Fixed Bias adjustment
 

Hello DIY Audio! its been a while...

I took my El34 SE amp, and converted it to a fixed bias, in all honesty, I didn't look back once I changed it. The Thing just flat out sounds better.

My EL34s are now getting to be 4 years old (its hard to believe its been that long). In that time, the abuse they have taken has been more than admirable, so I am deciding to retire them. I was thinking of getting a Pair of EH EL34s, and I was wondering if it was okay to adjust their bias voltage while the amplifier was on. whats the worst that could happen?

The bias circuitry is a UF4007 run into a rather large cap, and a voltage divider with a pot, if that makes a difference...

-Moose

yes you can, as a precaution, make sure the pot is turned way down, ie. as negative as you can make it.

Do you have a small resistor in the cathode cct? if you do, monitor the voltage across it to set the current as required.

Re: Fixed Bias adjustment
 

I don't see any other way you can do it. As billr said, start with your bias voltage at it's maximum negative value, power up, then adjust it until you get the current where you want it. If you don't have a resistor in the cathode circuit, you can measure the resistance of the transformer from plate to supply and monitor the voltage across it as you adjust. If you use this method, you must make sure your meter and test leads are rated for your B+ level. And remember you are working with lethal voltages.

Sheldon

Yep, there is a Resistor in there, I didn't think to start with the Voltage Really negative though, good idea

Thank you

Upon further reflection, let me qualify that a bit. I don't know your adjustment range, but I'm assuming that it can't go low enough to completely shut off the plate current. If it does, you want to make sure that your power supply components can handle the full unloaded voltage. I build my amps so that everything can handle the unloaded voltage, but can't assume that for other amps.

Anyway, you probably don't need to crank bias all the way down. I'd say that you are probably safe if you back it off so that you get about 50-70% of the target bias current with the tubes you have in place. That should give you a safe starting point for a new set. Nice thing about tubes is that they can tolerate brief periods moderately in excess of dissipation ratings, without damage.

Sheldon

On bias supplies, a very good hint to avoid disaster when perhaps the pot gets old and the slider 'disconnects'. Have a resistor (could be fairly high value) connected from the pot slider to the negative side of the bias supply, That way, should the pot slider connection to the track become suspect, one does not run the risk of sitting with an 'open' grid perhaps going positive and causing over-current.

I am not the originator of this; as far as I know KevinKR proposed it some time ago. (If I am crediting the wrong person, kindly advise!)

Johan is right. Pots can fail and you need a fail-safe backup that will send the grid negative. I've seen really dumb examples of a 'fail-safe' path to ground instead of the negative supply. :dead:

I set up the bias adjustment with the supply to a divider, consisting of a grid resistor in series with a pot to common, and the supply connected between them. If the pot wiper fails open, the full resistive value of pot makes the lower half of the divider, so the grid sees the maximum negative voltage.

Sheldon

The bottom schematic here is not exactly the same, but shows the same principle: http://www.people.cornell.edu/pages/...switching.html

Hey guys

I always use components that can handle the unloaded voltage...always...

mine is slightly different than the link, but yeah, I made sure to put an idiot resistor in the path to make sure nothing too bad happens

My comments (and I'm guessing Johan and Ray too) were intended also as broader observations on the issue. When I post, I want to keep in mind that there are many passive thread observers (as I often am). So I try and generalize my comments to be helpful to the wider audience, especially if we not addressing a defined schematic.

Sheldon