Poll

What is amplifier bias

I think it is very important to get just right
1 (6.3%)
Close is good enough
6 (37.5%)
Set the meter to mV
2 (12.5%)
The right bias comes from a design equation
3 (18.8%)
We are measuring what is traditionally called "bias"
3 (18.8%)
Set the meter to mA (milliamps)
1 (6.3%)
None of the above,  I have a question
0 (0%)
We are not measuring what is traditionally called  "bias" but something else
0 (0%)

Total Members Voted: 10

Voting closed: 25 Aug 2014, 03:17 am

Amplifier Bias

0 Members and 1 Guest are viewing this topic. Read 14911 times.

Roger A. Modjeski

Amplifier Bias
« on: 22 Aug 2014, 03:10 am »
Here is a little quiz on bias.

You get to vote for as many options as you want.

The poll closing date is noted and results will be shown and discussed then.

In the meantime lets hear what people think about bias.

bdp24

  • Full Member
  • Posts: 884
Re: Amplifier Bias
« Reply #1 on: 22 Aug 2014, 12:42 pm »
Roger, this seems like a good place to ask about the recommended bias for KT120 tubes when used in the RM200 amp. In the RM200 manual, the recommended and maximum values for all the tubes that can be used in the amp are listed, for both Ram Labs tubes and those tested by others. The list was written, however, before the introduction of the 120 tube. Can you specify the recommended and maximum bias values for the KT120 tubes you offer when they are used in the RM200? 

Roger A. Modjeski

Re: Amplifier Bias
« Reply #2 on: 22 Aug 2014, 05:00 pm »
Roger, this seems like a good place to ask about the recommended bias for KT120 tubes when used in the RM200 amp. In the RM200 manual, the recommended and maximum values for all the tubes that can be used in the amp are listed, for both Ram Labs tubes and those tested by others. The list was written, however, before the introduction of the 120 tube. Can you specify the recommended and maximum bias values for the KT120 tubes you offer when they are used in the RM200?

You can bias a KT-120 at a meter reading of 50-65 mV (50-65 mA idle current) as it has higher dissipation than a KT-88. Whether it sounds better there is up to you. I usually bias them at 40. Higher current does shorten life and run the amp hotter, though still much cooler than most amplifiers. I wonder if designers of high idle current amps bother to check if they are getting any advantage from high idle current or just assuming that it is better?

I am finding in my OTL amp that higher bias does not always result in lower distortion. One must be careful about what one assumes without checking it out.


BobRex

Re: Amplifier Bias
« Reply #3 on: 22 Aug 2014, 06:50 pm »
Well, you really need a "none of the above" based on how you worded your selections. 

Having a newbie user setting a meter to "ma" is asking for trouble.  Using the "mv" range might peg the meter, depending upon the resistor measured.  The actual bias range is as much a function of the design of the output stage (since few people worry about the bias of small signal tubes), so the exact reading is never really exact. 

That should hint at my vote.

Roger A. Modjeski

Re: Amplifier Bias
« Reply #4 on: 23 Aug 2014, 12:06 am »
Well, you really need a "none of the above" based on how you worded your selections. 

Having a newbie user setting a meter to "ma" is asking for trouble.  Using the "mv" range might peg the meter, depending upon the resistor measured.  The actual bias range is as much a function of the design of the output stage (since few people worry about the bias of small signal tubes), so the exact reading is never really exact. 

That should hint at my vote.
 

Ok,  I added none of the above however I think that invites a question. The question concerns traditional push-pull output stages.

bdp24

  • Full Member
  • Posts: 884
Re: Amplifier Bias
« Reply #5 on: 23 Aug 2014, 10:24 am »
You can bias a KT-120 at a meter reading of 50-65 mV (50-65 mA idle current) as it has higher dissipation than a KT-88. Whether it sounds better there is up to you. I usually bias them at 40. Higher current does shorten life and run the amp hotter, though still much cooler than most amplifiers. I wonder if designers of high idle current amps bother to check if they are getting any advantage from high idle current or just assuming that it is better?

I am finding in my OTL amp that higher bias does not always result in lower distortion. One must be careful about what one assumes without checking it out.

Follow-up question:

The instructions in the RM200 manual for setting the bias were (to quote the manual) "written for 4-ohm speaker tap use". If one has connected his speakers to the 8-ohm tap, I assume the bias is set off of those taps? If so, should the bias settings be different than those set off of the 4-ohm taps (I highly doubt it, but better to ask and be sure)?
« Last Edit: 25 Aug 2014, 03:42 am by bdp24 »

Roger A. Modjeski

Re: Amplifier Bias
« Reply #6 on: 25 Aug 2014, 02:24 pm »
Follow-up question:

The instructions in the RM200 manual for setting the bias were (to quote the manual) "written for 4-ohm speaker tap use". If one has connected his speakers to the 8-ohm tap, I assume the bias is set off of those taps? If so, should the bias settings be different than those set off of the 4-ohm taps (I highly doubt it, but better to ask and be sure)?

The bias difference to either the 4 or 8 ohm tap is so small as to not matter. I often put the meter negative on the center (ground tap) as it is available to clamp the probe and I can check all 4 tubes faster. Since it is ground I can leave it for both channels. Bias is not a critical value as most voters correctly stated in the poll.  Just ground the meter negative and check tubes 1-4.

Roger A. Modjeski

Re: Amplifier Bias
« Reply #7 on: 25 Aug 2014, 03:04 pm »
Thanks for your votes. I hope the results are shown now. As you see close is good enough. There is no equation for bias. I find it by careful experimentation.

We are not talking about what is traditionally meant by the term "bias". Unfortunately what we call "bias" is actually idle current measured across some (often unknown) resistor. The term likely became skewed by transistor amplifier biasing where one can only measure idle current. This got called bias, which I would not have done as the "bias voltage" (base voltage is fixed around 0.7 volts). Confusing, yes!!

When I am done with the move I can write up something more on the topic and in the meantime read your tube manuals.

In both transistor and tube amplifiers you are really measuring the idle current of the device and I would prefer we call it that. Bias is the negative voltage on the grid and it is an important distinction.

kenkirk

Re: Amplifier Bias
« Reply #8 on: 25 Aug 2014, 05:02 pm »
Roger,

Maybe you can expand on why bias will change as a tube amp warms up over 20 or 30 mins. Also why we see bias change as the power tubes age, and should one readjust the bias as the tubes age, or should we just leave it alone.

Thanks,

Ken

kenkirk

Re: Amplifier Bias
« Reply #9 on: 25 Aug 2014, 05:25 pm »
Also, could you explain bias run away as it relates to tube amps. I have seen this problem with KT90's that were probably a little gassy. It will get your attention.

Ken

Roger A. Modjeski

Re: Amplifier Bias
« Reply #10 on: 26 Aug 2014, 02:00 pm »
Roger,

Maybe you can expand on why bias will change as a tube amp warms up over 20 or 30 mins. Also why we see bias change as the power tubes age, and should one readjust the bias as the tubes age, or should we just leave it alone.

Thanks,

Ken

Now we really need to get particular about our terms. What I believe you are asking is why the idle current changes as the amp warms up. That is due to the cathode emission coming up in the first minute from zero. The remaining rise is due to elements changing shape due to heat. It takes perhaps 20 minutes for the grid side rods and other mechanical components to settle into their final shape due to simple expansion.

However in some amps the BIAS (negative voltage on the grid) may be changing due to circuit parameters and capacitors coming up to full charge. Here is why we need to correct our language. The thing you are measuring is not the bias on the tube but the idle current which is set by the bias on the grid.

Typical numbers are 20-100 mA for idle current which is typically read across a 1 ohm resistor. In the RM-10 I chose a 10 ohm resistor and in the Dynaco Stereo 70 David Hafler chose a 15.6 ohm resistor so that 100 mA=1.56 volts, the value of a fresh flashlight battery. Now why would he do that?

For BIAS typical numbers are as low as 10 volts for a EL-84 and as high at 60 volts for a KT-88. Again a big range. These are negative voltages placed on the grid through a large resistor of 47 K to 470 K ohms. They come from a regulated or un-regulated BIAS supply via a BIAS pot that you adjust. When you BIAS your amplifier you are adjusting one thing while measuring something else.

Also, could you explain bias run away as it relates to tube amps. I have seen this problem with KT90's that were probably a little gassy. It will get your attention.

Ken

Again terms: It's the idle current that is running away. However that is due to the bias on the tube actually being reduced by grid leakage or gas. In a good tube the grid current is zero. We actually measure this at RAM TUBES and anything over 1 micro amp is rejected. It is interesting you mention KT-90s as they are problematic in that way but the amplifier has something to do with it also. David Manley, the man behind that tube, did a strange thing in his amplifiers. If you read the VTL book you see that he tended to run large value grid (bias) resistors and in the bigger amps they were even larger. The reason for this is another topic of great length. In any event, the larger the grid resistor the more chance for a tube to run away. I used a 47 K ohm resistor in the RM-9 and similar in the RM-10. He used as high as 470 K ohm (the maximum allowed in the tube manual) to bias his grid. The short answer was to make the driver simple in his amps. He is not the only one who uses high value resistors.  A 470 K resistor has 1/10 the ability to hold the negative voltage on the grid. In an ideal world the grid bias resistor would be stated as a spec and a wise person would stay away from amps with high valued resistors. As you may see I take issue with designers who take things to the limit. Does anyone want a list?

As a tube ages grid leakage may go up and thus the negative bias on the tube is reduced, the idle current rises heating the tube and increasing the leakage. This is a regenerative process and the tube runs away. If a fuse or the user stops the process in time the tube may work fine for many months if kept in check. This also explains why a tube can run way (get a red plate) when played hard but be fine when played at low level. Harder play heats the tube to the point where the process may start.

If a tube is really gassy there is no hope for it as the gas in the tube prevents the grid from doing it job at all. However this is a rare case as the problem is usually simple grid leakage. ARC specified "Low gas 6550s" in their amps but what they really were saying is low grid leakage which is not usually caused by gas.

To make a really stable amplifier one either needs low value grid resistors (which are harder to drive) or do what I did in the RM-200 which is do drive them via a directly connected driver tube. The RM-200 can hold idle current 100 times better than most amplifiers. You can actually take tubes with high grid leakage that run away in other amplifiers and use them successfully in an RM-200.

bdp24

  • Full Member
  • Posts: 884
Re: Amplifier Bias
« Reply #11 on: 26 Aug 2014, 07:39 pm »
Roger, this is SUCH great information!

RDavidson

  • Full Member
  • Posts: 2860
Re: Amplifier Bias
« Reply #12 on: 26 Aug 2014, 08:00 pm »
As you may see I take issue with designers who take things to the limit. Does anyone want a list?


YES! The last thing I (or likely most other reasonable people) want, is to get into a tube amp that requires constant tube replacement or constant monitoring and repairs. Who gives a crap how great it sounds if you can't just turn it on and relax, over, and over, and over with consistency and reliability day in and day out for MANY months if not years.

bdp24

  • Full Member
  • Posts: 884
Re: Amplifier Bias
« Reply #13 on: 26 Aug 2014, 08:35 pm »

YES! The last thing I (or likely most other reasonable people) want, is to get into a tube amp that requires constant tube replacement or constant monitoring and repairs. Who gives a crap how great it sounds if you can't just turn it on and relax, over, and over, and over with consistency and reliability day in and day out for MANY months if not years.

That's one reason I finally got a Music Reference, after having owned ARC, Atma-Sphere, Dynaco, and Quad tube amps. Shoulda done it a long time ago!

kenkirk

Re: Amplifier Bias
« Reply #14 on: 27 Aug 2014, 04:52 pm »
Thank you Roger!  Yes that did clear up things for me.  I have witnessed idle current run away with new KT 90's in my CAT JL-1 amps. I also witnessed a bank of EL34's glowing so bright they were lighting up a large closet. They were not new, but I set the bias ( idle current ) correctly for the tubes in the JL-1's, then went about listening to music pretty loud. When I went to flip the record I was shocked at the bank of el34's lit up like light bulbs. No damage to the amp, but I am sure the tubes were toast. I pulled them. It was just an experiement with EL 34's anyway. I have seen KT 90's when new start to runaway in the JL-1's. If I kept the idle voltage around 5mv for a day, then gradually raised it to 25mv over several days, the tubes settled down and were good for 10,000 hours. I loved those KT 90 type II's. Anyway, my next set of CAT JL-1 tubes will come from you. But my RM 9 MKII and my RM 10 MKII seem to be stealing the show lately.

Ken

kenkirk

Re: Amplifier Bias
« Reply #15 on: 27 Aug 2014, 05:00 pm »

YES! The last thing I (or likely most other reasonable people) want, is to get into a tube amp that requires constant tube replacement or constant monitoring and repairs. Who gives a crap how great it sounds if you can't just turn it on and relax, over, and over, and over with consistency and reliability day in and day out for MANY months if not years.

oh so true. I think poor reliability has driven many a music lover back to solid state amps.

Ken

Roger A. Modjeski

Re: Amplifier Bias
« Reply #16 on: 28 Aug 2014, 03:16 pm »
Thank you Roger!  Yes that did clear up things for me.  I have witnessed idle current run away with new KT 90's in my CAT JL-1 amps. I also witnessed a bank of EL34's glowing so bright they were lighting up a large closet. They were not new, but I set the bias ( idle current ) correctly for the tubes in the JL-1's, then went about listening to music pretty loud. When I went to flip the record I was shocked at the bank of el34's lit up like light bulbs. No damage to the amp, but I am sure the tubes were toast. I pulled them. It was just an experiement with EL 34's anyway. I have seen KT 90's when new start to runaway in the JL-1's. If I kept the idle voltage around 5mv for a day, then gradually raised it to 25mv over several days, the tubes settled down and were good for 10,000 hours. I loved those KT 90 type II's. Anyway, my next set of CAT JL-1 tubes will come from you. But my RM 9 MKII and my RM 10 MKII seem to be stealing the show lately.

Ken

I have a good relationship with Ken Stevens (CAT designer) and have done some special testing for him. He wants very low grid leakage. The tubes I supplied him, and people he sends to me, have leakage below 0.1 micro amp which is 10 times lower than my usual limit. You have to be careful what you put in a CAT amplifier to prevent runaway idle current. As the tubes are connected in parallel in a bank, one tube with high grid leakage will take the others along on the trip to hell.

airhead

Re: Amplifier Bias
« Reply #17 on: 1 Sep 2014, 06:28 pm »
This is indeed very interesting, although I do not have the technical knowledge to really understand it.   I've been told that tube runaway can be a frequent problem when trying to set up NYAL (Futterman design) amps, and one of mine caught fire once!.  Most the time though, these amps  are indeed "turn them on and forget them."

enochrome

  • Jr. Member
  • Posts: 53
Re: Amplifier Bias
« Reply #18 on: 12 Sep 2014, 04:26 pm »
I am a new owner of RM-10 mkii amps that I have bought second hand. I am also new to tube amps and would like to set "bias" correctly for them but I have never done this before. Any help would be appreciated. I am concerned because I have set both amps up to be used as mono's. They are set at the 4 ohm tap for "light loading". When I play music I have to crank my Pass Aleph P preamp to almost 3 o'clock to get music to play at 75db. My speakers are 87db @ 6 ohms Revel M20's. I thought 70 tube watts bridged would be enough for my small room, but it isn't, and I think something is up that is not normal  :scratch:

Any advice would be greatly appreciated.

Roger A. Modjeski

Re: Amplifier Bias
« Reply #19 on: 19 Sep 2014, 04:27 am »
I am a new owner of RM-10 mkii amps that I have bought second hand. I am also new to tube amps and would like to set "bias" correctly for them but I have never done this before. Any help would be appreciated. I am concerned because I have set both amps up to be used as mono's. They are set at the 4 ohm tap for "light loading". When I play music I have to crank my Pass Aleph P preamp to almost 3 o'clock to get music to play at 75db. My speakers are 87db @ 6 ohms Revel M20's. I thought 70 tube watts bridged would be enough for my small room, but it isn't, and I think something is up that is not normal  :scratch:

Any advice would be greatly appreciated.

Something does sound wrong. The amps should play well above 75 db. They should hit over 100 at 1 meter. Which input jack are you using? try the other one. Also note that in mono the 8 ohm tap is actually 4 ohms and the 4 is 2. That is the way with all stereo amps when put into mono. It is a good idea to check each amp as a stereo amp to make sure both channels are working and have equal volume. When in mono, one channel can be out and it is not obvious.

Are you able to measure the bias between the two jacks of each channel as something close to 300 mV?

Do you have a manual?