Economy push pull audio output tubes

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I am looking to marry an existing preamp that I feel sounds good, to a push pull power amp of modest output. I'm speaking of guitar amps here. I want to use some of the easily procured small bottle output tubes found in the mass produced sets of the day. So rather than buy batches of each one, test various configurations etc. I am hoping that someone here has hands on experience with the various 25(a,b,c), 30(a,b,c) 35's,50's etc. Specifically, help me save a bit of money, and alot of time by suggesting where to start looking for the better sounding ones. Also, if any known consistency issues. I had come across a site about 10 years ago but can't remember where. Perhaps we've a few old school or that could help? Cathode voltage is not an issue.
 
There are only so many tube types readily available and likely to be around. As for small, EL84 and 6V6 are your main choices.

I don't know what you mean with 25s, 30s, 35s, and 50s, unless you mean power amp output levels. The difference between them is negligible. All else equal, 50 watts is only 3db louder than 25 watts. There are any number of existing circuits using a pair of EL84, or a quad. And a pair of 6V6 is the basis of the classic Deluxe Reverb by Fender.
 
Administrator
Joined 2004
Paid Member
What is your power target? Modest is undefinable. Some of the types you've mentioned are quite rare, and extremely expensive.

A single pair of 6BQ5 might be a realistic alternative, if you want dhts which seems to be the case from the types you mention perhaps a pair of 45s would do the job? (It's not very likely that any of the types you mention would have found their way into instrument amps in any significant way as they are a bit early.)
 
While I have played my guitar through a DHT amp and do like the unique tone. DHT's do not work for anything that gets lugged around and banged on.

As an exercise in insanity a dozen years or so, I put my 845SE amp in the trunk of my Ford Mustang, and plugged it into a power inverter. The fine details of a DHT amp are lost when sitting in a traffic jam in a convertible, and the filament in one of the 845's died from too many trips down a bumpy road.

Any DHT that is over 50 years old must be treated with care.
 
A said, what is modest output? 1W, 5W, 10W, 20W? I am guessing you are asking in the 5-10W range, the one watt thing has been extensively mined with the 12AU7 even though I think using a pair of 6AK5's is the better choice. I have a pair of 35L6's with transformer that came out of a radio which would make a nifty 6W amp rather than going single ended with a 6V6. While we are in the range there is the 6AQ5 which will make up to 8W in P-P, still cheap to buy and you don't have to do anything odd to use them, any 6V6 circuit but with reduced voltage. Not a lot cheaper but there is the 8 pin family of 12L6's, 12W6, 25L6, 50L6. About the same price as the 6AQ5's and power output.

Then there is the 12V6, might not be cheap as the above tubes but you can get them at reasonable cost. Use them in their spec sheet limits or below and you might never need to buy a second pair again. The odd pin or voltage versions of the EL84 is out there also. A quick search will get you more information.

You do not say what you are planing on using for HV supply or output iron, can make a bigger dent in the wallet than the tubes. What cathode voltage has to do with things I don't know.
 
What cathode voltage has to do with things I don't know.

When I first read this post it didn't make sense to me. then someone suggested that the 25, 30, 35 and 50 thing meant DHT's. But the thread title used the word economy, which doesn't jive with DHT. Power output maybe? By my standards any of these would be considered modest power output ratings. My guess on "cathode voltage" is heater voltage, and all of those numbers except maybe 30 fit that too.

family of 12L6's, 12W6, 25L6, 50L6. About the same price as the 6AQ5's and power output.

You don't want to know how much power I can squeeze out of these guys without melting them. Look at the data sheets, both the curves and Class A1 amplifier characteristics match, however the maximum plate voltage spec for the xxL6's is 200 volts, while the max plate voltage spec for the 'W6 is 300 volts. So if everything else is the same, why can't you feed a 25L6 say 300 volts......well you can. So if 300 volts works, how far can you turn the knob. Well I ran out of knob on my little 400 volt supply, so I hooked up a bigger one. All of these tubes can eat over 400 volts, Some tend to run away around 450 volts, or you can't run enough bias current to fix crossover distortion without melting the tube, but 40 watts from a pair on 300 volts works without any hint of redness, so make a 30 watt amp on 350 volts, and it will last.

There are at least two more members of this family that I won't mention here because they will disappear. PM me if you want the numbers. They have the same internals, same curves and characteristics, but smaller plates with smaller dissipation ratings to fit in smaller bottles.

Then there is the 12V6, might not be cheap as the above

The 12V6 can be found for $4, and it's 9 pin equivalent the 12AB5 is $3. 20 to 25 watts can be squeezed out of them if you bend the plate voltage spec a bit.

The odd pin or voltage versions of the EL84 is out there also.

Some of the xxL84's are not odd voltage EL84's. The PL84 and UL84 are odd voltage 6CW5's. They will crank out 25 watts per pair if you run the screen grid at 150 volts or so. Dropping them into an EL84 circuit with more than 200 volts on the screen is an instant meltdown.

There are plenty of common TV tubes, or old radio tubes that can be used to make a guitar amplifier, but more information is needed to figure out what is being asked for here. How much power?
 
Administrator
Joined 2004
Paid Member
Maybe he is talking about series string tubes like 25C5, 25L5, 50L5 and 50C5? Would be helpful if he spelled it out a bit more clearly.

Until and if the OP weighs in I guess we won't know exactly.

I would probably choose the 6AQ5/6002 over the list above.

I worry that he may be thinking of a very "economical" amp sans power transformer and hence the interest in these tubes?
 
It would also help to know what his existing preamp needs, 6V or 12V heater tubes, solid state?

very "economical" amp sans power transformer and hence the interest in these tubes?

That is a SHOCKING thought. I have determined that the series heater tubes can be used to make the cheapest possible guitar amp INCLUDING the cost of an isolation transformer, but this won't work if the existing preamp uses parallel wired 12AX7's, or need +/- 15 volts to feed opamps.
 
Pretty much know that you could get more out of the tubes than they are rated for George, even thought to mention you and the idea of modest power starts around 50W. :) From the OP's requirements it seems he might not have your disregard for seeing a little smoke. I would take a crack at seeing how much I could get out of some of those tubes one day, just not enough time to play around.
 
I think you're right George. As I read it again, I thought he probably is talking about trandformerless sets, and he came close to naming the 50C5, which is the king of transformerless consoles.

No reason to kill the thread though. A PP 50C5 pair using an isolation transformer for both heater and B+ should still be economical.
 
A PP 50C5 pair using an isolation transformer for both heater and B+ should still be economical.

I built a complete guitar amp for about $50 in parts cost for the original Hundred Buck Amp Challenge. That ended several years ago, and the cost of some parts has gone up a few $$$, especially the transformers. The schematic is enclosed. The power transformer is a Triad N-68X ($16.37 at Mouser). The OPT is a 70 volt line transformer from Parts Express (Part # 300-040 $4.90).

The little amp makes 4 watts at the edge of clipping, and can do anything from a sweet mellow clean tone, to high gain thrash by turning up R9, the saturator knob. The mosfet increases the gain if the input tube by bootstrapping its load resistance and R9 is the actual load resistor seen by the tube. Some tubes can be microphonic at full gain, so don't build this as a combo if you like a high gain full metal racket!

I used the 100 mA series string tubes because they suck the least heater current and cost the least $$$$. They were on both dollar lists, and still are. The amp rocks, and it is still the only working guitar amp I have, but that is about to change. It could indeed be built with 50C5's, a 12AV6 and a 12AU6, by shorting out R31.

The 50B5 is the same tube as the 50C5, with a different pinout. Essentially orphaned overnight by UL in an attempt to make transformerless radios safer, it is available in huge supplies often for less money than the 50C5 because there is little demand for them.

Diode D5 is important. The tube heaters run on full wave but UNFILTERED rectified AC. This allows the preamp tube to have it's heater grounded to eliminate hum, but allows heater current to flow continuously for 360 degrees of the AC line cycle. This improves the "power factor" of the amp allowing for a smaller isolation transformer.
 

Attachments

  • GTA_4T_11-10.pdf
    22.6 KB · Views: 220
LSR
The AA5 tubes will make a modest power amp.
I have an old magnavox push pull WITH power xfmr that I wired in a 1/4" jack to use as a guitar amp. It's got 1- 35w4, 2- 35c5 and 1- 12ax7.
It's a stock amp, and I haven't messed around with the circuit, so at best it's a lo-fi mono amp that was never intended for guitar work.
The 35c5 and 50c5 have a decent clean tone. (Mine run 186v at the plates). The single 12ax7 of which half is a phase inverter doesn't provide enough gain for much breakup.
(Nowadays, I use it with my phone, to play backing ttacks.)

I don't know what cheap is for your budget, but I'm into mine $40. (Auction) It's not a guitar amp by any stretch, but it's an easy way to get a testbed setup for testing the old tubes, and for testing separate pre and power amps.
 
You don't want to know how much power I can squeeze out of these guys without melting them. Look at the data sheets, both the curves and Class A1 amplifier characteristics match, however the maximum plate voltage spec for the xxL6's is 200 volts, while the max plate voltage spec for the 'W6 is 300 volts. So if everything else is the same, why can't you feed a 25L6 say 300 volts......well you can. So if 300 volts works, how far can you turn the knob. Well I ran out of knob on my little 400 volt supply, so I hooked up a bigger one. All of these tubes can eat over 400 volts, Some tend to run away around 450 volts, or you can't run enough bias current to fix crossover distortion without melting the tube, but 40 watts from a pair on 300 volts works without any hint of redness, so make a 30 watt amp on 350 volts, and it will last.

I always wondered about the xxL6 tubes, they are rated for 10W max plate dissipation
whereas the lowest rated 6L6G/GA/GB are all 19W. Were the XXL6's really closer to 19W?
Are they cylindrical plate or rectangular, I don't remember? These would also work in my
$100 amp.
 
The 12V6 can be found for $4, and it's 9 pin equivalent the 12AB5 is $3. 20 to 25 watts can be squeezed out of them if you bend the plate voltage spec a bit.

12AB5 is very interesting when looking for that Fender 6V6 sound, and it probably has a
rugged heater able to take a wide voltage range. Physically rugged since it is designed
for car radios, good for an amp that is hauled around. Small package, good for a small
amp - I like it when looking for a Beam Power tube. They also do not have the reduced
plate ratings with 315V and 12W. Tube curves go out to 500V on the plate so they were
at least tested that high.
 
Were the XXL6's really closer to 19W?

The xxL6's have a rectangular plate with no fins to help shed heat. The electron beam hits it in the middle of the large flat side, so that continuous operation above 10 watts can lead to a red spot.

The trick to squeezing lots of power lies in finding tubes that were properly assembled, and tweaking the operating conditions to maximize plate efficiency. A class AB1 push pull tube amp can exhibit a plate efficiency anywhere from 40 to 65% depending on loading and biasing conditions.

The tube being used has a lot to do with the plate efficiency as well. Take the 6V6GT, with 250 volts on G2, and 0 volts on G1, you can't get above 110 mA no matter how much voltage you put on the plate. Conversely, the 25L6 can hit 110 mA with only 110 volts on G2, but about 40 volts on its plate. This means it takes far less effort to pull the plate of a 25L6 down to 25 volts or so on signal peaks, so this type of tube will exhibit lower losses, and higher efficiency. An AB1 amp with a plate efficiency in the over 50% range is easy with a 25L6, but near impossible with a 6V6. Even higher efficiencies can be reached with AB2.

So, If I had an amp that made 60% plate efficiency, I could stuff 50 watts of DC into the plate circuitry, extract 30 watts of audio, leaving 20 watts of DC burned up in the tube plates, divide that between two tubes, and we aren't even busting the 10 watt plate dissipation spec. This amp should be able to operate at full power continuously for a couple of years before needing new tubes......

Enter Mr. Bob Carver. He made a career out of exploiting the fact that even today's over compressed noise music has at least 10 db of "crest" factor. In my world (linear RF power amps for LTE and others) we called this the peak to average ratio. Simply explained, if you turned up some highly compressed music to the edge of clipping on a 100 watt amp, the amp would be producing 10 watts of AVERAGE power. It is this AVERAGE that warms heat sinks, or melts tube plates.

So, in theory we could build a 10 watt amplifier, and extract 100 watts of music from it. In practice, we can't in both the RF and audio world. Our 10 watts amplifier would need 10 db of headroom to eat those 100 watt peaks without clipping, which involves higher B+ voltages, lower output impedances, and a few other tradeoffs.

There were a few tricks used by Carver and others, particularly class G and class H, where there were multiple supply rails or dynamically adjusted supply rails. I used some of this technology to create some high efficiency RF power amps, and have successfully applied it to tube audio.

Without these tricks, we can look for some tubes that have the ability to pull their plate down to near zero volts, preferably without needing AB2, eat lots of plate voltage (for that headroom), and operate on a fairly low idle current without exhibiting crossover distortion to allow a lot of B+ without over dissipation at idle.

TV sweep tubes fit this requirement, and so do a few others. The 25L6 is one. The old 6L6 types do not. The usual clue is a cathode structure that's bigger (physically and in heater power) than other tubes in it's plate dissipation class.
 
Tube curves go out to 500V on the plate so they were at least tested that high.

A tube operating into an inductive load like an output transformer connected to a speaker will see twice it's B+ voltage on the plate in NORMAL operation. This voltage can go far higher when the amp is driven to clipping and the plate current gets cut off. A guitar amp is worse case since the speaker's resonant impedance peak is usually inside the guitar's frequency range, and overdriving a guitar amp is an accepted practice. I have seen 2KV+ peaks on the plates of a guitar amp with a 430 volt B+, when driven well into clipping.

Most tubes will eat far more than their published plate voltage rating. Any tube that was originally used for TV deflection will have a "peak positive plate voltage" spec. That is 1200 volts on the 6V6 and 1500 volts on the 6W6 from which the 25L6 was made. The 12AB5 was never used in a TV, and has no such spec. It's plate is physically different than a 6V6, so it might not eat 1200 volts, but will eat far more than 315.

I tinkered with these in guitar amp duty years ago, and don't remember any reasons not to use them, I just never had many to play with.
 

PRR

Member
Joined 2003
Paid Member
25L6 and kin are not-at-all special heater 6L6.

The 6L6 came first, got very famous, and it does seem likely the "25L6" tag was chosen to echo the big tube's fame. Let the table-radio buyer think what he wants to think. But the 25L6/etc are clearly LOW-voltage tubes, with some but not all of the 6L6's advances.

As for "sound" (aside from the YEOWW! of transformerless shock), I bet the circuit, the OT, the speaker make FAR more difference than the tube.
 
I made a guitar combo using 12AB5's in the output. I used a Hammond isolation transformer 186D120, with 120:240 30 VA for the HV, used a doubler. I used the secondary for the primary and the two primary 120V windings in series to get 208V out. Theoretically it should be 2*291 = 582V dc. But with losses it turned out to be a manageable 420 Vdc with 415 Vdc on the plates. I had 30mA per tube at idle. At 410V*0.030A = 12.3W. The tube plate dissipation is 12W, say 2W of that is going to the screen, about 85% dissipation. Other measurement I took, playing at the edge of breakup the voltage drops to 370V, down to 355V with the amp feeding back and gain at a point where more really does not make much difference. All warmed up the plate voltage dropped to about 400V.

Hammond 125E for the OT and a switching voltage adapter to give me 12V dc for the heaters. I used a 12" Fender Special Design in this amp and clocked in at 24.5 lbs in a 5E3 sized cabinet. I did not select lighter wood for the cabinet, add a pound or two for the transformer, another pound or two for the speaker, it all adds up. I have a pentode up front and it does not like the vibration of the cabinet. I have the socket isolated with o-rings but it still runs into feedback with the gain up. I was going to convert it to a 12AX7 up front, I will get to it some day.

In the mean time I drew up a rough schematic of the amp I may build with the 12V6's.

DFqVeaw.png
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.