Tube Bias issue.

Hello all, I'm going to post my still unsolved problem again. I have a Rogue Zuess amp that you set the bias for each individual tube. The amp can be played in either 150 watts per channel in Triode mode or 225 watts per channel in Ultra-linear mode. I can set the bias in Triode mode at 40mv and it holds and works/sounds perfectly. If I switch the amp to Ultra-Linear mode both the left and right channel bias meter needles fluctuate wildly. The amp has been to Rogue and elsewhere and the result I'm told is that the amp is fine. I replaced the tubes from KT88's to KT120's which was a great upgrade but the Ultra-linear bias issue remains the same. I have Martin Logan Summits (2007) a Rogue Magnum 99 pre-amp, Sony XA9000ES SACD, Shunyata Hydra, all higher end Audioquest interconnects and speaker cables, mid level
Shunyata and Acoustic Zen powercords. I also put everything on 20amp dedicated circuits to no avail. I've tried switching cables, power cords etc.. unfortunately I haven't been able to switch out the speakers which seems like the only thing left. Could a crossover gone bad in one of the Summits or one of the speakers powered subs have an issue that could affect tube bias settings? The system sounds perfect in Triode mode but the speakers really bloom better with the extra power in Ultra-Linear mode that I miss. Any ideas at all would be so greatly appreciated. Thanks Robert.
The issue is in your amp, but the question is, does the bias actually fluctuate wildly in UL mode, or does your ability to measure the bias fluctuate widely. Please understand that the two are completely different things.

If you set the bias in triode and then change the switches to UL are you seeing any indication that the state of the tubes are changing, such as red plating in one tube? Or even a cold tube?

You should be able to directly measure bias across the cathode resistors, but if you are unsure of how to do this it is best to leave it to a tech or experienced hobbyst as it will entail removing the cover of the amp.
Marty (Viridian), you may find it helpful to take a look at the biasing procedure described in the manual for this amp. I would think that the wildly fluctuating meter reading is most likely a true indication that the bias is wildly fluctuating.

Rluthy, it's hard for me to envision how a speaker defect could be causing this problem. Bias current is DC, which flows on the primary side of the output transformers, while the speakers are of course connected to the secondary side. And transformers don't pass DC, aside possibly from extremely miniscule amounts of leakage. And in any event it sounds like the speakers don't have any defects, based on their performance when the amp is in triode mode.

My speculative guess is that some sort of oscillation is occurring, involving an interaction between the amp and either the speakers or the preamp. An oscillation involves a path that is a closed loop, and I'm guessing that part of that loop may include the effects of the AC current draw of either the Zeus amp or the speaker's bass amp on the AC line voltage being seen by either the other of those two components or the preamp. Some thoughts and questions:

1)Do the woofers appear to be moving when the problem occurs?

2)Are you following the manual's instruction to turn off the preamp when doing the adjustments?

3)Even though the manual says it's ok to change from triode mode to UL mode while the amp is powered up, have you tried changing into UL mode while the amp is turned off, and then turning the amp on and observing the bias meters?

4)Try plugging the speakers into a different AC branch than the one the amp and the rest of the system are plugged into, using a long extension cord if necessary.

5)Beyond that, all that occurs to me to suggest at this point is to go to Radio Shack and pick up either a couple of cheap speaker drivers, or alternatively a pair of 8 ohm resistors rated to handle substantial power levels (I think they sell one rated for 20 watts), and seeing what happens with those connected in place of the Summits. Although if the bias readings do not fluctuate with those dummy loads in place, per my comments above it would not necessarily be indicative of a speaker defect. My suspicion in that case would be that some kind of interaction between the amp and the speakers is resulting in an oscillation.

Good luck. Regards,
-- Al
Al, thanks so much. I think that using power resistors, as you suggested, is the easy proof of your theory.

The panels in those MLs are probably highly capacitive so that may support your supposition as well.

The Stereophile review of the M-180 (they did not test the Zeus), shows good behavior in triode mode but a single cycle of undamped ultrasonic ringing in UL. They did not test the amp with a capacitive load, and it is a different model so I am not sure that any inferences can be made from this.
does it occur on both output posts?(4 and 8 ohm)
Thank you all for your most appreciated knowledge and replys. To answer some questions, the tubes don't red plate, they still look stable and normal while the meters are crazy in UL mode.
Yes I've tried the 4ohm and 8 ohm taps, same issue.
Yes the preamp is off and I've also disconnected all but the speakers, same problem.
I've also unplugged the speaker power cords just leaving the speaker cables on, same issue.
Yes I've tried every variation with outlets, dedicated, not dedicated, the Hydra, no Hydra etc...
I will try to get some BS speakers if you think that would tell me something.
Thank you all again for your thoughts and suggestions, to be continued. Happy Holidays.
Fluctuating bias is huge problem that can destroy tubes very quick. At least the fluctuation should stop after short period of time less than a minute before you can play the music through(which is acceptible but not perfect).
There's definitely something bad when you connect to UL power supply(second grid). My wild guess would be going through applicable electrolytic caps that are part of UL supply circuit. I'd usually ID second grid pin# and trace all the way back to rectifier examining visually and then testing with DMM.
If the amp is still on warranty, use your old tubes in UL mode with music till they start blowing and destroying amp before you send it to the rebuild service back to Rogue(this way they will definitely know that the amp is NOT OK). You can use dummy load resistor to save your speakers while 'cooking' your amp. Repeat procedure while the amp is perfect or demand another unit.
If the bias is OK in triode mode and without a preamp attached, then we can assume that it is not noise coming into the amp from the preamp.

We can also assume that its likely that its not noisy driver tubes.

This leaves the switching means itself. My theory is that the switch that is used to go from Triode to UL is dirty. This might be an easy fix if this is the case. Get a spray can of contact cleaner from Radio Shack, or a spray can of Deoxit D (you never ever want to use anything more powerful than these products on a tube device BTW!).

Locate the switch, and with the amp off, find a way to get a shot of the spray into the switch using the pipette that is supplied with the contact cleaner. Then operate the switch vigorously about 15-20 times. Then try it out.

If that solves it- great! If not, at this point unless you have a strong technical background (and from the posts here I am assuming you do not) then it must be repaired by a competent technician.

The $5.00 can of contact cleaner is worth a shot, if you can pardon the expression.

Please report back.
Thanks, again. I tried the contact cleaner on the switch but that hadn't worked either. I am still working on a another set of test speakers.
If the bias is fluctuating, it is likely due to noise in the circuit, if OK in triode mode.

Its not likely to be influenced by speakers.

If there is a bias control, it might be worth cleaning it with the contact cleaner as well, then working it back and forth. If there is a balance control, the same for that.

If neither has an effect, the unit has to be serviced by a qualified technician.
Thanks man I appreciate it and will try that also.
Hello all, I disconnected the Martin Logans and hooked up a small pair of 6 ohm Phillips 2 way speakers into the amps 8 ohm taps and the ultra-linear bias fluctuation problem went away. Wouldn't this tell me that the amp is ok and the problem is with the Martin Logan Summit speakers? If it's the speakers I still don't understand why they sound so good when the amp is played in triode mode. What would the speaker affect in the amp to fluctuate the bias so wildly in ultra-linear mode? Still going crazy.
What tap on the amplifier were you using with the MLs?

MLs generally have a very low impedance and are rarely suited to work with tube amplifiers! It might simply be that the impedance of the speaker is playing a role.

Here's a theory:
Like any other loudspeaker, they can act as microphones and pick up sounds in the environment. Normally this can't affect the amp, but if the impedance of the speaker is very low (and it is) and the amplifier impedance is higher (maybe?) you might be able to see the effects of sounds in the room on the bias meters. Easy enough to find out...
Thanks again for your help, the MLs are connected to the 4 ohm taps but I did try the 8 ohm taps to see if there was a difference and there was not. How would I know if the bias meters were reacting to sounds in the room through the speakers? The room is quiet while this is happening.
Ralph, note in an earlier post that the OP indicated he tried both of the taps the amp provides (4 and 8 ohms). Also, he indicated that the problem occurred even if the AC power plugs of the speakers were disconnected. That would seem to rule out the powered woofer section of the speaker as having anything to do with the problem. And I'm not sure that an unpowered electrostatic element can be microphonic.
12-15-14: Rluthy57
I disconnected the Martin Logans and hooked up a small pair of 6 ohm Phillips 2 way speakers into the amps 8 ohm taps and the ultra-linear bias fluctuation problem went away. Wouldn't this tell me that the amp is ok and the problem is with the Martin Logan Summit speakers? If it's the speakers I still don't understand why they sound so good when the amp is played in triode mode.
My speculation, as stated in my initial post, was:
... if the bias readings do not fluctuate with those dummy loads in place, per my comments above it would not necessarily be indicative of a speaker defect. My suspicion in that case would be that some kind of interaction between the amp and the speakers is resulting in an oscillation.
If an oscillation is occurring, at some inaudible frequency, the amp's feedback loop would presumably be involved, and therefore pretty much the entire signal path of the amp would presumably be involved. I see that the small signal tubes used in the amp are two 12AX7's and four 12AU7's. What type and make of these tubes are you using, and are they the tubes that were originally supplied with the amp? Just a guess, but perhaps changing the 12AX7's to lower gain equivalents might resolve the problem.

-- Al
Thanks for pointing that out Al!

Some amplifiers are unstable with capacitive loads. I generally associate that with transistor amps (with poorly designed feedback elements), and a tube amp can have troubles too for the same reason. Changing the driver tubes to ones with less gain (12AU7) will reduce the gain of the amp and its feedback. With feedback you either want a little or a lot. I'm sure it won't hurt anything to try- at worst it will just affect the sound.

Actually this seems the simpler explanation so I am going with a basic incompatibility (amplifier instability when used UL) with this speaker. 12AU7s in place of the 12AX7s may well solve it.
Gentlemen I don't know what to say, you have solved a problem that had tortured me for several years including thousands of dollars in upgrades of tubes, cables, dedicated circuits, shipping costs and upgrades by Rogue and Hollywood Sound, power cords and a Hydra. All made the system notably better at each stage but never solved the Ultra-linear issue. Luckily I had some spare 12AU7's so I was able to make the swap and the meters stayed stable at 40Mv in UL. I left the amp warming for 2 hours and kept checking the bias settings before I played any music still afraid that the sound would be affected. It sounds awesome most notably the subs with the increased UL power. I use Sonny Rollins Way Out West SACD as my reference disc and it was insane. Triode sounds great for normal listening but UL is a must with my set up for the volumes and type of music I listen to. I will upgrade all the 12AU7's as soon as I can. I appreciate everyones feedback and I thank you guys, you two geniuses A1marg and Atmasphere for solving this nightmare for me. Thank you both so much. Happy Holidays and all the best to you both and all that love this awesome thing of ours : ) Robert.
It's morning now and I'm pinching myself to make sure I'm awake and that this fix really happened last night. I never give up (super OCD/perfectionist/stubborn) but I could have never solved this out with out all your help and vast knowledge. I Can't describe in words how happy you all made me. Thank you all again.
Awesome! Glad to have helped solve this extremely frustrating problem. And thanks for the nice words :-)

Ralph, in the interest of optimizing sonics do you think it would be worthwhile for Robert to try a pair of 5751's in the 12AX7 locations? As you no doubt are aware, 5751's are commonly used as lower gain substitutes for 12AX7's, having a "u" (amplification factor) of about 70% of that of the 12AX7. That is much higher than the u of the 12AU7, of course, so there is no guarantee that the problem wouldn't return with them. But other parameters such as Rp and Gm are much closer to those of the 12AX7 than the corresponding parameters of the 12AU7, so I'm thinking that they may therefore be more optimal sonically.

Looking at an older GE tube manual I have, though, I note that the 5751 has a rated maximum plate dissipation of 0.7 watts, while the corresponding figures for the 12AX7 and 12AU7 are 1.2 and 2.75 watts respectively (for both the "A" and no suffix versions). I suspect that is unlikely to be an issue, but having no knowledge of the specific design I'm not totally certain.

Best regards,
-- Al
Hi guys, I'm taking a shot and ordered 6 matched 12AU7 Genalex Gold Lions (Upscale Audio/Platinum) to replace the older regular Electro-Harmonix signal tubes that are in the amp right now. I'm hoping it improves what we've already done to this point, if not I'll have some great spares : )
Ralph, in the interest of optimizing sonics do you think it would be worthwhile for Robert to try a pair of 5751's in the 12AX7 locations?

It might be worth a shot- of course an easier tube to find is the 12AT7, which has the same mu: 70 instead of 100 of the 12AX7.

I prefer both tubes over the 12AX7 as its a lot easier to get bandwidth. I would think that you want as much gain as possible without pushing the amp into instability. The problem here seems to be that the feedback loop does not have proper cutoff frequencies set up, so with a heavily capacitive load (and in UL operation, which will have a little more voltage output due to the increased power and reduced grid capacitance) it can become unstable. Sorry for the technical mumbo-jumbo.

Apparently the condition is not able to damage the amp over short duration, so I think its worthwhile to try the various tubes (12AU7, 12AT7 and 5751) to see which offers the best sound along with stability. The only trick is that depending on the tube brand, the sound might vary enough that the brand may supersede the type. That is good if that's the case, then its all just about the best sonic performance.

I should have the new 12AU7s Gold Lions by Monday or Tuesday and hopefully I'll have some positive feedback to report.
New tubes are in and I'm happy to report all is still fine. Best Christmas gift ever. Merry Christmas to all those at Audigon and those that gave their helpful advice and special super thanks again to A1marg and Atmasphere for actually solving my elusive compatability problem. Couldn't have done it with out you two. Merry Christmas : )
sorry at "Audiogon" thank you.