 
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
|
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
210.165.109.226
I'm using a push-pull amp and attempting to match up 6550 output tubes from my stash. I understand that both Plate Current and Transconductance are important, but finding tubes that match up with both those values is rather difficult. Question: which of those two values is more important in finding a match?
Follow Ups:
at a given bias plate current goes prety much with Gm, so matching for equal plate current at the same biasing voltage is sufficient.For used tubes that have partial emission loss, I would match for equal Gm and then balance currents by tweaking the bias.
"at a given bias plate current goes prety much with Gm, so matching for equal plate current at the same biasing voltage is sufficient."I'm afraid I can't agree. I have tubes with a wide range of Gm that have very similar bias currents and vice-versa. I have NOS GE 6550A that vary as much as 1250-1500 umhos but are within a few ma. on current, for instance.
New production tubes often show the same result.
That's why I match cathode current first (at 465 volts plate and screen and -50 on the control grid for 6550A, for instance), then while the tube is blazing hot I check Gm. I can't check just current or Gm and get good matches.
My limited experience is definitely no match to yours. BTW, if the PP stage is not operated outside of Class A, is Gm matching important?
Class of operation has nothing to do with it, sser. Gm is essentially a measurement of a tube's ability to amplify, so it always matters if you want more than one tube to amplify a given signal equally (or as close to equally as possible).
that because in Class A each tube reproduces the full wave, the summed signal will not be distorted even with gross mismatch of tubes' GM. Just one half will contribute more to the overall signal than the other half. In Class B, it is totally different: because each tube deals with only half-wave, mismatched Gm will cause a lot of distortion.
But in a push-pull amp operating in class A the signal has been inverted in 1/2 of the output section, just like it would be in any push-pull amp.So while both sections pass signal full time, one output section is "pushing" while the other half is "pulling". Clearly if the amount of "push" and "pull" isn't equal or nearly so, there will be distortion. And the amount of "push" and "pull" is at least partially a function of the tube Gm.
Did I make that clear with my explanation? If not I'll try again.
Oh, as an aside that may help, while we often lapse into calling the phase inverter a "phase splitter" that's not the case at all. The signal is not split, rather it is inverted 180 degrees. So one of the phase inverter's two outputs is not inverted and is applied to one side of the output, the other is inverted and is applied to the remaining side of the output. So when the output stage gets signal applied one side "pushes" the signal, the other "pulls".
You mean inherent tube nonlinearity - more current on upswing than on downswing with equal driving voltages of different polarity. So, with different Gm, less cancellation of the second order harmonic than with equal Gm?Thanks for explanations.
the 12 Ruby EL34s in my Cary V12R vary so much in Gm? When Ruby, or GT, etc. match output quads, sextets, etc. how do they "normally" match them, Jim?
I'm not privy to all the details, but they state a current and a Gm spec on each tube so one would expect they are tested for both.I do know that at least some of the USA distributors test for plate current, not cathode current. I've been told this directly from them. I also know the amount of warm up/burn in time varies a lot. The tubes in your amp simply may have changed from use. The longer the burn in, the less likely the tubes are to change significantly - but the longer it takes and the more expensive it is. So burn in is a compromise in any case. I know of one distributor who told me some time ago they only run the tube until it's warmed up (a minute or two) - and others like the New Sensor brand family (and probably some others) run them 24 hours.
In general, Ruby seems to do the best job of matching of the "big guys". I will say that it used to be better than it seems to be now. I have never purchased matched tubes from GT, so I can't tell you about the quality of their work, sorry.
Transconductance are important in matching tubes.
cheers
LT
Wrong!Plate current matching at real world levels is the single most important measurement you can make on power tubes.
am I to assume that Transconductance is more important than Plate Current regarding matching output tubes up? Thanks.
No, do not make that assumption.If you have the ability to set the bias on each tube separately then you can compensate for tubes that pass more CATHODE (not plate - I'll explain why in a minute) current. However, if you can't make individual adjustments it's quite possible you could have one tube working really hard and the other tube loafing. At a given bias setting the currents can differ significantly. So for non-adjustable bias, or less than one bias adjustment per tube then current matching is the most important. Even if the Gm matches great - if one tube is nearly cut off and the other is red plating you have a problem!
In an application with individual bias control for each tube my rules change. As long as the currents aren't HUGELY different then Gm is my primary match criteria.
I would recommend that if possible you measure cathode current rather than plate current. In tetrodes, pentodes, and beam power tubes the screen current also flows through the cathode; and screen current varies from tube to tube, no question. A common way of measuring current through a tube is either to measure a voltage across a small value resistor (1 ohm or 10 ohms is common) in the tube cathode (and use Ohm's Law to convert it from volts to amps) - the Citation amps and MANY others are set up that way. In a few amps an ammeter gets connected in series between the cathode and ground to directly measure current.
So when you adjust bias on those amps you set it by measuring the CATHODE current. It's important to realize that the cathode current is the sum of the plate current and the screen current combined. Differences in screen current will impact the cathode current and will therefore need to be taken into account. Measuring just the plate current ignores the impact of the screen current and therefore could cause biasing errors.
If my goal in matching tubes is to provide conditions that mimic the real world operating comnditions as closely as possible, then it is incumbent upon me to measure CATHODE, not PLATE current. Why? Because when I ship you the tubes you will install them in your amp - and it is highly likely you will be measuring cathode current to set bias.
A "plate current matched quad" of tubes is not likely to be cathode current matched - due to variations from tube to tube in screen current!
So anyway, which is more important - plate current or Gm? The answer is... it depends! But the safest match criteria is cathode current first, Gm second.
I is most important in PP to balance the currents in the OPT primary. If the tubes will be used in pentode mode, then only the plate currents will go through the OPT primary and so it is important to set the bias so that the plate currents are equal. However, in UL or triode mode, both screen and plate currents will go through the primary and both need to be in balance.
Ray,"I is most important in PP to balance the currents in the OPT primary"
Important? Yes. Most important? No - IMHO primary current balance is moot if the tubes keep blowing up!
It is true in a pure pentode mode amp the plate currents are the only current flowing through the OPT - but that is not true of the tube. It still has both plate and screen current flowing through the cathode (unless it's a triode of course!). So if the bias is set by measurement of current in the cathode then you must consider screen current.
I know what you are driving at, and your point is well taken. And what you say is true in a few cases where the bias is measured at another spot besides the cathode and no screen current is present in the OPT.
I'm not suggesting that you should match tubes depending on how they are to be used, Jim. I'm talking about how they should be set up once they are in the amp.
This post is made possible by the generous support of people like you and our sponsors:
Top of Page ]
[ Contact Us ]
[ Support/Wish List ]
[ Copyright © 2025, Audio Asylum, all rights reserved ]
