Matching tubes/ valves

[diydidi]

I recently acquired a Maxi-Matcher 2 valve tester.
Been playing around with my batch of valves from a reputable American tube store. I bought a couple of 5881's as matched.
Plug 4 of them into my matcher and check for both plate current (Ip) and Transconductance (mu).
These are the readings I get.
V1: 15,9mA/ 2,32mu
V2: 17,8mA/ 2,41mu
V3: 17,5mA/ 2,38mu
V4: 18mA/ 2,35mu
Now, are these considered matched or not?
What tolerance between Ip and mu is acceptable?
Any help will be appreciated.

[Enzo]

Matched is whatever the seller decides is matched. Those are all within 10% on the milliamps. Average among them is 17.3ma.

With an average of 2.37 among the mu readings, the widest variation was 0.05 or about a 2% spread.

This is a guitar amp we put them in, not laboratory gear. So 10% doesn;t scare me off.

But as to whether you consider it matched, well, that is up to you.

[jbltwin1]

AND, to add to what Enzo has stated, if you want them even closer, convert said amp to bias balance and balance them IN the amp. Unless it's cathode biased. Then you have to wiggle a little harder. It's a guitar amp, not a space shuttle. Back in the day (ok, Im old!), we just popped them in and MAYBE checked the bias control voltage to see if it was there. We didn't even MEASURE bias current. Mike.

[Randall]

I for one think matching is over rated. Kinda like how CBS engineers worried about all that distortion in those black face Fenders they "fixed".

[Chuck H]

You read all over the place about how matched tubes are great, important, superior, get you the girl, etc. You certainly DON'T want grossly mismatched tubes, but stringent matching isn't necessary or even desirable for everybody. There are good things about matched tubes both tonally and electronically. The amp will operate at it's best efficiency for frequency range, dynamics and power. Great! Some amount of mismatch can also sound good for some tones to some people BECAUSE of the reduced range, dynamics and power as well as some little things like reducing audible crossover buzzies and often more 2nd harmonic in a less symmetrical wave form. But no electronic advantage I know of. So it can be different strokes and all. I buy matched tubes but never check them. When I have a "batch" of unknown tubes I match them for current at idle and driven to three quarters output, then match them together as best I can (usually within a few percent). Faghetabowdit. Close is fine.

[J M Fahey]

Biggest problem caused by gross mismatching is unbalanced DC current component will partly saturate the core and make it lose inductance, so lose bass.
Since you paid a lot (big iron, lots of copper) to get is , it sounds stoopid to lose that just because of carelessness.

That said, any guitar OT designer worth his salt *must* design it to stand Real World at least 10% mismatch, and not suffering much with, say, 15 or 20%, at least audibly.

Same reasoning behind designing, say , bridges to stand 20 fully loaded trucks although "you will never have more than 10" or calculating for 130V or 250VAC mains instead of *just* 120 or 230VAC.

Won´t comment on increased distortion because that´s a bonus in the guitar world

[g1]

Tubes can be matched for many different things, but usually what you get is matched for idle current at whatever level they are using in their jig. So if your maxi-matcher is testing at a different idle current, the numbers may not be as close as they were in the vendors set-up.

[mhuss]

+1 to all of the above

Remember also, the only matching that really matters is under the actual conditions in the amp they're going into (B+, load impedance, etc.). Any tester (including the maxi-matcher) will give you numbers based on whatever /its/ conditions are. Not saying they're invalid, but it's highly unlikely they'll precisely match any specific target amplifier.

[nickb]

I'm fond of data. I've tried a few approaches to matching and the best I've come up with measures bias current around the desired operating point, gm1 (transconductance for grid 1) and gm2 (transconductance for grid 2). I sort the results giving by gm2, then gm1 and finally bias so that bias has the highest priority. I've found that gm1 is somewhat correlated to bias and has quite a narrow spread compared to the bias. Also that gm2 is really quite stable. Here is a chart of gm1 vs bias and you can easily see how stable gm1 is compared to bias for a batch of about 100 EL34s.

The bias was measured at Plate=425, Screen=425 Grid1= -38

[Enzo]

Much respect for all who have done the research and tested methodology. But I have yet to hear what the actual goal of the matching process might be.

In the game we tend to adjust the bias for idle current. But what does that achieve? it tends to balance out ripple at idle. And we don;t hear ripple when sound is happening anyway. We can match for gm but then again, so what, it is at idle. We use the tubes with substantial current, relatively speaking, and we might want to match them so they have similar shaped curves up where we use them. Does anyone ever do that? Certainly idle current is no guarantee that at 50 watts the curves line up.

It seems to me that perhaps the emperor's new suit is a trifle threadbare.

And as I like to chant, it is all about the circuit, not the parts. We focus laser like on the power tube matching, but how many of us ever "balance" their phase inverters? I don;t mean shelling out the extra buck for gain matched dual triodes, I mean the whole boat, blueprinting all the resistors and caps, and matching the gain curves of the two triodes in pair type PIs.

This is an honest question, I am not just carping.

Has anyone studied how the matching we all do at idle is reflected in performance at real world levels under use? I am always impressed with the research projects Loudthud does, scope pix and all, it sounds like something he might do.

the matching at idle strikes me a lot like trying out amplifiers at bedroom levels and expecting the comparisons to still be valid at stage volume. Maybe that IS valid, and maybe the tube matching we do indeed extrapolates accurately, but do we know that?

[nickb]

Completely agree Enzo.

The problem is (1) customers demand it (2) The spreads are significant - there is 2:1 range in the above data and that is after I removed 25% of factory supplied and tested tubes from this test as being out of range.

PS: What are you doing up at 4am?

[Enzo]

OK, but I take this as a technical discussion. Customers want standby switches on their Champs. Customers want a lot of things. Customers want all tubes biased at "70%". Customers want to know how to set the bias on their preamp tubes. If all I cared about was what the customers want, then I could mark all my tubes as matched. They'd never know. NO, I am not so lacking in ethics that I would do that. But I could. But talking among technicians...

Spreads are significant, but that makes my point, do those spreads really mean anything during operation? What effect does idle speed have on my car while driving? If I matched two same autos for idle speed, and maybe correlate exhaust pollutants. What does that say about driving 70mph down the highway?

[J M Fahey]

I'm fond of data. I've tried a few approaches to matching and the best I've come up with measures bias current around the desired operating point, gm1 (transconductance for grid 1) and gm2 (transconductance for grid 2). I sort the results giving by gm2, then gm1 and finally bias so that bias has the highest priority. I've found that gm1 is somewhat correlated to bias and has quite a narrow spread compared to the bias. Also that gm2 is really quite stable. Here is a chart of gm1 vs bias and you can easily see how stable gm1 is compared to bias for a batch of about 100 EL34s.

The bias was measured at Plate=425, Screen=425 Grid1= -38

IM-PRES-SIVE

Can I double like this?

[nickb]

OK, but I take this as a technical discussion. Customers want standby switches on their Champs. Customers want a lot of things. Customers want all tubes biased at "70%". Customers want to know how to set the bias on their preamp tubes. If all I cared about was what the customers want, then I could mark all my tubes as matched. They'd never know. NO, I am not so lacking in ethics that I would do that. But I could. But talking among technicians...

Spreads are significant, but that makes my point, do those spreads really mean anything during operation? What effect does idle speed have on my car while driving? If I matched two same autos for idle speed, and maybe correlate exhaust pollutants. What does that say about driving 70mph down the highway?

IMHO the goal is to avoid having premature failures as a consequence of the current being set high enough to avoid crossover distortion on the 'cold' side while the other 'hot' one bakes itself into an early grave. Sonically, I doubt there is anything much in it.

[Justin Thomas]

Biggest problem caused by gross mismatching is unbalanced DC current component will partly saturate the core and make it lose inductance, so lose bass.

So, one more way to get the Fender Fart out of my Bassman!!!???


Justin