I have a SET amp that I have used in the past to drive a compression driver. It has been sitting idle and I decided to check it on the bench before putting it back in the system. The input tube is a 404a, wired as a triode, capacitor coupled to a 6L6GC output tube, also wired as a triode. The B+ for the 6L6s is 300 volts. It is cathode biased with a 400 ohm resistor and a 1000 mfd capacitor. The target current is 40 ma. One channel is right at 40 ma. The other channel is at 26 ma. BUT, if I pull the input tube for that channel, the current goes to 40 ma. No problem, it's a leaky coupling cap, right? Nope, I changed both of them out and it still is at 40 ma without input tube, but 26 ma with. The B+ for both input and output tubes stays the same either way. I reflowed all of the soldered connections in that channel. No change. What am I missing?
Member
Joined 2009
Paid Member
Sometimes a dc problem is actually an rf problem, a parasitic oscillation problem. That 404a looks like an absolute race horse of a tube, does it have grid stoppers right at the socket ? A well decoupled supply, careful grounding, etc.
i dabbled in 6cl6 before and dropped it, just like Bigun said, prone to hf oscillations, probably due to the very high gm, grid stoppers are a must....also grid leak resistors like to be on the low side...too much for me to handle at the time, i simply chose the 6bq5 instead...
Does each channel has a seperate heater supply, shared by a 404A and a 6L6GC? Maybe the heater of the 404A causes a short, pulling the heater voltage of the 6L6GC down with it.
With a B+ of 300 V and a cathode resistor of 400 Ohm your 40 mA seems very low to me.
The cathode voltage would be V = I x R = 0.04 x 400 = 16 V. Even when the dc-resistance of the primary would be 350 Ohm (which seems unusually high to me), the voltage between plate and cathode would stil be 300 - (16 + 14) = 270 V.
When looking at these curves from the last page of the STC Application Report on the 6L6G, I would expect a current of at least 60 mA.
With a B+ of 300 V and a cathode resistor of 400 Ohm your 40 mA seems very low to me.
The cathode voltage would be V = I x R = 0.04 x 400 = 16 V. Even when the dc-resistance of the primary would be 350 Ohm (which seems unusually high to me), the voltage between plate and cathode would stil be 300 - (16 + 14) = 270 V.
When looking at these curves from the last page of the STC Application Report on the 6L6G, I would expect a current of at least 60 mA.
I have already swapped both the input and output tubes. I will check for oscillation next. I don't understand why I am getting 40ma with that B+ and cathode resistor either. It doesn't agree with the calculations, but there it is. Maybe because the output stage is actually parallel feed, with a choke and a RatShack 70v TX as the output transformer? Meanwhile, why would oscillation affect the bias of the output tube?
Because of rectification effects due to the even-order non-linearity of the valve.
For example, if the positive peaks are high enough to drive the valve into grid conduction, the grid DC voltage will decrease until the DC current through the grid leak resistor matches the average rectified current.
For example, if the positive peaks are high enough to drive the valve into grid conduction, the grid DC voltage will decrease until the DC current through the grid leak resistor matches the average rectified current.
Last edited:
Eureka! The correct answer is "oscillation"! I put a shorting plug on the input and an 8 ohm resistor on the output and Shazaam! , I am getting 40 ma with the input tubes in place. Thanks to everyone who responded. This is only a temporary fix. I had an 82 ohm resistor at the grid of the input tube, but apparently this was insufficient. I will be replacing it with a 1k carbon film as close to the pin as possible, to see if it calms everything down. Thanks again. diyaudio is a tremendous resource thanks to all you lovely people.