[nnck]

Per Roger's suggestion, I have been trying out the 1ohm speaker taps on my RM200II. I may have more to say about my impressions after I listen a bit more. For now, I have a question about the tube biasing when using the lower taps.

I decided to check the tube bias again today for the first time after switching from the 4ohm taps to the 1ohm taps. Should the bias of any of the tubes change when you change speaker taps? Are you supposed to re-bias when you switch speaker taps? Because it seems as though tubes V4 and V3 (the tubes tested by contacting the black probe of the voltmeter to the (-) speaker taps) have jumped up by about 3-4 mV. On the other hand, tubes V6 and V5 (the tubes tested by contacting the black probe of the voltmeter on the (+) speaker taps) have remained close to the same as they were before.

I may have run into some confusion about which TP+ test point to use for the testing of the (-) speaker taps, which are the readings which appear to have changed. This is a little confusing to describe, but should you always be using both TP+ positions for the red probe when checking bias of the 2 tubes on one channel even when using the 2ohm or 1ohm taps?

[Roger A. Modjeski]

The bias method or setting does not change with the taps. It is always as stated "from the test point to the nearest 8 ohm tap" You can also measure it from the center tap to the test point with only a 2 mV difference, which is nothing.

Bias is not critical. Most amplifier owners are too concerned about bias values. It really doesn't matter that much. Between 30 and 40 mV there is so little change in the sound.  There is a much, much bigger change in the sound as one explores the taps. Thank you for exploring and nevermind the bias.

[nnck]

Thanks Roger. I will try to not concern myself much with the bias adjustments.

Since I'm a scientist though, I just cant help but ask you a little more about the process of checking the bias. I feel like I might still be a bit confused about it, and I'd at least like to know the correct way to measure it. So when you say to check the bias "from the test point to the nearest 8 ohm tap" are you saying to use the nearest 8ohm tap as your contact point REGARDLESS of which tap your speakers are connected to? I was under the impression that you contact the negative probe of the meter to the taps that your speakers are connected to (so in my case, I'd be contacting it to the 1ohm taps).

FWIW, I dont think my manual actually has that exact instruction in it. But I'll look at it a bit more closely.

[Roger A. Modjeski]

Here is the scientific explanation. The bias flows through the windings of the transformer which are about 0.1 ohm, virtually nothing. If you try measuring to any tap you will get about the same reading. Perhaps you are thinking that the bias if flowing to the tap where the speaker is connected. It is actually flowing to the center terminal which is ground. The bias, and the measurement of it, has no knowledge of where your speaker is connected or if it is connected at all. You can adjust the bias with nothing connected. This is true of most amplifiers, though some may become unstable, mine won't mind.

I find it easier to use either of the center terminals (ground) for my negative probe and leave it there for all 4 tubes. I recommend that you put the negative probe under the cap and tighten it for a good connection. Correctness of readings requires a good connection. All metals, including the probes have a thin layer of oxide on them. That is why readings can fluctuate unless you press hard. That is why probes are sharp (or should be) to break through the oxide.

The test jacks are actually miniature banana jacks and we can make you a set of long probes or sell you the mating plugs if you want to go that route. It allows a more stable connection which is good if you want to monitor the bias over time which you might do if you suspect a bad tube.

One more thing about bias... Where do you think we (designers) get the value?   Answer....We make it up. It's not in a book or data sheet unless you are building a published application, which is what most of our competitors do. When a designer is developing an amplifier he will usually notice that the amp has less distortion as he rasies the bias. Some are tempted to go beyond the book's recommended bias or dont consider what will happen when one owner has high line voltage.  That's what people mean when they say "most tube amplifiers come right out of the book". The "book" is the RCA, GE, Mullard, Genelex, or Telefunken tube manual.  The applications are usually written by engineers from the company that invented the particular tube. Interestingly enough, the other manuals simply copy the original applications word for word (and value for value). Genelex published a whole book on the KT series with some interesting applications that I have never seen in a commercial product.  The RM-200 application is not in the book, I made it up from my understanding of tubes, transformers and circuits. Have a look at this thread currently under discussion by ttan98 where we are discussing the RM-200 vs typical KT-88 amplifiers.  http://www.audiocircle.com/index.php?topic=125760.0
 
What he is referring to is a typical KT-88 application like the Dynaco MK III. That is an Ultralinear application right out of the book. In this case I think David Hafler and Herb Keros wrote the application and the data books picked it up due to its popularity. (Hafler claims to be the inventor of the Ultralinear connection, but the Brits disagree.)  What happens is owners, and designers, see lots of 50-60 watt amplifiers with those tubes producing that power and assume that's just how it is. Well, there are an infinite number of applications, as many as we can think up. Some good, some bad. Some applications are modified and pushed beyond safe limits. Some have very short tube life and are fussy about the tubes. The RM-10 gets over twice the power that previous applications have yielded. Every now and then the DIY or Asylum community gets upset about my power claims and says "wow Roger must really burn up the tubes". However the tubes last a very long time, often longer than in the typical lower power applications from Dynaco, Fisher, Scott, etc.

At the risk of making this too long perhaps I should explain what an "application" is.  It spells out the specific voltages, currents, negative bias and transformer impedance that will result in a particular power output. It does not include transformer losses or consider the effects of feedback because those are specific to the parts chosen by the user of the application. However it does give enough information so that someone need know very little about tubes and get an amplifier circuit running with parameters that won't blow up the tube. David Manley and I got into a discussion in Stereophile some 25 years ago where he insisted that the applications were written in stone and that other applications were heresy. We had a lively discussion which I should someday reprint.

In my designs I run lower bias (current) than most amplifiers because I want to tubes to last 5-10 times longer than Audio Research or CJ. It's easy to make a good sounding amp running the tubes at their max but then they only last about 1000 hours. It's much more challenging to make a good sounding amplifier (and one that is not embarrassing when John Atkinson measures it for Stereophile) http://www.stereophile.com/content/music-reference-rm-200-mkii-power-amplifier.

[nnck]

Great detailed explanation Roger. I appreciate the thought and time you put into our questions.

I will try not to obsess over the overall bias setting on the amp. But I can see that I probably will obsess over the using the correct method to make the measurements and/or do the adjustments. You did confirm one thing for me that I concluded this morning, and that is that the bias reading from any particular tap doesnt seem to change regardless of whether a speaker load is connected to it or not. However, I do still have a concern over something else you are saying in your last post.

Your comment that "If you try measuring to any tap you will get about the same reading" just doesnt seem to hold true for my amp, unless I am still not understanding something. If I leave the red positive probe in one TP jack and then contact the black negative probe to the 3 nearest speaker taps to it, the readings actually do change by about 3-4 ohms from one tap to the next.

Here's one real example from measuring one tube this morning: leaving the red positive probe in the same TP nearest the tube, if I contact the black negative probe to the 8ohm tap I get 36.4mV; moving the black probe to the 4ohm tap gives me 40.2mV; and then moving the black probe to the center tap gives me 44.4mV. I've tried doing the same thing for all 4 tubes with similar results- the readings go up by about 3-4ohms at a time as I move from the 8ohm to the 4ohm and then to the center taps.

To me, those readings dont look the same from one tap to another, as you suggest they should be. And I think this has been the source of my confusion all along. Because I see that the back of the amp says to measure the bias to the nearest 8ohm tap. And the manual is actually written in a way that led me to the incorrect assumption that you should measure the bias to the tap that the speaker is actually connected to (i.e. the manual instructions have a note saying "written for 4-Ohm speaker tap use", suggesting that you should measure to the speaker tap that you are actually using). And your latest post in this thread says you actually prefer to use the center tap (ground) for all of the measurements. So that covers all 3 speaker taps that you could possibly use to test from any particular TP jack. If the readings didnt change from one tap to the next, as you suggested, then I could see that it shouldnt matter. But as I indicate above, it seems to me that the readings actually DO change.

[Roger A. Modjeski]

I said they don't change much, yes as you go toward ground you pick up some of the transformer secondary winding resistance which adds to the measurement. By the way in several places you say 3-4 ohms when I think you mean to say 3-4 mV. Perhaps I am being too casual for you is stating I don't care about 3-4 mV out of 30 or 40, its only 10%. While the RM-200 will hold bias within a few percent of set value over a 15% change in  line voltage (I actually make the idle current go down slightly as line voltage rises so the dissipation remains constant) many amplifiers have idle current that rises faster than the line so that a 10% rise in line causes more than a 10% change in idle current and thus a 20% rise in dissipation (10% for the voltage and 10% for the current). Most traditional and some modern amps have this problem so that from time to time as our line voltage fluctuates the bias is never the same. One day you set it at 40 and the next minute, hour or day later it might be 30 or 50.  I am at the end of the street and on the last pole and my voltage goes from 106 to 120 all the time. Sometimes the needle on my line monitor moves back and forth like a metronome about once per second.  My amplifiers simply shrug this off. I was in Berkeley recently and measured my host's line at 128 Volts. A new record for me. However when I test my amps I turn the variac all the way up for several minutes. That's 140 Volts. I would never do that with a Dynaco, CJ, Rogue or ARC amp. The point I am making here, once again, is that the RM-200 has very stable and forgiving bias. I think those amps previously mentioned have caused all this concern about "precise bias". With those amps you really have to watch out.

In my bench tests the parameters of the amplifier that I feel are sonically important don't change much from 30 to 50 mV (mA of idle current). You are welcome to set the bias to any value you like. With Genelex KT88s you can safely go up to 50 mV but that will shorten the life of the tube. (KT-120s can go to 60). If you find the amp sounds good at 30 mV then you can save some electricity, heat your room less and the tubes will last longer.  At any bias level, lower taps make the amp sound better, have more control and expand the class A region. Those numbers are measured as stated on the amp "from the test point to the nearest 8 ohm tap" because that's where the resistor is.  If you want to be precise when using the ground post just subtract 20% from your reading. The point I am trying so hard to make here is THERE IS NO PRECISE NUMBER FOR BIAS.

Hoping I don't add to anyone's consternation is the fact that the bias in the two tubes of one particular channel should be the same within 20%. That does matter. If you have a larger imbalance than that the output tubes should be replaced with a matched pair. Keep in mind that since the driver tube is direct coupled, it influences the bias balance also. Often swapping the output tubes within one channel will make an imbalance either better or worse depending which is out of balance. If the output tubes are well matched the numbers will stay the same when you swap. Usually there is a bit of imbalance in the outputs and a bit in the driver which can either cancel each other or add to a greater imbalance. Of course choose the set-up that is best balanced.

[nnck]

I said they don't change much, yes as you go toward ground you pick up some of the transformer secondary winding resistance which adds to the measurement. By the way in several places you say 3-4 ohms when I think you mean to say 3-4 mV.

Yes, thanks for pointing that out. I absolutely meant to say 3-4 mV, not ohms. (just too much typing today).

Perhaps I am being too casual for you is stating I don't care about 3-4 mV out of 30 or 40, its only 10%.

Keep in mind what I was saying Roger: From any one TP jack, there was a 3-4mV increase by taking the measurement from the 8ohm tap to the 4ohm tap, and then another 3-4mV increase by going to the center tap. In the end there was about an 8mV difference by taking the measurements from the 8ohm vs center tap.

But regardless of that, based on everything I am learning today I am just going to start taking all of my measurements using the 8ohm speaker taps instead of taking the measurements from the taps that my speakers are connected to. That was really the source of my confusion. Thanks for all your help here.

[Guidof]

Excellent and informative discussion about bias and its measurement! I learned a lot from this thread and as a result have a better understanding of how my RM 200 II works.

Regards,

Guido F.

[Roger A. Modjeski]

Excellent and informative discussion about bias and its measurement! I learned a lot from this thread and as a result have a better understanding of how my RM 200 II works.

Regards,

Guido F.

Thanks, I took special effort in this to explain the much misunderstood topic of bias.