Tube preamp heads: Gimmick??

[seamonkey]

Exactly. But Seamonkey keeps insisting that they're some kind of fuzz circuit.

Good for the iggy feature - this is all I see.
But if you're referring to diodes. Then yes that are used in signal paths in this configuration as clippers - a form of fuzz.

 

[seamonkey]

McIntosh decided on green LEDs - It's an artistic visual statement.
In the audio, No diodes in their signal path.
They've been using a dedicated winding for feedback for better isolation from the drivers.
Class-AB with negative feedback.
In a DB test they are audibly the same as any power amp.

 

[JimmyM]

Good for the iggy feature - this is all I see.
But if you're referring to diodes. Then yes that are used in signal paths in this configuration as clippers - a form of fuzz.

Hmmm...do I believe you, a guy who never played a tube amp in his life, or Jerrold, the former chief engineer at Ampeg under SLM?

Jerrold by a landslide.

 

[R Baer]

I would love to hear feedback on that one.

The tube section on the Valkyrie is basically a mini 2 tube, push/pull design that feeds and output transformer and is located between the all solid state preamp and class A/B power amp. It was never really designed to be an overdrive circuit. I've always felt that "overdrive" is best left to pedals, so the tube section in the Valkyrie was meant to add a lot of tube harmonics and overtones, while fattening up the sound and giving your bass a wider spacial field. The sound is more three dimensional and gives you a bit of tube compression and transformer saturation that helps the class A/B power amp feel a bit more "tube like", but the reality is that the only thing that performs like an all tube amp is, well... an all tube amp. Modern modeling is quite good, if your'e into that sort of thing.

 

[Wasnex]

The tube section on the Valkyrie is basically a mini 2 tube, push/pull design that feeds and output transformer and is located between the all solid state preamp and class A/B power amp. It was never really designed to be an overdrive circuit. I've always felt that "overdrive" is best left to pedals, so the tube section in the Valkyrie was meant to add a lot of tube harmonics and overtones, while fattening up the sound and giving your bass a wider spacial field. The sound is more three dimensional and gives you a bit of tube compression and transformer saturation that helps the class A/B power amp feel a bit more "tube like", but the reality is that the only thing that performs like an all tube amp is, well... an all tube amp. I often hear guys refer to amps with tubes in the preamp as "tubey", but a HUGE part of why an amp like the SVT sounds like it does, is the tube power section.

To be fair, the preamp of an SVT has a significant impact on how the amp sounds as well. If you push an SVT output section with a GK 700RB preamp you will not get close to capturing the sound of a vintage SVT unless you are pushing the output section so hard that the distortion becomes the primary characteristic of the sound. IMHO an SVT 8Pro would sound a lot closer to a vintage SVT for all conditions except heavy overdrive of the output section.

In general I don't think it's possible to attribute tubiness solely or completely to either preamps or power amps. In a given design, the preamp and power amp each brings a set of specific qualities that are perceived as tubey. Some of the qualities are best achieved with the preamp and some are best achieved with the power amp. Some of the qualities can be captured by either or both. The specific mix and balance of tubey qualities in each preamp and power amp varies from one design to another.

If we are specifically talking overdrive, I prefer a condition where the preamp and power amp simultaneously transition to overdrive. IMHO stacking multiple low gain stages typically sounds richer and more harmonically complex than driving one stage or the other into hard overdrive. Unfortunately this condition is only achievable with an all tube design at a specific gain setting, so it becomes impossible to maintain the desired OD quality at different volume levels required for proper stage balance. I had hoped the Valkyrie circuit would overcome this shortfall.

If I understand what you have written, the Valkyrie tube circuit was intended to transition into soft overload in a way that is designed to capture some of the tubey characteristics that are uniquely produced by a tube output section. It doesn't sound like the circuit is intended to model a tube output section in a heavy overdrive condition. That is not to say the design could not have been optimized for this sort of operation.

Thanks for pushing the envelope with the Valkyrie design. I must say the amp peaked my interest and it would be on my bucket list if I were still collecting.

 

[BadExample]

Hmmm...do I believe you, a guy who never played a tube amp in his life, or Jerrold, the former chief engineer at Ampeg under SLM?

Jerrold by a landslide.

I'm going for seamonkey on this one, Jimmy. Why did all my seamonkeys die? Shattered my childhood. They were the only pets I was allowed.

 

[ThisBass]

If I recall correct, in SVT power amp overdrive and overdrive of the input diodes takes place about simultaneously. When power amp begins to clipping distort the diodes will also clipping limit the input signal, which - as mentioned - prevents harder overload of the output section.

The diodes are there.

But they don't cause distortion, they are really there to prevent excess overdriving. The amp will distort before the diodes conduct.

What they do is to smooth out the distortion somewhat.

The Trace Elliot V4, V6, and V8 all have clipper circuits after their master volume control.

For the TE V-type amplifiers there is little to no effect to the total sound cause the amplifier output saturates way before the clipper diodes reach saturation.

For the SVT the clipper diodes start to "smooth" saturate at about 200 Watt output power yet the power amp does reach full saturation before the clipper diodes can fully saturate.

 

[paulraphael]

I have a fully solid state Aguilar Tonehammer that I would go along with sounding "tubey". I also have an all-tube purist Mesa Buster that could 'fool' many as a transistor amp.

My experience as well. My Tone Hammer sounds more "tubey" ... in the most standard ways ... than my Demeter VTBP 201, which has 4 triode stages and no solid state circuitry in the signal path. This isn't a gimmicky tube preamp. Jim Demeter just chose to design a tube circuit that doesn't emphasize qualities we simpletons like to associate with tubes.

The Aguilarians, on the other hand, designed a solid state preamp to emphasize just those qualities.

I think designers today know their trade well enough to design a circuit that sounds any way they want, with either sand or glass. The choice comes down to their personal preferences, their geeky proclivities, and probably also their marketing departments.

FWIW, I've also owned tube preamps where the tube was 100% gimmick, as far as I could tell. You could switch between a modern Chinese tube that would be bright and brittle sounding in a serious tube preamp, and a vintage Mullard that would sound warm and wooley, and you couldn't tell the difference even if you sprained your ears trying.

[The Demeter let you hear all the differences between tubes, tonewise. But if you tried to overdrive it, it would sound as crappy as an overdriven transistor boom box].

 

[Wasnex]

For the TE V-type amplifiers there is little to no effect to the total sound cause the amplifier output saturates way before the clipper diodes reach saturation.

For the SVT the the cliper dopdes start to "smooth" saturate at about 200 Watt output power yet the ouput can reach full saturation before the clipper diodes can fully saturate.

The TE clipper diodes are arranged in series with a 100K resistor so it's not a surprise that they have less impact than the SVT diodes which are connected directly between the signal and ground.

I am surprised the SVT diodes begin to impact the signal at 200 watts. The circuit is: Input jack, series 1K resistor, clipper circuit, tube control grid. Per the SVT service manual the output section will reach full power with 0.25 volts at the input jack. The 0.25 volts will be dropped a bit by the 1K resistor before it hits the clipper circuit and the tube's control grid. This means the signal at the clipper will be less than 0.25 volts when the amp is producing full power. My understanding is the diodes are silicon 1N456 and will transition quickly into forward bias at around 0.65 volts and hard limit the signal to 1V.

Another part of the puzzle would be the bias voltage that develops at the control grid of the tube. This could conceivably shift the DC offset at the clipper up enough for it to be activated by the input signal at 200 watts.

 

[ThisBass]

The TE clipper diodes are arranged in series with a 100K resistor so it's not a surprise that they have less impact than the SVT diodes which are connected directly between the signal and ground.

I am surprised the SVT diodes begin to impact the signal at 200 watts. The circuit is: Input jack, series 1K resistor, clipper circuit, tube control grid. Per the SVT service manual the output section will reach full power with 0.25 volts at the input jack. The 0.25 volts will be dropped a bit by the 1K resistor before it hits the clipper circuit and the tube's control grid. This means the signal at the clipper will be less than 0.25 volts when the amp is producing full power. My understanding is the diodes are silicon 1N456 and will transition quickly into forward bias at around 0.65 volts and hard limit the signal to 1V.

Another part of the puzzle would be the bias voltage that develops at control grid of the tube. This could conceivably shift the DC offset at the clipper up enough for it to be activated by the input signal at 200 watts.

The schematic is (likely) a little bit misleading cause the vintage SVT needs little more voltage than 0.257 Vrms to push 300 Watt. May be Ampeg did not revised the rms numbers (the voltage numbers in brackets) with every revision of the amplifier.

The schematic reads 0.257 Vrms for 34.6 Vrms at the output to push 300 Watt into 4 Ohm load.
The schematic also notes full ouput power of the amplifier is 330 Watt.
At the onset of clipping my 70's SVT measures about 360 Watt ouput power into 4 Ohm (resistive) load.
So the input level for 360 Watt equals 0.282 Vrms.
Peak voltage then equals 0.398 Volt which was just enough diode (foreward) voltage to notice some impact to the signal on the grid of the PI.

Either way at least my "real world" 70's SVT demands for some little more of level at the input to push "full" power at the output. No, the amplifier don't sucks on any problems.

edit,
of course sometimes in the past I had this idea to remove the clipper diodes to check out the sound without the diodes. At the other hand as long as it sounds good there is no reason to touch a design. So I left the amp alone in this regard

 

[agedhorse]

If you calculate the voltage gain of the power amp, the diodes have a very small impact until the power amp is several dB into saturation. The variance is that the voltage gain is not exact, there is variation based on the gain of the tubes (as used in that circuit topology), so I would expect a 20% variation on the threshold. IMO, the main purpose is to minimize the deep saturation of intermediate elements of the power amp, similar to how a Baker's clamp operates. IIRC, this is an UL output stage, so entering and exiting clipping gracefully is an important feature.

This concept has been used in many designs for decades. I have seen it used in cinema power amps, servo amps, SelSyn amps (a form of positioning servo), and other places, and it's use is not limited to tube amps.

 

[Wasnex]

The schematic is (likely) a little bit misleading cause the vintage SVT needs little more voltage than 0.257 Vrms to push 300 Watt. May be Ampeg did not revised the rms numbers (the voltage numbers in brackets) with every revision of the amplifier.

Awesome! If I read your post correctly, your real world readings confirm that the clipper is set up to activate at approximately full output.

FYI. The parts list for the service manual I referenced is dated 5/1/77.

 

[agedhorse]

Awesome! If I read your post correctly, your real world readings confirm that the clipper is set up to activate at approximately full output.

FYI. The parts list for the service manual I referenced is dated 5/1/77.

Correct, plus you need to use peak voltages in the calculations.

 

[Wasnex]

Correct, plus you need to use peak voltages in the calculations.

This is why @ThisBass wrote, "0.282 Vrms. Peak voltage then equals 0.398"

In other words 0.283Vrms x 1.414 = 0.398Vpeak

 

[xk49w]

There is nothing very mysterious about these diodes. When the output stage is overdriven, the output feedback signal no longer corresponds to the input signal. The feedback difference amp will then drive the output stage even harder to try to make the feedback signal match the input signal. In the case of the SVT, the diodes prevent the input signal from achieving a level that would drive the output stage into bizarre levels of non-linearity. Result? It was thought to sound better. I might even agree.

 

[seamonkey]

The schematic is (likely) a little bit misleading cause the vintage SVT needs little more voltage than 0.257 Vrms to push 300 Watt. May be Ampeg did not revised the rms numbers (the voltage numbers in brackets) with every revision of the amplifier.

The schematic reads 0.257 Vrms for 34.6 Vrms at the output to push 300 Watt into 4 Ohm load.
The schematic also notes full ouput power of the amplifier is 330 Watt.
At the onset of clipping my 70's SVT measures about 360 Watt ouput power into 4 Ohm (resistive) load.
So the input level for 360 Watt equals 0.282 Vrms.
Peak voltage then equals 0.398 Volt which was just enough diode (foreward) voltage to notice some impact to the signal on the grid of the PI.

Either way at least my "real world" 70's SVT demands for some little more of level at the input to push "full" power at the output. No, the amplifier don't sucks on any problems.

edit,
of course sometimes in the past I had this idea to remove the clipper diodes to check out the sound without the diodes. At the other hand as long as it sounds good there is no reason to touch a design. So I left the amp alone in this regard

Interesting information. I appreciate you posting.

Even if the amp perhaps reaches distortion before diodes conduct, Nothing really stops the pre-amp from driving more voltage to the diodes beyond that point. Amp distortion plus diode clipping distortion is yet more distortion. If it's like most amps there is several more volts from the pre-amp headroom after driving the amp to full power. If this were some kind of protection circuit then Ampeg knows this. What designer wouldn't put in a couple of extra volts of headroom out of the pre-amp. Effects loops, driving other amps, it'd be smart to add some headroom.

Other threads mention the change in sound when the diodes are pulled. Many who seem to have equipment to physically measure the amp seem to have done so.

The IK Multimedia Certified amp models in Ampeg SVX are modeled down to the component level. They are accurate representations of the circuitry. Read the blurb on their site of finding a mistake in the amplifier schematic that the Ampeg Engineers confirmed. This was fixed in SVX before it shipped.
Ik states:

So accurate it even shut down
Actually, this is true. Funny story… during the first phase of testing with the Engineers at Ampeg, one particular software model was being driven so hard that it "shut down." Completely stopped working. Nada, nothing. Oh crap — we thought we were in trouble. But the lead engineer at Ampeg approved the model on the spot. You see, there was a slight design flaw in the actual amplifier being modeled that would cause a power fault shutdown if the amp was driven too hard for too long. The software model was in fact so accurate that it even replicated this design flaw… that's how accurate it is. If you think about it, it's the software engineer's equivalent of the "no brown M&Ms" rider clause in a performance contract — so it looks like we're playing tonight. But don't worry. We took took the "shutdown" feature out of our models so you can rock all night. Really.​

When they modeled to component level, the model followed the schematic. Down to the diodes. There is an extra critical listening phase after the circuit modeling, and it passed Ampeg engineers certification there also. Ampeg SVX 2 has even more models.

If anybody wants to know if diode clipping is for them it is easy enough for anybody who can't solder to add a Rat Tail cable to their bass or effects loop to add some smooth soft diode clipping to any brand amps.

 

[bklee]

While technically correct based on the notion that one component uses tubes and the other uses transistors, I think you're the only person I've ever seen refer to them as anything but tube amps since the power section was tube.


Which is why I referred to them as rare birds. The only other one I can think of were those Crank guitar amps that were popular in metal circles in the mid-00s.

Peavey Vintage uses SS pre-amp and tube output stage. Interesting complex sound profile (heavy midrange but not honky or clanky). I had the 6x10 converted to a bass amp by some having some electrical work done and changing the speakers and enclosing the back.

 

[agedhorse]

There is nothing very mysterious about these diodes. When the output stage is overdriven, the output feedback signal no longer corresponds to the input signal. The feedback difference amp will then drive the output stage even harder to try to make the feedback signal match the input signal. In the case of the SVT, the diodes prevent the input signal from achieving a level that would drive the output stage into bizarre levels of non-linearity. Result? It was thought to sound better. I might even agree.

The diodes cause a more compliant transition into non-linearity. This tends to be even more important with UL topologies, but overall it generally improves the performance and feel, helping prevent the amp's response from becoming "sticky" as the feedback tries to track the input.

 

[agedhorse]

This is why @ThisBass wrote, "0.282 Vrms. Peak voltage then equals 0.398"

In other words 0.283Vrms x 1.414 = 0.398Vpeak

I missed that. Take .398 x 1.2 = ~.5V which is usually into the beginning of knee territory.

 

[catcauphonic]

I love my 2 Genz hybrids & my Ampeg V4b! Streamliner comes close (enough) to the V4b, & the Shuttle can almost cop the SL, & go where the others can't.

Shuttle gets most of the gigs lately, as it bites and cuts thru 2 guitars better live for me