martin manning wrote: ↑Tue Sep 07, 2021 8:37 pm
The consensus seems to be generally on the cool side, around 55-60% Pa max based on cathode current.
When you bias this way, do you measure and subtract the screen current? I'm assuming it might only be as much as 5% of the cathode current total.
Strictly speaking yes you should subtract screen current to calculate actual anode dissipation, but most people just calculate it using cathode current. Screen resistor voltage drop is an easy way to get the screen current.
Understood. So it's like six of one, half a dozen of the other I'll have to really listen to it with the speaker, with the cab and live with it for a while.
I have to figure out the best way to bias this amp. Initially I got some strange readings from DCV measurements from the OT CT to GND and plate voltage to GND - a more than 10V difference! But tonight, tried it again and it read less than 1V which is what I got when I took the riskier route and measured between the CT and plate to get the voltage drop. This last is probably the most accurate as its more stable and unless I can get a stable voltage reading from Ct to GND and P to GND it is the one which when considering the OT resistance on both sides of the CT gives me the more realistic figure. I tried the bias test points but got only 0.27 mV on V8 and next to zero for the others so my bias circuit design is not fit for purpose if this is all it can do. I will try it with probes attached to the cathodes and with the bias probe between the valve and socket for comparison. There will be an optimum way to do it.
The reverb tank was installed and it makes a huge difference which side the return goes. I.e. away from the PT is the quietest. A shame because when it is installed in the head it means the in and out jacks will be at the back (facing forwards) where they can't be seen. Hey ho! Not a huge problem.
Stephen1966 wrote: ↑Wed Sep 08, 2021 9:00 pm
Initially I got some strange readings from DCV measurements from the OT CT to GND and plate voltage to GND - a more than 10V difference! But tonight, tried it again and it read less than 1V which is what I got when I took the riskier route and measured between the CT and plate to get the voltage drop. This last is probably the most accurate as its more stable and unless I can get a stable voltage reading from Ct to GND and P to GND it is the one which when considering the OT resistance on both sides of the CT gives me the more realistic figure. I tried the bias test points but got only 0.27 mV on V8 and next to zero for the others so my bias circuit design is not fit for purpose if this is all it can do.
I don't see how the bias circuit design could be at fault. You should get decent readings from either of the methods you describe, with the difference being due to the screen current. Maybe put a fresh battery in your meter, and make sure the probe connections are clean?
It is not dangerous to measure voltage across the output transformer primary. A battery powered multimeter is isolated, and it has high impedance in voltage mode. Measuring the plate current by shunting the output transformer primary with the meter in current mode is the riskier method since there is near zero impedance from one probe to the other.
Stephen1966 wrote: ↑Wed Sep 08, 2021 9:00 pm
Initially I got some strange readings from DCV measurements from the OT CT to GND and plate voltage to GND - a more than 10V difference! But tonight, tried it again and it read less than 1V which is what I got when I took the riskier route and measured between the CT and plate to get the voltage drop. This last is probably the most accurate as its more stable and unless I can get a stable voltage reading from Ct to GND and P to GND it is the one which when considering the OT resistance on both sides of the CT gives me the more realistic figure. I tried the bias test points but got only 0.27 mV on V8 and next to zero for the others so my bias circuit design is not fit for purpose if this is all it can do.
I don't see how the bias circuit design could be at fault. You should get decent readings from either of the methods you describe, with the difference being due to the screen current. Maybe put a fresh battery in your meter, and make sure the probe connections are clean?
It is not dangerous to measure voltage across the output transformer primary. A battery powered multimeter is isolated, and it has high impedance in voltage mode. Measuring plate current by shunting the output transformer primary with the mater in current mode is the riskier method since there is near zero impedance from one probe to the other.
Hmm, it could be the probe connections - I replaced the battery just before starting the build. I suppose it's also possible it could be a mains supply voltage drop between taking the measurements. I'm going to start by calibrating the variac and check it again after all the measurements. When I measure voltage from CT to plate I get a more or less stable figure and if put the meter in max mode it will freeze the highest voltage in 2 second time frames. I think the problem may be that I am not getting good connections with the probe. I'm going to try clipping one to a plate pin and have another to probe the CT. Swapping the leads I don't have to get too near the high voltages. Transformer shunt method - honestly, never tried it. I can replace the fuses in my meter, but it will be very costly if I burn out the windings on the OT using this method. If I can get a good CT voltage reading to GND and a good plate voltage reading to GND I can confirm the Vd against a measurement taken of Vd from CT to plate. I'll be happy if I can figure out why I'm not getting any sort of meaningful reading of the cathode current. The amp is still in a cold bias setting, until I am absolutely confident in my methods, I'm reluctant to decrease the negative voltage on the grid but maybe this could explain why I am getting close to 0mV on the cathodes at the moment.
I have that bias probe kit with a 1R resistor in series between cathode and GND as well. I think I will put it together and see if I get any different results using that. If it were a problem with the solid core wire I used (a break for instance) it's unlikely all four would not work! And anyway, I get a positive resistance between the jack test points.
remove all cathode connections and permanently install 1 ohm 2W resistor in each cathode (pin 8 ) ? i'm doing that by default in all amps, reading in DC mVolts is equal to mAmps in that tube. pairing of the tubes is easy to do, also is a fast method compared to all others (45 seconds to check all the tubes)
bepone wrote: ↑Thu Sep 09, 2021 11:47 am
remove all cathode connections and permanently install 1 ohm 2W resistor in each cathode (pin 8 ) ? i'm doing that by default in all amps, reading in DC mVolts is equal to mAmps in that tube. pairing of the tubes is easy to do, also is a fast method compared to all others (45 seconds to check all the tubes)
45 seconds... that's the plan! I have 1 ohm resistors in there already, what I can't figure out is why I'm getting close to 0 mV readings off them.
I think the problem there, might very well be in the circuit design. Essentially the 1 ohm resistors are in parallel with the cathode and GND, not in series as everyone else does it. I was aiming to maintain the ground potentials of the two pairs of output tubes. Tubes 7 and 8 are grounded at GND point 2 and tubes 9 and 10 at GND point 3 in the chassis. Really I should have included a second COM test point. Doing so, would have allowed me to put the resistors in series and then directed the ground wire to the relevant areas of the chassis. I considered deviating from the original design by combining the two GND points - a sort of GND 2.5 allowing me to use a single common ground for all but I didn't like that idea. So with the parallel circuit, the resistors are in an open circuit when the test jacks are in and the circuit is closed when the jacks are out. It creates a ground loop, which I can live with when it only takes 45 seconds but this ground loop may well be the cause of the false readings.
As the resistor is acting as a divider it's not really dividing anything if the cathode resistor sends 30 mV to ground and the permanent cathode wiring also sends 30 mV to ground creating a ground potential of 30 mV, maybe a fraction of a mV difference between the two as I measured. That's really the only thing I can think of. The amp works and I can measure a plate current. It doesn't make sense that a positive grid current however small would cancel out a positive cathode current.
this is what im using by standard in everything, catch black probe to the ground and just pass with red probe in one hand to the pins 8 of all output tubes,
CamScanner 09-09-2021 21.47__01.jpg
You do not have the required permissions to view the files attached to this post.
Nice drawing. Exactly, resistors in series in a closed circuit. That's the way it should have been done. I think I will need to do some work on the circuit to achieve this.
[Added] Simple enough I guess, just disappointed I missed it if I'm right about the permanent and parallel cathode wiring creating a ground potential equivalent to the output of my 1 ohm parallel resistor.
bepone wrote: ↑Thu Sep 09, 2021 12:55 pm
45-50mV is fine (this is anode and g2 currents together , summed through cathode resistor, so if 50 total, anode is cca 45mA which is fine)
Yeah, I was thinking of something like that. I am curious about the screen current though. Some people say cca 5 mA and some say cca 5% - they are not the same thing obviously so I think I will be measuring the voltage drop across the screen resistors to factor in the equivalent subtraction from cathode current. It's more for curiosity than anything else, I'm sure you are probably in a very similar area with mA and like a lot of the exercises I've done already it will probably only confirm what others already know. Still, I learn a little bit more, each time. So not a wasted effort.
g2 current/voltage drop over g2 resistor / is also good to measure, also for other reason.. if one tube is not ok and it is pulling more g2 current/because faulty/or oscillating/ it is possible to detect it through bigger voltage drop,
If I am somehow, through trying to measure the cathode current using the bias circuit, and the bias circuit is rendering the cathode current = 0 Vdc through ground loop/ground potential (when it should be between 25 and 30 mA - it is biased max cold) could that induce a 10 Vdc drop in voltage on my plates? Not the CT, that has the backing of a full military band in the form of B+1 voltage and current
I don't trust my parallel bias circuit design and I can easily change it by putting the 1 ohm resistors in series with the cathode lines out. The cable to each test point can simply go before each resistor in series not after as they are now. It wont isolate the 1 ohm resistors when they are not in the test circuit. Not that that makes much difference, but it will still make a ground loop when I plug the test jacks in.
I guess what I would also like to know is if I am somehow damaging my tubes using this bias test circuit?. The voltages in the cathode in normal operation are really small but a 10 V difference between the plate and the CT is [or rather suggests] a tenfold increase [or decrease - I can't get my head around it!] of current.
Thank goodness, it still plays but less and less - this thing scares me. That is probably quite healthy
martin manning wrote: ↑Thu Sep 09, 2021 7:49 pm
Are your 1Ω cathode current sense resistors located as shown below?
No, that's the problem - or part of the problem - I think. Your sketch is how I will have to rewire them. I thought I was being smart by placing them after the TP arrows in a parallel configuration and isolating them from the operational cathode-to-ground, i.e. the normal cathode-to-ground circuit when you don't get bright ideas like this one!
Bias test points.jpg
Getting at them is a bit trickier now but I only need place them in series in the normal cathode-to-ground and run the test point cable from between the cathode pin and resistor to the test point jack. It will result in the configuration shown in the schematic you marked up.
You do not have the required permissions to view the files attached to this post.
I'll have to really listen to it with the speaker, with the cab and live with it for a while. 
45 seconds... that's the plan! I have 1 ohm resistors in there already, what I can't figure out is why I'm getting close to 0 mV readings off them.