PDA

View Full Version : Mismatched power tubes


gkelm
01-09-2008, 06:06 AM
In Dec '07 Vintage Guitar, Terry Kilgore says that matched tubes are not necessary, unless "you want that completely even-keel clean sound." He states that tubes of different ratings get "real depth of tone and character" (see article for context, p. 112). This seems to contradict everything I've heard, and certainly the way tubes are marketed. Thoughts?

Greg

John Phillips
01-09-2008, 06:16 AM
He's right, depending on what you mean by 'depth of tone and character'. Unmatched tubes will produce more harmonic distortion, and a more complex tone. Matched tubes give what I would call a deeper tone, though - clearer and cleaner, and with relatively stronger fundamentals.

Mismatched tubes also produce more hum, and you can only go so far before you get to the point where you can't bias both (or all) of them in the right range - without separate bias controls, anyway.

But there are two parameters commonly measured for tube matching - plate current (the current the tube idles at with a given bias voltage) and transconductance (the gain, crudely). If you get a set with matching plate current but different transconductance, that would enable you to get the increased harmonic distortion without the hum and with both tubes biased optimally.

Why are matched tubes so heavily marketed? Because for most amps - without separate bias controls - it's at least useful to be sure all the tubes will run correctly, and hum will be minimised. For cathode-biased amps with a single bias resistor it's important to ensure that one tube doesn't 'hog' current at the expense of the others, too (which will shorten its life disproportionately). But there is a lot of hype about it too - and certainly with matched preamp tubes, which are unnecessary in any amp.

baron55
01-09-2008, 06:17 AM
The matched tube thing comes from the Hi Fi realm. Actually perfectly matching "Push Pull" power tubes effectively cancels out even order harmonics. So a little mismatch is desired tone wise. Tubes that are greatly mismatched will be less desirable since one tube is really cold and the other is too hot. Usually 5 and even 10ma apart is acceptable. So what Terry said was correct to a point

Blue Strat
01-09-2008, 06:33 AM
1) There's no consensus on what sounds best. If there were, there would be only 1 amp, one guitar, one pickup, one speaker, one overdrive box, one set of strings, etc.....

2) Does the article say "how mismatched" tubes should be to sound "better"? If not, he's basically not saying anything. It's very rare to be able to match tubes to less than 1/2 a mA. So does mismatched mean 1/2mA, 1mA, 2mA, 5mA, 10mA, or more?

3) Tube matching, at least closely, IS over rated. To be within 5 or 10mA is good because it keeps hum at a minimum and allows the tubes to wear out somewhat evenly.

4) The only reason I attempt to match tubes closely is that it's easier to do than to convince some people that it isn't necessary :)

5) Even with proper burn in before testing, tubes will drift in use and likely drift apart. So a "perfectly matched" pair of tubes won't stay perfectly matched for long. Starting off as closely matched helps to insure that they won't become "grossly mismatched" over time.

deeval
01-09-2008, 07:06 AM
Mike if I have tubes that are Old Ge Black plates and they dont match in my amp,without a bias pot,is it best just to try different ones till I find a pair that are within 5ma,that should be ok is that correct?
But will those tubes also dirft apart after a few hours use,even though they where never new to begin with?
Thanks

Blue Strat
01-09-2008, 08:27 AM
Mike if I have tubes that are Old Ge Black plates and they dont match in my amp,without a bias pot,is it best just to try different ones till I find a pair that are within 5ma,that should be ok is that correct?
But will those tubes also dirft apart after a few hours use,even though they where never new to begin with?
Thanks

5mA is good. Tubes drift...they can drift closer together or further apart or in the same direction staying the same distance apart. It's a roll of the dice. Well used tubes are less likely to drift than new ones.

gkelm
01-09-2008, 09:56 AM
1) Does the article say "how mismatched" tubes should be to sound "better"?

He does give an example...but I don't have it in front of me, can check it later.

Thanks guys, for all the comments...quite informative.

PFCG
01-09-2008, 10:44 AM
so when you buy a quartet, are they just closely matched pairs? same with a sextet etc? Ive always seen the tubes i want in matched pairs but not quartets. Would it be safe to buy two matched pairs?

Blue Strat
01-09-2008, 10:56 AM
so when you buy a quartet, are they just closely matched pairs? same with a sextet etc? Ive always seen the tubes i want in matched pairs but not quartets. Would it be safe to buy two matched pairs?

Pairs are pairs, quartets are quartets, sextets are sextets. 2 or more pairs may or may not match each other.

Where are you ordering tubes that doesn't offer quartets?

PFCG
01-09-2008, 04:05 PM
i actually havent changed my poweramp tubes yet, i have the TADs that came with the amp and have been happy, but ive been searching around online for what i might like instead of the TADs, and i saw the NOS JAN Phillips that the two rock guys use and will probably end up getting a quartet of those, but i meant tubes like the GEC KT 66 on your website shows only matched pairs.

Once i get a little more play money ill be retubing my preamp first, then my poweramp, so im sure if i need anything ill be coming to you first, since you have the widest selection and seemingly the most happy customers.

Blue Strat
01-09-2008, 04:32 PM
i actually havent changed my poweramp tubes yet, i have the TADs that came with the amp and have been happy, but ive been searching around online for what i might like instead of the TADs, and i saw the NOS JAN Phillips that the two rock guys use and will probably end up getting a quartet of those, but i meant tubes like the GEC KT 66 on your website shows only matched pairs.

Once i get a little more play money ill be retubing my preamp first, then my poweramp, so im sure if i need anything ill be coming to you first, since you have the widest selection and seemingly the most happy customers.

The GECs are sold out (look at the bottom on the listing). There are so few of these left in the world that finding a matched quad will be a major challenge.

PFCG
01-09-2008, 06:36 PM
ah, sorry. there has been a few of those kindsa instances where ive been looking around for stuff.

im sure youll be hearing from me sometime in the near future!

JackButler
01-09-2008, 06:42 PM
Interesting.
So, would be safe to assume and possibly fun if an amp I have )Twp Rock) has sperate bias controls for each tube, as long as I stay within spec, I could try some different combinations?
Tis is two fold:
Obviosuly, two different brands, obviously not matched of the same octal type? This is a no brainer if I read all of the above correctly.
Secondly, I know the answer is no, but HAVE to ask, what about say one 6l6gc and one kt66?

PFCG
01-09-2008, 09:44 PM
yes ive also read the diezels all have separate bias points so they suggest mixing and matching tubes. so you could just buy a bunch of high quality singles?

JES1680
01-10-2008, 07:40 AM
Ok, I'll throw a wrinkle in here. How about cathode biased (class A) amps. How important are closely matched pairs in them?

John Phillips
01-10-2008, 07:47 AM
Ok, I'll throw a wrinkle in here. How about cathode biased (class A) amps. How important are closely matched pairs in them?Do you mean cathode biased or Class A?

:)

It's more complicated...

In a cathode-biased amp with shared cathode resistors, matching is very important, because if the tubes are mismatched the hotter one will pull more current, raise the bias voltage for both, and so make the cooler one run still cooler. It will eventually stabilise, but not before the difference is much larger than it would normally be. This is called 'current hogging'.

In a cathode-biased amp with separate resistors per tube, it doesn't matter much at all, because each tube will automatically bias to the same operating point, since its bias voltage will adjust independently of the other tubes.

This is the same no matter what the Class (which is almost always AB).

In theory, matched tubes are essential for Class A because otherwise you cannot ensure that all the tubes reach the maximum possible voltage swing without clipping one of them before the other at either forward or cutoff, even in a separate-cathode-resistor amp.

Teahead
01-10-2008, 08:06 AM
I haven't the technical knowledge to offer an informed opinion either way. All I can say his Terry has been providing me with outstanding tubes at excellent prices for a few years now. I don't question anything he says about valves.

Often, I'd order a quad and get five or six tubes, initially I freaked, which are the matched ones? Terry would say, just try them, swap them around and above all else, use your ears. He was right of course, which makes me think he's most likely 100% correct on this issue too.

Blue Strat
01-10-2008, 08:15 AM
yes ive also read the diezels all have separate bias points so they suggest mixing and matching tubes. so you could just buy a bunch of high quality singles?

Yes, as long as there's a bias pot for each tube and you're not trying to mix weak tubes with strong tubes.

Blue Strat
01-10-2008, 08:20 AM
Interesting.
So, would be safe to assume and possibly fun if an amp I have )Twp Rock) has sperate bias controls for each tube, as long as I stay within spec, I could try some different combinations?
Tis is two fold:
Obviosuly, two different brands, obviously not matched of the same octal type? This is a no brainer if I read all of the above correctly.
Secondly, I know the answer is no, but HAVE to ask, what about say one 6l6gc and one kt66?

There's nothing technically wrong with using mismatched brands of the same tube type/number. Some people have found, by accident, that they actually prefer this. Back when you could buy US made tubes in drug stores, this was common practice. If one 6L6 went, the customer would buy another not necessarily of the same brand. They also wouldn't bias or check for matching. That said, there were occasional mysteries like "my amp sounds better/worse than before. What happend?". Or tubes wore out prematurely, etc.

Mixing a KT66 with a 6L6GC is ok too as long as they bias similarly and the amp can handle the added current requirements for the heater of the KT66.

Echoes
01-10-2008, 08:24 AM
I have a set of mis-matched Tesla EL34 running in my Komet head...they are running at 37ma and 22ma (one hot and one rather cold)...

they sound GREAT! probably won't last as long as 2 tubes running at 30ma...but...

JackButler
01-10-2008, 01:51 PM
Excellent info guys!

Roccaforte Amps
01-10-2008, 02:15 PM
In Dec '07 Vintage Guitar, Terry Kilgore says that matched tubes are not necessary, unless "you want that completely even-keel clean sound." He states that tubes of different ratings get "real depth of tone and character" (see article for context, p. 112). This seems to contradict everything I've heard, and certainly the way tubes are marketed. Thoughts?

Greg





You can only match for idle current. As soon as
you strike a cord on your guitar the tubes go all over the place.
The do not conduct identically,
so anyone who offers matched tubes claiming anything more than
idle current match is BSing you. A guitar produces a complex signal.
Personally, I like new sets of output tubes to match as close
as possible. The reasons are "hum", and tubes do not wear
at the exact same rate. So, if they're close to start with they'll
maintain their matching as a set much longer.
That's it, nothing more about it.

PFCG
01-10-2008, 05:41 PM
sorry to hijack, but whats the shtick on the cryo treatments?

Faustine Amps
01-10-2008, 06:51 PM
In Dec '07 Vintage Guitar, Terry Kilgore says that matched tubes are not necessary, unless "you want that completely even-keel clean sound." He states that tubes of different ratings get "real depth of tone and character" (see article for context, p. 112). This seems to contradict everything I've heard, and certainly the way tubes are marketed. Thoughts?

Greg

That statement is a real generalization. It may or may not be true, depending on far too many unpredictable factors.

When I put tubes in my own amps, I start with matched pairs and bias them to sound "good" (quotations are very intentional) by my own standards, which are totally subjective. And they don't produce a "completely even-keel clean sound" unless I adjust the controls for a "clean" sound. When I crank up my amps, those matched tubes are anything but "clean"! However, they do have plenty of "depth" and "character". Terry's statement isn't false, it's just not true!:messedup

That is not to say that you must use matched tubes to get a "good" sound (again, notice the very intentional quotation marks). It's a crap-shoot. Maybe they'll sound good and maybe they won't!

I think it's a good practice to install matched tubes in my own amps because tubes can drift any which way they please over time, so I like to have a known baseline to start from. I've tried lots of un-matched tubes and more often than not, I find I prefer the matched tubes. It's a shame, because I have a bunch of NOS RCA Black Plate 6L6GCs that aren't within 10mA of each other. Maybe I should burn them in for a few weeks and re-measure them.;)

As for cryogenically treated tubes, I'm of the opinion that they belong in the "snake oil" category. I've seen the heated debates with the dog-eared audiophiles versus the material science specialists and thermal engineers and I'm not sure that subject will ever be settled as long as there is some audiophile or guitarist who "definitely hears a difference" and is willing to pay for it. It always boils down to whether you choose to believe objective data or subjective opinion, since they are both "right".

But, that's just my 2 centavos.

Regards,
Tim Gregoire

wopr
01-10-2008, 07:20 PM
Ok, matched tubes are overrated, but what about the Biasing issue? If you can't independently bias each tube, can't mismatched tubes become a problem? Let's say for a fender blackfaced amp.

Thanks,

steve

Blue Strat
01-10-2008, 09:05 PM
Ok, matched tubes are overrated, but what about the Biasing issue? If you can't independently bias each tube, can't mismatched tubes become a problem? Let's say for a fender blackfaced amp.

Thanks,

steve

How mismatched? If they're out by 10mA, bias one 5mA colder than normal and the other 5mA hotter than normal.

Mr. Kite
01-19-2012, 09:16 AM
like

'58Bassman
01-19-2012, 09:35 AM
I would bet that on most recordings that have what we consider "classic" guitar/amp sounds, the tubes would be shown as "mis-matched" if they were tested before or after the recordings. While session players have always needed reliable equipment, the designers of older amps knew that if one tube went bad, the player would replace only that one tube. If the amp had a "hum cancel" pot, the person may/may not use it to decrease the hum but that's about as far as it went. Some amps, like the 5F6A Bassman, are called "self-biasing" and have no provision for this- it's fixed.

Aspen Pittman is the first person I remember to make a lot of noise about matching & testing tubes for guitar amps and IIRC, it was after people decided solid state really wasn't what the manufacturers said it was but, unfortunately, most of the original tube makers had already shut down and the only brand-new tubes that were available were being made in Eastern European/Asian plants with equipment sold off by RCA/Phillips/Sylvania/GE, Tung Sol, Mullard, etc but the people who knew how to make tubes with any consistency were retired, not available or dead. The new tubes at that time were unreliable, bad and inconsistent- until Aspen Pittman and Mike Matthews (Sovtek) started looking at how tubes were being made and sorted, all we had was NOS or used tubes and most people didn't have a good way to test them because they A) didn't know how or B) weren't going to buy a good tester.

Being picky about matching tubes is best for HiFi applications, not great guitar amps, IMO. If someone wants sparkly clean sound, fine but for a bit of hair, no.

The question I have is, What do I do if my brand new, unused pair of Radio Shack 6L6 tubes goes bad? They have a lifetime warranty.:huh

Guinness Lad
01-19-2012, 09:57 AM
Let's put it this way I have a 74' Vibrolux with hum balance. I moved it around and could never get the tubes to hum. A day later I went to a friends and checked the current with a bias probe and one tube was 22ma and the other was 52ma, the amp did not hum. I was stunned, I quickly moved the balance and got the tubes to within 1.5ma of each other.

Mr. Kite
01-19-2012, 10:04 AM
Let's put it this way I have a 74' Vibrolux with hum balance. I moved it around and could never get the tubes to hum. A day later I went to a friends and checked the current with a bias probe and one tube was 22ma and the other was 52ma, the amp did not hum. I was stunned, I quickly moved the balance and got the tubes to within 1.5ma of each other.
I haven't seen a hum balance pot. Does the amp have a bias as well?

Guinness Lad
01-19-2012, 10:10 AM
I haven't seen a hum balance pot. Does the amp have a bias as well?

Silver face amps were hum balance. By changing this it does alter the bias but only to even it out. The final value I got was 36ma. Once it was here I could not raise or lower both.

jcs
01-19-2012, 11:19 AM
Off topic a bit but here goes;

Imo, the most important issue is to NOT bias too hot.

70% dissipation is the maximum for most push/pull circuits and for years i read about folks setting the initial dissipation to 70%.......the problem with this is eventually one or more of the tubes will likely drift (possibly upwards) and/or the danged wall volts will creep upwards from when the bias was set which we know will increase current draw.......the result is one or more of those tubes is now burning itself out prematurely.

Now of course many will argue that 70% dissipation (or hotter) sounds the best with certain tubes (EL34's anyone?)

This is allegedly how King Edward Van Halen at least partly got his famous "Browntone", even though Edward lowered the wall volts with a variac, i am fairly certain he ran the tubes beyond 70% dissipation!

Read the very early EVH interviews....Edward claimed to be burning down power tubes to the tune of one set a week and he was likely using Sylvania or GE EL34/6CA7 big bottles which are a very robust tube.

Anyhow, my point is over-dissipation is the major issue and yes it is okay to run mismatched tubes....just keep in mind the hotter running tube is doing most of the work and will wear down more quickly.