'69 JMP SL100, Convenient Bias Control PCB

[bordonbert]

Hi. Don't know if anyone has gone down this route this far before. We know that the early SL100s need valve socketed bias monitors to set bias current. It's simple to do that but I always find that approach pretty bothersome. The usual recommendation is to add 1R resistors to the cathode which is great and, at 1R, has no effect on amplifier performance. The voltage based sockets have the 1R resistor in them and it doesn't matter whether your meter is connected or not. The current based ones require you to keep the meter monitoring the current direct connected or the valve is out of circuit. This means you have to switch off and swap over to read each valve, (unless you have 4 of the things). To my thinking there had to be a better way.

I added the 1R resistors to each of the cathodes making sure to simply unsolder the grounding wires and leave them in their original length and solder the resistor to them. I knocked up an extremely simple PCB with only direct connections for wires from the cathode/resistor junctions, two terminals for the meter reading voltage, and a 4-position slide switch. Now I can simply clamp the meter on the two terminals, run the amp up and switch from valve to valve to monitor and set the bias current.

It is a bit of a luxury but it is penalty free. The switch is neither carrying any voltage of note nor any current at all as it only feeds the high impedance voltmeter. There is no issue with oscillation as it's the cathode nor pickup as it represents a 1R resistance to ground. It now means I don't have to probe around the high voltage areas of the valve base to select which valve to read. And the whole job is absolutely reversible in about 15minutes as the PCB is firmly mounted on two 4BA brass standoff pillars screwed on top of the existing cap clamp screws. That's why the PCB is the size it is.

It may not float everyone's boat but I prefer a working amp which can be easily reversed to original and it definitely makes reading all of the valves' bias a doddle. Anyone else tried anything on these lines?

Bias_Control_PCB.jpg

[bepone]

looks eeeh. technically if you think deep enough, you will conclude that is possible to make oscillations that way.

what was the problem in checking with multimeter (one probe clipped to the gnd, and other in the hand) voltage to each cathode?

[bordonbert]

You are absolutely right Bepone, nothing wrong with the old fashioned method of metering the voltage, it's just a little bit fiddly to do it by probing around the base of the valves and if you do make a slip... With the anode voltage being around there, there is always the chance of a mistake causing some "unintended event". Not a big problem almost all of the time - until it is!

It was just a way of speeding up the process. Fiddling around to fit new valves during a gig is not exactly the right environment for reliability and anything which makes that quicker or less prone to error has to be welcomed. And it does come with no penalty as the whole setup, with only a few de-soldering actions, is completely reversible. No drilling, no cutting of any wires, no component changes are needed.

As to oscillation, I would suggest it's virtually impossible with the frequencies to which our stages are limited. And as I said, you have that ultra-low 1R impedance to ground all along those wires and that means no induced signals of any significance whatsoever, neither voltages nor currents. If that was a high impedance point then things would definitely be different. As a pointer, despite a couple of them running alongside the heater wires for some inches there is absolutely no increase in hum with this in place. And anything untoward would only be added onto the cathode which is effectively the negative input to the stage, (with the grid being the positive input). Oscillation really is not going to happen due to those wires from the cathode to the PCB.

I'm certainly not suggesting that this is a "must do" mod for everyone. There is nothing wrong with the standard method or any of them, valve base plugins included. It was just an idea as a way of streamlining the process.

[sluckey]

Nice job. But it's only part of the process. You still gotta poke the plate pin for every tube to complete the process. Be careful.

[bordonbert]

Hi Sluckey. I don't want to ruffle feathers here and this sort of discussion always seems to but you have posted a reasonable point. I certainly would not disagree with you per se but I'm afraid I'm unusual in issues like this. The 70% rule is a rule of thumb, nothing more. It is somebody's idea of a safe region to stick within below the 25W maximum anode dissipation. And the idea that maximising bias to the level it predicts is also unnecessary. I think of it this way.

Let's say my anode voltage is going to be around 470V. It WILL vary according to local conditions, what is nearby on the grid, time of day, state of supply lines in different locations etc. Why should I bother reading it every random time I reset my bias, what is the point? It's only going to be correct at that moment in time and it definitely will drift and change making my measurement inapplicable. Now the amp tech working on any number of amps would have to start by measuring it but as an owner of a particular amp why do I need to check it at the start of every re-biasing? Is it going to have changed significantly, (barring component failure which is a different scenario), particularly upwards?

From what I have always seen the thinking goes; "I must measure my anode voltage, I must take 25W as my maximum, 70% of that is 17.5W, divide my anode voltage into 17.5 and voila, I have the bias current I want to set my amp to". The feeling always seems to be that higher is better so get it on the brink of the 70% prediction. It doesn't make sense to me.

I prefer to say, "my rule of thumb is anode voltage of 470V, (just based on previous measurements and it will vary), same 17.5W maximum, bias current can be UP TO 37mA, set it to somewhere around 30mA and listen". Dissipation there will be 14.1W. If the HT rises to say 500V it is still only 15W. Now that is classed as biasing it cold but soundwise I can hear no significant difference in my amp when played at normal sensibly high volumes.

The bias is there to eliminate crossover distortion, ("reduce" in the real world), we all know that. What non-technical people miss is that crossover distortion has a curious characteristic over other forms of distortion. It more or less stays constant in level no matter what the signal level is. It is pretty much a fixed size "dead spot" at zero crossing. That means that, assuming it is minimised reasonably by biasing the output, at high drive levels it makes a small contribution to signal but as the signal level drops its contribution becomes greater as a proportion of the overall signal. I know from my hifi work, (yes it's not the same but it is much more demanding than our guitar environment), where reducing crossover distortion is a real major source of problems, it is the quiet passages that it affects most and these are used as the benchmark to test against and improve. What is inaudible at high volumes is grainy at very low ones. Guitar work is different. We play at higher volumes as the norm and we generate much greater levels of distortion deliberately than the crossover will contribute at those levels. It is swamped. I suspect that almost everyone, definitely most non-technical players, sets their bias according to what they are told with anode voltage as the starting point, the 70% rule as a rigid benchmark, and then sets to the maximum bias permitted by that calculation because that will sound best gauged against listening at low clean volumes. I feel that is totally unrealistic and unnecessary.

I don't bother measuring anode voltage for a run of the mill bias check or reset as I did it a while back and I know I am working at less demanding levels than any reasonable increase of anode voltage could demand. 70% dissipation is a useful rule of thumb, nothing more, and the idea that 34mA will sound very different to 35mA is another area which I would question.

Ok, I'll be careful here because that attitude is not shared by many, and it is religiously hated by a few. Don't want to poke anyone in the eye - but maybe it is worth considering. Over and out.

[pdf64]

I agree, as for a given mains voltage and output stage current draw, anode voltage isn’t a variable.
Plus probing output valve anodes can easily invoke oscillation, which will corrupt the reading and if ultrasonic, may not be readily apparent.
Superleads especially having a low to non existent margin of stability.
Better in my view to go with the HT voltage, ie at the OT CT. No risk of oscillation and the reading will be within a gnat’s todger of the actual anode voltage.
But if the voltage at nominal mains and typical current draw is already known, yeah, it seems kinda pointless anyway.