Hoffman Amplifiers Tube Amplifier Forum
Amp Stuff => Tube Amp Building - Tweaks - Repairs => Topic started by: silverfox on March 26, 2013, 01:35:34 am
-
I'm going to put together a bias circuit for EL34 or 5881 (6L6). I'm confused about the correlation between voltage and current. I see different voltages listed on bias adjustments- -37 volts or even 57 volts. When I have adjusted the bias those few times so far, I attempt to set the milliamp per tube values- 26ma for 5881 or 42 ma for EL34..
Is the voltage relevant to the tube function or is it strictly the tube current?
The only other option I can think of is the builder of the amp designed the bias circuit to properly bias the amp at a particular voltage setting and states that as the bias setting instead of a current value.
Silverfox
-
The way I set bias is the shunt method , I first check the plate voltage to know where I want to adujust the grid voltage( the negative 37 or negative 57 volts you speak of) to get the proper plate current (the 28mA or 42ma you speak of).
both the grid voltage and the plate current on all amps are set to the supply volltage , what is on the data sheet is not always what you find in a actual amp.
-
I think the Bias voltage, in relationship with a plate voltage
is given as to supply the builder with a start reference
what is really important (as told) is the current flowing through the tube
K
-
Is the voltage relevant to the tube function or is it strictly the tube current?
The control grid voltage is the thing you manipulate to change the plate current. It is the 'valve' part of the tube, that you variously 'open' or 'shut' in order to change the plate current.
In a triode, the plate current is the same thing as the tube current.
In a tetrode or pentode, the tube current is constant (for most of the range of available grid voltages) because of the screen. When the plate current decreases, the screen current increases, and vice versa, so that the overall tube current remains constant.
In the sense that the grid voltage is negative w.r.t. the cathode voltage in order to bias the tube, then the 'relative negativity' of the grid voltage affects the plate current in so far as the negativity of the grid voltage repels electrons away from their trajectory towards the plate.
Whats important for biasing is the measure of plate dissipation. Plate dissipation (Wa) is a factor of the plate voltage (Va) and the plate current (Ia). (Wa = Va x Ia). The plate voltage at idle is set by the HT (B+) voltage and the plate load. The plate current is set by the grid voltage. The absolute grid voltage is not as important as the plate dissipation. You aim for a grid voltage that will result in the desired plate dissipation. (In a pentode or tetrode, you subtract the screen current (Ig2) from the tube current which is typically either measured at the cathode (Ik) in order to get the plate current (Ia), or you measure the voltage drop across the known (measured) resistance of the plate load in order to get the plate current.
-
Silverfox
The bias grid volgage ( - xx volts ) is the way to have the proper cathode current . Negative grid bias voltage is not useful to know ; Cathode current ; YES
-
I'm an electronic retard, but the way i do it is the 1 ohm cathode method. As i understand it when you measure voltage like that it correlates to current. For example, if you read 36 Mv across the resistor it means the tube is pulling a current of about 36 Ma. Tho actually i believe you need to subtract about 5 Mv from that due to the screens pulling about 5 Ma. So 36 Mv across the resistor would mean the tube is drawing about 31 Ma. I may be wrong in some way because as i said, electronically i'm barely able to tie my shoes. But thats the way i understand it and do it and it seems to work just right. Of course you need to get very close tolerance resistors.
-
The only other option I can think of is the builder of the amp designed the bias circuit to properly bias the amp at a particular voltage setting and states that as the bias setting instead of a current value.
Sometimes people will cite a given bias voltage setting for a particular amp but that is not a best practice. It's misleading. It implies that if you set the bias voltage at the grid for -XX volts then the amp will be properly biased and that may or may not be true. A new set of tubes may draw more current or less current at that bias voltage setting than the original set of tubes did. That bias voltage setting they cite may be only totally accurate in their shop (with their 120VAC supply voltage) and with those particular tubes in new condition. Sometimes it's an attempt to simplify the tube biasing process but I don't think it's a good idea. It will probably get you in the ballpark. Hopefully. That's why we always measure the plate voltage of the tube and the current through the tube and calculate dissipation from that. Also bear in mind that somstimes a manufacturer will intentionally set the bias on the cold or conservative side to minimize the stress on the amp so it won't come back during the warranty period. Their setting may be perfectly safe but may not have the tubes running as optimally as you would like. You may like the amp biased hotter or colder.
As long as the bias circuit will provide a negative voltage range of adjustment that allows us to set the tube for the idle dissipation setting that we desire then the actual bias voltage setting itself in volts is not really relevant. It may end up being -30 VDC. It may end up being -50VDC. We don't use that voltage value in any of our calculations so the final value isn't important.
-
I think I'm getting a better idea now, but this leads to other questions.
For illustrative purposes please use the following image:
http://www.weirpowerindustrial.com/images/quinshan%20cut_v_Variation_1.jpg (http://www.weirpowerindustrial.com/images/quinshan%20cut_v_Variation_1.jpg)
Are we setting the range of travel on the valve stem?
If I were to mount two springs on the valve stem, one stronger than the other resulting in a preset minimum-maximun, does this represent what I'm doing in setting the bias?
Another way to phrase it would be: to connect the valve stem to some particular point on a limiting spring. Any position other then half way between would result in an unbalanced range of travel.
If so, where is the weaker spring? Bottom of range or top?
Am I choosing a particular value of spring?
Is this a bad analogy and doesn't work?
I'm pondering all this as I write...
For Tetrode-Pentode operation it sounds like there is a feed back network connected to the output of the valve for bias purposes.
Studiously,
Silverfox.
-
It's like the screw on a carburetor. The instructions say "2 turns", you start and warm the engine, then turn back and forth for best idle. May be 1.4 turns, may be 2.6 turns.... the '2 turns' is just a starting point.
Or:
You want to wash your dog with the garden hose. There's a handle on the spigot (valve) which adjusts water flow.
Voltage is like "half turn". It's a setting on something that controls the current, not the actual current.
But what you really want is a ceratin water-current at the end of the hose. Enough water to dig into the fur. Not so much current that the dog is blown away.
> -37 volts or even 57 volts
Even using the same spigot, on different water pressure, I would need different "turns" to get the right water current for my dog. If I had 60psi, I might only need 0.37 of a turn. With my 30psi I need more like 0.57 turns to get the same current.
When biasing a tube, you want to know the *current*.
You need an *estimate* of the voltage to design your bias supply. If a tube needs -12V, it might be silly to build a 100V bias supply. OTOH if the tube needs -85V (some do!), then a 50V bias supply is not going to do the job.
Some tricks:
You want to be able to "cut-off" the tube, at least for initial smoke-test.
You want to be able to set the tube "half on" for a good rich bias.
The "cut-off" is nominally the Screen voltage divided by Mu(g2). For most common audio power tubes, Mu(g2) is 10. EL84 is 18.
"Half on" is about half that. (0.6 may be a better approximation.)
So a mild 6V6 at 250V (which is hardly worth fix-bias):
250V/10 = 25 V, the most you could need to stay cold.
25 * 0.5 is 12.5V; 25V * 0.6 is 15V; this will be a good minimum voltage (hot tube) bias.
-
You need an *estimate* of the voltage to design your bias supply.
You also need an estimate of bias voltage to design the phase inverter, and know that it can easily drive the output tubes to full output power.
That full (cleanish) power point will be something near or just below a peak signal voltage that equal the bias voltage. For an output stage with -37v of bias, you want the phase inverter to cleanly make at least 37v peak output signal (and preferably a bit more), while taking a signal from the preamp that is well within the preamp's capability to deliver.
Radio men (like Leo Fender) probably looked at the designed output stage and said, "-45v bias is where we designed it (and the phase inverter/preamp), so always set the output tubes to -45v for this model and verify the tubes aren't smoking."
That said, if you never measured output tube current, you might not recognize easily some failure modes.
-
I may understand this better now see if this is correct: Adjusting the bias means adjusting a voltage on a screen inside the tube until a certain ma value flows from plate to cathode.
I thought adjusting bias was adjusting a current that flowed from the bias circuit into the tube and out of the tube. If as above is the case- perhaps you can see how I had all those misunderstandings.
Once the bias is set, the signal superimposes upon the bias and drives the tube to various levels of conduction. This conduction causes a flow of current in the output transformer.
???
Fox.
-
Adjusting the bias means adjusting a voltage on a screen inside the tube until a certain ma value flows from plate to cathode.
Your very close.
Not on a screen, on the control grid (G1). The screen (G2) is an element in a tetrode, pentode and other tubes we don't use for audio. But that's not important right now.
In a triode you have 3 elements, cathode (K), control grid (G1) and plate (A). K is (usually) referenced to ground through a R and a high dcv is placed on the plate. When the heater heats up the K a cloud of electrons are liberated from the K. The high dcv on the plate attracts them so current flows up from ground through the R then K then to plate. Adding a control grid we can now control this flow. Make the control grid more positive with respect to the cathode, more current flows, make it more negative with respect to the K, less current flows.
I thought adjusting bias was adjusting a current that flowed from the bias circuit into the tube and out of the tube.
Once the bias is set, the signal superimposes upon the bias and drives the tube to various levels of conduction. This conduction causes a flow of current in the output transformer.
No and yes.
No current flows from the bias circuit. In a fixed bias output stage the K is grounded with no R between ground and the cathode. To control the current flow you have to make the control grid more negative than the K which is at 0 dcv. So a negative dcv is applied to the control grid.
Yes about the signal superimposing upon the bias and drives the tube to various levels of conduction.
Brad :icon_biggrin:
-
I thought adjusting bias was adjusting a current ...
Current is like the flow of water or air. You can't directly cause water or air or electrons to flow in a specific direction without applying an external force.
You might think of that force as water pressure in your home's pipes. For electrons, we normally talk of "electromotive force," also abbreviated "EMF" and one of the several names for voltage. Volts are the water pressure that results in current movement.
The way the Navy made me learn Ohm's Law was Current = Voltage/Resistance, because you can directly change resistance (opposition to current) or voltage (force causing current) but current can only be changed by changing one of the other two.
I wrote it correctly this time Sluckey. :)
-
Still fighting a bias problem but perhaps closer now.
Question: as the bias voltage on the control grid goes up, the mv reading on the meter connected to the cathode resistors goes down?
Cutoff is the bias voltage level that prevents idle current from flowing through the tube? So the meter will read 0 mv then.
Silverfox.
-
Question: as the bias voltage on the control grid goes up, the mv reading on the meter connected to the cathode resistors goes down?
Cutoff is the bias voltage level that prevents idle current from flowing through the tube? So the meter will read 0 mv then.
Yes, but be more specific.
As the bias voltage gets more negative, tube current decreases. With a large-enough negative voltage, tube current will be zero.
I know you meant a bigger negative number as seen on your meter, but if someone that is very math-oriented read your sentence they may think "bias voltage goes up" = "bias becomes more-positive".
-
Yes, thanks. One of my goals is to improve my writing abilities.
I also figured this out since at the time I was trying to debug my bias circuit. I did and the amp sounds fantastic. I'm posting in the Power Amp Topic. There's a good chance I understand how the power amp works now after re-reading this topic.
Silverfox.