Output Tube Temps! Advice Please!
 

I just retubed the output tubes on my MC275 Mk VI (see my recent post of Psvane tubes).

I just checked tube temps and they are all over the place. V8 us 130F; V9 is 230F; V10 is 290F; V11 is 321F.

Is this normal? Do I have a bias problem? Or is this a tube problem? The vendor tested these tubes before sending out and they are within 5% for Ip.

I would check the grid bias of each tube ( if the 275 has provisions for consumers to check it). and set them as close as possible after the amp is warmed up. Also put the original tubes in and recheck their temps and bias.

I just put the original mac tubes back in waited 10 minutes and measured their temps. All were in the 250 degree F range--very consistent.

Then put new tubes back in but in different positions and after about 10 mins the tube that tested cold last time did again at 102 F. So the problem is moving with the tube. But the other three tubes were not terribly consistent either, with temps of 310, 356 and 373 F.

I then replaced the cold new tube with one of the factory tubes and left the other three in the same positions. As before, the factory tube stayed cooler than the new tubes at 240 degrees F. But the three remaining new tubes were now somewhat more consistent in temp but still showing the same progression. The tube that was 310 now 313, the one that was 356 now 328 and the tube that was 373 now 333 F.

My recollection is that the 275 does not have user adjustable bias controls. It still doesn't bar you from having matched tubes. Personally I would have the replacement tubes tested
to see how matched they are.
That said I understood the 275 is quite forgiving on tubes however as temperatures are a concern I would be tempted to exchange them for a better matched selection.

That's very good advise, I wasn't sure about the biasing scheme of the 275. A tube tester will tell all!!!! Being a cautious person I would stick with the factory tubes. I've heard this amp many times and thought it sounded great with original tubes. :thumbsup:

Sticking to factory tubes makes absolute sense, they're preselected by the manufacturer regarding their characteristic measurements.
For peace of mind have you considered a tube tester, many are of automatic operation. Two benefits apart from the peace of mind is that occasionally removing and replacing a tube is quite beneficial for base contacts.

Hi Rob,
Sorry to see you are having some trouble with the new tubes, especially since your report on the sound was really compelling and positive.

The 275 is a cathode bias amp, meaning there is no manual bias adjustment for each tube. Grid bias is developed by the voltage drop across a resistor in the cathode circuit.

Tubes can be matched at different operating points - an operating point is a specific combination of plate current, plate voltage, and grid voltage.

It is likely that your tube supplier uses a tester that checks and matches the tubes at a different operating point than they see in your 275.

At this point it would be worth sending the quad back and maybe even your good tubes to see if he can achieve a closer match to them.

I can send you a set of oem or “winged c” kt-88’s to use in the interim if you’d like.

There is an overall bias adjustment for all the tubes, the service manual has the adjustment procedure.

Sorry to geek out on you - just hoping for a great outcome.

Tom

Hope this tid-bit is interesting and provides some insight. Mil spec parts may have some different requirements (gold pins, metal case, etc.) but are simply selected from a very large batch to meet the tighter mil spec requirements. That said I would see quite a spread in specs when using a tube tester or transistor curve tracer. :yes:

Think I've Come to a Good Place
 

Thanks to everyone for the input and thanks especially to Tom for your kind offer! But I couldn't stop trying to figure this out. So now at 8:30 PM on a Sunday night after messing with this all day, I'm gonna declare victory!

Here's the Executive Summary: The Psvanes are now running 9 degrees cooler than the Factory tubes and the variance in their temperatures is the same as the Factory tubes. I think I'm gonna call this normal, exchange the one bad Psvane and be happy!
:beerchug:
If anyone's interested in how I got here, OR if you have a keen interest in tube temps in a MC275 VI OR you're obsessive like me, here's what happened: I decided to check and reset the bias and see the effect. While I had the amp lying on its side and out of the shelf I found I could get a clear shot at the center top of each tube with the infrared thermometer. And this gave more consistent results.

I had it running with 3 Psvanes and 1 Factory tube (in place of the dead Psvane) and took temps before and after resetting the grid bias from where I found it at -54VDC up to -61VDC where it should be (Mac specs -60 to -62VDC). I have NO IDEA why, but the 3 Psvane tubes dropped in temp an average of 18+ degrees after the bias adjust! The Factory tube stayed the same temp. :confused-2:

After the bias adjustment I took the temps of each Psvane a few times 5-10 mins apart, averaged across the temps I got for each tube, then averaged across the three Psvanes. Average temp was 141 degrees F. The spread between the coolest and hottest Psvane was 8+ degrees with the coolest tube running at 94% the temp of the hottest tube.

I then put all 4 factory tubes in place and did the same with them. Average temp across the 4 Factory tubes was now 150 degrees. And, with the more reliable way of taking temps, the Factory tubes no longer looked perfect. The spread between the coolest tube and hottest was--guess what!?--8+ degrees with the coolest tube running at 94% the temp of the hottest tube.

There you have it. Based on a set of 2, I think I'm gonna go out on a limb and say some variance in output tube temps is normal and OK. :thumbsup:

Oh almost forgot. Jumping off from Tom's mention of the operating point as a variable, I checked voltages at the output tube sockets. Have no way to check current. Control grid voltage was slightly different than the grid bias as measured at the bias checkpoint. It varied from -61.2 to -63. Plate voltage varied closely around 500 volts, from 499.3 to 500.9. The screen grid voltage was only a volt or so less, and spot-on consistent across all four tubes at 498.5. I tried to correlate the small differences in bias voltage and plate voltage with the temp differences but saw no coorelation.

Congratulations and well done for your perseverance. You can now listen in a wholly well earned relaxed and contented state.
Allan