• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

do pp output tubes have to be matched?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Bonjiorno, The guy who shipped me my amp didnt separately package the tubes and one of my 2 output tubes broke...he is sending me another one, but it wont be matched or anything...will I need to do something with the tubes...or should I just tell him to get me a matched pair?
 
If they are both new and of the same manufacture it should be ok.

Matched tubes are best but hard to properly do, and they often become different after a few months use anyway.

If your amp is adjustable bias then maybe bias them all the same, otherwise don't worry too much, it shouldnt make too much difference.
 
diyAudio Senior Member
Joined 2002
Hi,

If they are both new and of the same manufacture it should be ok.

What's the manufacture to do with the tubes being matched?

If they're not from the same manufacturer they can still be matched but I don't recommend doing so as they're usually different enough on other parameters to not sound the same anyway.

Unless the guy at amp's manufacturer actually did his homework and matched a pile of tubes for that particular model of amp, the odds are against a new tube arriving that would match the rest of the set you have left....

For best performance tubes in a PP amp should be matched in the amp itself, no tubetester can match tubes for your amp unless it was specifically calibrated for that task.
To the best of my knowledge no manufacturer does that unless they send out an amp for a review or prepare for an exhibition but I haven't told you that....:angel:
Needless to say they only go through the trouble because they want the amp to sound its best..............

Matched tubes are best but hard to properly do, and they often become different after a few months use anyway.

If they're properly matched they usually stay that way.

Cheers, ;)
 
fdegrove said:


If they're properly matched they usually stay that way.



Hi there....Thats the evidence I'm finding after so-called matched tubes been in high power amp circuits ....the quiescent currents do actually stay pretty stable..... but it's all down to who's turning the handle on the tester.
It's an illusion that matched pairs or quads are absolutely necessary as nfb can take good care of distortion variations. Put a screen in front of twp sytems one with matched tubes and the other unmatched and get people to listen and I guarantee they can't tell the difference.

richj
 
Hi there...others can round me up better on this one... adjusting the cathode current (s) is only one of several DC parameters....It doesn't do anything for the AC gain of the tube.....ie the gm (transconductance) will vary per tube. A common mistake is to simply adjust the I cath for quies draw..but if you looked at the full signal draw in fixed bias AB1/2, that figure could be wildly out. Cathode bias does automatically adjust this...a benefit ...however the AC dynamic control used on the anode stage drivers on some tube amps actually trimmed out the thd so each half p-p got exactly the right driving voltage.
A tube tester will check for dynamic gm and Icath for a matched pair otherwise there ain't much point doing it.

The old spec is to adjust for min thd by using the AC balancing potty on the driver stage when o/p stage is driven roughly 90% of full o/p when using a thd analyser. Okay for many without tools there's noway of measuring it....but another getaway' is to use closely matched resistors for the driver stage anode loads. A different example is the orig Williamson.. but better examples are the GEC 88-50 ; 30-100Watt amps.

richj
 
Hello,

While I advocate matched tubes... I do not think it is critical in many applications. Here's why I think so:

1) Many P-P amps are cathode bias and run at 60-70% of full potential power. Therefore the tubes do not get near their peak specifications where a mismatch would cause damage.

2) Many tube P-P amps are run class AB. Which means about 40-60% of the time they are both class A.

3) Most tubes specs should be within 10% of ideal design specs, or 10% of each other. I guess at most that could mean up to a 20% variance. But for the most part a variance of 5-10% is not bad.

4) A mismatch sounds better because there is less even order harmonic cancellation due to the mismatch. Actually, not better but more euphonic, which is better to those who like that kind of distortion.

Where matching is important:

1) In fixed bias amps running the tubes at close to full power potential. That is, a P-P amp running at a clas AB where it is mostly B. So the tube is biased at very close to the knee in the transfer curve (whatever that means). That way one can have up to a 95% efficiency of output power versus B+ supply (not counting the power drawn by the heaters). That is why fixed bias amps are adjustable.

2) Where one wants to make sure a close to perfect balance occurs for close to perfect signal symmetry (lowest distortion). This is why I get them for my amps. This symmetry also counts where the output transformer is concerned, where a slight difference will cause the OPT to see some positive or negative current flow (from the point of view of the secondary winding), causing unwanted residual magnetism in the core.

3) I guess there is a benefit for longevity also, where one tube is not causing the other to work harder, as it were.

Can't think of anything else not already covered here by others. I do know that when I matched pre and driver tubes also between channels the imaging and center voice or instrumentation was very precise. I also got a matched 12AX7 for an experiment, and again, it is definitely an audibly good experience.

My :2c:.

Gabe
 
I now realize after reading richwalters post that the tubes should be transconductance matched AND the DC bias point should also be same for both tubes.

9am53, what this means for you is that get the matched tubes and make sure their DC bias point same. It depends on your amplifier circuit, and how comfortable you are doing it(lethal voltages). It wont help if the tubes have only one shared cathode resistor. Basically measure the voltage across the cathode resistance for each tube and adjust either resistance until both voltages are same and as per spec of the design.

To go even further you should assure that the opposite phase AC signal swings coming from the driver and feeding each tube in the P-P pair is also of equal amplitude - thats where the matched anode loads for the driver stage comes into picture.
 
Thanks for the breakdown Percy...but I am still not getting this stuff...there is a lot to learn when it comes to the theory of these beasts and I have only but read the table of contents in the text. I will keep researching, but for the meantime I will use the tubes I (will) have, till they fry (which from what I understand will happen) then I will get a matched pair. thansk for the input though
 
Hi there. I'll throw in another variable...to what already has been discussed....All tube testing is done with the heater emission at 6.3V (I've forgotton why this weird 6.3V was got at) but I've come across power tubes where most characteristics seem matched at this nominal voltage........then change heater volts +/- 10% and gm and Icath loose track of each other. In general there is considerable running tolerance.......an example; the 6550WB Svet is very emission sensitive, a 0.3V drop in heater volts can see Icath sag whereas the effect on gm won't be so dramatic.
Do the same on an orig mullard EL34 and GEC 88's both sit steady like a duck...I'm implying that there is probably some merit in running tube heaters at somewhat higher voltage but within the spec allowed for. I've always argued that 5.7V (at the lower spec end) is really too low for power tubes and low emission can do more harm than running too high. I may stand to be corrected....

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