Power pentodes as preamp tubes?

[Austin]

I am thinking of building my amp with all power tubes from the input. Is there any reason I shoudn't? I have all these octal slots available and alot of power tubes at my disposal along with huge supply transformers.. Like 6g6g 6f6g etc, I could triode strap them still have plenty of gain I think. Like using the 6g6 as a pentode as input voltage amplifiers and using a 6f6 as a cathodyne phase inverter then using two more 6f6 as dc coupled cathode followers to drive two 6L6 as power stage. Makes sense to me but why doesn't everybody do it? I saw a data sheet listing the curves for ab2 operation when looking at the 6L6 data sheet last night.. I guess I will try it and see how well it works. Any thoughts anybody?

[Steve Conner]

Well, the classic tube amps were being manufactured and sold at a profit, they were subject to economics. You wanted the most gain per dollar in the preamp, and preamp tubes were designed to deliver that.

Now tube amps are an artform and you can do anything you want.

My Ninja Corvette essentially uses an EL84 as a preamp stage in high power mode.

[Austin]

Well, the classic tube amps were being manufactured and sold at a profit, they were subject to economics. You wanted the most gain per dollar in the preamp, and preamp tubes were designed to deliver that.

Now tube amps are an artform and you can do anything you want.

My Ninja Corvette essentially uses an EL84 as a preamp stage in high power mode.

Hi Steve, thanks for the reply. What do you mean exactly by "high power mode"? Pentode mode? I have this cool choke that was configured on the cathode of a 12ax7 with a bias tap. I was considering using it as a plate load for the 6g6 and using the bias tap as the screen supply, but what do you guys think?

[Gregg]

I've seen Hi Fi amps using EL84 and other power pentodes as preamp tubes so it's been done already. Whether they are suitable for guitar amps is another story.

[Austin]

I've seen Hi Fi amps using EL84 and other power pentodes as preamp tubes so it's been done already. Whether they are suitable for guitar amps is another story.

This is intended to be a clean P.A. system I can play mp3s and sing through with my gal.. Nobody has a tube pa system and I imagine tubes might sound sound nice with the vocals but we will see... I already have a guitar amp that I like.

[Steve Conner]

It's worse than pentode mode. The EL84 drives the bases of a couple of hefty power transistors through a Valve Jr. OT, which is used as an interstage transformer in this case. It makes about 40W and sounds remarkably good for what it is. I got the idea from RG's writings on the Thomas Vox amps. You can google "ninja corvette" for more details.

One problem I had is that the transistor output stage also amplifies the microphonics of the EL84. Preamp tubes are designed for low microphonics, power tubes aren't, as they are usually not operated with gain between them and the speaker.

I tried a few EL84s to select a quiet one, and also ended up with a silicone gasket from a coffee maker wrapped round the tube to damp it.

[Alan0354]

One question. What about the noise performance? I was just reading about the grid current( from RDH4 where there is current even at -ve bias), that would be a source of noise

In^2=2q(Ig)(BW) where In is the noise current, Ig is grid current and BW is the bandwidth.

I don't have the noise model of vacuum tubes, since the electrodes are all metal, I assume there is no significant resistance inside the tube to generate thermal noise. So my assumption is the grid current is the main source of noise refer to the input. Sounds like the power tubes has more grid current and has more shot noise. Question is would that be a problem. This is particular important if there is a large grid stopping resistor at the front end as was talked about in a post not too long ago in radio interference and how to stop it.

Also, the electrode capacitance and the inter electrode capacitance must be a lot higher for power tubes that might affect the performance.

[loudthud]

This is intended to be a clean P.A. system I can play mp3s and sing through with my gal.. Nobody has a tube pa system and I imagine tubes might sound sound nice with the vocals but we will see... I already have a guitar amp that I like.

Vocal harmonies through a tube PA sound awesome. I read that Elvis used five Dynaco Mark III's for a while in his PA.

[Austin]

One question. What about the noise performance? I was just reading about the grid current( from RDH4 where there is current even at -ve bias), that would be a source of noise

In^2=2q(Ig)(BW) where In is the noise current, Ig is grid current and BW is the bandwidth.

I don't have the noise model of vacuum tubes, since the electrodes are all metal, I assume there is no significant resistance inside the tube to generate thermal noise. So my assumption is the grid current is the main source of noise refer to the input. Sounds like the power tubes has more grid current and has more shot noise. Question is would that be a problem. This is particular important if there is a large grid stopping resistor at the front end as was talked about in a post not too long ago in radio interference and how to stop it.

Also, the electrode capacitance and the inter electrode capacitance must be a lot higher for power tubes that might affect the performance.

I was wondering about the capacitance too, I guess I just have to try it and see how it sounds. Shoudn't I get lots of voltage gain like 300 or so using a pentode? I could use one pentode as a ccs plate load for another pentode, and dc coupled cathode folowers using power tubes might sound pretty good.

[Alan0354]

I was wondering about the capacitance too, I guess I just have to try it and see how it sounds. Shoudn't I get lots of voltage gain like 300 or so using a pentode? I could use one pentode as a ccs plate load for another pentode, and dc coupled cathode folowers using power tubes might sound pretty good.

You are asking the wrong person here!!! I am new to the tube stuffs. I am going to try working on this a little. I look at 6L6GC, the gm is 4700umho vs 1600umho for 12AX7. So the gain is definitely higher. The plate resistance rp of 6L6GC is about 30K vs 62K of the 12AX7.

Gain of the stage u'=gm(rp//RL)

Where RL is the load resistance. So

Gain =u'= gm (rpXRL)/(rp+RL)

Using RL=100K

For 12AX7, (rpXRL)/(rp+RL)=37.5K

For 6L6GC, (rpXRL)/(rp+RL)=23K

So gain for 12AX7 is gm(rpXRL)/(rp+RL)=.0016X37500=60

Gain for 6L6GC is 0.0047X23000=108.1

From my calculation, the 6L6GC is about double the gain of 12AX7. Please few free to check and challenge my calculation, I don't know any of these 3 days ago!!! Just learnt it, fresh out of the oven!!!!

[Alan0354]

Still back to my original, how about the grid current. I have no data on this, this cause noise. Common sense tells me that the grid of a big pentodes is going to be a lot bigger, if there is any electrons from cathode or plate or others hitting the grid, you get more surface area and therefore more grid current. That will raise the noise. If you have a high gain design, that might give you a lot of hiss.

[Austin]

Still back to my original, how about the grid current. I have no data on this, this cause noise. Common sense tells me that the grid of a big pentodes is going to be a lot bigger, if there is any electrons from cathode or plate or others hitting the grid, you get more surface area and therefore more grid current. That will raise the noise. If you have a high gain design, that might give you a lot of hiss.

Why would grid current cause noise?

[Alan0354]

Why would grid current cause noise?

Because if you look at current at any given time, it is not smooth. Current is a flow of electrons, say you have average of 100 electrons per second passing through a junction A, it is not as if there is evenly one electron pass through A per 1/100 of a second. It is totally random. You can have two or three electrons bunch together and pass through at the same time, then there might be a dead spot for say 1/10 of a second, then 20 electrons passing in the next 1/10 of a second...............It is totally random. The only thing you can say is on "average" you have 100 electrons pass through junction A per second.

So by itself, it is not smooth, it is bumpy. So there is a noise involve in it as long as there is a current flow through. This is called shot noise ( vs thermal noise that only from resistance). and it is represented by a noise current In where:

In^2=2q(Ib)(BW)

Where Ib is the current flow ( grid current in this case), q is charge per electron which is 1.6X10EE-19 coulomb per electron. BW is the bandwidth of interest. In your case for guitar amp, BW=5KHz.

When you calculate noise, it is not at all obvious what is the cause of the noise problem. It all depend on the input impedance, the input current etc. That's the reason I kept challenging in the post about the grid stop resistor. Because until you have all the data, it is not obvious. You cannot lump everything into thermal noise generated by resistors.

In your case here, you still need to find the voltage noise of the tube. Just because I don't see where the resistance can be from a tube don't mean there is none. What do I know about tubes? That's the reason I asked more than once whether anyone has the noise model of the tubes.

Also, the calculation I gave on the gain was based on the gm given in the operating condition of the 6L6GC at the given plate current. At the low plate current that you are trying to run, is the gm still 4700? If it drops off, then you loss the advantage.

[Austin]

Well I didn't decide how much current to let the tubes run at but isn't it true that it will be more linear working into a higher resistance? I have no shortage of tube slots or tubes or power supply so gain is not an issue at all. Right now it has 1 6v6 1 6g6g 3 6f6g and two 6l6gc in there per channel and I was going to use the 6v6 as a plate load ccs and use the 6g6 as the preamp and use one 6f6 as cathodyne phase inverter and two 6f6 as driver tubes into tthe 6L6 power stage. Everything triode strapped for now just to get it working until I make a regulated screen supply. I will try and post some pics tomorrow.

[J M Fahey]

Not to rain on your parade, but 6L6 (or any other power pentode) transconductance at, say, 1mA plate current (you are feeding it through 100K after all) is WAY lower than at usual power tube current levels.
You want a very rough, no graphs voltage gain estimation for a 6L6 ? (EL34/84 can be calculated in a similar way):
Just consider this:
Idle: plate voltage=V+ , say, 430V ; grid voltage (bias)= -52V
Saturating: Plate, say, 60V ; grid: 0V .
(Delta means "variation")
Plate variation (delta voltage)= 430-60=370V
Grid variation (delta voltage) = 52-0=52V
Effective gain: 370/52=7 ..... incredible, isn't it?

So, what about that huge transconductance, is it a lie? Not at all, but is is the name to describe a *type* of Physical phenomenon: that when voltage changes, current changes.
In this case , it does not mean huge voltage gain but huge current swing for a relatively small (grid) voltage swing.

So, in a nutshell, use small triodes (or small pentodes if you wish) for gain and big pentodes for power.
Can't go wrong.

Hasn't been wrong for the last 90 years, go figure

[Alan0354]

Not to rain on your parade, but 6L6 (or any other power pentode) transconductance at, say, 1mA plate current (you are feeding it through 100K after all) is WAY lower than at usual power tube current levels.
You want a very rough, no graphs voltage gain estimation for a 6L6 ? (EL34/84 can be calculated in a similar way):
Just consider this:
Idle: plate voltage=V+ , say, 430V ; grid voltage (bias)= -52V
Saturating: Plate, say, 60V ; grid: 0V .
(Delta means "variation")
Plate variation (delta voltage)= 430-60=370V
Grid variation (delta voltage) = 52-0=52V
Effective gain: 370/52=7 ..... incredible, isn't it?

So, what about that huge transconductance, is it a lie? Not at all, but is is the name to describe a *type* of Physical phenomenon: that when voltage changes, current changes.
In this case , it does not mean huge voltage gain but huge current swing for a relatively small (grid) voltage swing.

So, in a nutshell, use small triodes (or small pentodes if you wish) for gain and big pentodes for power.
Can't go wrong.

Hasn't been wrong for the last 90 years, go figure

That's what I was worry about, that's why I did mention whether you get any gain in this low current level as all the spec are for running at optimal current and load.

[Merlinb]

Still back to my original, how about the grid current. I have no data on this, this cause noise. Common sense tells me that the grid of a big pentodes is going to be a lot bigger, if there is any electrons from cathode or plate or others hitting the grid, you get more surface area and therefore more grid current. That will raise the noise. If you have a high gain design, that might give you a lot of hiss.

Grid current is negligible as long as the brid is biased negative, which is normally the case for audio amps. You can safely ignore noise due to grid current- all that matters is shot and flicker noise in the anode current.

[Alan0354]

Grid current is negligible as long as the brid is biased negative, which is normally the case for audio amps. You can safely ignore noise due to grid current- all that matters is shot and flicker noise in the anode current.

It all depend, the spec of the grid 1 impedance of 100K in fixed bias and 500K in cathode bias of 6L6GC implies you cannot count on it's like an open circuit. Also it is explained in RDH4 about the electrons hitting the grid from cathode and gas etc. That's current. If people worry about thermal noise of a 100K grid stopping resistor, you better look into this. 1uA grid current is not small by any stretch. We worry about bias current of 1nA or smaller in low noise designs.

[Alan0354]

Grid current is negligible as long as the brid is biased negative, which is normally the case for audio amps. You can safely ignore noise due to grid current- all that matters is shot and flicker noise in the anode current.

It all depend, the spec of the grid 1 impedance of 100K in fixed bias and 500K in cathode bias of 6L6GC implies you cannot count on it's like an open circuit. Also it is explained in RDH4 about the electrons hitting the grid from cathode and gas etc. That's current. If people worry about thermal noise of a 100K grid stopping resistor, you better look into this. 1uA grid current is not small by any stretch. We worry about bias current of 1nA or smaller in low noise designs. Better get the grid current spec first.

[Austin]

So I should keep something like 30 ma through the tube to keep the transconductance up then?

Also the 100k grid stopping resistors are alot bigger than I intend to use. Why shouldn't I use smaller less noisy values so a bit of grid current wouldn't be a problem? I am not sure but it makes sense that the grid of power tube has some grid current even while at less potential than the cathode if even the 12ax7 does.

[Austin]

[Steve Conner]

Any parameter not specified on the datasheet will do whatever it takes to ruin your day. That's one of the corollaries of Murphy's Law of electronic design.

It follows that power tubes used as preamp tubes are likely to be somewhat noisy, probably for reasons similar to what Alan0354 set out. Noise figure wasn't on the datasheet, so the factory wouldn't have had quality control in place for it.

I say this with a hint of sadness as I've been looking for discrete transistors specified for low noise in the audio band, but I can't find any in current production, except for the high priced dual arrays, MAT01 and so on.

[Austin]

It won't hurt to try and see what happens I spose.. I'm cobbling something up right now....

[Steve Conner]

Whoa, I can feel the vintage mojo right there!

[Alan0354]

Any parameter not specified on the datasheet will do whatever it takes to ruin your day. That's one of the corollaries of Murphy's Law of electronic design.

It follows that power tubes used as preamp tubes are likely to be somewhat noisy, probably for reasons similar to what Alan0354 set out. Noise figure wasn't on the datasheet, so the factory wouldn't have had quality control in place for it.

I say this with a hint of sadness as I've been looking for discrete transistors specified for low noise in the audio band, but I can't find any in current production, except for the high priced dual arrays, MAT01 and so on.

Yes, I gone through looking, the price of MAT01 is shocking to put it politely. I did a little search before, for low noise, high beta and low cost, MPSA18 for NPN and 2N5087 for PNP is hard to beat.

[Alan0354]

Any parameter not specified on the datasheet will do whatever it takes to ruin your day. That's one of the corollaries of Murphy's Law of electronic design.

It follows that power tubes used as preamp tubes are likely to be somewhat noisy, probably for reasons similar to what Alan0354 set out. Noise figure wasn't on the datasheet, so the factory wouldn't have had quality control in place for it.

I say this with a hint of sadness as I've been looking for discrete transistors specified for low noise in the audio band, but I can't find any in current production, except for the high priced dual arrays, MAT01 and so on.

Yes, I gone through looking, the price of MAT01 is shocking to put it politely. I did a little search before, for low noise, high beta and low cost, MPSA18 for NPN and 2N5087 for PNP is hard to beat. If you have better ones, let me know.

You need high beta to lower the base current Ib.....which lower the shot noise.

Then the FETs which has the advantage in noise performance in RF, sucks in audio frequency as the 1/f noise is going to swamp all the other sources. Still the good old tried and true BJT.

In fact, I am waiting for 4 of my pcb coming back, those are for pedals that I have been working on. I came here to talk about tubes few months ago and totally dropped it. It was only because last month my Marshall JCM900 blown!! $#X&!!! I have no choice but to talk out my old beat up Fender Bassman 100 and rip out the whole front end and designed my own clean and dirty channel in. That's got me started again on the tube stuff. I like the result which promp me to start looking into the power amp. So far all I did was lower the grid voltage to bring the power tube more towards class A by drawing more quiescent current. Then I changed out the big OT with a Magnetic Components 40W OT for a pair fo 6L6GC instead of 4.

[Alan0354]

This is a good explanation of grid current:

Grid current characteristics

[Austin]

This is a good explanation of grid current:

Grid current characteristics

Where do you guys think grid current goes when it leaves the driver and goes into the grid of the next tube?

[exclamationmark]

I'm fairly sure it just sums with the plate current and can be measured through the cathode. I recall measuring the characteristics of some 6AS7G tubes and wondering why the voltage was different (not a huge difference but still significant) across each respective 1 ohm plate and cathode resistor when the grid was driven hugely positive. The DC cathode follower arrangement found in a ton of amps also seems to support this, as the cathode follower 'steals' current from the previous stage to achieve proper bias.

[Merlinb]

MPSA18 for NPN and 2N5087 for PNP is hard to beat. If you have better ones, let me know.

BJT noise at LF is also dependent on base spreading resistance. Devices rated for high collector currents usually have smaller spreading resistance. Obviously there is a balance to be struck between minimising base current (implying high hFE) and in minimising spreading resistance (implying a high current rating and therefore probably low hFE). The ZTX653/753 and BC337/327 represent good examples, as they both have low spreading resistance and healthy hFE >200.

[Steve Conner]

Good idea Merlin. The transistors have to be surface mount, but we already use the FZT751 in other places, a surface mount version of the ZTX751. I'll try those out for noise.

It doesn't really address my original gripe, though: nobody makes a transistor actually specified for low-noise audio amplification any more. Diodes Inc. could decide to add some excess noise mechanism to the FZT751 tomorrow, it would still pass their quality control, and there would be nothing I could do about it. Not very convenient for mass production.

I wonder what makers of mixers and mic pres do nowadays.

[R.G.]

I wonder what makers of mixers and mic pres do nowadays.

In the popular jargon of the southern USA, they go fishin' and take pot luck.

The thing that helps is that on average, transistor processing is much better than it was in the 70s, which was pretty much the last hurrah of the discrete transistor, and the poor processing that made the worst of noise issues in discretes has been washed out with the semiconductor tide that's lifted all boats.

Still not a pointed design for low noise, but better in general.

[Merlinb]

I wonder what makers of mixers and mic pres do nowadays.

Douglas Self recommends the 2N4403, with rbb at 40 ohms.
This guy apparently measured the BC337 at 30 ohms http://www.janascard.cz/PDF/Design%2...amplifiers.pdf

Modern opamps are getting pretty darn good though, with noise specs that are approaching those of discrete transistors. Eventually there may be nothing to choose between them. (I recently picked up some free samples of the OP07D for a high source impedance application. It has some impressive specs without being absurdly expensive.)

[Alan0354]

BJT noise at LF is also dependent on base spreading resistance. Devices rated for high collector currents usually have smaller spreading resistance. Obviously there is a balance to be struck between minimising base current (implying high hFE) and in minimising spreading resistance (implying a high current rating and therefore probably low hFE). The ZTX653/753 and BC337/327 represent good examples, as they both have low spreading resistance and healthy hFE >200.

You know the two you point out are medium power transistor, I went and search the data sheets in digikey and there is no noise information on either one. They are not meant to be low noise front end application. Here is the data sheet of MPSA18:

http://www.onsemi.com/pub_link/Collateral/MPSA18-D.PDF

Look at the audio frequency performance, if you can find one better, I would love to use it.

There are three sources of noise, base spread is only one of the three. At audio frequency, 1/f noise is more important than others. There is no particular way to predict 1/f noise, you look at the data sheet on the referred input noise voltage and current to determine, not just by base spread resistance. Base spread only tell you about the thermal noise. High beta lower the base current and lower the shot noise. You have to look at your application to determine which is the best.

I look at 2N4403, again there is not noise spec. How come both of you keep looking at transistors that are not specialized for low noise application and keep talking about noise?

[Alan0354]

Good idea Merlin. The transistors have to be surface mount, but we already use the FZT751 in other places, a surface mount version of the ZTX751. I'll try those out for noise.

It doesn't really address my original gripe, though: nobody makes a transistor actually specified for low-noise audio amplification any more. Diodes Inc. could decide to add some excess noise mechanism to the FZT751 tomorrow, it would still pass their quality control, and there would be nothing I could do about it. Not very convenient for mass production.

I wonder what makers of mixers and mic pres do nowadays.

I think so. They might not say for audio, but you can look at the noise graph, you can get the information. Here is the data sheet of 2N5087. If you look at Fig. 1 to Fig.5, you have all the information you needed for design already. This is a through hole, you can easily find the surface mount equivalent.

http://www.onsemi.com/pub_link/Collateral/2N5087-D.PDF

This one has very good low frequency performance.