[Amps] Matching GU-74Bs

[Roger]

Michael Tope wrote:
> I just got a note from Alex UR4LL that is out of GU-74Bs. When I went 
> searching the internet for alternate suppliers, I found a tech note from 
> Alpha indicating that use of anything except "matched pairs" of tubes 
> voids the warranty for the 91B. This seems strange since the tubes are 
> run in parallel. It would seem that the anode current sharing might not 
> be exact in an unmatched pair, but that it would be way off before there 
> was an impact on performance. Also, what would be the danger to the 
> amplifier. Seems like the worst thing that would happen is that the tube 
> with more gain would run hotter than the other tube, but that would only 
> be a danger to the tube, not the amp. Moreover, seems like you could 
> detect a really bad mismatch by comparing the exhaust air temperature 
> coming off the anode chimneys. I found a thread on QRZ.com where W8JI 
> dismisses "matching" as mostly a "sales ploy". It seems pretty rare that 
> Tom isn't spot on with his technical advice, so I am guessing that is 
> what Alpha is up to with their tech note (or perhaps their overarching 
> concerns is people plugging in dud tubes rather than mismatched tubes). 
> They are advertising matched pairs of 4CX800As tested and warranted for 
> 1 year for $675.00:
>   

Are these Eimac, or GU74s?  There is a difference and they appear to be 
using the old surplus GU74's by their own statements
.
 I agree with Tom that it makes little difference unless there is a 
major difference between the tubes.
Then there is the approach that Emtron uses with their multi tube amps.  
They have a switch and adjustable bias for both tubes so you just plug 
'em in and match them up there.   It's not perfect, but apparently works 
and makes more sense to me than having to find matched tubes. 

The only way to really match tubes is with a curve tracer (that's the 
name IIRC), but I doubt any of the tubes we see sold as a matched pair, 
three some or 4 some, are done with anything other than idle current.

 Just plug in one at a time and check the idle current. If you have a 
pair that comes within Alpha's spec that should be good enough.  Alpha 
must have a spec for "matched pair" that the idle current will be within 
so many %. If they are within that it should meet the spec. OTOH they 
may insist on a *purchased* as a matched pair instead of a 
do-it-yourself matched pair. At any rate, if you match your own, be sure 
to document the current for each tube.

However...A lot depends on quality control.  If the tubes are 
significantly different inside  THEN  they should be matched according 
to both idle current and gain, but that would be some pretty sloppy 
quality control.
> http://www.rfconcepts.com/s.nl/it.A/id.4266/.f?sc=15&category=15030
>
> Here is the tech note:
>
> http://www.rfconcepts.com/Alpha-Support_2
>   
Note they say purchasing non tested (not non matched) and leave that for 
the user to imply. Second the way I read it, it's not just non matched, 
it's non warranted and I would take that to mean they expect you to 
purchase the tubes from them (They specifically say so in the last 
sentence of the second paragraph. Use their tubes or no warranty!).  BUT 
this is only during the warranty period. Once the warranty has expired 
on the amp you are on your own anyway.  They also refer to a possible 
catastrophic failure from non tested tubes. In my opinion: This is not 
due to tubes being mismatched, it is due to internal shorts or gassy 
tubes.  Also IF they are really concerned about the tube(s) causing a 
catastrophic failure in the amp I'd say their protection circuitry is 
lacking.

BUT they never explain what they mean by catastrophic failure. A failed 
tube, or something failed in the amp.   A failed tube is no biggie 
(depending on how much you paid and your budget), but something major in 
the amp is something else.
> I could see how checking each tube individually (perhaps by running it 
> by itself in the 91B at reduced drive) would be a good idea just to make 
> sure the tube isn't bad, but beyond that is it really necessary to run 
> "matched pairs", and if so, what is the matching criteria and how close 
> is close enough?
>   
Alpha should have a criteria *available* that says when tubes are close 
enough to be considered matched. I'd bet they do nothing more than check 
the idle current at some given bias and pair the tubes up with the 
closest to any given number being considered matched based on what they 
had for idle current.

Possibly they do it with some drive applied, but I doubt it.

Maybe they just used a poor choice of words (Many engineers have that 
problem), or this is not the same company philosophy that brought us the 
original amps.  Just remember you use their tubes for replacement during 
the amp's warranty period or no warranty according to that linked statement.


73 and good luck,

Roger (K8RI)
> Thanks,
>
> Mike W4EF..............
>
> _______________________________________________
> Amps mailing list
> Amps at contesting.com
> http://lists.contesting.com/mailman/listinfo/amps
>
>