Could current null meters like these be a useful addition to Williamson PP stages that have one of those balance potentiometers? I've seen fixed bias amps with a handy switched meter for verifying the bias setting on each side, but does that alone verify everything DC-wise that should be correct balance wise.
A difference in DC balance will tell you that the power tubes need to be matched again (replaced). As de DC value is derived from the voltage over the cathode resistors. If there is a different voltage then there is a different current. This is an unbalance and will lead to distortion and maybe hum (as the output transformer now will no longer be able to nullify this).
Looks like a 40,s/50,s meter .
You will of course have to take into account the series resistance of the meter plus any measured inductance if moving coil if you intend to fit it permanently.
What are the measurements ?
You will of course have to take into account the series resistance of the meter plus any measured inductance if moving coil if you intend to fit it permanently.
What are the measurements ?
No measurements, I'm just wondering in general if this kind of null meter could have an application in a PP amp to indicate proper DC current balance. I look at hundreds of Williamson schematics and don't recall seeing a null meter in one, it would interesting to see if and how one can be used though. Like the bias meter I would have it switchable in or out, to verify/adjust things occasionally.
The reason I asked is you are introducing an inductance into a tube used in an audio circuit .
Quite common in all types of RF receivers for various reasons including introducing a time constant and induced emf/phase angle.
Some redesign /component changes would need to be applied.
Quite common in all types of RF receivers for various reasons including introducing a time constant and induced emf/phase angle.
Some redesign /component changes would need to be applied.
Understood about the inductance, that's why I was thinking switched meters, you'd only use them to verify or adjust bias and balance pccasionally.
I think you need to sense in the anode lead for pentode connected tubes. There can be several mA mismach in screen grid current between two tubes with identical cathode current.
You need to know 2 things about the meter:
The DCR of the meter coil.
The full scale sensitivity of the meter.
Suppose the meter is 500uA from center to one side (1mA full scale).
Suppose the meter coil DCR is 55milliOhms (common for 1mA full scale movements).
Suppose the output tubes self bias is 30V.
1. If you use individual self bias with individual bypass caps, for the most part, the signal current is bypassed by the bypass caps.
And, the inductance of the meter's moving coil will not affect the circuit.
Connect the meter in series with a 10k resistor. Connect that series meter and 10k resistor, from one cathode to the other.
If the voltage difference from one cathode to the other cathode is 1V, the current will be
0.1 mA. The meter will read 1/5 of the way from center to one side of full scale.
The only danger here, is if one tube has 30V self bias, and the other tube goes bad, and draws no current, then the meter will receive 3mA current, and will peg the meter to one side (a 6X overload).
2. For fixed bias circuits, with individual 10 Ohm resistors in series with each cathode:
With 50mA of cathode current in each 10 Ohm resistor, the voltage across each 10 Ohm resistor is 0.5V.
A 100 Ohm resistor in series with the meter will give a sensitivity of 0.05V from center to one side.
A difference of 5mA of the two cathode currents, will cause a differential voltage of 0.05V.
That will cause the meter to swing to one side or the other.
1mA will cause the meter to go 1/5 of the way from center toward one side.
The problem with this method is that the meter movement may be beat to death, because of the large differential AC currents in the two 10 Ohm resistors.
. . . So, I guess that means you have to go to Individual Self bias resistors, and Individual self bypass capacitors.
Or . . .
3. Use a meter switch, a resistor in series with the meter. That makes the current meter act as a voltmeter. The problem is, you are going to get readings from the center to one side, not from the left side to the right side.
Again, this will work for individual self bias resistors and individual bypass caps.
But it will beat the meter to death if you use fixed bias and individual 10 Ohm resistors (well, you could bypass the 10 Ohm resistors with 10,000 uF across each one).
4. Your choice.
5. For reference to using a meter to match output tube plate currents, take a look at the Heathkit W5M Manual (original manual is free on the web). Look at the schematic first. Then look at one of the last pages for a graph of distortion versus un-balanced current.
That is a real education about the effect of un-balanced output tube current on early lamination saturation can do to distortion when global negative feedback from the output transformer secondary is used.
6. Ooops. I finally enlarged the picture of your meter. It has 1mA 1/2 scale, 2mA full scale.
The DCR of the meter coil.
The full scale sensitivity of the meter.
Suppose the meter is 500uA from center to one side (1mA full scale).
Suppose the meter coil DCR is 55milliOhms (common for 1mA full scale movements).
Suppose the output tubes self bias is 30V.
1. If you use individual self bias with individual bypass caps, for the most part, the signal current is bypassed by the bypass caps.
And, the inductance of the meter's moving coil will not affect the circuit.
Connect the meter in series with a 10k resistor. Connect that series meter and 10k resistor, from one cathode to the other.
If the voltage difference from one cathode to the other cathode is 1V, the current will be
0.1 mA. The meter will read 1/5 of the way from center to one side of full scale.
The only danger here, is if one tube has 30V self bias, and the other tube goes bad, and draws no current, then the meter will receive 3mA current, and will peg the meter to one side (a 6X overload).
2. For fixed bias circuits, with individual 10 Ohm resistors in series with each cathode:
With 50mA of cathode current in each 10 Ohm resistor, the voltage across each 10 Ohm resistor is 0.5V.
A 100 Ohm resistor in series with the meter will give a sensitivity of 0.05V from center to one side.
A difference of 5mA of the two cathode currents, will cause a differential voltage of 0.05V.
That will cause the meter to swing to one side or the other.
1mA will cause the meter to go 1/5 of the way from center toward one side.
The problem with this method is that the meter movement may be beat to death, because of the large differential AC currents in the two 10 Ohm resistors.
. . . So, I guess that means you have to go to Individual Self bias resistors, and Individual self bypass capacitors.
Or . . .
3. Use a meter switch, a resistor in series with the meter. That makes the current meter act as a voltmeter. The problem is, you are going to get readings from the center to one side, not from the left side to the right side.
Again, this will work for individual self bias resistors and individual bypass caps.
But it will beat the meter to death if you use fixed bias and individual 10 Ohm resistors (well, you could bypass the 10 Ohm resistors with 10,000 uF across each one).
4. Your choice.
5. For reference to using a meter to match output tube plate currents, take a look at the Heathkit W5M Manual (original manual is free on the web). Look at the schematic first. Then look at one of the last pages for a graph of distortion versus un-balanced current.
That is a real education about the effect of un-balanced output tube current on early lamination saturation can do to distortion when global negative feedback from the output transformer secondary is used.
6. Ooops. I finally enlarged the picture of your meter. It has 1mA 1/2 scale, 2mA full scale.
Last edited:
Ah, I see. Its simply a matter of reading 0 volts when its balanced, nulled. So this is the appropriate type of analog meter for this application as I would need it to read both negative or positive, without pinning it of course, and for volts not milliamps as seen in the photo above. Of course any old digital volt meter could be used as those inherently swing positive or negative with a sign.
I could install any meters with a switch to disengage it when listening I suppose? Beautiful circuit. Maybe I can scrounge up an old panel volt meter with a centered Needle.
Thanks!
Last edited:
to add... The numeric value reading of the analog meter isn't important, just indicating zero is, so I suppose I could get a sensitive meter and put large value precision series resistors on each side to increase the impedance to that of a digital meter. then any zero-center meter will work.
I always put a couple diodes anti-parallel across the meter movement to help protect it.
Also a small ceramic cap if it's going to be exposed to an RF field.
Also a small ceramic cap if it's going to be exposed to an RF field.
Demonkleaner's diode protection is really essential - these meter movements are fragile. Also, you don't need any precision resistors; only the center point matters. You might even want a variable resistor (a pot) in series instead, so you could adjust sensitivity on the fly.
In DTN Williamson's day, meter movements were too expensive to "waste" permanently attached someplace, so provision was made for external ones. Today we can afford all kinds of luxury items, including beautiful made-for-WWII meters. Splurge. Be happy.
YOS,
Chris
In DTN Williamson's day, meter movements were too expensive to "waste" permanently attached someplace, so provision was made for external ones. Today we can afford all kinds of luxury items, including beautiful made-for-WWII meters. Splurge. Be happy.
YOS,
Chris
A typical classical 1mA meter movement has a DC resistance of 55 milli Ohms.
1mA into 55 milli Ohms = 55mV
No diode I know of turns on hard with 110mV input (a reasonable meaning of 'turns on hard' is a diode that is 55 milli Ohms, so that it shares the current 1:1 with the meter DCR.
A possible example that may apply:
Of course, a 1mA meter that has a 400 Ohm resistor in series, will read 1mA when you apply 0.4 Volt.
You can put a diode across the meter and resistor that are in series, if the diode does not conduct significant current at 0.4V, but does conduct significant current at 0.6V.
If we have enough RF field to affect the meter movement so that it needs a capacitor across it, then we better have other methods to keep the RF out of the audio amplifier, or the amplifier will be affected worse than the meter.
The typical home does not have a 5kW AM transmitter next door.
1mA into 55 milli Ohms = 55mV
No diode I know of turns on hard with 110mV input (a reasonable meaning of 'turns on hard' is a diode that is 55 milli Ohms, so that it shares the current 1:1 with the meter DCR.
A possible example that may apply:
Of course, a 1mA meter that has a 400 Ohm resistor in series, will read 1mA when you apply 0.4 Volt.
You can put a diode across the meter and resistor that are in series, if the diode does not conduct significant current at 0.4V, but does conduct significant current at 0.6V.
If we have enough RF field to affect the meter movement so that it needs a capacitor across it, then we better have other methods to keep the RF out of the audio amplifier, or the amplifier will be affected worse than the meter.
The typical home does not have a 5kW AM transmitter next door.
Last edited: