Every now and then I get confused about how to set the bias correctly on a tube amp, especially when there are changes in B+ and plate voltages (usually higher than designed for).
Can anyone recommend a great article on how to set the bias for different amps?
Here are some articles that I've read so far
http://www.eurotubes.com/euro-l.htm
http://www.tone-lizard.com/Biasing.htm
Thanks!!
Can anyone recommend a great article on how to set the bias for different amps?
Here are some articles that I've read so far
http://www.eurotubes.com/euro-l.htm
http://www.tone-lizard.com/Biasing.htm
Thanks!!
I would disagree with 90% of biasing articles on the internet....
What it really boils down to is efficiency vs. distortion..........
Real simply..the hotter you bias, the lower the distortions will be....thus more dissipation to burn known as in-efficiency and tube life shortened...
This is at the discresion of the listener as well as the condition of his wallet.....
But keep in mind that there is a point you reach in biasing where it is "optimal"...meaning that "most" of the distortions are reduced and any further biasing "hotter" will only lead to diminished returns...since the distortion curve never really reaches 0, but is instead asymptotic....
Biasing by means of observing DC idle cathode current, although a popular method is useless...it only serves the purpose of checking to see if your DC dissipation is within safe limits...you should really be looking at the plate current...
The cathode current is going to be a bit higher than the plate current and through your numbers off... For a real example...I am looking at some EL34M tubes... I have a mA meter on the PLATE, SCREEN and CATHODE..... Both tubes show 34mA on the cathode....One tube has 28mA passing the plate with 6mA passing the SCREEN.... The other tube is passing 32mA in the PLATE and 2mA in the SCREEN..... So go figure what the Worst-Case Analysis is when you use 4 of these tubes in a Push-Pull amp.... The whole point of matching the DC currents is to maintain balance/cancelation of the net DC flux in the OPT....
Well, keep in mind it's the PLATE current you need to concern yourself with not the cathode current to the OPT , unless your dealing with Macs or other Cyclotron type stuff....
That is only DC balancing...what really matters is the AC signal balance which happens to be gm, transconductance, matching of the tubes....
OK, so you buy your matched tubes from some online store...SO the tubes arrive and they are well matched, but you don't seem to make much power output...well you probably have low gm matched tubes, unless you specified what gm you wanted...
Tube sellers love ignorant people...they get to dump all thier low measuring matched tubes on them....since experienced amp builders will refuse and turn them back.....
Another to keep in mind....When you drive signal into an amp..the DC voltages will usually take a major droop....and the operating points are all different now...so that idle current bias check you did at idle is telling you what now????
Also depending on the plate load, this will push the tube into different operating regions which can cause the AC plate dissipation to exceed durring operation, but return to a safe DC plate dissiaption at idle...this is the classic case of "my tubes red-plate when the amp has signal but when I check the idle currents the tubes are biased very concervatively"..... This is usually due to too small a plate load.......
The word "Cross-Over Notch" has somehow made it's way from the transistor world to the tube world.... and it has been re-defined...
"Cross-Over Notch" by definition is a CLASS-B phenomena from the transistor early days .....And relates to when BOTH sides of a push-pull stage have a dead-time....ie, BOTH sides of P-P are in cut-off at the same time.... This leaves a flat line between the waveform....as well as nasty turn-ON, turn-OFF transients.....
Vacuum tubes circuits such as Class AB1 are technically not exhibiting "Cross-Over Notch" ...since neither tube is in 100% cut-off at the point of inflection, 0-axis.... The cut-off occurs later by some number of degrees....
Instead the curvature one sees in a cold biased Class-AB1 stage is the tube's non-linearity of it's transfer curve, right 'before' cut-off.... If you were to place "IDEAL" perfectly linear tubes in a Class AB1 stage, you would not see this curvature NOTCH regardless of where you biased the tubes and the power output would remain constant...you would only see this flat line Cross-Over Notch the instant you break-over into Class-B ......
Ocassional I will see a real crappy set of new tubes that always will have a non-linear notch in the waveform, regardless where these are biased, even past thier plate dissiaption...most likely these tubes are defective and have poor grid geometry....
Biasing in Hi-End audio was always about getting the strong gm matched tubes and biasing with signal gen and a distortion analyzer to bring the distortions down to a specific requirement or percentage and or just using your ears provided you know what your listening for...hopefully you get the distortions down to where they are acceptable and you are in a safe range of plate dissipation, then thats a win-win.... As for efficiency,who cares???.. ..thats for the marketing guys to worry about for advertising purposes...
For guitar amps.... the lower gm tubes are less brutal and break up softer.... the "notch" tends to add a fuzzi-ness to the sound that is just plain annoying to listen to when the amp is cranked... Bias out that Notch by ear or with a scope and things usually sound sweeter.... no need to continue biasing hotter after the notch is 98% gone....you will only get diminishing returns...remember it's only Class AB1....if your that anal about removing that last few percent of that notch then go get a Class-A amp..otherwise if you continue biasing hotter, then you will soon find yourself in Class-A and the tubes will be to having a major melt-down....
just me 2 cents....
cerrem
What it really boils down to is efficiency vs. distortion..........
Real simply..the hotter you bias, the lower the distortions will be....thus more dissipation to burn known as in-efficiency and tube life shortened...
This is at the discresion of the listener as well as the condition of his wallet.....
But keep in mind that there is a point you reach in biasing where it is "optimal"...meaning that "most" of the distortions are reduced and any further biasing "hotter" will only lead to diminished returns...since the distortion curve never really reaches 0, but is instead asymptotic....
Biasing by means of observing DC idle cathode current, although a popular method is useless...it only serves the purpose of checking to see if your DC dissipation is within safe limits...you should really be looking at the plate current...
The cathode current is going to be a bit higher than the plate current and through your numbers off... For a real example...I am looking at some EL34M tubes... I have a mA meter on the PLATE, SCREEN and CATHODE..... Both tubes show 34mA on the cathode....One tube has 28mA passing the plate with 6mA passing the SCREEN.... The other tube is passing 32mA in the PLATE and 2mA in the SCREEN..... So go figure what the Worst-Case Analysis is when you use 4 of these tubes in a Push-Pull amp.... The whole point of matching the DC currents is to maintain balance/cancelation of the net DC flux in the OPT....
Well, keep in mind it's the PLATE current you need to concern yourself with not the cathode current to the OPT , unless your dealing with Macs or other Cyclotron type stuff....
That is only DC balancing...what really matters is the AC signal balance which happens to be gm, transconductance, matching of the tubes....
OK, so you buy your matched tubes from some online store...SO the tubes arrive and they are well matched, but you don't seem to make much power output...well you probably have low gm matched tubes, unless you specified what gm you wanted...
Tube sellers love ignorant people...they get to dump all thier low measuring matched tubes on them....since experienced amp builders will refuse and turn them back.....
Another to keep in mind....When you drive signal into an amp..the DC voltages will usually take a major droop....and the operating points are all different now...so that idle current bias check you did at idle is telling you what now????
Also depending on the plate load, this will push the tube into different operating regions which can cause the AC plate dissipation to exceed durring operation, but return to a safe DC plate dissiaption at idle...this is the classic case of "my tubes red-plate when the amp has signal but when I check the idle currents the tubes are biased very concervatively"..... This is usually due to too small a plate load.......
The word "Cross-Over Notch" has somehow made it's way from the transistor world to the tube world.... and it has been re-defined...
"Cross-Over Notch" by definition is a CLASS-B phenomena from the transistor early days .....And relates to when BOTH sides of a push-pull stage have a dead-time....ie, BOTH sides of P-P are in cut-off at the same time.... This leaves a flat line between the waveform....as well as nasty turn-ON, turn-OFF transients.....
Vacuum tubes circuits such as Class AB1 are technically not exhibiting "Cross-Over Notch" ...since neither tube is in 100% cut-off at the point of inflection, 0-axis.... The cut-off occurs later by some number of degrees....
Instead the curvature one sees in a cold biased Class-AB1 stage is the tube's non-linearity of it's transfer curve, right 'before' cut-off.... If you were to place "IDEAL" perfectly linear tubes in a Class AB1 stage, you would not see this curvature NOTCH regardless of where you biased the tubes and the power output would remain constant...you would only see this flat line Cross-Over Notch the instant you break-over into Class-B ......
Ocassional I will see a real crappy set of new tubes that always will have a non-linear notch in the waveform, regardless where these are biased, even past thier plate dissiaption...most likely these tubes are defective and have poor grid geometry....
Biasing in Hi-End audio was always about getting the strong gm matched tubes and biasing with signal gen and a distortion analyzer to bring the distortions down to a specific requirement or percentage and or just using your ears provided you know what your listening for...hopefully you get the distortions down to where they are acceptable and you are in a safe range of plate dissipation, then thats a win-win.... As for efficiency,who cares???.. ..thats for the marketing guys to worry about for advertising purposes...
For guitar amps.... the lower gm tubes are less brutal and break up softer.... the "notch" tends to add a fuzzi-ness to the sound that is just plain annoying to listen to when the amp is cranked... Bias out that Notch by ear or with a scope and things usually sound sweeter.... no need to continue biasing hotter after the notch is 98% gone....you will only get diminishing returns...remember it's only Class AB1....if your that anal about removing that last few percent of that notch then go get a Class-A amp..otherwise if you continue biasing hotter, then you will soon find yourself in Class-A and the tubes will be to having a major melt-down....
just me 2 cents....
cerrem
Cerrem,well put.
Just to mention, those using p-p output stages with toroid trannies must be wondering what happens to the performance with the inherent misbalanced flux with a closed core ? The minute lamination gap and extenstion of the B/H curve created by the standard E&I laminations is a distinct advantage.
Since the tube currents v.s gm/v on AC signals cannot track tobe perfectly identical, I cannot see the light of day of those ranting about the advantages of toroid output transformers in output stages.
richj
Just to mention, those using p-p output stages with toroid trannies must be wondering what happens to the performance with the inherent misbalanced flux with a closed core ? The minute lamination gap and extenstion of the B/H curve created by the standard E&I laminations is a distinct advantage.
Since the tube currents v.s gm/v on AC signals cannot track tobe perfectly identical, I cannot see the light of day of those ranting about the advantages of toroid output transformers in output stages.
richj
- Status
- Not open for further replies.