Is matched and balanced tubes generally an hoax?

[pnp]

I'm asking this because I need 8 matched and balanced 6350 and I've bought four sets from diferent but reputable sources (aparently, at least...), being one set bought directly from a major high-end brand that uses this tubes.
The problem is that before using them, I allways have them tested by a technician in whom I've absolute confidence to do it properly.
The results are appalling: not one of this sets was matched and balanced. Worst, in all of them diferences in mA were huge, not only on individual tubes but also between diferente tubes.
On all sets I've paid premium to get them properly matched and balanced and none was.

A few years ago I've read on a French professional áudio magazine an article about this problem. They bought a few lots of the same tubes from sellers in Europe, China and USA. Not one lot was matched and balanced!!

There seems to be a common pattern here....

What's your experience on this?

 

[Steve Williams]

where I buy my tubes I always get the measured results of each tube I am buying

 

[pnp]

Tom lex,

I have no knowledge on tubes or electronics at all to demand anything from a tube seller. I can only trust on their words about balance and matching. I usualy ask them about procedures they use and test equipment used. I then ask for advice about it to my technician and if he feels that something more is needed I ask the dealers for it.
I completly trust my technician. He is a second generation technician and is working on the field for more than 40 years. He only works on amps and preamps, either professional or high end áudio. He has a vast knowledge on tube gear (he works with the best guitar players and bands on the country) and vast colection of tube manuals, technical papers and tube testers.
Further, he only works with two tube supliers in Germany, the only ones he trusts. They ship their tubes with a detailed spec sheet and 99% of the cases the values are spot on with his own tests.
Unfortunatley neither one of this supliers have 6350 in stock.
Do you know of a trustworthy dealer where I can get a good set of this tubes? If you do, pease let me know...

 

[Steve Williams]

can you share with us where you bought your tubes and the measurements weren't what they said

 

[Keith_W]

My audio engineer believes that there is no such thing as a pair of matched tubes. He says that the only way to get around it is to buy an amplifier with individual bias for each tube.

 

[puroagave]

This is what roger modjeski of RAM has to say: http://www.ramlabs-musicreference.com/virtues.html

i'm a believer and buy most of my tubes from him. i've owned few VTL amps that have no bias adjustments (tiny triodes come to mind) and matched quartets are a must.

 

[zztop7]

checker will not load the tube the same as your amp

My audio engineer believes that there is no such thing as a pair of matched tubes. He says that the only way to get around it is to buy an amplifier with individual bias for each tube.

I agree with Keith, but I will even go further. Checking the bias while each tube is operating in the amplifier is NOT taking into account changes that happen during Load of actual operating conditions.

Even on SET amplifiers, the bias is properly checked with NO load [signal] from the preamplifier [most of my amps are SETs, therefore that was not a slam].

Further >>> no matter how well they are matched on a checker, the checker will not load the tube the same as your amp.

Further >>> once you operate the tubes, even if they were totally matched in your amp, the tubes are going to change at different rates and NOT match.

I have seen tubes very accurately matched on High End tube checkers, then put in amps and checked in the amplifiers only to find them running Totally Different [hot & cold comparatively].

zz.

 

[MylesBAstor]

I agree with Keith, but I will even go further. Checking the bias while each tube is operating in the amplifier is NOT taking into account changes that happen during Load of actual operating conditions.

Even on SET amplifiers, the bias is properly checked with NO load [signal] from the preamplifier [most of my amps are SETs, therefore that was not a slam].

Further >>> no matter how well they are matched on a checker, the checker will not load the tube the same as your amp.

Further >>> once you operate the tubes, even if they were totally matched in your amp, the tubes are going to change at different rates and NOT match.

I have seen tubes very accurately matched on High End tube checkers, then put in amps and checked in the amplifiers only to find them running Totally Different [hot & cold comparatively].

zz.

Isn't that why many a manufacturer, with cj coming to mind, actually select their tubes based upon their performance in the actual circuit? Does the equipment meet its specs with this particular set of tubes or not? I'm sure ARC has to do something similar.

Then of course, how and if the tubes were burned in will affect their measurements. Then again, what equipment are they using to match their tubes?

 

[cjfrbw]

I am a contrarian to the audiophile false precision theme again. Came to believe that if tubes are within plus or minus ten percent, they are fine as pairs. There are so many variables with tubes, when they burn in and age, they change anyway at variable rates. Same date and batch of manufacturer and plus or minus 10 percent.

 

[microstrip]

There are tens of technical papers explaining why you should match the tubes in push-pull designs or when you put in parallel. Some amplifier designers optimize their circuits to allow some imbalance, others not.

As a general rule tubes should be matched and balanced after burn-in and in conditions similar to the operation point in the amplifier where they will be used - 99% of tube testers doesn't allow it. Manufacturers and a few specialist tube sellers do it properly, most sellers who advertise it do not care about it. If they do not ask you what is your amplifier brand and model you should be cautious.

 

[microstrip]

I am a contrarian to the audiophile false precision theme again. Came to believe that if tubes are within plus or minus ten percent, they are fine as pairs. There are so many variables with tubes, when they burn in and age, they change anyway at variable rates. Same date and batch of manufacturer and plus or minus 10 percent.

+1!

Although new ARC tubes (that have been burned in) usually come from the factory with a matching better than 5% the official ARC allowance for matching is about plus or minus 12%.

 

[microstrip]

Isn't that why many a manufacturer, with cj coming to mind, actually select their tubes based upon their performance in the actual circuit? Does the equipment meet its specs with this particular set of tubes or not? I'm sure ARC has to do something similar.

Then of course, how and if the tubes were burned in will affect their measurements. Then again, what equipment are they using to match their tubes?

Burn-in will affect the stability of the measurements. I think that they use some circuit similar to what I use - a circuit supplying a constant current that allows measuring grid voltage and the transconductance injecting a signal. Some extra points allow to check for grid current.

 

[c1ferrari]

An Amplitrex is an interesting and useful device; however, the final word, for me is predicated upon results from a Tektronix 576 Curve Tracer and curves found in an RCA tube manual.

 

[Barry]

For the benefit of those that might read this old thread, the Amplitrex AT-1000 is certainly a useful device and CAN run tube curves like the Tektronix in "computer mode" at any parameters you select, can provide simultaneous results at multiple bias voltages on one graph like those found in an RCA tube manual, and give you a print out. It can also compare different tubes' curves and multiple section tubes on one graph. The program's a little balky at times and some times hangs up. The instructions could be much better.

Based on a limited sample I've seen, I'd concur that ARC does an excellent job matching as micro states. The half dozen suppliers I've used seem to match within 10% or so but it depends on the tube. The 40 individual SED winged C 6550C's I tested last weekend had bias voltages and Gm values over a range of more than 25% so I don't think I'd want to take a chance on not getting those matched for example. There's that good article on RAM Labs site on tube matching that's worth a read and referenced above if you're not familiar with this topic. From what I have seen, if you burn tubes in for a few days and match them closely for both iP and GM at a given bias point, they should match as closely for any other operating points assuming you use a tester that employs real world operating voltages.

 

[Mendel]

Some interesting points here. I believe that the important thing to "match" in output tubes is the current draw of the tube in milliamps, GM matching is less important. That said, I agree that amps with individual bias pots for each tube are best and allow you to match the current draw for push/pull pairs very closely. Never liked the idea of biasing a "bank" of tubes as one tube in the bank may end up drawing far more current than another so the push/pull pairs become unbalanced.
As for suppliers that truly try to match output tubes by current draw, I recommend Jim Mcshane in the U.S. and thetubestore.com in Canada. They both mark the current draw of each output tube right on the box so the is no doubt. That said, plus or minus 10% is probably OK for most circuits although with individual bias pots you can equalize current draw very closely for better performance.

 

[MtnHam]

Wolcott amps (unfortunately out of business) auto bias each of the 8 output tubes (EL34's) every time the amp is switched on. The special computer chip adjusts the bias of each tube repeatedly until all are in spec, and then is switched out of the circuit. An LED for each tube blinks until all are balanced, and then stays on constantly, unless a tube fails to come into spec in which case it remains dark, indicating which tube that should be replaced. Matched sets are not necessary. IMO, all tube amps should be designed in this way, and some are.

 

[jazdoc]

One great piece of advice is that you never know how a tube will sound until you put it in the actual piece of gear....

 

[MMF]

Both Ia and S are very important for the matchining, but most of the sellers on Ebay don't have tube testers, that can measured Ia and S. Some other just have no conscience.The other problem is that they use tubetesters, that shows only "good" , "very good","bad" or other units that are incomprehensive. I don't belive to people, that don't have actual readings of Ia and S in their ads,the parameters are measured at and actual picture, that shows the tube with glowing heater on the tester. It is very difficult to find 2 tubes with the same parameters, and almost impossible 4 ones. I think, that you probably have to choose from over 100 tubes for 4 matched ones. Some months ago I started selling tubes on Ebay and all of my clients are happy with their purchases.

 

[GaryProtein]

I'm asking this because I need 8 matched and balanced 6350 and I've bought four sets from diferent but reputable sources (aparently, at least...), being one set bought directly from a major high-end brand that uses this tubes.
The problem is that before using them, I allways have them tested by a technician in whom I've absolute confidence to do it properly.
The results are appalling: not one of this sets was matched and balanced. Worst, in all of them diferences in mA were huge, not only on individual tubes but also between diferente tubes.
On all sets I've paid premium to get them properly matched and balanced and none was.

A few years ago I've read on a French professional áudio magazine an article about this problem. They bought a few lots of the same tubes from sellers in Europe, China and USA. Not one lot was matched and balanced!!

There seems to be a common pattern here....

What's your experience on this?

where I buy my tubes I always get the measured results of each tube I am buying

The sources are telling you they are balanced. One gives a measurement sheet, another may not.

I'm not a tube guy, and I'm just throwing this out there, but the cynic asks...

How do you know that your source is any better than another?

If you don't have the ability to test the tubes yourself, how do you know they aren't lying to you?

 

[c1ferrari]

Personally, I inquire/request triode section values for emission, Ia, and mutual conductance, Gm. Often, Ia is unavailable.
Based upon what is reported, I determine whether values constitute well-balanced sections. I also inquire about noise and microphonics.
All tubes are are evaluated upon receipt using an Amplitrex AT1000 and/or Tektronix 576 Curve Tracer. Tubes which do not satisfy parameters are returned or a price adjustment agreed upon with vendors.