I'd ask LTA, and thinking while tubes might bias close when measured manually within the amp itself. However, the slight variation in thickness of tube glass thickness at the top dome of each tube (i.e. hand made tubes) could yield some variation in temperature, a 20 degree spread though, hard to say. LTA should know something more. Good luck.
Power tube operating temperature a good proxy for power output???
Background:
Recently got a lovely Linear Tube Audio ZOTL40 Reference tube amp. It runs 4 pcs EL34 power tubes, autobiased, 2 driving each channel. Being auto-biased, the user does not see the actual biasing condition at each tube (or pair of tubes). Amp sounds wonderful, works perfectly. One of the reasons I like this Berning design amp so much is that it runs the power tubes very easily, thus allows for very extended tube life -- especially attractive to me as I still have several quads of UK made EL34's and KT77's collected in the 80's and 90's, not the current Russian reissues... the real stuff. I can run these with long tube life.
Just for fun, I used a laser thermometer on hand (from track driving, checking tire temps) to check the operating temperatures of each tube... found there is a decent variation in temperatures (160-185-185-178 degrees F) at the exposed top domes of the 4 tubes in the amp...
Question:
So this raises a question in my mind... would such a 10-15% variation in individual tube operating temps be a fair indication or proxy for output level of each tube? The 4 tubes do sit in a row, so the two end tubes running cooler makes sense as they see cool air on one side, while the middle tubes don't... and of course, even identically constructed tubes have meaningful variances... we can also test GM in a tube tester -- but those test parameters is not close to real operating voltages and currents...
Input welcome... thanks in advance. No problem here -- just curious to know, from those very experienced with tube amps and tube amp design, if you have an informed point of view on this.
Recently got a lovely Linear Tube Audio ZOTL40 Reference tube amp. It runs 4 pcs EL34 power tubes, autobiased, 2 driving each channel. Being auto-biased, the user does not see the actual biasing condition at each tube (or pair of tubes). Amp sounds wonderful, works perfectly. One of the reasons I like this Berning design amp so much is that it runs the power tubes very easily, thus allows for very extended tube life -- especially attractive to me as I still have several quads of UK made EL34's and KT77's collected in the 80's and 90's, not the current Russian reissues... the real stuff. I can run these with long tube life.
Just for fun, I used a laser thermometer on hand (from track driving, checking tire temps) to check the operating temperatures of each tube... found there is a decent variation in temperatures (160-185-185-178 degrees F) at the exposed top domes of the 4 tubes in the amp...
Question:
So this raises a question in my mind... would such a 10-15% variation in individual tube operating temps be a fair indication or proxy for output level of each tube? The 4 tubes do sit in a row, so the two end tubes running cooler makes sense as they see cool air on one side, while the middle tubes don't... and of course, even identically constructed tubes have meaningful variances... we can also test GM in a tube tester -- but those test parameters is not close to real operating voltages and currents...
Input welcome... thanks in advance. No problem here -- just curious to know, from those very experienced with tube amps and tube amp design, if you have an informed point of view on this.
7 responses Add your response
So, the thing to keep in mind is that the power dissipated by a tube is a function of tube efficiency, signal output AND idle power. So, a variation in tube temp may be nothing more than seeing how each tube is biased, but not an indicator of their relative output to the speaker. My guess is that if you tested the amp near full power the temps would be a lot closer. Of course, it can also vary due to proximity to other surrounding components. Would be cool to know if you swapped the tubes, does the temperature variation follow? |
Tubes should be equally hot across the amp for even tube wear. If the inside tubes are hot that's too much camber. If the front tubes are hot that's too much roll stiffness, you're pushing, and this can be solved with some combination of springs, bars, or air, er I mean bias. Wait a minute. Are these bias ply tubes or radials? |
OP, I would wonder on that one, myself. I’ve used thermal guns, sense they became available to mechanics. Take your readings at several different spots. The top may get the hottest, but that’s not the actual plates, wires or glass. I suggest you put 6 dots on the valves, with a marker. I would aligned the marker dots with the pins. Upper, mid and lower on the valve in two marked spots per, Upper/Mid/Lower. I suggest a GOOD warm up AND playing at a given volume. I’d do a Low, Mid and High volume, test, too. When you collect your data, average, and see if thing look more in line. I suspect "Idling", is one set of numbers. Different loads, will change the temps, and the numbers will start looking closer.. Remember YOUR hand position too. If it's deck height and 6 " away. do the same for all. It's the most important part if you really want to know. "The reference" points, have to be the SAME. The expensive valves you have, for the sake of keeping them looking good. Be careful removing your marks (if you do) Isopropyl can take off the silkscreen, or information on the glass valve, GEs, RCAs, Tellies, Volvos, Mazdas, (you know the powder wipes right off.) Happy holiday, happy listening Regards |
The output tubes are designed to run hot. In a typical operation, I measured the turkish proxy output of the 7591 and EL84 types running north of 300 degrees Fahrenheit on a glass jacket, even if it seems a bit daunting compared to the world of solid-state components. In AB1 two-cycle mode, even under no-load conditions (no signal), at 34 mA cathode current, your 7591 tube plates dissipate about 13 watts per piece, and some standard Fisher devices have heated lamps much more than that. That's like having a 60-watt bulb under the hood, not even considering the heat from the shields and other tubes and components such as power resistors. |