How to properly bias an express......

[marcoloco961]

Very interesting. I had read about the screen current contributing to the overall mA at the cathode, but was under the impression that the screen current was minimal enough not to make a big difference. After reading this I got curious and fired up the express and did some measuring. I ended up with a 4.9v drop across screen resistor on v4 and a 5.2 drop on v5. 1005 and 1008 ohm respectively on the resistors. So approx. 5mA per tube of screen current. Meaning the 45mA I was calling plate current was actually only 40mA.

Having tried to pay attention and making note of the difference between 40mA and 45mA of total cathode current when setting it originally, I was thinking that it was running pretty far on the cool side of the tube capabilities. So I cranked it up 5mA per tube.

WOW. Thanks Paulster. There is the bite I had been looking for. I had just been afraid to go over the 45mA specs I had read somewhere for the recommended plate current. I didn't want to fry new tubes. No red plating after about an hour of playing, and the response was much more dynamic than before.

You ought to write a book Paulster. I would buy it. Great tutorials. I learn something new each time I read your posts. How about a cathode follower tutorial for us novices........LOL Or the PI section, that still screws me up every time.

[ampgeek]

Hey Rooster,
Screen grids always draw current from the PS. Just the way they work! How much they draw has a lot to do with plate voltage, input signal level, etc..etc..

You will always see ratings for g2 (screen grid) power dissipation in the tube spec sheets. Power dissipation is a function of voltage drop across and current through any circuit. The screen resistors (which are put there to limit current draw and, ulimately, power dissipation) are just a convenient place to measure voltage drop which, if you know the resistance, can be used to calculate current draw.

Tubes deemed "robust" enough by the designers/engineers/users wont include resistors to prolong tube life. But...the screens still draw current.

The amount of current through any resistor in a circuit is a function of the upstream and downstream voltage or...voltage differential which is the true V in ohms law. It isn't the static voltage at either end.

When there is a differential voltage across a resistor...you can be reasonably well assured that there is current flowing through it.

Cheers,
Dave O.

[rooster]

ampgeek - Hm, not questioning if the screens draw current, that's a given. My question was the intended operation/use of a bias probe. Well, you can use it to sit on yur desk, I suppose.

I can see where paulster is directing things and it makes sense. However, I haven't seen this idea presented in the use of one (a bias probe) and they have been around for a relatively long time, and made by many companies over the years. Its a simple formula, subtracting the screen resistor current, I just haven't seen this method/idea presented before to arrive at the given cathode current number when using a bias probe device.

ampgeek, do you have a reference document that presents this? I would like to see the text is all. Hell, some people just adjust the plates to cherry and then back the bias off a bit. Sure, done that. Its all good and nothing is wrong with any method if the amp operates. They're only Mullards, you know?

[paulster]

My question was the intended operation/use of a bias probe. Well, you can use it to sit on yur desk, I suppose.

A bias probe is essentially all pins on the plug wired to all pins on the socket except that the cathode is connected via a 1ohm resistor and the two wires coming out of the probe are the two ends of the resistor.

This means that you can measure cathode current without being exposed to any dangerous voltages and also that you're not likely to over-stress your tubes since there'll be a margin of safety built in by not considering the screen current. It means you'll always bias slightly on the cold side.

Far better in terms of accuracy would be to have the 1ohm resistor in the plate connection instead but, although, you'd still have a millivolt drop across the resistor, the plate reference voltage would be several hundred volts above the chassis and ground, so it would be unsafe to expose outside of the bias probe. Even with silicone leads and insulated 4mm plugs you wouldn't be able to take the liability issue on, especially when the target market is for people who aren't comfortable with working with high voltages inside their amps.

I still think they can be useful if you don't want to mod an amp to have the 1ohm resistors fitted internally, and the limitation can be worked round very easily if you've already measured the screen resistor values and voltage drops.

[rooster]

Paulster - So....... they exist, bias probes. Got all that.

My question was: Where did this idea you presented - the part about subtracting the current across the screen resisitor - from the current seen at the cathode, to arrive at the *correct* bias current?

Can I read this specific info somewhere? Or is it something you have extrapolated from your experience?

[Firestorm]

KOC covers it somewhere in TUT 2 (and concludes that you may as well just bias to the cathode current since that will make the bias setting a little cooler and prolong tube life). Also suggests that once you know what percentage of cathode current is screen current for a given tube in a given amp (say 10%), you can just assume it.

[Structo]

The plate current can be indirectly measured by determining the cathode current in the tube. Since no appreciable current flows in the control grid of the tube, all of the plate current must also flow in the cathode. The cathode current, however, will also have all of the screen grid current flowing in it as well, since the cathode current is the sum of the plate and screen currents. Typically, the screen current in most commonly used pentodes is around 5mA (this varies, of course, depending upon the class of operation, the bias point, and the type of tube used). The screen current can be accurately measured by determining the voltage drop across the screen resistor and dividing by it's value (for example, 5V across a 1K screen resistor indicates 5mA of screen current). If you install a 1 ohm resistor in the cathode lead of each output tube, you can measure the voltage drop across it to get the cathode current, as described above in the plate resistor measurement. The advantage of this method is that there are no high voltages involved, since there will only be a few millivolts difference between ground and the other side of the 1 ohm resistor. The disadvantage is that you must subtract the screen current in order to accurately determine the plate current. However, since the screen current is only a few mA, it can usually be ignored, and the error will be in the conservative direction, i.e., less plate current than expected, which is good for tube life. This method of biasing is the most highly recommended. There are a few bias meters on the market that use this method, including the SwAMProbe (it doesn't measure across a resistor, rather, it breaks the cathode circuit so you can insert a milliammeter for direct cathode current measurement), and the Bias King. These devices have socket adapters that go in between the tube and the socket to make it easy to measure the bias current on an amplifier without having to modify it.

http://www.aikenamps.com/Biasing.html

[ampgeek]

Hey Rooster,
Forgive me for not seeing your bias probe perspective here. I will admit that I have never used one in person but I do understand the principle well.

I don't have any specific articles/references to pass along on this topic. My knowledge/understanding of this topic is more an accumulation of vacuum tube theory articles/snipets and lots of hanging around forums and discussing it all with good folks like yourself.

In the end, it comes down to the physics. The screen grid is used to soak up and sequester the secondary emission of electrons that occur when the stream of electrons that boil off of the heated cathode impact the plate. Think slo-mo video of a drop of milk hitting the surface of a standing puddle. A small drop produces a significant splash of secondary material! If they were not sequestered, they would "fall" back onto the plate and reduce the fidelity of reproduction of the original signal present on the tube's control grid.

To remain charge balanced, every electron that is "captured" by the grid must be offset by a unit of charge in the opposite direction. In our case, from the power supply. Thus, current flows into the grid from the power supply. Note that in our vernacular, current flows in the opposite direction of electron flow.

Ultimately, the current flows toward the supply of electrons in the first place. That being the cathode and, ultimately the ground.

Thus, we can measure the current exiting the cathode on it's way to ground. This is identical to what happens with the primary stream of electrons boiling off of the cathode whose flow rate is controlled by the signal on the grid. It, too, exits at the cathode.

Hope that this helps!
Cheers,
Dave O.

[paulster]

My question was: Where did this idea you presented - the part about subtracting the current across the screen resisitor - from the current seen at the cathode, to arrive at the *correct* bias current?

See Tom's post from Randall Aiken's site as that explains the mechanics quite well.

In terms of biasing we usually say that you shouldn't go above 70% plate dissipation at idle for a class AB push-pull amp, so we've immediately referenced the plate dissipation rather than the total power dissipated in the tube.

The practical ways to measure the plate current are the transformer shunt method (which reads plate current directly) or the cathode resistor method, which is generally considered safer and more commonly used.

The cathode resistor method includes the current being drawn by the grid (effectively zero), the suppressor grid (also effectively zero), the plate and the screen grid as it all has to flow via the cathode connection to ground.

Since only the plate and screen are of any significance and the screen current can be directly measured across its resistor you should technically subtract this from the cathode figure to get a true reading of plate current, which is what you are using as your bias target.

[Structo]

I think once you are familiar with your amp and how much screen current is drawn, you can safely ignore it when setting your bias.
It provides a safety margin and if you do include it in your measurement you will have a better idea what the tube is doing real world.

But if it is under 5ma then it doesn't hurt a thing to ignore it.
Most amps aren't biased up to the 70% rule anyway for the best tone.

The Dumble 6L6 amps for example seem to behave and sound best on the cold side at around 55-60%.

[redshark]

Correct me if I'm wrong Structo, maybe those bias settings in a D-style amp is because they try to avoid power tube distortion, is all preamp...right?
Besides that going back to the Trainwreck world whoever draw the Kelly schematic measured -33vdc which makes me think also that the amp was biased on the cold side.
I wonder that since the express amp is on the edge of stability maybe biasing a little bit colder could help a complicated build to be stabilized.

[paulster]

Besides that going back to the Trainwreck world whoever draw the Kelly schematic measured -33vdc which makes me think also that the amp was biased on the cold side.

Assuming Kelly was using Siemens EL34s then with a B+ of 395V and a bias of -33V that would be pretty hot. Right about 70% I'd suggest, based on mine at 400V on the plate and a bias of -32.9V.

[redshark]

well, mine uses siemmens el34-s, my negative voltage now is -32vdc and my bias probe gives me 42.5
If we substract those 5-6mA makes me be around 36-37 with a 409vdc on plates. That is cold so kelly should be colder.
Paulster maybe in the UK u get diferent voltages.

[paulster]

My plates are at 400V and cathode currents are 49.5mA at -42.9V with Siemens EL34s. The screen currents are about 5.5mA each, so my plate current is 44mA, which equals 17.6W, or just over 70%.

Those figures are all slightly rounded so I'm actually right on the nose of 70% with my tubes.

It sounds like you've got a pair that need bias a bit colder than average because I'd say the pair I've got in mine at the moment are pretty typical of the batches of Siemens EL34s I've got.

I'd recommend getting your bias hotter and see how it sounds. It will also drop your plate voltage a little bit, so keep an eye on that too.

Being in the UK won't make any difference because we're talking about the rectified DC at this point. Your plate voltage is close enough to mine to be down to fluctuations in wall voltage rather than a fundamentally different set of secondary voltages.

[rooster]

OK, well thanks for all the input. Tom, yes, this is what I wanted to see. I also appreciate your take on things regarding the 70% number. ampgeek and paulster, comments duly noted as well. This has been interesting, yes.

I think if I have one comment to make about biasing an amp it would be that my main interest is in setting the current value in the area that the particular circuit was designed for. In other words, I am not interested in making my BF Fender sound like a Marshall by underbiasing it, nor am I interested in making my Express sound like a BF Fender by overbiasing it. This type of behavior is a total waste of time to me because I don't want my BF Fender to sound like a Marshall - I have a Marshall for that, etc..

Now, getting back to redshark and his interest on this subject. He and I are on the same page I think. We both want to know that our Express's are sounding/operating in a manner that is typical for what Ken intended. And this is not an easy thing to do. .....Unless you are Glen, of course, or somebody else who has a real deal to compare things to. And then you still need the right speakers, and the right guitar perhaps, etc., etc. (Oh, and Glen needs those original tubes!!!)

I think we may all have to agree that this is always going to be a grey area, this knowing exactly how our particular clones would sound if Ken biased our amps with the tubes that he chose. Regarding the *correct* biasing method, Ken has left the building in a trail of crumbs. redshark, you get the 'Best Detective' award today by pointing out the -32 number. I hadn't heard that before.