Tube Testing (Trust)

Discussion in 'Amps and Cabs [BG]' started by Basslice, Sep 29, 2020.


  1. Basslice

    Basslice Supporting Member

    May 11, 2008
    Western Massachusetts
    I have been moonlighting over at Antique Radio Forums. I have been interested in testing power tubes. I've been using a lot of JJ tubes and bought a bunch that were supposedly matched from tube depot only to have a local guy near here with a old tester to tell me that they were not matched.

    That got me thinking. We Valve Amp people are at the mercy of trusting people we buy from. Who is really to say if that NOS GE 7027A is any good. Tube Depot JJ7027A's that I bought came with a transductance number - but no indication of how it was measured. Considering we pay so much for these increasingly rare tubes, it would be good to have some trust in numbers and not be at the mercy of the seller.

    Most tube tester rigs are subject to a lot of error and each model is designed differently and tells you different things. Just bringing a tube to someone and having them stick it in a tester really doesn't guarantee anything.

    So I am designing a tube tester that is built from lab equipment and will take look at all my power tubes. At this point, I am only designing it to test octal power tubes and compare the measurements to RCA data sheets.

    Heck, maybe I will start a side business to scientifically certify tubes?
     
    bobyoung53, arbitrary, Low8 and 6 others like this.
  2. agedhorse

    agedhorse Supporting Member Commercial User

    Feb 12, 2006
    Davis, CA (USA)
    Development Engineer-Mesa Boogie, Development Engineer-Genzler (pedals), Product Support-Genz Benz
    Do you know what parameters you want to test for and at what points along the curve(s)?

    Matching curves with a single point or value is a fool’s errand unless that particular value has meaning.

    2 differently shaped curves could match at a single point yet be wildly different at other points.

    Do tube vendors cheat? Absolutely. Do those accuse tube vendors of cheating with their different measurements know what they are talking about? Often not. Do buyers of such tubes demand rediculous things from tubes? Yes, quite often... therefore feed into the whole mystique of the tube underworld.
     
    quickfix, FretFree, PillO and 10 others like this.
  3. Basslice

    Basslice Supporting Member

    May 11, 2008
    Western Massachusetts
    Well, with the way I am setting it up I can get a reliable measurement of Gm at real operating voltages and supplying a 1 VAC 1 kHz signal.

    I will also be able to scan plate voltage and screen voltage with bias and recreate tube spec curves if I want.
     
  4. Passinwind

    Passinwind I know nothing. Commercial User

    Dec 3, 2003
    Columbia River Gorge, WA.
    Owner/Designer &Toaster Tech Passinwind Electronics
    When you get there be sure to update your profile to Commercial User and get ready for a pretty different posting experience here. ;)

    You might want to ask Tube Depot about their matching system, I wouldn't be quite so quick to vilify them with getting both sides of the story.
     
    PillO, equill, superblues and 7 others like this.
  5. Basslice

    Basslice Supporting Member

    May 11, 2008
    Western Massachusetts
    Yes, true. I figured once I test some tubes in a known way I can figure out if the mismatch measured by the local guy is real or not. I don't mean to vilify anyone, but there is some cause for concern. Especially when the average buyer really has no idea about the characteristics of any particular tube. A case for caveat emptor, right? Unfortunately, 90% of the tube testers out there in basements or shops are over 50 years old and different model (or even same model different unit) have been shown to make very different measurements. Interpreting the results is difficult.

    Again, if I was buying a $5 tube, who cares? But when people are listing tubes for $50 or more per tube, it would be useful knowing if it is really worth the money. Same for paying additional for a matched set.​
     
    FretFree, Happy Face and agedhorse like this.
  6. Passinwind

    Passinwind I know nothing. Commercial User

    Dec 3, 2003
    Columbia River Gorge, WA.
    Owner/Designer &Toaster Tech Passinwind Electronics
    It can be a real can 'o worms, hence my suggestion to make sure you and Tube Depot are looking for the same criteria. FWIW, I went to school with the designer of the Apex matching system: How Apex Matching® Works | Apex Tube Matching. I've yet to try any tubes matched on the Apex system though, as my usual vendor only offers it on other brands than what I've been using recently.
     
    Basslice, Jim C and wave rider like this.
  7. agedhorse

    agedhorse Supporting Member Commercial User

    Feb 12, 2006
    Davis, CA (USA)
    Development Engineer-Mesa Boogie, Development Engineer-Genzler (pedals), Product Support-Genz Benz
    So matching points and matching curves at various operating voltages and currents and screen voltages yield entirely different sets of information. I think you will find that communicating this information to the consumer will be many times harder than generating the information.
     
    PillO, equill, M0ses and 8 others like this.
  8. Basslice

    Basslice Supporting Member

    May 11, 2008
    Western Massachusetts
    Whoah! Agreed. I am not jumping into commercial testing. I was just saying that it could be an interesting side gig, albeit just from a scientific point of view.

    My interest in all of this is fairly just academic. It is also being fed by Covid time expanding my hobby time.

    I read "Getting The Most Out of Vacuum Tubes" by Robert C. "Bud" Tomer. It is clear that testing one tube characteristic probably won't give you a good idea how well it works. However, you can start to exclude tubes.

    My first goal is to get the tester running. What information I get I will share. I am not going to change, impact or dent the last 100 years of tube knowledge. None of this stuff is magic, but it is becoming a lost art and harder to make comparisons.
     
    FretFree, Happy Face, Bent77 and 2 others like this.
  9. Basslice

    Basslice Supporting Member

    May 11, 2008
    Western Massachusetts
    I agree, but up till this point it is kind've useless for me to talk to Tube Depot about this because I don't have any data other than the guy that tested my tubes verbally saying they were not matched. I have no reason to believe that his testing method was good or poor, but he uses it. Now if I can make some measurements and see things wildly different than the specs provided, I could say something.

    I am not even saying there is a problem. Mostly this is an academic endeavor for me. I find this stuff interesting and cool.
     
    Happy Face and BassikBrad like this.
  10. Passinwind

    Passinwind I know nothing. Commercial User

    Dec 3, 2003
    Columbia River Gorge, WA.
    Owner/Designer &Toaster Tech Passinwind Electronics
    Roger that. By "specs provided" do you mean the published generic characteristic curves, or? You mentioned TD giving a transconductance number with no qualifiers, why not at least ask them for that much, and the guy who tested them for you too for that matter?
     
    BassikBrad and agedhorse like this.
  11. Basslice

    Basslice Supporting Member

    May 11, 2008
    Western Massachusetts
    You know, I just read your link to Apex. Thanks. It rang a bell. The tubes I got were Apex matched. I recognized the logo. How the local guy tested was on some bench tester.
     
    Passinwind likes this.
  12. Passinwind

    Passinwind I know nothing. Commercial User

    Dec 3, 2003
    Columbia River Gorge, WA.
    Owner/Designer &Toaster Tech Passinwind Electronics
    First things first: how do they sound in your amp? The last quartet of "matched" EL34s I bought have around a 5% variance in idle current and they sound fantastic, so I never bothered to go any deeper. I might have to ping the Apex guy though, like you I find this stuff pretty interesting. The very experienced local amp repair guy here built his own computerized matching rig and I was going to help him with some modeling correlation stuff on a work barter, but so far the timing has never quite worked out. Looking forward to seeing what you come up with.
     
    Happy Face and BassikBrad like this.
  13. Basslice

    Basslice Supporting Member

    May 11, 2008
    Western Massachusetts
    That is a good point. I have a pair of unmatched vintage 7027A tubes lying around (my V4-B requires four so I bought 2 new matched pairs). They had measured poor before I mothballed them. Then I rebuilt a B-25 and bought new JJ 7027A's for that (it takes two).

    The rebuilt amp sounds great. For kicks, I put in the old mismatched, "poor" tubes in and guess what? It sounds great....
     
  14. Passinwind

    Passinwind I know nothing. Commercial User

    Dec 3, 2003
    Columbia River Gorge, WA.
    Owner/Designer &Toaster Tech Passinwind Electronics
    I'm pretty well known for my motto: if it ain't broke fix it anyway.

    The corollary is kind of inportant though: don't try this at home. :cool:

    But seriously, asking the right questions is really key to getting the best results, IME and IMO.
     
    FretFree, BassikBrad and BassmanPaul like this.
  15. seamonkey

    seamonkey

    Aug 6, 2004
    Here's some great information from Mesa Boogie on how they test, it's quite amazing.
    How are MESA tubes tested?
    It's how they can offer the longest warranty in the industry on their tested tubes.

    You might want to take a look at mbrennwa/PyPSUcurvetrace
    There's some tube examples down the page.
    "Python" is good to learn these days. There's a huge user base, and a library for practically everything.
     
    keith rosier likes this.
  16. BassmanPaul

    BassmanPaul Inactive

    I have two tube testers. Now a Heathkit that I use for go, no go tests and the other a Hickok that I've not used. The paper roll of characteristics fas fallen to pieces and I need to find a new one or a book to replace it.
     
  17. beans-on-toast

    beans-on-toast

    Aug 7, 2008
    Tube Depot keeps coming up with complaints related to their matching and customer service.
     
    Wasnex and Basslice like this.
  18. agedhorse

    agedhorse Supporting Member Commercial User

    Feb 12, 2006
    Davis, CA (USA)
    Development Engineer-Mesa Boogie, Development Engineer-Genzler (pedals), Product Support-Genz Benz
    Did the guy’s bench tester test at operating plate voltage and load?
     
    Wasnex and DavidEdenAria like this.
  19. beans-on-toast

    beans-on-toast

    Aug 7, 2008
  20. My V-4 has 4 7027As that when I stuck the Eurotubes bias probe , the tubes were mismatched maybe 20%. I forget now. My Sunn 2000S > 6550s were closer, but still off. But, both amps sound awesome. So my conclusion is that matching may be a little more anal than need be.
     
    96tbird and MYLOWFREQ like this.
  21. Primary

    Primary TB Assistant

    Here are some related products that TB members are talking about. Clicking on a product will take you to TB’s partner, Primary, where you can find links to TB discussions about these products.

     
    Dec 1, 2021

Share This Page