Mark V bias issue

The Boogie Board

Help Support The Boogie Board:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

bandit2013

Well-known member
Boogie Supporter
Joined
Feb 22, 2013
Messages
4,067
Reaction score
506
Location
North Carolina
When I took the amp into work Friday, I wanted to run a test on the bias circuit with an oscilloscope in order to confirm what I measured with a DVM on the bias voltage. Back story: a screen resistor failed some time ago and after the repair the problem was still present. Mesa 6L6 STR440 tubes would red plate when using the 45Wmode using the variac power setting as it was preferred over full power and 90W operation due to the ice pick. The way I was running the amp may have added to the issue (playing loud). Since I ran through three sets of tubes I changed over to Tung Sol 7581A tubes which held up and lasted for 6 months of use. Original tubes failed within the first two months of use. Any other tubes following lasted at least an hour. The last set I was using full power at 90W with no issue but when I tried my favorite arrangement 45W on Variac the center tubes redplated within 20 minutes. So when I changed to the Tung Sol I kept the amp at full power at 90W. 6 months is not long enough and that is when the tubes took a dramatic change in tone. Observing the plates when in use the tubes were operating at the onset of red plating since the plate seams were red hot but did not expand farther than that. SED =C= 6L6GC were the savior tubes, I got a NOS set that were sold by GT in the 90's and those lasted almost a year before I changed to a new set. I still have them and they are still good to use as is the other set. However, since SED (Svetlana Electron Devices) stopped selling consumer grade products since the glass supplier either closed shop or the quality of glass just was not suitable, SED shut down production of audio tubes.... what a bummer. It was time I needed to figure out what the problem was with the amp. It is related to the bias circuit. The other issue was the chassis was getting so hot I could not touch the back side of the amp, speaker cable insulation was getting really soft due to excess heat as well as the power transformer. The thermal issue changed when I started using the SED tubes. It got better when I converted the head to a combo as the speaker provided additional air movement.... the point of the story.... I took the liberty to change the bias voltage that should have raised the -51V (calculated value should be -51.096V) to -51.300V. I have seen some videos regarding the Mark V on diode replacement and one of the suspect didoes was the one associated with the bias circuit. (may have been an early model as there was issue with some diodes at one point). I swapped the 82.5K resistor with a 91k resistor and my woes with red plating Mesa 6L6GC tubes were over. So why the long story..... the amp is now running too cold due to the change in bass response. I had assumed there is still an issue that has been left unresolved since the screen resistor repair. Now to the point... I measured the bias voltage with an oscilloscope to ensure the diode was operating properly and that there was no notable ripple in the bias circuit. No ripple at all and was very stable but not at the correct value. Bias voltage was confirmed to be -59.01 Vdc. When the amp was taken out of stand by the bias amplitude would drop to -57.39 Vdc. Not the expected value of -51.3Vdc. This would indicate the 33k resistor is either open or there has been trace damage to the PCB, or it may be something else.

Perhaps it is time to work backwards and measure voltage drops on the complete bias circuit (at the PI tube and on the grids of the power tubes) before I replace any components. The 1.62V change when the amp is taken out of stand-by may be a concern if I am able to correct the bias voltage to where it should be, that change would definitely run the tubes too hot (-49.47Vdc) unless this is typical when the amp is out of standby. What is causing the 1.62V droop in voltage is something I need to figure out. At this point getting a bias probe kit may be a good idea just to verify that once the amp has been fixed the tubes will not red plate shortly after running the amp.

I do have another amp I can use to verify if bias shifts when out of standby. It may not be a Simul-class amp but should provide a clue to what I am looking for. Could be just a waste of time but I have plenty of that.
 
Never mind..... the schematic is incorrect. Totally misleading. Everything checks out fine.

I did some calculations based on the components on the PCB as well as the schematic, also confirmed with LTSpice at what I was measuring with the DVM is accurate. At first I thought that the 33k resistor may be open but all is good.

Bias input voltage at full power measures 60.1VAC(RMS) and at variac power it is 52.5VAC (RMS). Schematic has some flaws in the leaked version as the capacitors are not 470uF but are 47uF.

What is stated in print on the schematic is not accurate. -51Vdc is not valid unless the amp is in variac mode.

Measured across the 47uF cap that is parallel the 33k and 82k resistor (mine has a 91K resistor instead).
Full power: -58.5Vdc
Variac power: -51Vdc

There is a slight change in voltage with the power and PI tube pulled but not by much and remains about the same (in standby or out of standby).
What I have seen regarding measurements using bias probes, if the bias voltage measured is at -46V that may account for about 3mA on the current limiting resistor 2.2k.

Next thing to check but will require a bias probe to confirm plate current on both pairs of tubes. At leas the amp no longer red plates the power tubes. If something is out of whack it would have to be on the power transformer as the calculations match the measured values (including the total circuit resistance with no power applied and after the 47uF reach a steady state supplied by the DVM.

While I have the amp chassis out I may stick the C39 back in and run another recording comparison using the load attenuator.
 
The reason behind this thread is that I felt the tubes sounded like they were not biased properly. Bottom end was a bit flabby or weak. Well since I had a 91k resistor in place of the 82.5k resistor on the bias circuit, I felt the sound was lacking any boldness I had prior before the change but cannot afford the SED =C= tubes at the current market price assuming you can find them. I still have two sets to use but actually prefer the tone of the Mesa 6L6GC tube a bit better now that I am using the V4 mod =12AT7.

Before I installed the chassis back into the amp, I recalled I had some resistors I had ordered for other repairs if needed in the event something went on a different amp (Mesa fixed it so no longer need the parts). I had the resistor that would work in my favor in parallel to the 91K, a 1M 1% would roughly bring the 91k down to 83.4k. Much closer to the original 82k. I wanted something close but do not want to be replacing tubes every month or less. Installed it and compared total impedance which was about 500 ohms less. With just the 91k resistor the bias impedance measured 23.75k, with the additional 1Meg resistor that dropped to 23.29k. In short the effective substituted resistance is 83.06K since the 91K resistor measured a bit lower (89k). Odd that the measured bias voltage after the resistor was added did not change. The tone did change a bit as the bottom end was a bit bolder and not sounding as weak, brightness also increased a bit (based on perception). For all I know, the resistor network on the power tube board may not even match the schematic. At this point I do not care as the amp passed the red plate test (at least for the short trial period since the last post).

What may be on paper may not match empirical measurements as tolerances have to be taken into account and many other factors that are not in print (schematics are missing many components so I would not plan on building an amp based on them). It is what it is.... My task was not to blueprint the amp to match the schematic but to understand what need to be changed to improve the tone quality. Get a bias probe you idiot (I rude reminder to myself).
 
Interesting read this,

Will be interesting to see if your red plate issue does rear it's ugly head again. Will the 2 resisters be staying in for a while or will you put it back to your 91k value?
 
I have the 91k resistor in place of the 82.5k. The additional 1M in parallel reduces the impedance as if I had replaced the 91k with an 83k resistor. Of course I did check the resistance measurement as the 91k is a 5% vs the 1% components. So far no red plating and I would have seen it by now considering the age of the tubes that are in the amp. (mix and match of red color coded Mesa tubes that did not have internal vibration issues that were loud enough to annoy me). Reason for addressing things is I felt the 91k resulted in a colder tube sound, bass was not very bold and seemed to be lacking the punch I recall having before (along with the ice pick tone). Since the ice pick is more of a preamp issue, getting the amp closer to what it should be would improve performance. Still need bias probes to measure plate current and voltage. I would like to find one that will also provide bias voltage as I am curious if it is balanced or not. I forgot how much fun the 45W mode on Variac power is with elevated volume levels. That was the tube killer set up as it always happened on Variac power at the 45W mode setting on CH3 using pentode. I have been checking the plates every so often and so far no change in color at the seams (which is good). Even the Mark IV was capable of pushing the plate seams to dark cherry but never went past that point. If there was an issue with the original 82.8k resistor that would not have mattered much as that would result in a higher negative bias voltage resulting in cold tubes.

I am sure anyone can look up my issues on the Mark V (still have the same amp too). Perhaps some things do improve with age.
 
why put soooo much work into the red plating ? i would send it back to mesa boogie .and tell them to fix it.
I own one . never had any problems. and after paying nearly 3000.00 for the mark v head. if i had any problems. i wouldnt be wasting my time fixing it. mesa boogie would..and if they cant fix it. i be getting my money back.. and off to diezel vh 4, i go !
 
Back
Top