Confusion over biasing VM2266

[Adey1981]

Earlier in the year I bought all new tubes for my 2266 (Shug pre's and matched jj kt66s). I decided to bias the tubes myself following posts on this forum and from Steve Dawson on the other place that I can remeber the name of right now.

Amp on, speaker connected, wait 10 mins, dmm set to 200 DCV, measured pins 1 and 3 to ground at con10. They were already close and all it took was to tiniest possible adjustment to bring them close. I ended up at 47.5 and 48.5 so was happy.

I was still getting some buzzing that increased with the MV, so took it to my tech. He has emailed me and I'm very confused, it says..
"Firstly, the Biasing was correctly set. Min & Max Bias range for this Amp was calculated to be between 21mA ( min ) and 31mA ( max ). Biasing set to 27mA +/- 0.5 mA for the perfectly Balanced output KT66's. The original setting of 46mA per valve was considered quite a severe threat to the continued function of the output Transformer. Waaayy too high !Valves probably wouldn't have lasted very long"

So first thing is, he mentions mA not mV. I measured in mV.
Secondly, I thought that the point of biasing is to get 2 tubes that are already close, i.e 'matched' and then to balance or bias them so then even out as much as possible. How then could it be, that an amp can be "waaay" over biased? Is there another adjustment pot somewhere that controls the overall plate voltage? When people talk about under and over biasing, how are they doing this? because the intructions i've read talk about balancing power distribution rather than the total power level itself.

Hopefully im making some sense, cuz I sure is confused right now. Its not that im questioning my tech at all btw. I'd just like to understand the technical side of things a bit better.

Many thanks,

Ade

 

[Kinkless Tetrode]

Mv = Ma if measured across an 1 ohm resistor. Modern Marshalls with fixed biasing usually have that built in. Your probably okay there.

A Ma figure is meaningless without knowing the approx. plate voltage. It's plate dissipation rating (in this case 25 watts)/ plate voltage = gives you the ma that is 100% plate dissipation. You then must decide if you want it to be 60%-70% of that. For example, 46ma per tube might be okay if the plate voltage is 400v, but if its 480v that would be way too high. 32ma would be about right for 480v but it would probably be too low if the plate voltage is 400.

 

[Adey1981]

I assume all 2266 have the same plate voltage right? Everything I read about biasing the 2266 just said to measure the mV at 48. I still dont understand how the bias would be too high in this amp, it dosent have independant controls for each tube just a trim shared by both. I have emailed him to clarify, but thought I ask here just incase.
I'm still confused

 

[Kinkless Tetrode]

There will be a little variation in PV from amp to amp. I don't know what PV a 2266 typically runs. I know the 2246 runs high plate voltages, though. One would need to know the plate voltage to calculate the proper safe range of adjustment. And it is a range of acceptable settings rather than being a hard number. Some like it closer to 60% and some like it around 70%. Cooler usually prolongs the life the tube.

There will also be some variation depending on the bias setting. With a hot bias setting the plate voltage will be slightly lower because there are more electrons flowing through the tube.

48 doesn't seem to work for the typical range of probable PV for a 2266. It doesn't work for a cool setting for two tubes assuming a very high plate voltage of the 2246 either.

 

[Adey1981]

Ok I think that makes sense to me, my tech did also mention the 60-70% thing so it seems like he knows what he's doing.

I still dont understand how the bias could be too high or low, when the 2266 has no apparent way of raising or lowering the plate voltage. You've helped me enough though, so i'll ask my tech to clarify when I see him.
I appreciate your help KT, so thank you for your input. Cheers.
Ade

 

[paul-e-mann]

Ok I think that makes sense to me, my tech did also mention the 60-70% thing so it seems like he knows what he's doing.

I still dont understand how the bias could be too high or low, when the 2266 has no apparent way of raising or lowering the plate voltage. You've helped me enough though, so i'll ask my tech to clarify when I see him.
I appreciate your help KT, so thank you for your input. Cheers.
Ade

Read this: https://www.eurotubes.com/store/pc/how to bias videos.htm

And watch the generic bias video, they explain everything.

 

[Adey1981]

Thank you pedecamp, I will check that out.
Cheers

 

[Chris-in-LA]

I assume all 2266 have the same plate voltage right? Everything I read about biasing the 2266 just said to measure the mV at 48. I still dont understand how the bias would be too high in this amp, it dosent have independant controls for each tube just a trim shared by both. I have emailed him to clarify, but thought I ask here just incase.
I'm still confused

Setting the bias means adjusting how hot or cold the tubes are running, not making them even or matched. It would be nice to know what your plate voltage is. My kt66’s are biased at around 35mv with a plate voltage of 435. 27mv seems low unless your plate voltage is 555v.

 

[XTRXTR]

I am saying this to provide another version of the same facts that have been explained so far. Sometimes it helps to hear or read another reference to the same information before it finally clicks. I know I have had many reference books over the years.

The power transformer HT winding will drop in voltage as more current that it supplies is demanded from it. So as a power tube has its bias setting increased from cold to hot more idle DC current flows in the tube and that current is provided by the power transformer. As that current increases the B+ Voltage will decrease.

A point about transformers. Most people generally think of a transformer as a thing that takes a primary voltage and turns it into a secondary voltage. Or similarly a primary current turning to a secondary current. This while true is not a complete picture of what it does. The main function of a transformer is to transmit power from the primary to the secondary. Power in watts is equal to Volts multiplied by Amps: P=VA And it does this by changing a primary voltage and current rating to a secondary voltage and current of the same rating. If you have a 10 watt Power Transformer PT then on both sides, the primary and the secondary; 10 watts will be transferred.

The power equation P=VA tells the whole story. 10 Volts on the primary at 1 Amp is 10 watts; 5 Volts at 2 Amps is 10 watts on the secondary. Now imagine the tubes are what is dissipating the 10 watts on the secondary. The secondary B+ at 500V and the plate current at 20mA = 10 watts. But you want a hotter bias so you turn the bias pot until you see the plate current you want, the Voltage must come down to supply higher current in order to keep the equation at 10 watts. 400V at 25mA = 10 watts.

In reality there are more physics involved and this was just a simplification I wrote to help understand the relationship of the Plate Voltage and Plate Current, a tube and a PT while adjusting the bias. The PT in no way will limit current for you, the PT has a point of failure if too much current demand is placed upon it; it will burn, as does the tubes you place in your amp. Failure modes are almost always due to a lack of understanding of the devices used.

 

[Pete Farrington]

But you want a hotter bias so you turn the bias pot until you see the plate current you want, the Voltage must come down to supply higher current in order to keep the equation at 10 watts. 400V at 25mA = 10 watts

The previous paragraph about HT winding voltage dropping a little (as more HT current is drawn) was great, but the 2nd went a bit off course. The misapprehension being that power will necessarily be limited to 10W.
In reality, a hotter bias point will happily draw more power from the PT, below magnetic saturation (or the HT / mains fuse blowing, or mains breaker tripping etc) there’s nothing to prevent the secondary and primary drawing more current.

To expand on the HT winding voltage dropping as the HT current increases, it's a characteristic referred to as regulation.
For guitar amps, it's typically between 5 and 10%, ie if at 0mA HT the winding voltage was 100V, then for 10% regulation, at full load current, the winding voltage would be 90V. The drop is caused by copper and iron losses, ie winding resistance and inefficiencies in the magnetic circuit. Both will result in the PT getting warm.

For Hammond's JTM45/50 PT, the 290 MAEX, the HT winding voltage is 729.1 at no load, 690 at 150mA; unfortunately at the full rated load current of 213mA, the winding voltage hasn't been noted, but it will be somewhat less than 690V.
https://www.hammfg.com/files/parts/pdf/290MAEX.pdf

Of course the HT VDC applied to the valves will have losses in addition to the winding regulation, eg the voltage dropped across the rectifier diodes will increase as load current increases, and the VDC average voltage on the reservoir cap will fall from its Vpeak level as load current in increased from 0mA.

 

[RLW59]

I still dont understand how the bias could be too high or low, when the 2266 has no apparent way of raising or lowering the plate voltage.
Ade

You don't set bias by adjusting the plate voltage. You set bias by adjusting the bias voltage, which changes the amount of current the tubes draw. You measure the current draw (using 1 ohm resistors you can indirectly measure the current draw in volts instead of amps, but it's the amperage you're actually concerned about), then turn the bias trim pot to raise or lower the bias voltage to change the current draw. (The 47.5/48.5 mV you measured was actually telling you current your tubes were drawing.)

The plate voltage changes slightly when you change the current, but that's a "side effect". You take that into consideration when setting bias, but you don't set bias by adjusting plate voltage.

The schematic gives a nominal plate voltage -- a theoretical ideal. In reality, transformers vary a bit and may put out slightly higher or lower voltage than what the schematic says.

And the output voltage of a transformer is directly dependent on the input voltage. And that can vary a lot. Your electric company calls it "120 volts", but again that's "nominal" -- it can be as high as 130 volts. (Your transformer turns "120 volts" into 400+ volts, so an extra 10 volts from the wall outlet can add some 30 volts to the plate voltage.)
--------------------------
Ask your tech to tell you the plate voltage he measured in your particular amp. From that, people here can calculate what the proper current draw should be.

 

[XTRXTR]

@Pete Farrington I did state a similar precaution you mention. I put it in terms of the current since current is the thing that burns everything in electronics. I was simplifying the idea intentionally for the OP.

@RLW59 I think you are right to point out that biasing a fixed bias amp is about setting the negative bias voltage on the grids which controls the amount of electron flow from the cathode to the anode aka plate where it dissipates. Bias is on the grid not the plate. It is not technically correct to say you set your bias at 37mA. That statement is concerning the plate current at 37mA not a bias current. In fact there is no bias current, there is a bias voltage and it is on the grid. You can have a grid current but that statement is about AC signal clipping.

You could correctly say I set my negative bias at -35 VDC in order to get my plate current to 37mA and with a plate voltage of 473Vdc, for a plate dissipation of 17.5 watts, 70% of Max. Pc*Pv=.7Max, 0.037*473=17.5

 

[RLW59]

I was trying to say the same as you, I just wasn't as clear.

When I said "then turn the bias trim pot to raise or lower the bias voltage to change the current draw. (The 47.5/48.5 mV you measured was actually telling you current your tubes were drawing.)" I wasn't implying anything about "bias current".

You measure plate voltage, calculate the plate current that will put you in your desired bias range, measure the plate current, then adjust the plate current by changing the bias voltage until the plate current reaches your calculated desired plate current.

If your desired value gives higher than 70% dissipation (some people like them hot), recheck the plate voltage to see if changing the plate current affected the plate voltage significantly. If so, recalculate the current at that voltage and readjust.

 

[RLW59]

Just because I'm a contrary fellow:

Huge numbers of people assume that the plate voltage is reasonably close to the schematic, say within 10% and never actually measure it.

And a lot of those people don't even calculate it themselves based on the nominal plate voltage.

They just read what plate current most other owners of that model set theirs to. And as long as they don't listen to that one crazy guy who's willing to burn up tubes every 2 months because "hot gives the tonez", setting it to the "group average" is usually fine.

Google turned up the Steve Dawson reference. He designed the amp. And he recommends about 47mA. Lots of owners have measured their plate voltages and calculated that 47mA is about right.

So it's perfectly reasonable that Adey1981 is confused about what the tech told him. There are bad techs out there. And good ones make mistakes. Maybe the tech absent mindedly plugged it into a variac set high.
-------------------------
Setting the bias too high can hurt things, but setting it too low won't damage the amp -- and the tubes will last longer.

Too low can sound bad, but that's the worst it can do. If it sounds good when you get it back from the tech, it's all good.

It would still be interesting to know the plate voltage. Maybe we can figure out why your amp wants a different plate current than most 2266's.
-----------------------
Did changing the bias fix the buzzing? Or something else was causing it?

 

[Spanngitter]

As per the information I have the Marshall Bias Spec for the VMs are
72 to 80mVdc per side (pins 1 and 3 of CON 10) with respect to chassis ground for the 2466 (2 Valves are measured via common resistor!)
43 to 48mVdc per side (pins 1 and 3 of CON 10) with respect to chassis ground for the 2266
The 2266 has a average B+ of 440V whilst the 2466 slaps the big heads with a stunning 530Vdc hence I consider these specs a bit on the high side and usually recommend to not exceed 70mA/mVdc for the 2466 and 42mA/mVdc for the 2266 as it might shorten your valve lifetime quite measurable...

 

[Pete Farrington]

What folk seem to overlook is that the 25W limit of the KT66 is under the design max system, eg it leaves no margin for tolerances in the mains voltage, and the OT / PT windings.
As such, it’s a lower rated valve type than the EL34, whose 25W rating is under the design centre system, which takes that stuff into account.
So the KT66 is probably more properly thought of as being equivalent to a 5881, which has a 23W design centre anode dissipation limit.
https://tubedata.altanatubes.com.br/sheets/086/k/KT66.pdf

https://tubedata.altanatubes.com.br/sheets/129/e/EL34.pdf

https://tubedata.altanatubes.com.br/sheets/127/5/5881.pdf

 

[Chris-in-LA]

So, with a b+ of 440, 34ma would be 60% dissipation for a single kt66, right? OP’s original setting of 47ma would be quite high. His tech’s setting of 27ma would only be about 50% dissipation, a little on the low side.

 

[Pete Farrington]

So, with a b+ of 440, 34ma would be 60% dissipation for a single kt66, right? OP’s original setting of 47ma would be quite high. His tech’s setting of 27ma would only be about 50% dissipation, a little on the low side.

440 x 0.034 = 15W
15 / 23 = 0.652
How many mA of the cathode current to assign to the screen grid? I suppose it depends on the particular valves fitted, my old but very serviceable GT KT66 only draw about 1mA g2 current at idle, with a HT of about 450V and about 30mA cathode current.
Going from that -
440 x 0.033 = 14.5W
14.5 / 23 = 0.631, so 63%.

 

[Chris-in-LA]

440 x 0.034 = 15W
15 / 23 = 0.652
How many mA of the cathode current to assign to the screen grid? I suppose it depends on the particular valves fitted, my old but very serviceable GT KT66 only draw about 1mA g2 current at idle, with a HT of about 450V and about 30mA cathode current.
Going from that -
440 x 0.033 = 14.5W
14.5 / 23 = 0.631, so 63%.

I was looking at the Robinette bias calculator 34ma = 60%, that’s basically what your figures are showing, right?

 

[XTRXTR]

@Chris-in-LA
I think Pete was saying that the Op's tube Max dissipation should be thought of as More like a 5881 which is 23 Watt Max. And going by the voltages you had discussed with plate current of 34mA that would make the tube at 15 watts which is 15/23 = 0.652 = 65.2% of Max dissipation. But he then amended it taking into account screen current. With a final of 63%