SO ~ what gives an AMP it's character or tone ???

Kebabkid

monquixote said:

The main bit of confusion seems to be that because Fenders tend to have 6L6 or 6V6 and Marshalls have EL34s that all amps with those valves sound a certain way when that isn't really true.

Well done @monquixote. Yep, sure there might be "a specific sound expectancy" with certain valve types which probably rings true at times but at the end of the day, people must use their ears.

ICBM

ecc83 said:

Yes! Should have said "almost all" use a common cathode resistor. I did ask once a B's why two are so rarely used but they could not tell me either! I doubt that the saving of one R and cap would be it!

They don't "sound right" if the aim is the traditional cathode-bias sound.

A friend of mine once had an AC30 which had been modified by a proper electronics engineer with separate cathode resistors and caps - actually each with a small trimmer as part of the total value, so each valve could be biased perfectly - neatly done and theoretically correct, but the amp just didn't sound right - it sounded 'hard' and too 'hi-fi'.

I had a hunch it was this mod, so without doing anything else I removed it and put back the standard crude single resistor and cap, and… instant great AC30 tone.

As ever, what's right for hi-fi is wrong for guitar amps and vice versa! This is one of the incidents that led me to develop this rule of thumb .

That slightly harder, crunchier (more fixed-bias-like) sound is apparently why Cornell uses separate resistors on the Plexi 18/20 as well - he was deliberately trying to make it sound more like a small Plexi Marshall... and it does, more so than the Marshall version which uses a shared resistor.

monquixote

So when are we going to see the @ICBM signature head

ICBM

monquixote said:

So when are we going to see the @ICBM signature head

lol

http://upload.wikimedia.org/wikipedia/commons/a/a4/Mesa_Boogie_Dual_Rectifier_Tremoverb_100w_Head.jpg

guitarfishbay

ChrisMusic said:

Thanks for that @guitarfishbay.  I remember you saying that you tried the Zilla 2x12 cab but the removable rear panel flexes too much under pressure.  That should be fairly easy to cure with a bit of batten, foam strip, and some T nuts, which is the way they should have done it IMHO, or at least to have braced the back panels to stop flexure when used as close backed.

The Rectifier cabs use a Mesa spec V30 as far as I am aware, so I guess that voicing is a big part of how they sound.

What was your experience with other speakers with that amp ?

How would you say that the Roadster compared to a clean Fender if you were to put them in the same room ?

The Road King does look very tempting with all the progressive linkage, EL34 & 6L6 switchable power sections etc, but I could buy a small country for the money they cost, so it's not really an option for now.

Well I don't know if Zilla have changed their construction techniques since.  The one I had was bought second hand, so I have no idea how old it was.  Personally I wasn't impressed with the 3 piece back at all.

Mesa V30 cabs sound a bit smoother than regular V30 cabs.  Regular V30 cabs have a bit more bite and brightness to them, whereas the Mesa V30s sound quite different.  It feels like the dip in the mids is in a different place, plus the high end is a bit smoother in response.  Personally I prefer the Mesa V30s with Mesa amps.

I've tried the Roadster through the following cabinets, if I owned it I'll say as that means I've had a decent amount of time with that cab and tried it with a band at some point -

Mesa Rectifier 2x12 V30s (owned) - This was my favourite cabinet for the amp.  It had the best balance, and all the channels sounded great through it.  I especially like the cleans through this cab, seems to have the best balance of low end warmth and top end sparkle when set up for a brighter sound.

Mesa Rectifier Oversized Straight 4x12 V30s (owned) - This sounded huge and all the channels sounded great, but IMO the channel 4 modern setting, which has huge amounts of bass, was almost too loose in response for my tastes.  What I was impressed with on this cab was how great you could get the Tweed channel to sound with the gain turned up for a warm blues sound.

Marshall 1960B T75s (owned) - Not bad, but it seems to accentuate the bass and sizzle in the top end on the gain channels.  OK for a cheap second hand option but didn't blow me away with any channel.

EVH 4x12 Greenbacks (owned) - Really nice for classic rock if you avoid the Modern channel.  The vintage channel sounds really nice in spongy/valve rectifier modes.  The clean channels also sounded good, but didn't have the same depth to the bass as the V30 cabs.  I liked it, but at high volume I didn't think the Rectifier was a good choice for this cabinet as in full 100 watt mode the bass response got very soft.

Blackstar S1 Oversize 4x12 Straight V30s (owned) Slant - This sounded really quite different to the Mesa Oversize 4x12, despite being very similar in size and specification.  The two biggest differences were the baffle was slightly slanted (the Mesa OS Straight is completely straight) and the V30s were regular V30s.  All the channels sounded good with this cabinet, but the highest gain sounds were a bit more bright/harsh compared to the Mesa cabinets.

Orange PPC412 (used in recording studio) - Only used CH4 Red Modern Bold/Diode with this cab, and it was ok but felt a bit stuffy in the mids.  Not my favourite.

Mesa Stiletto slant 4x12 (used in recording studio) - Again only used CH4 Red Modern/Bold Diode with this cab, but it sounded pretty good.  It had less bass than the Oversize 4x12 and didn't seem quite as loose.  It was less stuffy in the mids than the Orange Cab.  I liked it better than the Oversize 4x12 cab to be honest, but I really could not find one used and they seem quite rare.

Overall, based on my own experiences the cabinet I'd personally recommend is the Rectifier 2x12 with the Mesa V30s.  Or if you didn't want a closed back cabinet I'd still suggest looking at an open backed Mesa cabinet with V30s, or possibly the Black Shadow 90s which I haven't personally tried but are supposedly a good match for Lonestars if cleans are your main criteria.

Again I have not compared a Fender directly with the Roadster, but I personally feel the Roadster cleans were by far the best cleans I've heard through a closed back cabinet and were very high quality.

The main caveats with the amp IMO are:

- Dialling them in for non Mesa cabinets is slightly harder, and seems to require slightly less conventional settings.  But for the most part it can be done.  The channel that seems pickiest is Ch4 Modern.

- Ch4 Modern sounds better with a bit of volume behind it, again IMO.  I think the cleans and even mid gain sounds can be quite nice at lower volumes, but you seem to need to get the volume up a bit to get the most from the highest gain sounds.

- The reverb is very very nice, but there is a gap in reverb if you switch channels with the reverb on.

I never tried the amp with EL34s, but people don't seem to think it makes a huge amount of difference.  If you didn't need the option to have both 6L6 and EL34 at the same time a Roadster would be a cheaper option than the Road King.

Hope that helps!

monquixote

ICBM said:

monquixote said:

So when are we going to see the @ICBM signature head

lol

http://upload.wikimedia.org/wikipedia/commons/a/a4/Mesa_Boogie_Dual_Rectifier_Tremoverb_100w_Head.jpg

Maybe a 200W version made of depleted uranium?

ChrisMusic

Thanks for clarifying that @monquixote.  Thats a great set-up you have there, sounds really good.  You are right about the Xotic SP compressor, it's on my "to buy" list, does the job brilliantly.  It's also good to see some of the accepted wisdom shaken up a bit

ChrisMusic

Wow, that's a very comprehensive, well constructed and knowledgeable reply @guitarfishbay.  Thank you for taking the time to put all that together for me.  Really valuable information there, which hopefully will be of use to others on here too.

The information on the Roadster is really useful, I have to admit I was rather tempted by it.

The whole post will be really helpful in narrowing the choices when it comes time to get the speakers and cabinets sorted, cheers.   

ChrisMusic

Moving on with the exploration:

Presence, resonance, negative feedback.

I assume these are all related to power section topology.

What is happening with these, how and why is it done, and importantly for this discussion - how does that affect the sound ?

ICBM, is that a Stiletto hiding under the Trem-o-Verb ?

John_A

I've done a similar experiment to ICBM  with home built 18 Watters, one using 6V6's and the other EL84's, the difference is very subtle, it's the circuit that shapes the sound not the valves, everything matters though, but not as much as many people think

ICBM's what makes an amp list is pertty accurate, personally I'd change it a little bit as follows:

Circuit design
- preamp topology
- power amp topology
Speakers
OP Transformer type
Component types
Cabinet design
Valve type
Cabinet material
Circuit layout
Wiring dress
Valve brand
Grille cloth
Cabinet covering
Chassis material

ICBM

ChrisMusic said:

Moving on with the exploration:

Presence, resonance, negative feedback.

I assume these are all related to power section topology.

What is happening with these, how and why is it done, and importantly for this discussion - how does that affect the sound ?

ICBM, is that a Stiletto hiding under the Trem-o-Verb ?

Not sure, it's not my pic! I think it is.

Negative feedback is where a small part of the output signal of a stage is fed back into the input of it, out of phase, so it counteracts the amplification to some extent. This is useful because it helps to smooth out non-linearity in the stage, so although you get less gain, you get a cleaner sound, and wider and more even frequency response. It also reduces the output impedance of the stage and increases its damping factor so it becomes less dependent on the load.

In most amps it's done on the power stage, so the loss of gain is not very important anyway, but you get closer to a 'perfect' output stage. (As you may have guessed, this comes from hi-fi design originally.) Usually the signal is fed back in at the phase inverter.

The result is a cleaner, clearer, 'bigger' sound but also more hi-fi, ie less 'characterful' in some ways. When the power stage distorts, it gets more complicated because the NFB is abruptly cut off at the point of distortion, so the transition to distortion becomes sharper and the distortion 'harder' in character. A perfect example is the difference between a Vox AC30 (no NFB) and a Marshall Bluesbreaker (heavy NFB). Both amps are fairly similar-sounding clean (not exactly obviously), but the AC30 has a very bouncy, compressed overdrive where the Marshall is much tighter and crunchier.

Usually a presence control removes high frequencies from the NFB loop, which takes away some of the damping at the top end and makes the highs more dynamic and lively, which makes the amp more 'present' in a mix, hence the name. A resonance control is the same but works on the low-end, or sometimes just on the whole amount of the NFB - it makes the amp sound looser or more resonant - at the expense of making it much more speaker-sensitive. A great example is the Modern High-Gain mode on the Mesa Dual Rectifier, which turns off the NFB entirely just in that mode - that's what gives the huge boomy 'Recto' tone, but makes it very fussy about using the right cabinet. (The presence control on that channel works in the preamp instead.)

martinw

ICBM said:

There are some cathode-biased guitar amps which use separate resistors - the Orange Rocker 30, several Matamps and the Cornell 18/20 are a few I know of - but it's fairly rare. It also makes the amp "sound more fixed-bias" compared to a shared resistor, I'm not quite sure why!

The MJW Orion and Electra use separate cathode resistors (when cathode biased), but I'll forgive you for not being aware of those!

I'm using it as a way of reducing power output, ideally without altering the tone too much. There is a small change to the tone and feel of the amp, but it's hard to say how much of that is because of the power reduction and how much the  change of bias method.

jpfamps

ecc83 said:

ICBM said:

There are some cathode-biased guitar amps which use separate resistors - the Orange Rocker 30, several Matamps and the Cornell 18/20 are a few I know of - but it's fairly rare. It also makes the amp "sound more fixed-bias" compared to a shared resistor, I'm not quite sure why!

The Mesa Simul-Class is unusual in that the 'Class A' section is fixed-bias, but I'm not sure if it's true Class A or not - it is low-powered enough that it's possible. It also runs in triode mode which sounds quite different from tetrode/pentode, which may be more important than the class!

Yes! Should have said "almost all" use a common cathode resistor. I did ask once a B's why two are so rarely used but they could not tell me either! I doubt that the saving of one R and cap would be it!

The HT-20 biasing is unusual. About 12 volts of it comes from cathode bias* but there is also a negative supply with setting and balance control. Not sure why they went that way? One advantage is of course that the V loss in the cathode circuit is only about a third of full CB.

One operational advantage is that you can swap the 34s pretty much with impunity and the bias will come up about right. Best of course to get it set and balanced soon as...

*At first glance it looks as tho' the 20 has separate cathode Rs but in fact it is two 7 watters in parallel.

Dave.

One advantage of separate cathode resistors is that if a valve fails open circuit then the bias of the other valves is not affected.

In an amp with a shared cathode resistor then a valve failure means the remaining working valves become biased hotter, which especially dangerous in amps with over dissipating EL84s.

In a true class A amp driven below clipping the sum of the current draw through each side of the power stage is constant, and thus the voltage across a common cathode resistor will be constant. In practice there will always be a slight variation and thus a cathode bypass cap is used to shunt ac current to ground.

In a class AB amp (ie virtually all "class A" guitar amps), when the current draw from the power supply is NOT constant as the amp is driven harder. The cathode bypass cap does shunt some of this extra current to ground but when the amp is driven hard the extra current causes the cathode bypass cap to charge up increasing the cathode voltage, ie the amp becomes biased colder as it is driven harder.

This is actually stated on the JMI AC30 schematic, which indicates a power valve cathode voltage of "12.5V AT 30V, QUIESCENT 10V", which of course indicates class AB operation........

As most of the current flow (if not all) with be from the positively driven side, the current through one side of the push pull pair affects the bias of the other side, and, in my view, is part of the cathode bias sound.

At best this will increase crossover distortion, and at worst will cause blocking distortion where amp is driven into cut off.

The amp will only return to "correct" bias conditions once the cathode resistor can discharge, and this will depend on the value of the capacitor and the impedance it sees to ground.

In a fixed bias amp, the bias is set by a negative voltage applied to the grids of the power valves, and this will vary very little, if at all, with level of power valve drive.

Using separate cathode resistors for each valve to bias the output stage, then the current draw on one side of the push pull pairing with not affect the biasing of the other side (in fact the other side will be biased hotter as the cathode bypass cap discharges). Thus the relationship between the biasing and drive levels with be more akin to fixed bias, and the opposite of cathode bias with a common cathode resistor.

For those of you how have managed to read this far, I haven't mentioned grid conduction; suffice to say this phenomenon exacerbates the differences between bias methods; however I think that this post is long enough already!

The use of separate cathode resistors seems to have been popularized by Mullard in their application notes, although I don't know if they were the first, and my guess is that is this was either a tacit emission of class AB operation, or a method of temporary overloads more benign.

jpfamps

Yes! Should have said "almost all" use a common cathode resistor. I did ask once a B's why two are so rarely used but they could not tell me either! I doubt that the saving of one R and cap would be it!

The HT-20 biasing is unusual. About 12 volts of it comes from cathode bias* but there is also a negative supply with setting and balance control. Not sure why they went that way? One advantage is of course that the V loss in the cathode circuit is only about a third of full CB.

One operational advantage is that you can swap the 34s pretty much with impunity and the bias will come up about right. Best of course to get it set and balanced soon as...

*At first glance it looks as tho' the 20 has separate cathode Rs but in fact it is two 7 watters in parallel.

Dave.

The AC50 used EL34 and combination of cathode and fixed bias with bias adjustment for both sides of the push pull pair, so allow balance.

Regardless of the sonic properties of cathode/fixed biasing, If you want to incorporate bias balance with pure cathode biasing, then you will need a fairly hefty pot to cope with the bias current if you what the amp to be reliable, so a small element of fixed bias is a sensible approach.

Also, adjusting a cathode bias pot would create a ton of noise, and you can imagine the number of communications you would have about this!

ICBM

jpfamps said:

As most of the current flow (if not all) with be from the positively driven side, the current through one side of the push pull pair affects the bias of the other side, and, in my view, is part of the cathode bias sound.

At best this will increase crossover distortion, and at worst will cause blocking distortion where amp is driven into cut off.

The amp will only return to "correct" bias conditions once the cathode resistor can discharge, and this will depend on the value of the capacitor and the impedance it sees to ground.

In a fixed bias amp, the bias is set by a negative voltage applied to the grids of the power valves, and this will vary very little, if at all, with level of power valve drive.

Using separate cathode resistors for each valve to bias the output stage, then the current draw on one side of the push pull pairing with not affect the biasing of the other side (in fact the other side will be biased hotter as the cathode bypass cap discharges). Thus the relationship between the biasing and drive levels with be more akin to fixed bias, and the opposite of cathode bias with a common cathode resistor.

Aha! Thank you . That explains it - I had been trying to think of the mechanism but had a bit of a mental block for some reason.

ChrisMusic

ICBM said:

ChrisMusic said:   Moving on with the exploration:

Presence, resonance, negative feedback.

Negative feedback is where a small part of the output signal of a stage is fed back into the input of it, out of phase, so it counteracts the amplification to some extent....    (edited for brevity: please read the full explanation in the post above)

What an excellent, clear and concise explanation, with plenty of sonic descriptive references, thanks ICBM, that explains a lot.

@John_A, thanks for your revision of the hierarchical list, I have added OT and component specs to my question below.

Also thanks to @ecc83 @martinw and @jpfamps for your contributions.  That will take a bit more digestion, although I surprised myself as to how much I followed.  I am loving all the learning, and will return to these comments as the jigsaw picture slowly crystallises, thanks.  There are a diverse bunch of people on here, so it's nice to add some depth to the conversation, and push the learning envelope too.  I hope we will all be a little wiser about amp design and how that translates to the tones we are trying to achieve, that's got to be a win all round IMO.

A couple more questions come to mind:

•  What other attributes of power amp design and component specification give a tangible sonic signature ?

•  What about the interaction between speaker and amp ?

•  How about the role of the power supply in all this ?

jpfamps

ChrisMusic said:

ICBM said:

ChrisMusic said:   Moving on with the exploration:

Presence, resonance, negative feedback.

Negative feedback is where a small part of the output signal of a stage is fed back into the input of it, out of phase, so it counteracts th

•  What other attributes of power amp design and component specification give a tangible sonic signature ?

•  What about the interaction between speaker and amp ?

•  How about the role of the power supply in all this ?

There are several aspects of power amp design that need to be considered: valves used (and number there of), configuration (single ended or push pull, pentode, triode, ultra-linear) bias method, class of operation, HT voltage, screen supply voltage, screen supply impedance, loading, output transformer construction, use of negative feed back.

Of course changing one parameter will often affect the others. 

For example increasing the HT voltage will mean that the amp has to spend more time in class B loading to keep dissipation within limits, and that for optimum output power the loading on the primary of the output transformer needs to be increased (although many manufacturers don't do this and use the "standard" values".

Negative feedback will linearize the output and reduce somewhat the differences between valves. 

Negative feedback is not mandatory in a guitar amp, and although most use it, there are plenty of examples of amps that don't employ it, eg the AC30, and 5E3 Tweed Deluxe. Incidentally we don't use negative feed in our guitar amps either (shameless plug).

The output transformer also has a significant effect on the sound of the amp. The low frequency response of the transformer is predominantly determined by the size of the core. A larger core also means thicker wire can be used, and less power is lost to I squared R heating (termed copper loss).

A larger core does mean that the HF response is reduced; this can remedied by using grain orientated transformer laminations, and by interleaving the winds; ie winding a bit of the secondary and a bit of the primary etc. 

The HF response is important not so much because there is much (if any) sonically useful information in an electric guitar in the HF region (in fact probably not much above 8k), but if you are applying negative feedback around the power stage you can run into problems with phase shifts in the feedback path causing oscillation.

Briefly, for negative feedback to work the signal being sent back to the input should be 180deg out of phase with the input signal. However, wherever there is an element in the feedback path that has frequency dependent roll off (eg an RC filter), then there is a phase shift in the signal as the filter starts to attenuate. If you have enough gain and filters in the feedback path, then you can end up with positive feedback and the amp with "ring" or worse oscillate. Of this oscillation or ringing is ultrasonic, so is not audible on it's own, but has a detrimental effect on the guitar signal.

This is a real problem with some commercial designs. The Blues Junior in particular is prone to utlrasonic oscillation.

Anyhow, getting back to transformers, a larger core, higher grade steel, and interleaving all cost more, and as the output transformer is one of the more expensive components in the amp, it is an area where significant costs can be saved.

The interaction between the speaker and the amp is important, as the amp does not see a constant impedance (far from it in fact). Furthermore the speaker also "sees" a relatively high and non constant impedance looking into the amp. This is a problem for attenuator manufactures, who ideally would want preserve the interaction between the amp and speakers. Most attenuators fail in this regard.

Assuming that the power supply is adequately rated, then the mains issue with the power supply is how well regulated it is. 

Regulation in power supply terms means how well the power supply voltages hold up when the load on power supply is varied. A power supply with perfect regulation will maintain the same voltage regardless of the current drawn from it. Of course the perfectly regulated supply doesn't exist, but something that to all intents and purposes is "perfect" over the range of current drawn in an amp can be made, either using a linear power supply, or switching mode.

Most guitar amps don't have a regulated power supply, and this means the voltages in the power supply will drop as more current is drawn from it. In a class AB amp, more current is drawn as the amp is driven harder. This is often referred to as "sag".

The effect of this is the reduce the power of the amp in a signal dependent fashion, ie it acts a bit like a compressor. 

The amount a voltage drop will depend on 1) the impedance of the power supply. This is the sum of the secondary wiring resistances and the rectifier impedance. Valve rectifiers have significant impedance and hence more sag 2) the size of the main filter caps. Larger caps will hold the HT voltage up better when more current is demanded. However larger caps can stress the tranformer, and valve rectifiers have limits as to the maximum filter cap that can be used with them.

I haven't discussed screen grid supplies, but I think this post is long enough as it is......

ChrisMusic

A very detailed and authoritative post @jpfamps , I have been waiting for the alluded to, part 2, so please feel free to continue.

Anything you can add to the previous (of future) descriptions about how it translates from architecture and operation, into the sonic palette we understand as musicians, would be greatly appreciated.

If there is anything you can add about the nuances of adding the variable of speaker type and construction into the discussion, that would be good, as the interactions between power amp and speaker obviously have a great affect on the sonic signature and the tone we perceive.

57Deluxe

According to Mark Bartel (Tone King amps) it seems to be ALL about the cabinet design and the way the speakers interact with it...

ChrisMusic

57Deluxe said:  According to Mark Bartel (Tone King amps) it seems to be ALL about the cabinet design and the way the speakers interact with it...

Yes, I had that in mind too.  Another can of worms with cab design plus speaker characteristics and selection.  Obviously an important and interactive part of the jigsaw with the power amp.

I was saving that for the next question, as it takes a bit of digestion with each new bite into valve amps...