High OR low plate voltage question

[SILVERGUN]

Can anyone give me a simple explanation of what the advantages/disadvantages are in using a higher plate voltage (with lower current) as opposed to a lower plate voltage (with higher current) in power tubes?......if it would equal the same dissapation, whats the difference?

Is there a noticeable sonic difference?,,,,,,i.e.--why do so many amps have the pt plates running so high?

Sorry if you've answered this question a thousand times,,,,but I'm really just looking for the Cliff's notes 

[John]

why do so many amps have the pt plates running so high?

My opinion is because the transformers need to run at a higher voltage to supply the required current. I only say this after searching for something like a 250-0-250 that supplies 150mA. This is only a guess! I think PRR had said you can make a lower voltage/higher current tranny, but it would cost more? Again, just a guess.

Regarding your first question, 65Amps came out with a low voltage/high current design a little while back. I think most guys hear agreed that it sounded good, just like a ton of other amps do. 

I think one of the questions that gets asked most (and one I've googled most) is "how can I lower my B+?".

[SILVERGUN]

Thanks John,
Yeah,, I remember watching that video from 65amps, and still questioning the purpose, because Dan was a little vague about the reasoning and the result of that design concept......(like most, I assume he's protecting his design, by not just coming out and saying what it accomplishes)

From what I'm gathering,,,the rule of thumb regarding the answer I'm looking for is,,,,higher plate voltage is used to provide more clean headroom,,,,whereas lower plate voltage (w/ higher current) is more likely to be used to push the tube into distortion earlier

Limiting the plate voltage limits the total plate dissipation, which in turn limits the output of the tube, which is contradictory to most traditional designs which are trying to maximize output per tube (and that's why it's not as common)...........so if it doesn't matter to me that my EL34 will only be putting out 5 watts in class A (or if that's what I think I want), I can lower the voltage AND lower the current to get that result

My question is more- "what does that mean to my ears?"......will certain tube types sound better running in one condition over the other
Does a hot-biased 6V6 sound as good or better than a cold-biased EL34 (putting out similar wattages) in an overdriven SE amp?

[SILVERGUN]

And while I'm throwing hypothetical questions around....

Wouldn't it be advantageous to the output tube's life cycle to run it way under max specs?.......why do most people design around max specs...is it for volume? 

[DummyLoad]

why do most people design around max specs...is it for volume?

usually for headroom and tone quality. as general rule, tubes sound better when you're "kicking the snot of them..." so to speak.

you trade tube life for tone. that's what fender and marshall did/do. others play follow the leaders. mostly, we do so as well...

--pete 

[Ed_Chambley]

SG, check out Little Walter amps.  He does exactly what you are talking about and also uses ptp wiring.  Ben Fargen is doing it with a pair of EL34 in his new SE, 2 tube EL34 with DC heaters and Marshall tonestack.

If you have a variac, you can drop the supply voltage to give you an idea.  It is not exactly the same, but it is similar.

[phsyconoodler]

I've had a couple of 65 amps Soho's come in for retubing and they don't sound all that good to me.In fact my clones sound better to me and a few friends that have played them.Not really that impressed by them overall.
  The boost is nice but even the point to point approach on terminal strips leaves a lot to be desired.Likely done that wat to make it difficult for people to copy,but frankly I would not want to copy it.

 i have built both high voltage/low current and low voltage/high current EL84 amps and they both have some nice things going on. the low voltage amps were spongy and thick,breaking up early and having a nice transition from clean to dirt right at your fingertips,just by using string dynamics.
  The high voltage ones I've built were way punchier and had a nice attack and were more present.Also very nice.
I like both as much as the other,they are simply different.
  I use the Sluckey 1/4 power switch on both versions to add/subtract that 'feel' that both modes give you.I especially like the lo-fi sound and feel of the lo power switch on.And of course the power of the amps in full power mode.
   It's all about your playing style I like to think.I have to play an amp for a while to get used to the way it responds best before I evaluate it.A softer approach or a more percussive approach can bring out the best in a number of different amps.
  all I know is is that some amps just don't have what it takes.The Soho was one such amp that really didn't quite have what I need to make me all excited about an amp.
  

[phsyconoodler]

Also,Allesandro has made amps in the past that were true class A with a quad of EL34's run at about 250v on the plates.I have not heard one but I have a felling they sound great.
  I did build a princeton circuit in on combo amp with a pair of 6
V6's running at 285v and the overdrive was spectacular sounding.It had a sweet tone to it that was hard to describe.Not a lot of volume but sweet,sweet overdriven tone.
  Maybe 10 watts?
I had a power transformer I had laying around that I used just as an experiment. It was very nice overall.Just had very little headroom.Rowing the guitar volume was the key to using this little bugger.

 i want to build another low voltage amp with a pair of 6550's or EL34's.I bet it would be sweet too.
Go for it! Experiment. Hammond has a wide variety of transformers to choose from to make it happen.And they will custom make one for you too at a very reasonable price.

[SILVERGUN]

as general rule, tubes sound better when you're "kicking the snot of them..." so to speak.

Now thats what I'm talking about....great "rule of thumb"  
... and I realize that this is all subjective....but I do appreciate opinions

Coming from welding machines to guitar amps is a bigger stretch than I initially thought,,,,current is much more important to holding an arc than pushing a speaker

As always,,,thanks ED, I hadn't seen those amps before.......that's where I'm going with this----SE EL34 @ low plate V with high gain preamp

And thank you phsyco,,,I was typing while you were typing...

[HotBluePlates]

... the advantages/disadvantages are in using a higher plate voltage (with lower current) as opposed to a lower plate voltage (with higher current) in power tubes?...

That isn't the issue that drove the old amps to move to higher voltage. The issue was power output.

If you run class A, the tubes dissipate the maximum power at idle. Tube current must flow 100% of the time (definition of class A), so the tubes idle at a given current and only have a peak at double the idle current (and a negative peak at or near zero current). In this scenario, plate voltage cannot rise above a certain amount or the tubes overheat.

If you run class AB, the tubes can have a peak current more than double idle current (maybe 3-4 times idle current). That's because during part of the cycle, the tube is shut off with no current flowing, so average dissipation during the entire cycle is reduced. As a result, plate voltage can be made higher than allowable in class A, which further increases power output.

Along the way, OT primary impedance is lowered because that allows more current to flow through the tubes during peaks (remember the higher peak current in AB?).

For the most part, the old amps' preamp voltage was just a byproduct of how the output stage was run. The class of operation of the output stage partly dictates the bias voltage, which also dictates the needed drive voltage from the preamp/phase inverter, which then dictates the design of the phase inverter in conjunction with the available supply voltage.

Everything else is a byproduct of these choices (intentional or not). Along the way, you might notice a little different quality to the distortion of the output stage. Even harmonics should be cancelled by push-pull operation (unless the OT is grossly unbalanced or the output tubes are extremely mismatched), so you'll really just be hearing differences in the amount of odd harmonic distortion in the output tubes. More odd harmonic generally equals a crisper, brighter sound.

Overall, 3rd harmonic distortion tends to increase with increase load impedance. Assuming you keep the same power output, you'd have to increase load impedance to offset an increase of supply voltage. But this isn't what the classic amps did (they switched to class AB). So there may be increased distortion and/or increased odd harmonic with increasing plate/screen voltage. Unfortunately, the old data sheets generally don't show this trend, but keep a fixed supply voltage as load impedance is changed.

[kagliostro]

Very interesting tread

I was just wondering around this matter but for one other reason and with one other purpose

may be High OR low plate voltage & current question

Is possible to use lower <Va - V2g - Bias setting> to tame a tube as to use it in a smaller amp than it will be usually ?

In other worlds, not to lower voltage and increase current but, instead, lower voltage and current

as to build a small amp (with the consequence of low cost OT) using "big" tubes

Will it work in terms of sound/tone result ? and what will happen to the OT primary impedance ?

Using a VVR the power of the amp is decreased and the OT primary impedance is maintained at the level of the standard,

deliberately using low voltages and currents can't be considered the same thing ?

K

[HotBluePlates]

Is possible to use lower <Va - V2g - Bias setting> to tame a tube as to use it in a smaller amp than it will be usually ?

In other worlds, not to lower voltage and increase current but, instead, lower voltage and current

as to build a small amp (with the consequence of low cost OT) using "big" tubes

You can always use less voltage or current than typical for any tube. The exceptions to this are found in radio transmitter practice, not guitar amps.

... what will happen to the OT primary impedance ?

If you're designing from scratch, then nothing "happens to" the OT impedance. You select an impedance based on the chosen supply voltage and desired output power. Normally, you then select an output tube capable of delivering the implied current the other choices lead you to. But if your desired power output is way below what is normally demanded of the tube (1w from a KT88), then the OT primary impedance doesn't really matter.

Using a VVR the power of the amp is decreased and the OT primary impedance is maintained at the level of the standard,

deliberately using low voltages and currents can't be considered the same thing ?

When you use a VVR and reduce the plate/screen voltage, then tube current is also reduced. So it is exactly the same thing as what you first asked. Or if I'm misunderstanding you, then yes, using a VVR is the same as reducing the voltage and current while keeping the OT impedance the same.

[kagliostro]

The exceptions to this are found in radio transmitter practice

Interested ( ) clarification question

you mean in radio transmitter units, not inherent with the use of radio transmitter tubes in guitar amps or not ?
(at the moment I've a pair of QQE04/20 tubes laying around ..............  )

I wasn't considering a KT88 at 1W output, something a bit more high in W, say near 1/3 - 1/2 of the tube ability

(my QQE04/20 within 15W instead of 26W as to use an Hammond 125E or better 10W as to use a 125D)

When you use a VVR and reduce the plate/screen voltage, then tube current is also reduced. So it is exactly the same thing as what you first asked. Or if I'm misunderstanding you, then yes, using a VVR is the same as reducing the voltage and current while keeping the OT impedance the same.

So I'm a bit confused now, I must select a new OT primary impedance in a <lowered Voltage and Current project> or I can use the standard primary impedance indicated in the datasheet for standard voltages and current ?

K

[alerich]

Good question and great thread topic. I have pondered over this topic and the whole 70% dissipation quite a bit lately. I have a Sovtek Mig60 head. It runs a pair of EL34 tubes. The amp has 4 ohm sense resistors from pins 1/8 to ground. The plate voltage is 640VDC. It has a separate screen supply of about 350VDC. The schematic directs you to set the bias so that you read 80mv DC across the sense resistors. That calculates to 20ma through each tube and at 640VDC on the plates it comes in at a conventionally paltry 12.8 watts per tube at idle. Based on everything I have learned about tube amps and bias this thing is overbiased if you follow those instructions.

The kicker is - it sounds amazing. The amp was modded for more preamp gain by Scott Splawn and further tweaked by me but the stock Mig60 is a pretty close rendition of the JCM 800. No mods have been done after the master volume. The PI and power amp are completely stock. I know the added gain will enhance the tone for my playing style but this amp is the best Marshall type clone I have ever heard. I'll never sell it. What puzzles me is how good it sounds when you factor in all of the math in the previous paragraph. Admittedly the amp is not quite as loud at my JCM 800 clone at similar master volume dial settings but that's part of the charm of this amp. It delivers the goods nonetheless.

It's one thing to achieve 70% dissipation with either higher plate voltage/lower plate current or lower plate voltage/higher plate current as this thread explores but this particular amp has me questioning the whole 70% rule altogether.

[John]

As I understand it, for fixed bias you shoot for 70% at idle, so that at full roar you're not red plating. With cathode bias, it sort of takes care of itself. It's also why changing the cathode resistor doesn't change the bias as much as you think, because with a larger value K resistor, your plate voltage then rises.

I forget exactly the numbers, but I went from a 150R on a pair of EL84s to something like 300, but after doing the math they were still dissipating almost the same wattage... around 14/15 I think? The only way to get them lower was to lower the B+, which I did with tube rectifier.

And most likely, I haven't explained it correctly in my first paragraph, but it's the way I keep it straight in my head, so good luck with that. 

[HotBluePlates]

The exceptions to this are found in radio transmitter practice

Interested ( ) clarification question

... the use of radio transmitter tubes in guitar amps or not ?

I mean there are tubes generally meant for use in RF transmitters which have graphite plates instead of getters. In these tubes, you have to red-plate them as that is how these tubes absorb stray gas molecules. So if you reduce the voltage/current in these, you might have issues with gas in the tubes.

But that doesn't apply to anything you'd reasonably use in a guitar amp.

So I'm a bit confused now, I must select a new OT primary impedance in a <lowered Voltage and Current project> or I can use the standard primary impedance indicated in the datasheet for standard voltages and current ?

The OT impedance is all about obtaining a desired maximum clean power output.

If you need a certain amount of clean power output, then there is a "right answer" for OT impedance and supply voltage given that there are only so many different transformers to pick from. So you pick the OT for the power you need and supply voltage you have, then choose output tubes which can deliver the current needed.

If maximum clean power is not important, you can pretty much use any primary impedance with any tube while realizing that power output will either be a little or a lot less, and distortion will be a little or a lot more.

I wasn't considering a KT88 at 1W output, something a bit more high in W, say near 1/3 - 1/2 of the tube ability

Then do what you want. If you're not trying to wring every last watt out of the circuit, OT impedance isn't critical.

[jazbo8]

Here is a graphical look at output power & distortion vs. output load (for 6L6) in SE & PP operation., as the graph shows when the load is below the optimum, the 2nd harmonic dominates, and when the load is above the optimum, the 3rd harmonic dominates. The optimum load is defined as the load that delivers the maximum power output with minimum distortion (~3.5k in the graph).

[EDIT]
For SE Operation (6L6 - 250V Plate, 250V Screen)


For PP Operation (2x6L6 - 360V Plate, 270V Screen)

[kagliostro]

Many thanks HotBluePlates

I mean there are tubes generally meant for use in RF transmitters which have graphite plates instead of getters

Now I understand, I've seen those big tubes redplating in RF equipment

If you need a certain amount of clean power output, then there is a "right answer" for OT impedance and supply voltage given that there are only so many different transformers to pick from. So you pick the OT for the power you need and supply voltage you have, then choose output tubes which can deliver the current needed.

And if is correct to chose the tubes looking to the PT and OT, often we want to start from a given tube, and we base the choice of the transformers (PT & OT) using the data the tube requires we find in the datasheet

this is, as you say, for max performance, using all the tube ability to give power

if I remember well PRR told Voltage / current = impedance
(as example 400v / 0.04A = 10.000ohm impedance)

If maximum clean power is not important, you can pretty much use any primary impedance with any tube while realizing that power output will either be a little or a lot less, and distortion will be a little or a lot more.

So we can use an arbitrary impedance and we can expect a lower power and more distortion that I think will be due to the impedance mismatch from the tube and the OT

Having a given tube that can express his better performance at a particular voltage and current (one option for QQE04/20 say 425v | 0.026A | Bias -60v  = Wo 2 x 16W http://www.r-type.org/pdfs/qqe04-20.pdf see page #5) of course there is a better suited OT primary impedance for this voltage and current values

If you're not trying to wring every last watt out of the circuit, OT impedance isn't critical.

If I want to have only  Wo = 2 x 5W instead of 2 x 16W is correct to say I can:

lower voltage keeping current the same (Bias voltage used to control the current)

lower current keeping the Voltage the same (Bias voltage used to control the current)

lower Voltage and current (Bias voltage used to control the current)

At this point, if what I told is correct I've a voltage and a current to be used as to establish an OT impedance

the question, if possible to be presented, is:

is better to apply the previous formula Voltage / current = impedance to the voltage and current I've chose to use or it will be better to use the "standard" impedance given by the data on the datasheet for the max performance

If I've an existing Power Amp with a certain OT and I use a VVR to decrease power (assuming I decrease Va - Vg2 & Bias) what I know is that with the same OT primary impedance the reduced performance in Wo of the tubes will have the same tone with less headroom and the tube will act as a smaller tube (I can obtain distortion/compression from the tube at a lower level, as if I was using a 5W power tube instead of a 15W power tube)

So, hard question, is better to use the OT primary impedance given by the "max performance" in the datasheet or the OT primary impedance using the arbitrary chosen voltage and current ?

K

p.s.: Jazbo8, you posted when I was writing my replay, I've seen the plot, but this didn't help me to understand what is better to use to compute the OT impedance in this particular instance

[PRR]

> the plot, but this didn't help me to understand what is better to use to compute the OT impedance in this particular instance

That plot was taken for 250V 70mA. The dip in the THD curve is 3.5K load, which is near-max power with low THD. If we changed to 400V at the same dissipation we would have to reduce to 42mA current and now the near-max power with low THD would shift to about 9.5K load.

> Using a VVR the power of the amp is decreased and the OT primary impedance is maintained

AND the current is (usually) reduced. Either because lower volts with same cathode bias resistor makes less current, or because the fancy VVR scales the fix-bias voltage nearly the same as B+ is scaled. If both V and I are reduced, R (best load) is about the same.

> impedance mismatch from the tube and the OT

You never match the *tube*. Pentodes have two plate resistances, you only want to stay well away from either extreme.

You then "match" the V/I in the tube to the load. And for a given V your selection of I is limited by dissipation.

Windings 10 ohms to 1K are cheapest. Higher Z means more turns of fine wire. Fine wire costs more per pound for the cost of making it thin, then more time in the winder for more turns of more-delicate wire.

[jazbo8]

@kagliostro - You already got some good advises, but I think your last sentence summed it up well - "... or the OT primary impedance using the arbitrary chosen voltage and current". If you are not going for the maximum output power/minimum distortion but instead are interested in getting the best tone. Then basically anything goes as HPB said, however, it might be easier to start with commonly available OPT (perhaps higher than the optimum primary impedance if you are looking for more odd harmonics), then adjust the plate/screen voltages and bias accordingly.

But I think this sort of tweaking should still be done within some limits, e.g., you probably would not want to get 25-50% less power output by using a really mis-matched OPT/load - since that's just wasting tubes (they are not free!) Just my 2c...

[kagliostro]

Thanks PRR

I think I'm getting closer to understanding this whole thing

AND the current is (usually) reduced. Either because lower volts with same cathode bias resistor makes less current, or because the fancy VVR scales the fix-bias voltage nearly the same as B+ is scaled.

If I've an existing Power Amp with a certain OT and I use a VVR to decrease power (assuming I decrease Va - Vg2 & Bias)

As you say I was assuming that using a VVR Voltage and Current where both decreased

If both V and I are reduced, R (best load) is about the same.

this clarify me that I can use an arbitrary OT impedance, but I can safe use the OT primary impedance that is better suited for max power and if I want I can also play a bit around this impedance as to get a more easily available OT

...

Thanks Jazbo8

it might be easier to start with commonly available OPT

Yes, that is a good point

you probably would not want to get 25-50% less power output by using a really mis-matched OPT/load - since that's just wasting tubes (they are not free!)

not via an OPT/load mismatch (intended using datasheet voltage an current with an arbitrary chosen OT) but via a reasoned choice of Voltage - Current - OT

However most of the cost are not the tubes, but the transformers, and there are valves that for some reason may be available at low cost, as an example I got my QQE04/20 at less the half the price of a pair of EL95 tubes and I've some GU50 tubes I got really on the cheap

so using this tubes for small amp may be a way to use it with a financial commitment not very high, also just for building satisfaction

K

[SILVERGUN]

Thanks for carrying on without me here guys...
I had a vacation day yesterday and chose to not look at the computer
I always learn stuff from you guys when I go back and read the posts,,,,and I enjoy the fact that we dont hesitate to expand on a topic
Good stuff here 

I have always felt like voltages are just a guide, and a lot of guys spend way too much time trying to match the voltages that someone else was getting,,,and I understand that from a cloning perspective......but when you're building from scratch, AND don't care about sucking the maximum decible per tube,,,it changes the options and opens up the use of some different PT secondaries
And, as you guys have noted above, it opens up our ability to use different OTs than just what is "recommended"
Merlin's got a nice free page on his site that is in line with what you guys are saying:
http://www.valvewizard.co.uk/se.html

For me personally, I'm now looking at using more common 120 > 230 VAC xfmrs that I have more access to (on the cheap),,,,,,for certain situations.....and just using separate filament supplies