Tubes... and Biasing

[Schu]

[Fjd]

I dont know of any fairly modern consumer amp that requires one to bias from under the chassis....................if they exist I would certainly stay away from that type design

Well, there is my problem and why I don't know how to use the socket bias item as none of my amplifiers would be considered modern consumer amps.

I had an old Fisher that you needed to measure the voltage on pin 3 of the output tubes.

My other difficult amplifier is a pair of Electra-print original 300b/6A3 DRD mono blocks. Those amplifiers have a big 5K ohm 50W power resistor that is adjustable by using two wipers. The biasing is done by moving the wiper on the variable power resistor while measuring the voltage drop change across a 10 ohm resistor. Another twist to this is that everything is inside the chassis where it is diffiult to reach unless the amps are turned over.

Here are a couple of pictures of the big Electra-prints.

Edited June 20, 2014 by Fjd

[minermark]

My fellow bud's and I had a little beer party at my place last weekend, this subject was discussed, and the general consensus was, ah, a little voltage now and again aint going to hurt ya

[NOSValves]

Screwdrivers and pins and multi-meters is the old way of doing it. You use a voltmeter, and put the test probes into test sockets or pins and then turn a pot on the chassis until the voltmeter reads the right value. However, today with logic chips costing pennies, you can employ a digital bias meter in the amp which requires no tools, no guesswork, and no knowledge by the user. Typically it has three lights: Red LED is too high, yellow LED is too low, Green LED is Exact bias. One per channel and the user is never out of bias. The cost for this kind of setup is peanuts in a new amplifier.

Bias in General

"Bias" means putting the tube into it's steady-state DC operating condition, so that an incoming AC signal will be amplified. It means arranging the DC voltages and currents on the tube's primary terminals. For a hypothetical power tube conditions for bias might look something like this: Plate voltage 450 VDC. Screen voltage 425VDC, Cathode voltage 0VDC, Grid voltage -45 VDC resulting in a plate current of 50mA. What to notice? The grid must be negative when compared to the cathode, and the relationship of grid to cathode determines how much plate current will flow in the tube. The more negative the grid become, the less plate current will flow. You can make the grid so negative that no current will flow. If you make the grid less negative compared to the cathode, the plate current will rise. A large amount of plate current is controlled by a small amount of grid voltage - - - that's the essential operation of any vacuum tube.

This plate current is called the "bias current" and is in milliamps. The grid voltage is called the "bias voltage," and is in volts. The bias and "operating point" determines how much signal can be handled without distortion, and how much power will be generated by the tube. And what class of operation the tube will operate in. The tube has a maximum allowable plate current that can not be exceeded. So, when biasing a tube, there are ranges and limits for each kind of tube.

How to create bias? There are two general ways to make the grid negative compared to the cathode. First is called "cathode bias" and the second is called "fixed bias." With cathode bias, you insert a resistor from the cathode to ground and when the plate current flows through the tube, a voltage is created by the current through this resistor. The voltage will be positive compared to ground. Since the grid is referenced to ground also, it is now more negative than the cathode. To make the bias more negative, just make the cathode resistor bigger. Very simple method of bias. The second method is called fixed bias because it depends on creating a fixed negative voltage source somewhere and applying it directly to the grid. I could tap the main transformer secondary, use a rectifier, and establish a supply of say -50VDC. Now, if you connect the cathode to ground and connect this fixed negative voltage to the grid, you have bias because the grid is negative compared to the cathode.

Play on words. The advantage of a "fixed bias" is that it is almost always adjustable! There is almost always a pot connected to the fixed negative voltage source, so that it can be swung from say -40VDC to -60VDC. By doing this, a variety of bias points can be achieved for the tube connected. Whereas the cathode bias resistors are actually "fixed" values and not adjustable. Because of these words, people often confuse "fixed" with "cathode" bias.

A disadvantage of cathode bias is that the voltage on the cathode is coming at the expense of plate voltage. Example: suppose your power transformer can only create a 450VDC supply. If you need a bias point of -50VDC and you elect cathode bias, you will have a cathode voltage now of +50VDC and a plate voltage of 450VDC for a NET (between plate and cathode) of only 400V. If, on the other hand, you apply fixed bias, your real NET plate voltage will be 450VDC as measured from plate to cathode. That 50VDC extra on the plate will mean more power output.

Cathode bias is the simplest for the user because there is nothing to adjust. Simply plug the tubes in. This is often called "auto bias" because the tube automatically achieves bias because of the resistor in the cathode circuit.

In a very general sense (not a strict rule) lower cost receivers and integrateds or consoles often had cathode bias (autobias) and higher end separate amps used "fixed" (meaning adjustable) bias!

yup they cost pennies and then fail requiring expensive service. They are also never perfectly accurate.. IMHO the old way is the best most reliable and accurate way to bias an amplifier. My 7 year old grandson learned how in about 5 minutes. Just how lazy do we really need to be?

50 to 60 year old amplifiers from the hay day of tube Hi Fi command big bucks in the range of 5X, 10X, 20X or more than original cost because they are way more reliable and repairable once the aged components need replacement then any new fangled product made today. How many of these circuit board based units with all the modern IC chips do you really think will still be in service 20 to 30 years from now? Not many when those chips become obsolete and can no longer be sourced or even worse the circuit board that you are forced to use to employ these tinny little delicate failure prone parts burns up from a tube failure from the failure of this miraculous chip allowing the output tubes to run away when it fails? Sorry to say IMHO you just cannot reinvent the perfectly round wheel. But you can over engineer it!

[minermark]

Screwdrivers and pins and multi-meters is the old way of doing it. You use a voltmeter, and put the test probes into test sockets or pins and then turn a pot on the chassis until the voltmeter reads the right value. However, today with logic chips costing pennies, you can employ a digital bias meter in the amp which requires no tools, no guesswork, and no knowledge by the user. Typically it has three lights: Red LED is too high, yellow LED is too low, Green LED is Exact bias. One per channel and the user is never out of bias. The cost for this kind of setup is peanuts in a new amplifier.

Bias in General

"Bias" means putting the tube into it's steady-state DC operating condition, so that an incoming AC signal will be amplified. It means arranging the DC voltages and currents on the tube's primary terminals. For a hypothetical power tube conditions for bias might look something like this: Plate voltage 450 VDC. Screen voltage 425VDC, Cathode voltage 0VDC, Grid voltage -45 VDC resulting in a plate current of 50mA. What to notice? The grid must be negative when compared to the cathode, and the relationship of grid to cathode determines how much plate current will flow in the tube. The more negative the grid become, the less plate current will flow. You can make the grid so negative that no current will flow. If you make the grid less negative compared to the cathode, the plate current will rise. A large amount of plate current is controlled by a small amount of grid voltage - - - that's the essential operation of any vacuum tube.

This plate current is called the "bias current" and is in milliamps. The grid voltage is called the "bias voltage," and is in volts. The bias and "operating point" determines how much signal can be handled without distortion, and how much power will be generated by the tube. And what class of operation the tube will operate in. The tube has a maximum allowable plate current that can not be exceeded. So, when biasing a tube, there are ranges and limits for each kind of tube.

How to create bias? There are two general ways to make the grid negative compared to the cathode. First is called "cathode bias" and the second is called "fixed bias." With cathode bias, you insert a resistor from the cathode to ground and when the plate current flows through the tube, a voltage is created by the current through this resistor. The voltage will be positive compared to ground. Since the grid is referenced to ground also, it is now more negative than the cathode. To make the bias more negative, just make the cathode resistor bigger. Very simple method of bias. The second method is called fixed bias because it depends on creating a fixed negative voltage source somewhere and applying it directly to the grid. I could tap the main transformer secondary, use a rectifier, and establish a supply of say -50VDC. Now, if you connect the cathode to ground and connect this fixed negative voltage to the grid, you have bias because the grid is negative compared to the cathode.

Play on words. The advantage of a "fixed bias" is that it is almost always adjustable! There is almost always a pot connected to the fixed negative voltage source, so that it can be swung from say -40VDC to -60VDC. By doing this, a variety of bias points can be achieved for the tube connected. Whereas the cathode bias resistors are actually "fixed" values and not adjustable. Because of these words, people often confuse "fixed" with "cathode" bias.

A disadvantage of cathode bias is that the voltage on the cathode is coming at the expense of plate voltage. Example: suppose your power transformer can only create a 450VDC supply. If you need a bias point of -50VDC and you elect cathode bias, you will have a cathode voltage now of +50VDC and a plate voltage of 450VDC for a NET (between plate and cathode) of only 400V. If, on the other hand, you apply fixed bias, your real NET plate voltage will be 450VDC as measured from plate to cathode. That 50VDC extra on the plate will mean more power output.

Cathode bias is the simplest for the user because there is nothing to adjust. Simply plug the tubes in. This is often called "auto bias" because the tube automatically achieves bias because of the resistor in the cathode circuit.

In a very general sense (not a strict rule) lower cost receivers and integrateds or consoles often had cathode bias (autobias) and higher end separate amps used "fixed" (meaning adjustable) bias!

yup they cost pennies and then fail requiring expensive service. They are also never perfectly accurate.. IMHO the old way is the best most reliable and accurate way to bias an amplifier. My 7 year old grandson learned how in about 5 minutes. Just how lazy do we really need to be?

50 to 60 year old amplifiers from the hay day of tube Hi Fi command big bucks in the range of 5X, 10X, 20X or more than original cost because they are way more reliable and repairable once the aged components need replacement then any new fangled product made today. How many of these circuit board based units with all the modern IC chips do you really think will still be in service 20 to 30 years from now? Not many when those chips become obsolete and can no longer be sourced or even worse the circuit board that you are forced to use to employ these tinny little delicate failure prone parts burns up from a tube failure from the failure of this miraculous chip allowing the output tubes to run away when it fails? Sorry to say IMHO you just cannot reinvent the perfectly round wheel. But you can over engineer it!

[NOSValves]

Another quick question.... when you see rust on a transformer is than an indication of something bad... like excess heat?

Nope not usually too awful bad.. it means careless owner allowed dust to build up which during the hot/cold cycle attracted condensation (moisture) which in turn after prolonged neglect ended up rusty the metal surfaces. Or stored the amp for years in a very moist environment. In reality it rarely ends up being of any consequence unless extremely bad. I've rebuilt thousands of vintage amps with many of them having surface rust on transformer lamination's and end bells over the last 13 years and rarely do I see transformer failure.

Edited June 21, 2014 by NOSValves

[NOSValves]

Oh...and if you have to do manual bias... Don't take any alcohol prior you do the bias setting...alcohol+small screw driver+some small holes+high current = are not good combination...

And how exactly, did you discover this fact?

There is no high voltage or high current involved. Please if you do not know what your talking about leave the responses to those that do. Creating unnecessary fear is going to do anyone any good. In reality you are measuring less then a volt with zero current delivery abilities. The only thing in danger when setting the bias is the actual tube you are biasing so too many drinks would not be a good idea! Also you would not want to be careless and touch one of the output tubes since they get pretty darn hot when warmed up! But honestly many here are making a mountain out of a mole hill.

Edited June 21, 2014 by NOSValves

[Schu]

Thanks craig, just trying to get my head around everything.

[Dr Morbius]

Craig is da man!!!

[NOSValves]

Mark you never cease to amaze me. The master of techno babble. Too bad it's all complete BS based on half truths driven by alterior motives...

[erik2A3]

...please keep this thread going in a positive, helpful, and informative direction.

[tube fanatic]

A question which I've not been able to have answered is whether modern push-pull tube amps with autobiasing also have that feature applied to the phase inverter. If it isn't, and the inverter goes out of balance (as it absolutely will over time, even if it is well balanced to start with), then the autobias feature is of dubious benefit. In fact, in vintage amps which don't allow for phase inverter balancing, it is easy to see (and sometimes hear) why excessive worry over output stage balance is unwarranted. A nice feature of the Scott 299, as an example, is the ability to balance the phase inverter as well as the output stage. If one is looking to minimize distortion, that's the only way to do it. Williamson addressed this issue with his use of the cathodyne phase inverter which, in spite of small differences in the grid to plate and grid to cathode capacitance which caused slight imbalance, worked beautifully and is fully self balancing. When I was designing and building push-pull amps I preferred it myself over the more conventional types. What I'm getting at is that is just isn't worth obsessing about output stage balance unless what feeds it can be balanced in addition. It's one reason why I embrace, and now spend all of my time designing, single ended amps which can be made to be truly "plug and play" and require no adjustments or maintenance. If high power (by which I mean upper teens) is needed, a pair of parallel 7591s, for example, will easily fill the need, so why bother with push-pull at all? Anyway, that's my view of the situation!

Maynard

[NOSValves]

...please keep this thread going in a positive, helpful, and informative direction.

Please mind your own business

[NOSValves]

A question which I've not been able to have answered is whether modern push-pull tube amps with autobiasing also have that feature applied to the phase inverter. If it isn't, and the inverter goes out of balance (as it absolutely will over time, even if it is well balanced to start with), then the autobias feature is of dubious benefit. In fact, in vintage amps which don't allow for phase inverter balancing, it is easy to see (and sometimes hear) why excessive worry over output stage balance is unwarranted. A nice feature of the Scott 299, as an example, is the ability to balance the phase inverter as well as the output stage. If one is looking to minimize distortion, that's the only way to do it. Williamson addressed this issue with his use of the cathodyne phase inverter which, in spite of small differences in the grid to plate and grid to cathode capacitance which caused slight imbalance, worked beautifully and is fully self balancing. When I was designing and building push-pull amps I preferred it myself over the more conventional types. What I'm getting at is that is just isn't worth obsessing about output stage balance unless what feeds it can be balanced in addition. It's one reason why I embrace, and now spend all of my time designing, single ended amps which can be made to be truly "plug and play" and require no adjustments or maintenance. If high power (by which I mean upper teens) is needed, a pair of parallel 7591s, for example, will easily fill the need, so why bother with push-pull at all? Anyway, that's my view of the situation!

Maynard

Good post! No almost no modern amps have any AC signal balancing abilities. My amplifiers do but as I'm sure you know test equipment and the knowledge to use it is required. I think your overstating the tendencies for a well designed inverter needing adjustment. In most cases tube wear is what will cause it. Swap out the tube with a fresh one and any major imbalance is gone.

[NOSValves]

Funny I had this engineer working with me at one time to help design a front end... He had me install a temporary pot to adjust the plate resistors for a long tailed pair to balance the AC...this engineer then told me to measure the value of the end resulting resistance once balanced and install that value plate resistors and forget about it. I refused and installed an AC balance control.... wonder who that engineer was

Those darn ulterior motives just keep cropping up.

Edited June 21, 2014 by NOSValves

[mike stehr]

http://en.wikipedia.org/wiki/Alan_Blumlein

[erik2A3]

Speaking of biasing (check out the battery of pots) and some very nice Japanese OTL amp point-to-point wiring -- for those who appreciate skilled vacuum tube wiring and craftsmanship:

http://www.miyajima-lab.com/e-amp.html

Quite beautiful work, I think. Stereophile's Art Dudley really liked these amps.

[Schu]

I do like the way the wiring is routed... I think ALL WIRING should be routed as such. AEROSPACE style.

I know it takes much more time, but it really is the proper way to do things.

Edited June 21, 2014 by Schu

[tube fanatic]

A question which I've not been able to have answered is whether modern push-pull tube amps with autobiasing also have that feature applied to the phase inverter.

I don't know which commercial products might have it, but it's an easy circuit to implement at the phase inverter stage. We can thank Bleumein again for the principles of how it is done, and then we can thank technology for providing inexpensive and very powerful parts for accomplishing it.

No question that there are devices available which make automatic balancing easy to implement. Personally though, I try not to use anything solid state in my amps (other than the diodes in the p/s) due to their failure rate when exposed to powerline surges and spikes. Back in my TV repair days I attributed many, if not most, failures to such events and sometimes went as far as installing MOVs for suppression (I have a friend, an ex-Army ET, who collects TVs which people put out with their garbage and repairs them for sale at his church's rummage sales- he too feels that the SS components which he finds fried took a powerline hit). The tube type TVs weren't susceptible to anywhere near the degree of the solid state types. I have to wonder why the manufacturers of amps employing numerous solid state devices don't spend the few cents to stick MOVs in there as well. Maybe it's because they want to have the built-in obsolescence by not doing so. I can't imagine it's because they assume that everyone uses an external surge/spike protector. The lack of a channel balance control (pot or solid state) in so many modern amps is another factor which amazes me!

Maynard

[paul79]

What's tha beef guys? (Mark and Craig)

Edited June 21, 2014 by paul79