Bias sensing resistors - aren’t?

[martin manning]

OK, measure the DC resistance of each side of the OT from CT to Plate lead. That would be from wherever the OT CT lead connects (fuse holder in your case) to the plate pins on the power tube sockets, where the plate leads are connected. Do this with power off and filters discharged.

[Raoul Duke]

Ok, got it.

Inside (V4) is 41.9R
Outside (V5) is 42.5R

Now I do the same measurement with it warm and running for the voltage drop, correct? Then the drop equals subtracting the plate voltage to ground from the CT voltage?

Then divide the drop by the resistance to get plate current (each tube)?

This is my understanding of the procedure. Please correct me if I’m wrong.

Thanks!

[martin manning]

No all you need is the DC resistance (measured with power off) and the voltage across the primary half, CT to plate lead, with power on. Then I = V/R.

You should measure the resistance with the OT warm because it will be a bit higher than when it’s cold.

[Raoul Duke]

And the “I” is the actual plate current/bias current then?

That “I=V/R” shows up again, lol. The more I learn the math side, the more I see why “everything affects everything” in these circuits.

This was simpler than I thought. Thanks as always for the lesson Martin!

[martin manning]

And the “I” is the actual plate current/bias current then?

Yes, and that is an advantage to doing it this way, no need to adjust for or measure the screen current.

[Raoul Duke]

Very much appreciated! I’m a fan of “simpler”.

This will be the focus of my day tomorrow when I can warm it up good and take my time with it. It will be interesting to see if the 1R measurement “nuance” will prove accurate.

I’ll post results, thanks again.

[lonote]

I have this triangle burned into my brain from many years back. And I still use it.

It's a simple & easy way to remember the relationships.

Part 2 would be Watts= V X I

[Raoul Duke]

Thanks lonote! I’ll print this out and stick it above my bench and let osmosis do its thing.

It’ll burn in eventually, but having the visual reference is great. Thanks again!

[Raoul Duke]

So today I went ahead and fired everything up and got it good and warm. Then I measured the voltage from each OT plate lead to the center tap while running and got:

V4: 1.358
V5: 1.276

Then I shut it down, drained the filter caps, and measured resistance from each OT plate lead to the center tap and got:

V4: 41.7R
V5: 42.8R

Then divided the voltages by the resistances and got:

V4: .0325
V5: .0298

Swapped the power tubes around to see if I could get them closer and got:

V4: .0304
V5: .0334

So no real difference.

As far as compared to measuring across the 1R resistor both resistors measured exactly 5mV higher than the mA calculations above each time I did this exercise. Example:

V4: .0325mA and 37.5mV across the 1R
V5: .0298mA and 34.8mV across the 1R

So, I’m inclined to think the resistors are actually doing their job; I just need to measure across them vs to ground.

Curious to hear what you guys think?


Then the play test - which sounded noticeably better than before! I’ll create a small thread in the Marshall section outlining what lessons I harvested and implemented from TAG and how they turned out. There were a few things I tried and pretty much everything worked out for the better thanks to the many knowledgeable and experienced Marshall aficionados here.

And… I learned more thanks to the trusted advisors here at TAG. Thanks guys, I appreciate all your help and lessons!

[Bergheim]

So, oddly enough after racking my brain and trying to see the problem from different angles - I thought: “Those 1Rs are like little cathode resistors, right? Maybe I should measure the voltage drop across them?”

So I did. I started getting readings in mV that made sense given the resistors I have in the bias supply. So I tweaked until I was at about 65% on one tube and 62% on the other and now it’s ready for a sound test…

Still bugging me why those 1Rs aren’t reading in the conventional manner. I cleaned and scuffed all the metal below those lugs (including the lugs) and used brand new screws with nylock nuts. I’ve seen some layouts where they have those tubes connected at the cathodes and then run to the B+ filter cap ground. Thought of trying that - but if I can prove what I have here is accurate - I’ll stick with it. Measuring across them is a small variation I can work with if it’s accurate.

A 1R resistor passing 40mA will drop 40mV. If the voltmeter is set to auto range or any voltage range other than millivolts specifically, it will normally read 0V.
Sounds to me there isn't really a problem?

[martin manning]

That all looks good.

V4: .0325mA and 37.5mV across the 1R >>> So (37.5-32.5)/37.5=0.133, indicating Ig2 is 13.3% of Ik
V5: .0298mA and 34.8mV across the 1R >>> So (34.8-29.8)/34.8=0.144, indicating Ig2 is 14.4% of Ik

which looks about right.

Still a mystery as to why you can't measure Vk to ground to get the Ik (in mV, indicating mA). They are obviously returning the current from the tubes.

Personally I like this Ohms Law / Watt's Law wheel. It'd make a nice refrigerator magnet.

[Bergheim]

So, I’m inclined to think the resistors are actually doing their job; I just need to measure across them vs to ground.

This is weird. And there isn't a chance you might have mixed the cathode and ground ends of the 1R? Ie hooking one probe to the ground side of the resistor and the other probe to the chassis?

[Raoul Duke]

Wow, this is embarrassing.

So I decided to double check the simplest things first and realized that Bergheim was absolutely right. My only explanation is that I had to do some creative bending to get the 1Rs positioned where I could reach them without shorting something else, and in the process of installing them and the heater ACT and flipping the chassis around on the bench - I installed them “backwards”.

I know, resistors aren’t polarized - but the sides that were bent for the probes were attached to the ground lugs, so I put them in backwards and just went right to the “probe bends” once running without really looking at the bigger picture around the tube sockets.

Bergheim and sluckey were both right: I was measuring from the wrong end of the bias sensing resistors due to the loose connection between the ears, lol.

The upside is that thanks to Martin, I learned an alternate way to check the bias - which is another invaluable tool for a hobby builder like myself - and allowed me to prove the 1Rs are working correctly.

Thanks Bergheim! I appreciate the nudge in the right direction.
And thanks for the new refrigerator magnet Martin - it’s already up on the wall.