KT90EH biasing discrepancy

[marshallguy]

While biasing a pair of KT90EH power tubes, I found a big discrepancy between the shunt method and the 1 ohm cathode resistor method. With 477vdc plate voltage, I read 54 mA shunt and 70mA resistor. Measured 8v drop across screen so minus that I still have 62mA. That’s still a 8mA difference between the 2. The tubes are rated at 46w each. I shoot for 60% so it’s in the ballpark but using the same methods with a pair of EL34’s on another amp, 50w JCM800 4210 combo, I’m only about 1mA off.

The KT90’s are in a JCM800 2203 head, a 100 watt normally 4 power tubes. I only been using 2 so the transformers can handle them no problem.  I also cut the output impedance in half to the speakers. Don’t know where the other 8mA is being derived using the cathode 1 ohm resistor method. I tried both sides of output transformer and all tube sockets.

Anyone have any clue as to what the difference is?

[HotBluePlates]

While biasing a pair of KT90EH power tubes, I found a big discrepancy between the shunt method and the 1 ohm cathode resistor method. With 477vdc plate voltage, I read 54 mA shunt and 70mA resistor. ...

You measured precisely 0.070v across the 1Ω resistor?  Not 0.069v?  Not 0.07v and the meter lacks resolution for an extra digit?

Now how about that 1Ω resistor?  Is it precisely 1Ω and not 1.2Ω?  Not 0.89Ω?  Can your meter resolve that? (I know mine cannot)

There's room here for a few-millivolt error measuring across the "1Ω" resistor, from maybe several sources.  And we haven't thought too hard about your meter's accuracy & calibration.

... I read ... 70mA resistor. Measured 8v drop across screen so minus that I still have 62mA. That’s still a 8mA difference between the 2. ...

The screen resistors are precisely 1000 ohms?  I'd be amazed if they were exactly on-value.

No rounding here of the number to get 8 volts?  Is it "8 volts" on a scale that goes to 100v?  (in which case, the meter is rounding digits without your awareness)

While biasing a pair of KT90EH power tubes, I found a big discrepancy between the shunt method and the 1 ohm cathode resistor method. With 477vdc plate voltage, I read 54 mA shunt ...

Resistance of the winding can change from hot to cold.  I didn't believe how much until I measured the resistance of some pickups, then put them in the freezer for a few minutes & re-checked.

I don't know a good method of checking the resistance of the OT winding while hot, except perhaps having an IR thermometer & a hair dryer, heating the OT to some temp and checking the resistance.  This CANNOT be done with the winding energized, as it risks meter-damage.


I don't have a good sense of the proportion of any of the possible error-sources I've mentioned above, but there's plenty of places for error to creep in.

[shooter]

So the math adds up using both methods with the EL34 but not the KT90EH.

put 10 hours on the tubes and the math won't add up the same as with zero hours either.  Play with the amp cranked and it all goes to hel-p
keep it in the ball-park, check as often as your education requires for a good nights sleep and you'll find, it all works like magic anyway 
6 yr Navy advanced electronic systems, Radar Tech, 2nd class FFC, radar endorsement, 28 weeks C.A.T. tech school, 24 week MRI tech school
12 week, 8 week.....I sure was happy when Friday came and I could beat up a jeep and climb rockfaces to detox 

[HotBluePlates]

... you  didn’t mentioned the EL34’s using the same methods were correct. ...

So the math adds up using both methods with the EL34 but not the KT90EH. ...

I honestly didn't get to the bit that mentioned the close agreement with the EL34s.

It would appear then that the plate current drifted a lot with the KT90s, at least much more so than with the EL34s.  Using a Eurotubes Pro One (that inserts an ammeter directly at Pin 3), it is common to watch plate current for power tubes drift around as the amp warms up.

... I just wanted to point out the wider difference between methods on this particular tube.

... I measure actual current not the resistance then voltage way ... I also use 2 meters, one in test points across one side of tyranny and the other from the resistor to grounded so I can see them both drift up or down in sync. A 3rd meter monitors the plate voltage. ...

What about moving that 3rd meter to be voltage across the screen resistor?  That screen current is the missing piece of cathode current, and is the one item potentially out-of-sync with your other measurements.

... My background, 4 yrs ground radio tech USAF, FCC General Radio license with Radar endorsement (passed each the 1st time too 😁) 30 yrs Communications, Navigational and Radar system specialist with the FAA. ...

No offense intended!  With your background you can probably appreciate that troubleshooting something remotely based on statements we cannot be sure are trustworthy means we question everything, because untested assumptions mean the fault goes undiscovered.

So when the statement didn't make sense (cathode current is 8mA more than plate current + screen current), the natural place to look is the measurements that lead to the statement.  (Unless you believe there is 8mA of G1 current contributing to cathode current, along with G2 + anode).

[shooter]

someone asked the other day on another site; "Did they fix the Bermuda triangle, haven't heard about it in so long?"


does the "high" current stay with the socket or follow the tube when you swap tubes and re-check?

[marshallguy]

someone asked the other day on another site; "Did they fix the Bermuda triangle, haven't heard about it in so long?"


does the "high" current stay with the socket or follow the tube when you swap tubes and re-check?

Stay.