SVT-CL: Mixing Power Tubes

[dimitris rigas]

Invalid Link Removed

These guys are in Germany. I've had good luck with their branded tubes in a number of amps as well, though I haven't tried their 6550. They do sell a matched sextet.

Invalid Link Removed

thanks!

 

[JimmyM]

Do we really think these 'rejects' get destroyed? I wonder how many get re-sold.

I am down to maybe 7 or 8 USA 6550 that test strong....GE, Sylvania and a few very rare real Tung Sol 6550....the 'Ei' brand was another with a bad reputation in power tubes though the 12ax7 can sound great but often go microphonic (a sort of Telefunken copy).

I'm down to the six JJ's that are in my HSVT-CL...the only 6 I ever owned And they've been fine for 6 years, though the last couple years haven't put much wear and tear on them. First 4 sure did, though!

But in all honesty, I have never heard a difference with power tubes in my life with any amp I ever owned, so I don't sweat it so much and I just look for the cheapest ones I can get with a 24 hour burn-in and matching. I expect that not to change with the HSVT-CL when the time comes.

 

[Wasnex]

Regarding the earlier conflict and subject title, here's my experience.

When you put a tube into a circuit there is an interaction between the tube and circuit. Because the tube impacts circuit conditions, it also impacts other tubes that are in the same circuit. This interaction seems to have been discounted by one side of the discussion.

Say you have two 6550s in one side of the push pull circuit of an SVT and they are biased to 24mA each. If you install another 6550 in the circuit, the bias of the first two tubes will not remain static at 24mA. The amount the current changes will depend upon the operating characteristics of the new tube that is installed in the circuit.

My observation is that if you put a tube that is not reasonably matched then it may not work out very well. Either one tube will hog the current and two tubes will barely conduct, or two tubes will hog the current and the third tube will barely conduct. Ideally you want all three tubes to be pulling the same amount of current within about +/-15%, so all three tubes have to be fairly closely matched.

So if you have five good tubes in an SVT, I don't agree with buying a pair of tubes and putting one in the push side, and one in the pull side. If one side has three relatively well matched tubes, leave them alone. If you have a basket of misfit tubes, you can roll tubes until hopefully you find one that biases up well with the two good tubes. You will need to open the amp and take voltage readings to determine the individual current of each tube in order to ensure a reasonable match. Otherwise, IMHO, your best bet is to install a matched trio.

I have observed this behavior in several different amps. For example I have an amp with four output tubes and one bias control.
The tubes are ordered (1,2) OT (3,4).
I had a situation where tubes 1 & 2 were matched within a few mA and tubes 3 & 4 were matched within a few mA, but the current across the transformer was not balanced, so there was hum. The tubes on the right were pulling about 15% more current each than the tubes on the left.

So the logical solution would be to re-sort the tubes so there was one hot tube and one cold tube on each side of the OT. If the matching was right at 15% this would be fine IMHO. For example (1,4) OT (2,3) Unfortunately sorting the tubes in the way resulted in about an 80% mismatch between tubes 1 & 4 and between tubes 2 & 3. I didn't change the bias I just re-sorted the tubes, and the hot tubes got hotter and the cold tubes go colder. IMHO the change was fairly drastic.

IMHO your are better off just replacing the one bad tube, than installing a mismatched tube in the trio that is still working. Leave the working trio alone. If one of the trios are not matched, use the bias lights to the total current in the proper range, and then adjust for minimum hum.

 

[agedhorse]

Do we really think these 'rejects' get destroyed? I wonder how many get re-sold.

I am down to maybe 7 or 8 USA 6550 that test strong....GE, Sylvania and a few very rare real Tung Sol 6550....the 'Ei' brand was another with a bad reputation in power tubes though the 12ax7 can sound great but often go microphonic (a sort of Telefunken copy).

Define reject... tubes that are non-functional or obviously defective are almost always destroyed (or turned into Christmas tree ornaments), but those that are a little off spec (and even this has many meanings) are rarely destroyed because one customer's "screen-outs" may be exactly what another customer is looking for (or believes is the holy grail).

 

[DavidEdenAria]

I'm down to the six JJ's that are in my HSVT-CL...the only 6 I ever owned And they've been fine for 6 years, though the last couple years haven't put much wear and tear on them. First 4 sure did, though!

But in all honesty, I have never heard a difference with power tubes in my life with any amp I ever owned, so I don't sweat it so much and I just look for the cheapest ones I can get with a 24 hour burn-in and matching. I expect that not to change with the HSVT-CL when the time comes.

The tubes need to burned in at full operating volts, thats the difference....know your vendor....for years 'Ruby tubes' had a good reputation of screening tubes.

Some vendors i know burn certain brands of power tubes in for 48 hours....some of it is slight bias drift....i have several tube testers to screen for transconductance, shorts etc but the tubes go into a live amp for many hours to monitor bias stability, noise etc....of course all power tubes will drift a bit in current draw as they wear but if you have a weak worn tube with weak mutual conductance that is idling over 70% dissipation it wont last long and you should never have a tube idling over 70% dissipation....despite what Eddie Van Halen did LOL!

 

[DavidEdenAria]

Regarding the earlier conflict and subject title, here's my experience.

When you put a tube into a circuit there is an interaction between the tube and circuit. Because the tube impacts circuit conditions, it also impacts other tubes that are in the same circuit. This interaction seems to have been discounted by one side of the discussion.

Say you have two 6550s in one side of the push pull circuit of an SVT and they are biased to 24mA each. If you install another 6550 in the circuit, the bias of the first two tubes will not remain static at 24mA. The amount the current changes will depend upon the operating characteristics of the new tube that is installed in the circuit.

My observation is that if you put a tube that is not reasonably matched then it may not work out very well. Either one tube will hog the current and two tubes will barely conduct, or two tubes will hog the current and the third tube will barely conduct. Ideally you want all three tubes to be pulling the same amount of current within about +/-15%, so all three tubes have to be fairly closely matched.

So if you have five good tubes in an SVT, I don't agree with buying a pair of tubes and putting one in the push side, and one in the pull side. If one side has three relatively well matched tubes, leave them alone. If you have a basket of misfit tubes, you can roll tubes until hopefully you find one that biases up well with the two good tubes. You will need to open the amp and take voltage readings to determine the individual current of each tube in order to ensure a reasonable match. Otherwise, IMHO, your best bet is to install a matched trio.

I have observed this behavior in several different amps. For example I have an amp with four output tubes and one bias control.
The tubes are ordered (1,2) OT (3,4).
I had a situation where tubes 1 & 2 were matched within a few mA and tubes 3 & 4 were matched within a few mA, but the current across the transformer was not balanced, so there was hum. The tubes on the right were pulling about 15% more current each than the tubes on the left.

So the logical solution would be to re-sort the tubes so there was one hot tube and one cold tube on each side of the OT. If the matching was right at 15% this would be fine IMHO. For example (1,4) OT (2,3) Unfortunately sorting the tubes in the way resulted in about an 80% mismatch between tubes 1 & 4 and between tubes 2 & 3. I didn't change the bias I just re-sorted the tubes, and the hot tubes got hotter and the cold tubes go colder. IMHO the change was fairly drastic.

IMHO your are better off just replacing the one bad tube, than installing a mismatched tube in the trio that is still working. Leave the working trio alone. If one of the trios are not matched, use the bias lights to the total current in the proper range, and then adjust for minimum hum.

You dont need to stick your hand inside an amp to check current draw, just buy a Bias Rite (the ones with plate volt option even better).....also, at least visually check the the bias circuit and screen resistors to see if any leaking, bulging caps or burned resistors (which can be hard to see without magnifying).....keep both hands out of an amp for sure...one hand behind your back

Best is a tech!

 

[Wasnex]

Were these tubes burned in at operating volts by the vendor and then matched?

Thats a shame and seems unusual but something tells me these tubes were not screened properly.....man, that is scary.

They came in used amps I bought. They were OEM in one of the amps for a short period of time. The amp was near mint when I received it, and had 5 out of the 6 original tubes. The original owner had a failure right out of the box. The company has gone back to JJ output tubes again. For context, I have 20 tube amps, and have bought all of them used. The ones that had a notable amount of tube problems all arrived with EHX 6550s or KT88s.

 

[Wasnex]

Do we really think these 'rejects' get destroyed? I wonder how many get re-sold.

I think there is a difference between a reject and a tube that has failed. I think the rejects are returned and sold to someone else.

 

[DavidEdenAria]

I think there is a difference between a reject and a tube that has failed. I think the rejects are returned and sold to someone else.

Im referring to what the 'tube vendor' does with tubes that fail certain parameters in the burn in....not tubes that were sold to someone and failed and returned but that too is a concern.

I trust that most big name vendors do NOT want their tubes (with often their labels printed on them) to go in circulation...the rumors of rejects getting 're-circulated' thru the decades is nothing new fwiw.

 

[agedhorse]

Retail tube vendors typically destroy for credit or return for credit anything that is truly defective. Tubes that "screen out" are often re-sold to another customer or re-labeled, or placed into a category of a lower spec.

High volume tube dealers may strip their name off of the tube and sell to another dealer at a big discount, or sell to an ebay dealer as a no-name or "your name here" part at a big discount. We may be talking a hundred tubes a week to "dispose" of.

Most big manufacturers of amps will screen in-house and return those tubes that do not meet their own spec. There may be nothing wrong for the average user or less critical application, and these often enter the retail chain as a brand name of one sort or another. They may be perfectly good, or they may have minor issues. With a large manufacturer, this may end up being a few THOUSAND tubes a month screened out.

 

[beans-on-toast]

There are a lot of companies selling what they call matched set that are not well matched. Not all companies burn-in the tubes before matching them. You really want the tubes matched under the operating plate, screen, and bias voltages of the amp. Some vendors do this, most don’t.

Pre matched sets that some vendors sell are matched at plate voltages of the “average” amp. This is more common these days. Vendors leave the matching another company.

Then there are the ones that simply use an uncalibrated tube tester testing at a low plate voltage.

How close the matching tolerance is needs to be considered, is the cathode current within 2%, 5%, 10%, 20%...it matters. Is the tube tested for noiseand microphonics? So many things to consider.

There’s the good, the bad, and the ugly out there. Finding a good reseller that matches tubes with a high attention to detail is a big part of the battle.

 

[Wasnex]

There are a lot of companies selling what they call matched set that are not well matched. Not all companies burn-in the tubes before matching them. You really want the tubes matched under the operating plate, screen, and bias voltages of the amp. Some vendors do this, most don’t.

Pre matched sets that some vendors sell are matched at plate voltages of the “average” amp. This is more common these days. Vendors leave the matching another company.

Then there are the ones that simply use an uncalibrated tube tester testing at a low plate voltage.

How close the matching tolerance is needs to be considered, is the cathode current within 2%, 5%, 10%, 20%...it matters. Is the tube tested for noiseand microphonics? So many things to consider.

There’s the good, the bad, and the ugly out there. Finding a good reseller that matches tubes with a high attention to detail is a big part of the battle.

How has your luck been with Apex matching? When they first started, they had quite a few models offered with burn in service. Now most of them don't seem to be available with burn in. Their description of the matching process reads really well. Basically they claim to match the tube at multiple operating voltages. But if the tubes are not burned in first, I don't think there is much efficacy to matching them.

I have only bought one set of Apex matched tubes, a quad of Sino EL34Bs. The matching was right at 15%, which IMHO is barely acceptable. The previous tubes in the amp were perfectly matched TADs, but I did something stupid and ganked them. Anyway, I haven't been in a hurry to buy any more Apex matched tubes based on my experience.

The last quad of JJ KT88s I bought from Eurotubes were perfect. The last quad of Ruby KT88V-STRs I bought were a mess and I should have sent them back. Looks like they may be out of production.

 

[beans-on-toast]

How has you luck been with Apex matching? When they first started, they had quite few models offered with burn in service. Now most of them don't seem to be available with burn in. Their description of the matching process reads really well. Basically they claim to match the tube at multiple operating voltages. But if the tubes are not burned in first, I don't think there is much efficacy to matching them.

I have only bought one set of Apex matched tubes, a quad of Sino EL34Bs. The matching was right at 15%, which IMHO is barely acceptable. The previous tubes in the amp were perfectly matched TADs, but I did something stupid and ganked them. Anyway, I haven't been in a hurry to buy any more Apex matched tubes based on my experience.

The last quad of JJ KT88s I bought from Eurotubes were perfect. The last quad of Ruby KT88V-STRs I bought were a mess and I should have sent them back. Looks like they may be out of production.

Their process is good, if they follow their own procedures. 15% is not that great, I would have sent them back. I have bought some Apex matched tubes from AES and they were within 5%. AES has been good for hard to find tubes. I buy most of my tubes from thetubestore these days. tctubes is excellent. Never had a problem with their products. Not to say that it can’t happen but that’s what the warranties are for.

 

[seamonkey]

I recommend and anybody wanting to see high end vacuum tube testing take a look at Mesa Boogies site:
Invalid Link Removed

Read and watch it. You'll notice tubes are not printed/labelled until they pass all their tests. I can only hope other high end manufacturers also do the same. Mesa does double the warranty period over other tube brands so I doubt that others do test as much testing. Other's don't show they do as much testing.

Also Mesa is not saying what factory they get their tubes from. For example: Invalid Link Removed
This is probably because Mesa can do a better job testing tubes than the actual manufacturers can, and tubes within Mesa's specs and quality actually are not dependent on a single manufacturer.

The makers of modelers, simulators, and profilers, are dependent on the reference amplifier used to reproduce the intents of the original designer. A good set of tubes that match the original designers intent are needed. I'm sure Mesa Boogies and IK Multimedia followed practices like this to create the Amplitube models, along with critical listening skills.