vacuum tube voltages wacky....why?

Thread Starter


Joined Apr 3, 2019
Restoring and a HG-9A Heathkit Signal generator. So I measured some voltages and I see that my 6 x 4 vacuum tube is being over driven yet the cathode is putting out less voltage than it should but the tube worked well on the tube tester does this mean it could be weak ?....and following that circuit the next two tubes Voltage isout of whack too....any thoughts?upload_2019-10-15_15-18-16.jpeg


Joined Jul 18, 2013
Your PI filter capacitors could be low capacitance, typical problem with old equipment.
If the rectified DC is low, it is going to also be low on V1 V2.
What about your AC supply?
Also no neutral or earth GND indicated.
What is your location?


Joined Jan 15, 2015
I agree with Max also. I would start at the beginning. Measure between Ground (Transformer CT) Red / Yellow and each Red AC secondary. Note your AC voltage. Next measure the DC voltage of the pi filter output. Since the 6X4 is a dual diode vacuum tube (valve) I can see it being over loaded but not over driven. Measure the plate voltage on V2 (pin 6). Then just remove V1 and V2 and measure the plate voltage pin on the V2 socket. Has it increased? If the voltage remains low then start looking long and hard at the pi filter 40 uF capacitors. The typical choke on these designs with a 6X4 was about 8.0 mH. Again, my best guess would be a cap or the caps.



Joined Apr 21, 2014
Extending the excellent points above, you can always remove the other two vacuum tubes and see if the load reduces and +B voltage increases. Especially the 6AU6, which has a very significant drop on its plate.

Also, if you are measuring all this with a modern DMM, keep in mind the original voltages may have been done with a 20k/V meter at 1kV range, which would give about twice the input impedance. That could account for a small reduction on the measured voltages.

To check if the Pi filter capacitors are dried, put the meter in AC mode and measure +B. You will have a strong AC component if the premise is true. (I would replace/change the capacitors anyways) **

** (edit) check the remarks below
Last edited:


Joined Jun 4, 2014
To check if the Pi filter capacitors are dried, put the meter in AC mode and measure +B. You will have a strong AC component if the premise is true. (I would replace/change the capacitors anyways)
Many (most?) small DMMs do not have capacitor coupling on AC volts ranges so you will get incorrect readings if there is significant DC present as in this case.


Joined Apr 21, 2014
Good point, which could be amended by adding a 4.7nF or 10nF high voltage (630V, 1kV) capacitor in series with the probes.

Although when I used DC coupled meters in these scenarios, the wild varying voltage shown was indicative of serious problems on the rail.
Getting there....but raising the voltage output brings distortion to the sine wave with flat spot at bottom. Reducing output voltage restores clean sine wave but at 1/3 voltage tubes installed...all new caps the 6au6 not able to be driven hard enough? Also raising the output frequency by the multiplier switch lowers output voltage ...inverse relationship....afraid I’m a nubbie runnin around a squirrel box! (Never bored tho!)



Joined Jan 23, 2018
Flat spots are either cutoff or saturation, and indicate that the control grid bias is not set for minimum distortion. And I think that the screen grid voltage has some effect as well. So measuring all of the voltages of the tube where the distortion appears can tel you a lot, and the manual has a table of voltages provided. It might also be a coupling capacitor breaking down, but that is less likely. I suspect the 16 mFD capacitor on the screen grid of the 6AU6


Joined Feb 20, 2016
Capacitors are usually the first culprit to look at. I would think about replacing all the electros. They do dry out over time and cause many odd faults.


Joined Jan 23, 2018
One moore possibility is that the flat-topping is beyond the max specified output, in which case that may be fine. Look at the specifications in the manual, nothing is promised above the specified max output. If it is below specifications then I suggest checking that lamp in the circuit that is used to stablize the amplitude, and also that 20 Mfd capacitor next to it, which may be leaky. And check the variable resistors and switch contacts, which may be dirty.