.

Not tricky at all. SS amps can more or less easily be tricked into having low damping factor. Here's one of several ways:
Variable Amplifier Impedance
Soft clipping helps as well:
http://web.archive.org/web/200906061.../soft-clip.htm

But as noted, there's been a lot of tries on this, in all kinds of ways. What happens is that the genius/inventor gets it to fit what they think they hear and then assume everyone else hears the same approximation.

[diagrammatiks]

Diagrammatiks wrote: "You could reamp the smaller power amp into a bigger amp. However, while the transformer doesn't saturate...the frequency dependent impedance of the speaker and it's interaction with the transformer will be lost."

I disagree, strongly. If the first amp is just a tone generator, with the signal tapped from the OT secondary, feeding into a slave amp, you can get the larger part of the first amp's tone, I have lined out amps where it has been impossible to determine the point at which the slave amp takes over...the tone remains the same, just gets louder, however the dynamic characteristics of the final stage, especially speaker array, do influence the overall sound.

I also have a 12A#7 powered amp...my findings are more in line with JMaf's figures, I can get useable cleanish tone from <0.2W by the time you're much past 0.5W it's driving hard. 2W RMS output? A tall order from a 12AT7 in PP...I settled on 12AT7 for apartment use due to it being by far the quietest, but with still with good tone compared to ECC99, 12BH7, 12AU7, 12AY7, 6SN7. 12AT7 also offers plenty of options regarding transformer choices, working fine from 17K to 34K primary.

6SL7 make good, low output power tubes too but more restrictive regarding primary Z, if you use 2 of them with triodes in parallel, one each end of the OT primary, this can open up OT options as 8K to 8.5K primaries are very common and a 8ohm or 16ohm speaker on a 4ohm tap can give a useable impedance match.

On a recording, or carefully lined out I defy ANYONE to be able to identify the difference between a small triode/twin triode used as a power tube, compared to a pair of high power tetrode/pentode.

Jon Wilder wrote: "Of course...lots have tried to get "that sound" at lower volumes...and all have failed. Why is this?" No they haven't. The reason it's not a popular course of action is purely expense & practicality. You buy a 0.5W to say 6-7W amp, but it's not loud enough to gig, so you have to mic up (& buy your own stage monitor) or slave out to a big amp so you can hear yourself on stage...So now you have the bigger system doing the work FOH and a very expensive preamp and all the associated interconnect leads/mics etc. I loved the tone of my old EL84 amp lined out through my Twin, I thought, "this is IT!"...for about 2 gigs, then knocked it on the head because of setting up time, stage space, amount of things to carry and the fact that many musicians have a "ball-park" sound in their heads that they play off of & you can usualy get close enough with a stage sized amp & only the most dedicated would carry all the extra clutter. Plus why have a larger amp/FOH system that you don't like the sound of in the first place...it makes no sense.

You're greatly oversimplifying the interaction between the variable load of the speaker across the output transformer.

If you're concerned about identifying the tone from a recording of course it will be difficult. However, a guitar amp isn't a playback difference.

I think a majority of players will be able to differentiate between an ultra-linear, true triode, triode strapped pentode, and a true pentode power section.

How many of the classic rock tones were made on true triode amps?

Of course if you think that a pp small signal triode sounds good in the power amp section that's fine. I was merely stating that I thought that it sounded like poop. On the other hand I'm not a big fan of most types of power amp distortion in the first place.

However, that doesn't change the fact that certain tones are associated and dependent on the overdrive characteristics of certain families of output tubes.

How many quad pushpull 300b guitar amplifiers can you name?

You can also get less then 2 watts with a pair of ef86's in the power stage.

[Steve Conner]

Sure, he is greatly simplifying it. The question is, can you get away with that? I think that should be an open question.

My own circuit doesn't actually interact as well as the Valve Reactor. I think the reflected impedance gets inverted, so the higher the speaker impedance, the lower the impedance the tube sees. But it seemed to work just fine: the transistors have curves similar to pentodes, so it drives the speaker with a high impedance at the end of the day. That was the closest I could get without making a custom interstage transformer, complicating the circuit a lot, or infringing the Korg patent.

I think there are no 300B guitar amps mainly because of the cost of 300Bs, resulting in a pathetic watts per dollar ratio.

[MWJB]

Diagrammatiks wrote: "However, that doesn't change the fact that certain tones are associated and dependent on the overdrive characteristics of certain families of output tubes."

I think it would be more accurate to say that certain tones are associated with the overdrive characteristics of certain tubes & families of tubes in certain circuits and at certain voltages.

"If you're concerned about identifying the tone from a recording of course it will be difficult. However, a guitar amp isn't a playback difference." By the same token if the amp is just a tone generator, relying on a slave to punch up the dB for FOH, so that the volume of the tone generator is moot then it will still be nigh on impossible to identify power tube type.

"Of course if you think that a pp small signal triode sounds good in the power amp section that's fine. I was merely stating that I thought that it sounded like poop. On the other hand I'm not a big fan of most types of power amp distortion in the first place." If the PP triode amp is just a tone generator and the dB at it's speaker largely irrelevant, why is there the need to push it into any kind of unpleasant distortion in the first place?

[diagrammatiks]

Hmm..I thought that was the point of the low watt amp in the first place? To get output tube distortion at low volume?
Err..Did I show up for the wrong party.

[MWJB]

No...not output tube distortion at low volume, but good, useable, pleasing tone at low volume. The amp should have some clean tone...after that it's up to how you set the controls that dictates how much distortion and from where it is prevalent. Hook up the low powered amp to a slave and the volume (dB) that the low powered amp makes at the speaker becomes pretty well moot...unless it is just about loud enough for on-stage monitoring (put a 0.5W amp on a stand at head height, by your ear & line it out to PA, you'd be surprised as to how well that works for low/moderate volume/coffee bar type gigs!).

I'm not sure focussing on preamp vs output distortion is very helpful...unless you are building some hi gain beast, they should compliment each other. Too much output tube distortion is still, well..."too much".

[Wilder Amplification]

I hear this argument a lot here. But I have an observation... How come it's possible to hear "that" tone when listening to a recording at much lower volume levels??? Kind of punches a hole in this avenue of thinking.

No it doesn't. The amp was cranked when it was recorded. Therefore, the recording mic captured all of those nuances that were created by a cranked amp and sheer volume, and those nuances all got recorded. The amp being cranked and the sheer volume created the recorded sound. You're just "reproducing" what was already created by a cranked amp. You're not doing that with a guitar amplifier.

Furthermore, most all recorded material is compressed...AFTER the microphone already picked up all the sonic nuances created by a cranked amp...so you can hear everything. Not so on a guitar amp so these two things alone trump your attempt at "punching holes in the above explanations".

The Fletcher munson curve isn't so dramatic that it can't be compensated for via post amp EQ. Feel is a different issue. Amps and guitars don't interact the same at low wattage as they do at higher wattages. But the actual waveform and compensation for the failings of the human ear certainly can be recreated at lower volumes. It's just so much trouble and expense that no one has done it.

But you have to get it up to sheer volume in order to CREATE the tone in the first place, then reproduce that created tone. Again, you're not "reproducing" sound with an amp, you're CREATING sound with it, and certain criteria must be met in order for the amp to do it. Sheer volume is one of those criteria.

[Chuck H]

No it doesn't. The amp was cranked when it was recorded. Therefore, the recording mic captured all of those nuances that were created by a cranked amp and sheer volume, and those nuances all got recorded. The amp being cranked and the sheer volume created the recorded sound. You're just "reproducing" what was already created by a cranked amp. You're not doing that with a guitar amplifier.

But if the original percption of the tone is partly due to the effect of sheer volume on the human ear, that gets lost. That was my point.

Furthermore, most all recorded material is compressed...AFTER the microphone already picked up all the sonic nuances created by a cranked amp...so you can hear everything.

Everything but the effect of volume on the human ear... Since the volume is removed from the formula. That was my point.

But you have to get it up to sheer volume in order to CREATE the tone in the first place, then reproduce that created tone. Again, you're not "reproducing" sound with an amp, you're CREATING sound with it, and certain criteria must be met in order for the amp to do it. Sheer volume is one of those criteria.

When did I say anything about reproducing cranked amp tone with an amp??? I'm talking about recreating the perception to the human ear via EQ and possibly some compression. Or rather, whether that's even needed.

How is it that part of the sound of a cranked amp is due the sheer volumes effect on the human ear, and yet you can record that same amp, replay the recording at a lower volume and the sound is correct. I think it's clear that in this example the effect of volume on the human ear has been removed from the equation.

The theory I was "punching holes" in was that volume is a needed part of the formula.

[Amp Kat]

There is also alot to say about why distortion sounds good and what variations of distortion and why. One of the reasons has to do with thd as in even order and odd order from the fundamental frequency. There are also lots of techniques like rounding the peaks to take the edge of of the square wave to get a more even order distortion. Obviously there is odd order too which transistors produce naturally and would think they are very close to getting them to produce even order but even SS pres can deliver and have been delivering very good distortion when running them through a well balanced clean tube output stage. Of course running an all tube distorted and compressed sustain tone and then force the tube output section into power tube distortion has probably the highest order of even order thd you can get which IMO is a big factor when talking about how good distortion sounds.

[blearyeyes]

Anyone hear the Quilter Labs solid state amp? Founder of QSC and builder of my first custom amp in the 60s Pat Quilter is back at it trying to make SS sound like we all want them to. I haven't heard them. Little bitty amp 8lbs? and 200 watts.

[blearyeyes]

There is also alot to say about why distortion sounds good and what variations of distortion and why. One of the reasons has to do with thd as in even order and odd order from the fundamental frequency.

Not to mention the wide variation of distortion that guitar players need for different genres of music. Personally I like old school, but there are a ton of new guys using a completely new kind of tone for metal. Plus I have guys show me stuff they love that sounds OK but it just isn't "that tone". But maybe what I think is good tone is just my dated opinion. So I have loosened up my call on what sounds good and am looking for something that will work with all of the different guitar sounds I love and the ones I am going to love. I've played guitar since before surf music, hootenanny, jug bands and Peter, Paul and Mary were big. But now I see more and more bands using the Fractal Designs front end. Listen to their MP3 demos and let me know what you think. I was personally sucked into equipment lust. I just hope we aren't going the way of Hi Fi: From records out of tube amps driving nice big speakers that moved air to MP3 compressed files out of blue tooth bull pucky speakers. Still I hope we can rise to the challenge and come up with equipment that can advance the electric guitar, be it 6 or 8 or 10 string.

Felt like opining to the great minds that lurk around here. 8-))

[Steve Conner]

I built an experimental hybrid amp a while back. A tube preamp and 1x EL84 output stage, driving a bunch of power transistors through an interstage transformer, kind of like the horrid old Thomas Vox amps but with a few tweaks to improve the reliability and allow use of an ordinary tube amp output transformer instead of a custom part.

It sounded remarkably good. I think the interstage transformer supplies some of the mojo that an output transformer would. My next project will be to replace the preamp with a SS circuit too just to see if it can be done.

My biggest problem with digital modelling is that you're not playing an amp, you're playing a programmer's mental image of an amp. I think Cliff Chase of Fractal Designs is an extremely smart guy and his mental image may well be almost complete, but I'm willing to bet the territory he has explored is mostly metal.

I don't much like MP3s or tiny plastic Bluetooth speakers either. I have a compromise at home, a digital hard disk player that plays lossless files through tube amps and big speakers made of some sort of wood-like substance.

[jazbo8]

This is an interesting thread as I am gearing up for my next build... I have been mulling over whether to build a low-power amp that could somehow emulate the tone of output power tube/OPT distortion (e.g., Firefly or the 1W Marshalls..) or a "regular" high-power amp that I can use with either a VVR or an attenuator, all this of course with the goal of playing at home at modest volume AND getting the "tone", granted it is not something that the pros should worry about since they play in clubs, in those case, I see little reason to use a low-power amp as many have already pointed out...

[Steve Conner]

Well, check out the schematics and demo video here. scopeblog » Ninja Corvette Hybrid

Here's one interesting consideration: A regular solid-state output stage uses split supplies and DC coupling to the speaker. Any even-order distortion in the output stage builds up a DC offset that pushes the speaker cone out of place. This was never a problem in tube amps because the output transformer can't pass DC. My design uses a simple trick to address this so I can crank out as much even order as I want.