How “clean” is an amps power section?

Nerine · August 2023

Was tooling around running my FM9 into the Fx return of my Friedman, so essentially using it as a valve driven power amp and cab. (Cab sims off in the Fractal)

Sounded pretty immense, and possibly a bit more “in the room” than through my usual FRFR setup. Plus, no IR option paralysis.

It got me to thinking about a Fryette Power Station and a 1x12 cab or something as a means of amplification.

But then I kinda took a step back and wondered how coloured the power amp section is in a typical amp?

I guess each amp is filtered slightly differently and the negative feedback will be different etc, but how MUCH different are we talking about here??

Is running into the FX return really colouring things that much? I’d assume the speaker will have more of an impact than the output section??

ecc83 · August 2023

Depends! Most FX returns go into the PI input, often top of the MV pot but, as friend ICBM will no doubt be able to tell us, not all FX loops are equal!

Assuming a feed to the PI grid then yes, the rest of the circuit is more or less "Hi-Fi" so long as you keep the power levels fairly low. A 15W amp (2x EL84/6V6) should be very clean at 6-8 watts and that will be quite loud through a decent guitar 12. A 50W amp (EL34s) would be hi fi ish at 20W and below.

Those statements need some qualification at very low frequencies, below 60Hz say as few guitar amps have expensive output transformers but then nor do they need them.

Dave.

Danny1969 · August 2023

The effects return on most amps is coming back just before the phase splitter on most amps ... so typically it would be running through a 12AX7 or similar, then the output valves, then the transformer and then the speaker. This is the main meat of a valve amp and it adds a lot of colour. If you plug a Marshall JMP-1 into the effects return of a Hot Rod is sounds completely different to plugging it into a Marshall valve power amp. So the power sections adds a lot of colour

ecc83 · August 2023

Danny1969 said:

The effects return on most amps is coming back just before the phase splitter on most amps ... so typically it would be running through a 12AX7 or similar, then the output valves, then the transformer and then the speaker. This is the main meat of a valve amp and it adds a lot of colour. If you plug a Marshall JMP-1 into the effects return of a Hot Rod is sounds completely different to plugging it into a Marshall valve power amp. So the power sections adds a lot of colour

Sorry Danny I don't agree. A PI and a couple of EL84s* will deliver 10 watts at about 3% THD (will look that up!) even without any NFB and given a half decent transformer. "Colour" only starts to occur when the power level gets close to maximum. Why folks by power soaks!

Of course, some amp designs include response shaping caps in the OP circuit but most don't.

I remember checking a 100W 4 x EL34 amp (you can guess whose!) on an AP test set and it needed driving to 120W to hit 10% THD and 10% is still pretty clean. At 50W that amp would have a THD figure below 1% IMHO.

*That is essentially my old Rogers 10+10 amp sans NFB and that was more than loud enough through far less sensitive speakers than 95dB/W/m git types!

Dave.

ecc83 · August 2023

http://www.r-type.org/pdfs/el84.pdf

That shows two valves in AB push pull cathode biased will deliver 17W at 4% THD so at half that power there will be less than 2% THD (mainly 3thrd harmonic) and that is barely noticeable.

No, that figure does not include transformer distortion but given reasonable balance of Ia that will not add much distortion at "guitar" frequencies.

Dave.

ICBM · August 2023

There used to be a 100W valve hi-fi monoblock power amp kit you could get from Maplin (can’t rememember the brand) - I looked at the schematic and was quite surprised to see that it was basically an Orange OR120 power amp!

A clean power amp well below its maximum clean output doesn’t colour the tone that much - certainly not compared to the speaker and the cabinet. The power valve type has very little effect, contrary to popular belief.

If you push it to the point of overdrive then it starts to have a large effect - the transformer, NFB, valve type, phase inverter circuit type and values all make a big difference.

Nerine · August 2023

Hmmm.

Interesting.

Thanks for the responses so far.

I wont be pushing the Power Amp into overdrive as that would kind of negate having all the modelling stuff beforehand.

i guess the main reason for the question is that this method still allows me to run direct to the PA with a cab simulated output in the Fractal, and have a more in the room feed to my amp behind me - negating the need for mics etc. I've found that I prefer most of the amp models in the Fractal when power amp modelling is switched on, so hence the question of cleanliness running into a non-saturated valve power section. It then means the models sound better going though the cab sims/IRs and hence PA, too. Win win.

Example, the Dumble model doesn't sound amazing running into the Friedman power stage with the power amp modelling turned off, but sounds rather awesome with the power amp modelling turned on, but seeing as I'm borderline anally retentive I don't want DOUBLE Power Amp going on, y'know.

Anyway, I could waffle on for hours about this stuff, and in particular, praise for the FM9. This method is essentially like changing the guts of my amp to the models I'm using, and it's all very amp in the room sounding seeing as you are getting all the filtering and directness of an actual guitar speaker and cabinet etc, rather than an IR

The feel is amazing, too, plus less faff with HPFs and LPFs when dialling in sounds for FRFR monitoring etc as the signal chain by proxy does all that for you.

The flexibility is huge.

ICBM · August 2023

Nerine said:

Example, the Dumble model doesn't sound amazing running into the Friedman power stage with the power amp modelling turned off, but sounds rather awesome with the power amp modelling turned on, but seeing as I'm borderline anally retentive I don't want DOUBLE Power Amp going on, y'know.

Why not? If it sounds good, it is good. The reason why 'double power amp' sounds better is largely down to playing at low volume, so the Friedman is almost completely clean and you aren't getting the sonic benefit of it. It won't be very different if you use an actual power amp, at least if it's going into the same speakers.

I know it can sometimes be easy to over-think stuff like this, but trust your ears instead.

Nerine · August 2023

ICBM said:

Nerine said:

Example, the Dumble model doesn't sound amazing running into the Friedman power stage with the power amp modelling turned off, but sounds rather awesome with the power amp modelling turned on, but seeing as I'm borderline anally retentive I don't want DOUBLE Power Amp going on, y'know.

Why not? If it sounds good, it is good. The reason why 'double power amp' sounds better is largely down to playing at low volume, so the Friedman is almost completely clean and you aren't getting the sonic benefit of it. It won't be very different if you use an actual power amp, at least if it's going into the same speakers.

I know it can sometimes be easy to over-think stuff like this, but trust your ears instead.

Oh, absolutely, 100% agree. I was kinda exaggerating above. I wouldn't not use something because the setup is a bit weird or convoluted. I'll only ever judge something on how it sounds.
In theory at least it kinda seems a bit fruitless emulating a big, dynamic 100w 6L6 power section for it to be potentially negated by running it into a "smaller" 50W EL34 driven one, and whether the compromise ends up being a Dumble preamp circuit into an EL34 power section - which is a bit weird, topology-wise and compared to the classic circuit..

And I guess, ergo, whether a Fryette Power Station (or similar) would have a flatter response than one you would typically find in a regular guitar amp, giving a slightly truer representation of the modelled amp.

Seems like that might be a more negligible difference than I was expecting, though, which is cool, 'cos hooked up to the Friedman it does sound killer however you slice it.

Essentially, if the valve power section of the amplifier is pretty clean, the power amp modelling in the Fractal is more useful and means the model is more accurate, so hence my question being about how coloured you can expect a typical valve power section in a guitar amp to be.

Modulus_Amps · August 2023

Hold on chaps, the amount of negative feedback a power section has changes the tone fairly dramatically. The friedman power section has bass boost built into it and the presence control option too.

All power sections are not created equal

ecc83 · August 2023

Modulus_Amps said:

Hold on chaps, the amount of negative feedback a power section has changes the tone fairly dramatically. The friedman power section has bass boost built into it and the presence control option too.

All power sections are not created equal

No, "all power sections ARE created equal" if they conform to the valve mnfctr's data. They will sound different if people insist on using frequency selective filters in an NFB loop* or Zobel networks across the traff primary. Remember this thread assumes the amplifier is being used well below its nominal power rating.

*Yes, 'flat' NFB will change the tone a bit because it will damp the speaker better and guitar speakers are very honky buggers. But that is making the amp MORE hi fi, not less so.

Dave.

Modulus_Amps · August 2023

Sorry Dave. Fortunately people do insist on using frequency selective filters in power sections so they can not be ignored. This is not a theoretical discussion about data sheet designs. When comparing real world products on the market that the average guitarist can buy they are designed differently, they have filters, they have varying amounts of feedback that does affect the final response and feel of the amplifier at low levels and high levels. The mentioned Fryette power station and Friedman amp both have filters in the negative feedback loops, so they will sound and feel different compared to plugging into an effects return on an amp that doesn't.

Point is his FM9 may sound more immense through his Friedman amp because it is boosting the bass and low mids.

ecc83 · August 2023

Modulus_Amps said:

Sorry Dave. Fortunately people do insist on using frequency selective filters in power sections so they can not be ignored. This is not a theoretical discussion about data sheet designs. When comparing real world products on the market that the average guitarist can buy they are designed differently, they have filters, they have varying amounts of feedback that does affect the final response and feel of the amplifier at low levels and high levels. The mentioned Fryette power station and Friedman amp both have filters in the negative feedback loops, so they will sound and feel different compared to plugging into an effects return on an amp that doesn't.

Point is his FM9 may sound more immense through his Friedman amp because it is boosting the bass and low mids.

Well yes but I did make that qualification.

Dave.

Nikc · August 2023

Surly this is where a good clean transistor amp comes into its own - I run my cr60C in a 4 cable set up - i've turned off IR's and power amp sections so when I use an amp model it's just the pre into my amps power section - loads of options and sounds great. Although i tend to just my modeller just for effects rather than amp modelling although its nice to have the option.

ecc83 · August 2023

Nikc said:

Surly this is where a good clean transistor amp comes into its own - I run my cr60C in a 4 cable set up - i've turned off IR's and power amp sections so when I use an amp model it's just the pre into my amps power section - loads of options and sounds great. Although i tend to just my modeller just for effects rather than amp modelling although its nice to have the option.

Well yes,...but only if the tranny amp is GENUINELY rated at at least three times the power output of a comparable valve design.

Transistor amplifiers need/use a lot of NFB*. The result is they have very low distortion up to close to clipping then WHAMMO horrible ***t!

Valve amps on the other hand, even those with a bit of NFB overload much more gracefully. Valves are also better at delivering 'volts' whereas transistors are good at current and you need V to beat thermal compression. Valves really ARE louder, watt for proper watt.

*Same goes to a degree for truly hi fi valve amps designed with high feedback. Never had one to try but I would think it would be hard to get an OD 'sweet spot' with a 12W Quadll?

Dave.

vasselmeyer · August 2023

Nerine said:

Hmmm.

Interesting.

Thanks for the responses so far.

I wont be pushing the Power Amp into overdrive as that would kind of negate having all the modelling stuff beforehand.

i guess the main reason for the question is that this method still allows me to run direct to the PA with a cab simulated output in the Fractal, and have a more in the room feed to my amp behind me - negating the need for mics etc. I've found that I prefer most of the amp models in the Fractal when power amp modelling is switched on, so hence the question of cleanliness running into a non-saturated valve power section. It then means the models sound better going though the cab sims/IRs and hence PA, too. Win win.

Example, the Dumble model doesn't sound amazing running into the Friedman power stage with the power amp modelling turned off, but sounds rather awesome with the power amp modelling turned on, but seeing as I'm borderline anally retentive I don't want DOUBLE Power Amp going on, y'know.

Anyway, I could waffle on for hours about this stuff, and in particular, praise for the FM9. This method is essentially like changing the guts of my amp to the models I'm using, and it's all very amp in the room sounding seeing as you are getting all the filtering and directness of an actual guitar speaker and cabinet etc, rather than an IR

The feel is amazing, too, plus less faff with HPFs and LPFs when dialling in sounds for FRFR monitoring etc as the signal chain by proxy does all that for you.

The flexibility is huge.

This is what I do with my Boss GT-1000 if I want an amp on stage as well as through the PA. I take a feed off the unit before it hits the cab SIM and put that into the FX return of my DV Mark head which runs a 2x12. I also take an XLR out, including the cab and mic SIM to the PA for FoH sound.

ecc83 · August 2023

"I wont be pushing the Power Amp into overdrive as that would kind of negate having all the modelling stuff beforehand."

Understood but you do need a lot more 'honest' power from sstate kit to stay clean and loudish. Valve OP stages can still sound very clean even when in fact they are driven to 5% or more THD. With transistors it tends to be ***T or Bust (sometimes very much the latter!)

To put some numbers on it I would say a WELL designed '15W' valve amp could hold its own against a nominal 50W transistor especially if class D.

Dave.

Howdy, Stranger!

Categories

In this Discussion

Become a Subscriber!

How “clean” is an amps power section?

Comments