Mark V 90 power tube voltages and bias

[Nandrewjackson]

Inspired by bandit's in-depth posts about the mark V, I've decided to try and keep my combo. I say try to because space in my apartment, finances, etc. I could use the money, and they sell a lot for $1800+ used, but I had mine cross posted and it didn't sell. Lol.

Also under inspiration about the mark V 90, I decided to get to the bottom of the questions about running 6V6 tubes.

Using the "output transformer resistance method" I have lots of data for the voltages and bias. I have done all modes for 6L6 and 6V6, I haven't done EL34 yet.

There are 2 modes that are spot-on for 6V6 operation.

I will post the details as soon as I organize my notes. Not sure if I'll type it out (only have phone) or just take pictures of the data written out.

 

[bandit2013]

I am interested to find out what you have discovered with the 6L6 voltages. More curious on the grid bias voltages on the Class A sockets as mine is way off.

6V6, interesting. The Mark IVb could do that but was not sure on the Mark V90. I would assume you need to run that in varaic power mode?

 

[Nandrewjackson]

Here's the first page of data. This covers 90 watt mode with a quad of 6V6.
There are 3 modes that will work, 1 at a nearly perfect bias. And 1 mode is a NOPE.
I used EH 6V6GT. I can't say how other brand 6V6 would bias, so be careful.

 

[Nandrewjackson]

All testing with FX loop hard bypass. No reverb tank plugged in. No footswitch plugged in. 8 ohm speaker connected to 8 ohm jack. Cooling fan on. GEQ on. All controls at half, except volumes at zero.

All mesa 12ax7s. Mesa 5u4gb. Matched quad of electro harmonix 6V6. 2 seperate pairs of 6L6, 1 pair mesa str440 grn, 1 pair "apex matched" JJ 6L6.

I used info from rob robinette "OT resistance method" for calculating plate current. Rob robinette bias calculator for percentage dissipation information.

 

[Nandrewjackson]

Also to note. My multimeter wasn't giving me decimals on triple digit voltages, so everything is rounded to the nearest whole number. When the pairs were off left to right it was never more than 1 V so I averaged them. Also, the OT primary halves were 46 and 41, so I averaged the halves to 43.5 to simplify.
I ran some numbers to more precision, the results are basically the same as presented.

 

[bandit2013]

Interesting. With the amp at full power, the class A bias voltage on the control grid seems a bit low in magnitude at -49V, that would be a hot bias for Class A, almost enough to run EL34 tubes, just not hot enough though. TBH, I am not really sure what the bias voltage on the Class A should be for safe operation. However, could be related to impedance of the 6V6 in the circuit. I would assume it would be around the -50.93 to -51.08V which is what Mesa states the bias voltage is (-51V) but in the wrong place. the Class A/B bias is what I found in simulations.

 

[Nandrewjackson]

It must be the characteristics of the 6V6 making it more positive. With the 6L6's installed, I was hitting -51. I'll get those written up and posted soon.
I was a bit surprised how each tube and mide changes the center tap voltage.

 

[Nandrewjackson]

I'm excited to try 90 watt, variac, set to EL34 bias, running 6V6 quad, but I don't even have an electric guitar here

 

[Nandrewjackson]

6L6 results.
I left out the specifics about 45 watt mode. I was very disappointed with the tube rectifier results. Hopefully tube rectifier with EL34 will be better.

 

[bandit2013]

I see. I thought that is where the 6L6 class A bias should be at. I have been trying to run PSpice simulations on the Mark V90 bias circuit that contains the 60VAC transformer down to the grid stoppers and voltage divider for the class A control grids. It is not much of a change in bias voltage when changing the 82.5k to something else. I was able to prevent instant red plate issue as it was from the factory. I will probably restore the resistor to its correct value but then change the 33k to something a bit larger in value. I need to take actual measurements to see where things are currently at. Forgot to bring home the oscilloscope to confirm the voltages and to see if there is any voltage ripple in the waveform. That would indicate I have a bad cap. The last time I looked at it, it was all over the place at full power but was stable at -51 in Variac power mode. That was the voltage across the 82.5k resistor. Would like to get the Mark V90 where it should be. It does sound good as it is but think the bias is on the cold side. I have a 91k ohm 10% resistor that measured 86.6k ohms in there now. I have not red plated any power tubes yet since the change. Better check to see if I have any sacrificial tubes I can use or just not bother doing this due to the difficulty getting replacement tubes.

 

[domct203]

Great thread.

I added a trimmer to my Roadster’s bias circuit as soon as the warranty was over, about 11 years ago.

I tried every 6L6 under the sun in that amp and always found myself settling between 55-65% max plate dissipation. My Stiletto Deuce also has the same trimmer circuit, that amp sounds it’s best with 6CA7’s running about 60%.

This is based on how I use my amps, everyone has different goals.

Mesa’s high gain offerings are really about pushing the preamp, which explains the cooler/cold output section, they are looking for the clean headroom that lies just beyond crossover distortion.

I didn’t mess with my MKV, I never really bonded with that amp. I had one of the ice-pick ones, and surmised the issue was deeper than just tubes. Ended up selling that head to fund my second TC-100 as I was hardly playing it (just enough to keep the caps happy).

Dom

 

[Nandrewjackson]

Page 2 of 6V6 info. 45 W mode.
Here's the other good 6V6 setting, similar to 90W mode. 45W, variac, EL34 bias, 75%.
I still have my chassis out, Bandit, if I could grab any other measurements for you, let me know. I only have a DMM, no scope or other equipment.

I guess I lucked out with the tone of my mkV, Dom, sounds like you've got some nice other models.

 

[tbonesullivan]

Are all 6V6 tubes able to withstand that high of a plate voltage? Very interesting to read this. I usually bias pretty conservatively. I found that my favorite sounding "vintage" sounding amp that everyone loves the sound of was biased at like 30% dissipation, and deep dove into the whole negative feed back thing etc, so I pretty much just set adjustable bias amps at 40-50% and forget it.

 

[rarebitusa]

My Stiletto Deuce also has the same trimmer circuit, that amp sounds it’s best with 6CA7’s running about 60%.

Not to vector the thread but always looking for Stiletto info since you mentioned that where are you running the chan masters and the master out to get to 60%? TIA

 

[Nandrewjackson]

Vintage deluxe reverb shows 415VDC. I would say either bias setting (6L6 or EL34) on variac power and its totally safe for 6V6.

 

[domct203]

Not to vector the thread but always looking for Stiletto info since you mentioned that where are you running the chan masters and the master out to get to 60%? TIA

I was referring to the % of max plate dissipation at idle, or where the output section is biased. There should not be any signal passing as you are setting idle current when referring to the bias point of these amps.

A rule of thumb for setting the bias point is to stay under 70-75% of the max plate dissipation of the tube, so a 25W tube at 60% would be making about 15W of power at 60% (25x0.6).

To figure out the plate current you would divide that 15W by the measured plate voltage (B+).

I don’t remember the exact plate voltage of my Deuce, but we can use 450V for a reference

15/450=0.033, or 33mA of idle plate current. To set the amps bias point using this method I would adjust my bias voltage applied to the control grid until I measured about 33-35mA of current at idle.

Dom

 

[rarebitusa]

Ahhh ok.. thanks that certainly makes sense, wasn't paying close attention as I should have...

But, but, but I find the Stiletto a very different beast then my Marks.. much more dependent on the power section for tone. Have run a set of 6CA7’s which were ok but perhaps I should have asked... how much did the trimmer and biasing effect the Stiletto? Worth doing?

I find it needs to be cranked to get it really going, hence the previous chan masters question.

 

[bandit2013]

Page 2 of 6V6 info. 45 W mode.
Here's the other good 6V6 setting, similar to 90W mode. 45W, variac, EL34 bias, 75%.
I still have my chassis out, Bandit, if I could grab any other measurements for you, let me know. I only have a DMM, no scope or other equipment.

I guess I lucked out with the tone of my mkV, Dom, sounds like you've got some nice other models.

No I am good, Thanks for the information. I have my chassis out and can get what I need. I will try to remember the scope for next weekend. Seeing the actual waveform is a good help to identify any voltage fluctuations, ripple or other unexpected conditions that a DVMM cannot provide. I may end up soldering some leads on the board to make connections easier. Not sure if I need to look into it at this time. Amp is sounding great. I also have the same amp issues that Dom has encountered with his. I never bonded with the Mark V90 either as it still has or shall I say had ice pick and boxy tone. I have been able to work most of that out with preamp tubes (mostly NOS that Mesa used back in 1990) and a blend of power tubes. STR440 and STR441. May try the STR448 (red bias color) and see if that is any good.

 

[domct203]

Ahhh ok.. thanks that certainly makes sense, wasn't paying close attention as I should have...

But, but, but I find the Stiletto a very different beast then my Marks.. much more dependent on the power section for tone. Have run a set of 6CA7’s which were ok but perhaps I should have asked... how much did the trimmer and biasing effect the Stiletto? Worth doing?

I find it needs to be cranked to get it really going, hence the previous chan masters question.

Agreed, the Deuce indeed needs to be cranked to sound it’s best, but so does my Roadster. I don’t remember exactly where I run the output control, most likely around 11:00. I treat it like any other good tube amp and put an OD into the front end as a boost to get the saturated tone at less than coliseum volume. I find my Deuce similar to my DSL, just thicker and smoother as it gets louder, and not as dependent on ear splitting volume to sound full.

Dom

 

[Wolfgar926]

Also to note. My multimeter wasn't giving me decimals on triple digit voltages, so everything is rounded to the nearest whole number. When the pairs were off left to right it was never more than 1 V so I averaged them. Also, the OT primary halves were 46 and 41, so I averaged the halves to 43.5 to simplify.
I ran some numbers to more precision, the results are basically the same as presented.