Interesting bias question

[scottl]

Here is a good one for you tech gods.....

On one of my amps, I have a switch for 3H choke or 500 ohm resistor. 460V on plates. Fixed bias. 1K screen resistors. KT77 tubes.

The amp has external bias points and individual trimmers. When I measure the bias w/choke ON, I get 34ma. If I switch to resistor, I get a measurement of about 30ma. I realize that using the 1ohm resistor method, you measure the screen current as well. Question is, which measurement would be closer to the actual plate current. If I were to set the bias, would you recommend using the lower or higher number as a guideline, only being concerned with accuracy and not "safety buffer" (ie. setting to the higher number even though it is not really that high due to the screen current)

I suppose I should measure it using the shunt method, but I really want to understand the reason why there is a big change in my bias reading. The voltage does sag a little when the 500 ohm is there.

Oh, I know my ears in the end win in this, I just want the tech answer.

Thanks!!!

[Blue Strat]

If your readings are correct, the only possible explanation is that the resistance of the choke is higher than 500 ohms (and I'm not sure that makes sense). If so, apply Ohm's Law I=E/R for screen grid current.

Try disconnecting the choke and measure it's resistance.

The simple answer is that the lower reading is closer to the actual plate current, but only because it's lower than the other reading. ;)

[scottl]

Thanks Mike,

My choke measures about 100ohms. It is a standard Fender choke.

I assume that the lower is more accurate, but I was wondering why the difference. I bet John Phillips knows!!!!

I'll also open the amp up and measure the drop across the 1K screen resistors and see if the screen current changes much.

[Blue Strat]

Thanks Mike,

My choke measures about 100ohms. It is a standard Fender choke.

I assume that the lower is more accurate, but I was wondering why the difference. I bet John Phillips knows!!!!

I'll also open the amp up and measure the drop across the 1K screen resistors and see if the screen current changes much.

Good idea. It could also be due to randomness of readings and/or line voltage changes between readings. To get the full picture, you should read your plate and screen voltages when taking the bias measurements.

[scottl]

Yep... Will do. But it is not random or line voltage. I can switch the choke on or off in real time with the probe in. Bias reading always changes the same amount. I have also noticed this exact situation for years (One of my Fuchs has a choke/resistor switch). I will say the change is a little less when using 6L6 tubes. Somehow this is related to the screens, me thinks....



Good idea. It could also be due to randomness of readings and/or line voltage changes between readings. To get the full picture, you should read your plate and screen voltages when taking the bias measurements.

[VacuumVoodoo]

Yep... Will do. But it is not random or line voltage. I can switch the choke on or off in real time with the probe in. Bias reading always changes the same amount. I have also noticed this exact situation for years (One of my Fuchs has a choke/resistor switch). I will say the change is a little less when using 6L6 tubes. Somehow this is related to the screens, me thinks....

You thinks correctly.
KT77 is somehow erroneously taken to be close equivalent of EL34 and often thought of as "EL34 on steroids". This is not so, KT77 is a beam tetrode while EL34 is a pentode. KT77 was designed to be primarily used in ultralinear PP configuration. Using it in a non-UL configuration requires that screen grid voltage be well regulated, fixed and not exceeding 300V. This is usually not observed in guitar amps. The original datasheet is HERE (http://www.mif.pg.gda.pl/homepages/frank/sheets/086/k/KT77.pdf)

Most probably (>95% probability) the plate supply to OT CT is taken before the choke/resistor, screen grid supply after. Lets assume, for arguments sake, that the premp circuitry post choke draws 12mA (6 x 12AX7) in addition to the two screengrids which will draw at idle ca 5-7mA each.

Current through choke: I(preamp)+2xI(sg)= 12+2x6mA=24mA
Voltage drop across choke: 100Ohmx24mA=2.4V
Voltage drop across screen resistor: 1kx6mA=6V
Screen grids are 2.4+6=8.4V below B+, in this case ca 452V

Switch over to 500Ohm resistor:
Voltage drop across 500 ohm= 12V which leads to screen grids sitting 18V below B+ i.e 442V (ok, average voltage, ripple with just a resistor will be significantly higher than with choke)

Now, if you gaze long enough at tube characteristics you'll see that plate current is strongly dependent on screen grid voltage besides bias and plate voltages : everything else being held constant a drop of 10V in screen dc voltage can easily cause a few mA drop in plate current, idle screen grid current will change too but proportionally less.

In other words you are observing a perfectly normal behaviour.

[Old Tele man]

...try "playing" with those numbers using the "Child-Langmuir 3/2's Law" equation for tetrode/pentodes:

Ik = (Ip+Is) = K*(-Vg + Vs/µ1 + Vp/µ2)^(3/2)

where:
Ik = Cathode current
Ip = Plate current
Is = Screen current
K = Tube Perveance coefficient, amps-per-volt^(3/2)
µ1 = Tube SCREEN (triode) amplification factor
µ2 = Tube PLATE (tetrode/pentode) amplification factor
Vg = Control grid voltage
Vs = Screen grid voltage
Vp = Plate voltage

[John Phillips]

You thinks correctly.
KT77 is somehow erroneously taken to be close equivalent of EL34 and often thought of as "EL34 on steroids". This is not so, KT77 is a beam tetrode while EL34 is a pentode. KT77 was designed to be primarily used in ultralinear PP configuration. Using it in a non-UL configuration requires that screen grid voltage be well regulated, fixed and not exceeding 300V. This is usually not observed in guitar amps. The original datasheet is HERE (http://www.mif.pg.gda.pl/homepages/frank/sheets/086/k/KT77.pdf)

Most probably (>95% probability) the plate supply to OT CT is taken before the choke/resistor, screen grid supply after. Lets assume, for arguments sake, that the premp circuitry post choke draws 12mA (6 x 12AX7) in addition to the two screengrids which will draw at idle ca 5-7mA each.

Current through choke: I(preamp)+2xI(sg)= 12+2x6mA=24mA
Voltage drop across choke: 100Ohmx24mA=2.4V
Voltage drop across screen resistor: 1kx6mA=6V
Screen grids are 2.4+6=8.4V below B+, in this case ca 452V

Switch over to 500Ohm resistor:
Voltage drop across 500 ohm= 12V which leads to screen grids sitting 18V below B+ i.e 442V (ok, average voltage, ripple with just a resistor will be significantly higher than with choke)

Now, if you gaze long enough at tube characteristics you'll see that plate current is strongly dependent on screen grid voltage besides bias and plate voltages : everything else being held constant a drop of 10V in screen dc voltage can easily cause a few mA drop in plate current, idle screen grid current will change too but proportionally less.

In other words you are observing a perfectly normal behaviour.If I'd got here first I would have said the same thing but with less math ;).

Basically, dropping the screen voltage a few volts relative to the plate reduces the conductance of the tube a bit, so with the same plate and bias voltages applied you get less current.

In fact, the higher reading (with the choke) is closer to the true plate current, at least proportionately, because the screen current remains about the same, so the error due to it is a smaller proportion... but not much :). The screen current is often as much as 10-20% of the total (but not known precisely), which makes any plate current measurement using the 1-ohm resistor method fairly inaccurate anyway... which is the main reason I don't like it. (At least the error is in the safe direction, since the plate current is always less than the cathode current.)

If you're concerned about finding the true plate current, it's much better to measure the voltage drop across the OT primary and calculate the current from the DC resistance of it (remember that the two sides are normally not exactly the same either). This method is a bit more dangerous, though.

[Blue Strat]

You thinks correctly.
KT77 is somehow erroneously taken to be close equivalent of EL34 and often thought of as "EL34 on steroids". This is not so, KT77 is a beam tetrode while EL34 is a pentode. KT77 was designed to be primarily used in ultralinear PP configuration. Using it in a non-UL configuration requires that screen grid voltage be well regulated, fixed and not exceeding 300V. This is usually not observed in guitar amps. The original datasheet is HERE (http://www.mif.pg.gda.pl/homepages/frank/sheets/086/k/KT77.pdf)

.

Yes but since nearly NO currently made tubes meet the original published spec, is this argument even relevant? You can call something anything you want, but that doesn't change what it is;)

[VacuumVoodoo]

Yes but since nearly NO currently made tubes meet the original published spec, is this argument even relevant? You can call something anything you want, but that doesn't change what it is;)

Well, so far as difference between a beam tetrode and pentode yes. OT: ask me about new production 7591, there are 2 brands available, only one is reasonably close to original specs, the other I suspect to be a 6V6 with 7591 pin configuration.

[Old Tele man]

OT: ask me about new production 7591, there are 2 brands available, only one is reasonably close to original specs, the other I suspect to be a 6V6 with 7591 pin configuration....welcome to the "new-world" order of STANDARDS, where you're guaranteed to find one (or more) that's suitable; but if not, just 'make-up' your own like everybody else does !!(tongue-in-cheek comment, of course)

[Carol-AnnAmps]

Obviously my telephone answer last night wasn't enough :)

There is no such thing as accuracy when it comes to correct biasing, there's sensible boundaries, thats all.

Try this simple experiment: check your bias in your own house. Now take it someone elses house or a club in a different town or neibourhood and check it again. You will soon realise you may be chasing rainbows.

Here's another simple answer, if you use NOS tubes use the higher value, if you use dodgy Russian ones, use the lower value. There that was easy and I didn't need to quote any formulas. I could do though if it makes you feel more empowered and makes me look credible......

Just busting ya Scott.....

Take care mate.