Trouble measuring fixed bias LTP bias voltage

[Gaz]

I'm trying a fixed bias LTP for the first time, using a voltage divider/filter off the B+ rail for my bias voltage.

It works fine, but I'm just trying to dial-in the bias voltage to about -2V, but am get some discrepancies in my readings.

For example, when I measure from grid to cathode I get -1.85V, but when I measure the from ground to cathode and ground to the fixed bias voltage, then subtract the difference I get -3.2V!

What gives? I have a Fluke 187 Meter, but it seems like it must be an error. I really just wanna know which measurement is the one I should go by.

[xtian]

Sorry, not following. You're tying to adjust the bias of the LTP, not the power tubes?

[Gaz]

Yes, nothing to do with the power tubes, just the LTP phase inverter.

Trying to set the bias with a typical voltage divider, but getting two different readings depending on if I measure between cathode and grid, or instead cathode to ground and grid to ground, and then measuring the difference between the two. The should be the same, right?

[JazzGuitarGimp]

Unless you are using a high input impedance meter, such as a VTVM, or a FET input meter, you cannot measure the grid voltage directly. Your Fluke's input impedance is too low, and it is pulling the bias voltage reading toward ground. Your best bet is to measure the grid voltage at the voltage divider node, and trust that the grid voltage (through the 1M-ish resistor) is the same.

[Gaz]

Unless you are using a high input impedance meter, such as a VTVM, or a FET input meter, you cannot measure the grid voltage directly. Your Fluke's input impedance is too low, and it is pulling the bias voltage reading toward ground. Your best bet is to measure the grid voltage at the voltage divider node, and trust that the grid voltage (through the 1M-ish resistor) is the same.

Thanks Lou, that really helps. Here's a drawing of the inverter as is. So you are saying I would measure from from ground to the 100K/1M junction, and at the cathode them subtract the difference?

Is there an alternate way I can check the bias as well?

Thanks!

[JazzGuitarGimp]

Yes, subtract the voltage at the cathode (with respect to GND) from the voltage across the 10uF cap. This should always produce a negative number.

For example, let's say the voltage at the cathode is 59V, and the voltage across the 10uF cap is 57V, so 57 - 59 = -2V

[Gaz]

Thanks Lou, that really helps!

[Stevem]

You need a Fluke with a relative button function !

[martin manning]

You need a Fluke with a relative button function !

Not at all! That function is an offset measurement feature for voltages and resistance! Not the issue here! The problem is loading the circuit with the meter! Just don't measure from grid to ground!

If you measure cathode to grid, does that agree with Lou's method?

Why the 100k off the first grid? Seems low.

[Gaz]

You need a Fluke with a relative button function !

Not at all! That function is an offset measurement feature for voltages and resistance! Not the issue here! The problem is loading the circuit with the meter! Just don't measure from grid to ground!

If you measure cathode to grid, does that agree with Lou's method?

Why the 100k off the first grid? Seems low.

Ok, Martin, that's what I was thinking too concerning the relative function. I really need one of those trick benchtop meters... If my bench only had the space!

The 100K is a copy from the late 70's Hiwatt:
http://hiwatt.org/Schematics/DR_Pre2Input.gif

I am using a voltage divider instead of grabbing the bias voltage off the cathode, however.