Tube Replacement: Single, Pair, or Quad?


Hello, can anyone help me?

I have a new (to me) pair of Cary SLM-100s. Without fail one of my amps will blow a "tube" fuse after approximately 45 minutes of listening. Immediately prior to the fuse blowing, two of the tubes (positions 3&4) become very bright. After this happened a few times, I switched Tube #4 in one amp with Tube #4 in the other amp with similar results in that amp.

I am using Tesla KT88 power tubes (not JJ Tesla.) I have them biased to approximately 150ma.

My initial diagnosis was that I have a bad tube. However, my question is, if the tube is OK, why would it continue to operate correctly in 45 minute increments? Could something else be wrong? If I need to replace the tube, do I have to replace tubes in pairs (e.g., 3&4) since this is a push-pull design? Do I have to replace all four? Can I just replace the one bad tube?

Also, the original owner did not have this problem. Further, when I installed the tubes I did so randomly not knowing which tubes he had in which amp in what order he had them in. Is this relevant?

Thanx in advance for your assistance.
pawlowski6132
I'm not familar with the operating aspects of the Cary M100, however if it allows for individual biasing of each tube then you can replace one tube at a time, they don't have to be match BUT 150ma per tube would be way off the mark. Depending on plate voltage I would expect something more in the 50ma +/- area. If you have no provision for individual biasing of each tube but must bias the who quad then you need to replace (at least this time) the whole quad. 150ma for a quad would sound ok (in general terms).

I suggest you get a copy of the manual, or biasing instructions from Cary, poste haste.
I think it may be something very simple like how you are doing the the bias adjustment. Especially if (you didn't say) the 45 minute "blow time" is unrelated to listening levels. In other words, something besides the driving voltage is causing the tubes to begin drawing too much power after warm-up. (I could be wrong, but.....)

You should adjust the bias with the speakers connected, but the inputs disconnected, and only after the amp has been running for 20 minutes or so. If you don't do it this way, you may be inadvertantly setting the bias too high. Also, the bias for 6550's and KI88's is generally somewhere around 35-45 ma, Does the manual specify 150? I understand what Newbee says about 150ma for the quads, but I had two sets of quad 6550's in each of my ARC monoblocks, and only biased each set of quads to 32ma!

If re-biasing doesn't solve the problem, it could be a bad tube. You could replace the one tube without much loss in performance. A matched pair is better, but matched quads are really overkill.

If you think it is the tube, and/or the bias suggestions aren't relevant, I'd recommend getting a matched pair, and saving the one good one from the previous pair for a spare.
IMO, you should go to the Cary website, get their phone number and call Kirk on Monday with your question.
Thanx Lak. I did that already. They want me to send all eight tubes in and they'll test them for $125 bucks. I'm 99.99% sure I know which one is bad. Newbee, unfortunately, I can only balance the whole quad. The manual states 175ma, Cary technicians suggested 150ma. I keep it somewhere in the middle. Nsgarch, thanx for your suggestions too. I think I'll try to replace just a pair although, I can't find Teslas KT88s ANYWHERE. I'm sure the tubes are the issue now becuase I had a spare set of eight Svetlana 6550s which I listed to for four hours last night without interruption. Thanx all.