Need For Matched Output Tubes?

[Beck-Ola]

A question for SteveD, Kevin, or any other experts out there. This is about the need for approximately matched sets of output tubes.

As far as amps with separate bias adjustment pots for each tube (e.g., JMD:50, recent VM 2266, etc.) is it at all important that the output tubes be closely matched as long as they can be adjusted to idle at the same level?

Of course I'm talking about the same tube type and brand, EL34 or whatever. For example, you have one EL34 that idles at 30mA and another at 40mA in a non-adjustable amp but you can get them both idling the same in these newer amps. Or do they need to be a "matched pair" to perform well under conditions other than idle? Thanks for any info!

[slowpokerhino]

I believe the seperate bias pots negates the need for matched tubes entirely. Could be wrong but I don't think I am.

[surfnorthwest]

I believe the seperate bias pots negates the need for matched tubes entirely. Could be wrong but I don't think I am.

You're right.

[SteveD]

Sorry guys but you're wrong.

Having the capability of matching the standing DC current through individual valves does not match their AC parameters in any shape or form. It simply balances the DC currents through the valves. The transconductance (think of it as gain) of the unmatched valves will still be significantly varied in which case the AC waveforms will be different amplitudes and therefore asymetrical. This is not particularly a problem in our world.

It's never been essential to have valves matched in a guitar amp. It's more critical to hifi buffs. Some people actually prefer unmatched valves and back in the day nobody cared. Opinions vary these days.

[surfnorthwest]

Sorry guys but you're wrong.

Shit, I hate it when I'm wrong, slowpoke was wrong first.

[slowpokerhino]

Yay, first!!

[Beck-Ola]

Thank you very much the replies (even Slowpokerhino's and Surf's...it was an admirable effort). But awfully glad Steve is around.

I see how they would be responding differently to the AC part. So it is probably ok to run them in that scenario as long as the sound is acceptable. What I do is hardly high-fidelity so it will probably work in "my world" too!

That's cool cause I have at least a couple of pairs around that I didn't use before since they weren't matched very well. Maybe this is THE tone secret. Ha-ha!

-- Tue Feb 08, 2011 8:58 pm --

Yay, first!!

Like George Carlin says, everybody is a winner.

[KevinOConnor]

Hi Guys

To those who thought they were wrong: you are actually right.

Steve's response was a bit heavy handed inasmuch as making a blanket "wrong-ness" statement.

The bias pots let you balance hum. They also allow a higher percentage of available tubes to be used in the amp, provided there is sufficient control voltage range designed in.

For an amp designed for EL-34, 6550, KT-88 or similar with a 1,600mA heater, it is easy to swap in low-heater-current tubes such as 6L6GC or anything related. If the bias control voltage range is wide enough, you can use alternative tubes AND mix tubes.

The reason matched tubes was ever important was because the first guitar amps were actually hi-fi circuits. Designers and techs followed hi-fi rules without being aware that that was what they were doing. Lethargy set in and we still battle with it today, and that is prabably why Steve's first response was that wrong thingy. Then he became aware that guitar amps are not hi-fi tone reproducers rather they are tube-safety-following tone creators.

Also, most companies do not want to deal with potential liabilities that may arise if they suggest to the equipment user that other tubes could be used, or that the user can bias them himself, or that mixing of tubes is possible, or that you can pull any tube from any position you want in a fixed-biased amp.

The asymmetric behaviour of unmatched tubes merely results in a fatter tone; you still get full output power and loudness.

My amps have always allowed the player to mix, match, pull tubes at his own choosing. It's not rocket science or even voodoo.

Have fun
Kevin O'Connor

[MKB]

A couple things to add to Kevin's post above: Santiago suggested in a video for the YJM at NAMM that a user could mix and match output tube types due to the auto bias circuit in the amp. I haven't read the manual though and do not know if that is an advertised feature.

If a pair of tubes are not matched in bias current, the difference in current can have an effect on the operation of the output transformer. If the misbalance is too high, it can saturate the iron from the standing DC current difference. EI transformers in general are not as sensitive to this as toroidal transformers.

[Beck-Ola]

If the bias control voltage range is wide enough, you can use alternative tubes AND mix tubes.

The asymmetric behaviour of unmatched tubes merely results in a fatter tone; you still get full output power and loudness.

I thought about trying that to see how it sounds. I'm glad I asked because now I can make good use of some pairs that aren't so well matched.

Based on this it seems like a balanced P.I. is a waste of money too unless it's for hi-fi.

[KevinOConnor]

Hi Guys

Actually... the risk of saturating the output transformer due to mismatched currents is not something to worry about.

Balanced splitter halves, balanced coupled caps, balanced dynamic performance of output tubes are almost all things that you don't want in a guitar amp. These are things that reduce even-order harmonic distortion and some IM distortion. However, that leaves you with only odd-order distortion. In a hi-fi amp, you are trying to reduce ALL distortions, but for MI you want a "pleasing" balance of harmonics.

There is a minimum symmetry or balance required to attain full output power, BUT... this is so far from needing anything matched that one needs ever consider matching to any great extent.

My previous post was no attack on Steve, rather a defense and explanation of the position he is in. His corporate yoke prohibits him from telling you about inherent capabilities of Marshall's amps - and for that matter, of any manufacturer's fixed-bias amps - so his first reaction has to be instinctively to protect the interests of his employer. My interests are different inasmuch as my work is to empower players and have them become more involved with their gear so they get the most use and pleasure from that gear possible - and for the longest time possible. Sound like "viagra for amps".

Have fun
Kevin O'Connor

[SteveD]

Hi Guys

To those who thought they were wrong: you are actually right.

Steve's response was a bit heavy handed inasmuch as making a blanket "wrong-ness" statement.

The bias pots let you balance hum. They also allow a higher percentage of available tubes to be used in the amp, provided there is sufficient control voltage range designed in.

For an amp designed for EL-34, 6550, KT-88 or similar with a 1,600mA heater, it is easy to swap in low-heater-current tubes such as 6L6GC or anything related. If the bias control voltage range is wide enough, you can use alternative tubes AND mix tubes.

The reason matched tubes was ever important was because the first guitar amps were actually hi-fi circuits. Designers and techs followed hi-fi rules without being aware that that was what they were doing. Lethargy set in and we still battle with it today, and that is prabably why Steve's first response was that wrong thingy. Then he became aware that guitar amps are not hi-fi tone reproducers rather they are tube-safety-following tone creators.

Also, most companies do not want to deal with potential liabilities that may arise if they suggest to the equipment user that other tubes could be used, or that the user can bias them himself, or that mixing of tubes is possible, or that you can pull any tube from any position you want in a fixed-biased amp.

The asymmetric behaviour of unmatched tubes merely results in a fatter tone; you still get full output power and loudness.

My amps have always allowed the player to mix, match, pull tubes at his own choosing. It's not rocket science or even voodoo.

Have fun
Kevin O'Connor

:cuss

All I said was that simply matching DC bias current through the valves will not stop them from being asymetrical in case anyone thought it would. I didn't even say this was a problem and even stated that some players prefered it!!!!!!!!!!!!!!

Nothing heavy handed about it at all.

[ryanmichael95]

Steve did say that back in the day nobody cared....that's code for "If it worked for Jimmy Page, it'll work for you." I've used matched sets of tubes many of times, but there's an unmatched quad of 6L6's in my Mesa right now. I biased it and it sounds great, just as expected.

[MKB]

There is a minimum symmetry or balance required to attain full output power, BUT... this is so far from needing anything matched that one needs ever consider matching to any great extent.

Many of the technical concepts discussed here on tube matching are a bit subjective. The engineer in me would like to see actual numbers to indicate when mismatching would cause problems. I've never really tried to find such data myself as I only work on amps part-time, and only for myself.

I know you can control the saturation point of an output transformer by adjusting the standing DC current through it (not the current to both tubes as this cancels out, the current in one side that would appear as the mismatch between the two halves). That would actually be a fun and interesting experiment. One heavyhanded way to try this would be to pull one tube out of a 100W power amp quartet. But this test may cause other problems, will have to think through this a bit. It should certainly only be tried under controlled conditions. Or maybe those of you that had a tube pop on a gig could chime in. :)

There is the old story of a guitarist (Paul Burlison with the Rock n Roll Trio) that reportedly pulled one output tube from his amp to get a distorted tone, and did some famous early recordings like that. That should have caused a large current mismatch in the output trasformer as well as perhaps some asymmetry/distortion in the output waveform.

[slowpokerhino]

There is the old story of a guitarist (Paul Burlison with the Rock n Roll Trio) that reportedly pulled one output tube from his amp to get a distorted tone, and did some famous early recordings like that. That should have caused a large current mismatch in the output trasformer as well as perhaps some asymmetry/distortion in the output waveform.

It's a very common "mod" for a Twin Reverb to pull 2 tubes and disconnect 1 speaker (pulling the tubes will cause the transformer to look for an 8 ohm load instead of 4). I've tried it, and while it does work, I prefer it "stock".