[FullRangeMan]

conrad-johnson amps have what______________?
seems simple to me...easily accessible, stays put a long time.
I've owned others where I had to use a meter fomr underneath...hassle.
As I understand it self biasing has some limitations, if they are audible I don't
know.

According Decware site self adjusting bias is suited to extract low power from a tube.
Of course Iam unlettered in tubes and biasing.

[FullRangeMan]

Thanks for the link. I am looking for new commentary as I was told some existed.  The tubes noted at the beginning of the link like the 845 and have such high bias voltages (greater than -100V) that developing that bias over a cathode resistor would produce a lot of heat...

Wow, Thanks Mr.Modjeski for this tutorial about Bias, I appreciated it.
Regards

[corndog71]

Roger,
Have you seen this mod for the SCA-35/ST-35? 

http://www.tronola.com/A_New_Look_At_An_Old_Friend.pdf

Also, Joe Curcio sells an auto-bias circuit for the ST70.

http://www.curcioaudio.com/BCS_3.htm

[Roger A. Modjeski]

Roger,
Have you seen this mod for the SCA-35/ST-35? 

http://www.tronola.com/A_New_Look_At_An_Old_Friend.pdf

Also, Joe Curcio sells an auto-bias circuit for the ST70.

http://www.curcioaudio.com/BCS_3.htm

I wish Tronola could tell his story in fewer pages. His first mod still requires matched tubes. He has an interesting concept tying the bias reg to the B+. Some of what he says is questionable.

I think Joe Curcio's auto bias is a servo. There aren't enough technical details for me to tell for sure. He ususally does good things.

[Ericus Rex]

The plate resistance of a pentode is so large that it doesn't matter and even a triode with low plate resistance it matters not. When we set the idle current (mV across our cathode resistor) in a tube we are done. All done and nothing else matters. By the way, we bias the tube to reduce the cathode current from its maximum value which occurs at zero grid volts which typically occurs at full output. By contrast bipolar transistors and MOSFETS are always off and we have to bias them on to their idle value.

Your second point is correct, most voltage gain stages are cathode biased because they run very class A where the idle current is so large compared to the signal that the signal can't influence it. There are some DIYers using LEDs in the cathode of a 12AX7 etc for bias. This idea demonstrates a lack of understanding on their part and it is a very bad idea. Perhaps they do it so they can see the tube is on. Regardless, this practice or any other method of fixing the bias on such a tube removes its ability to adjust itself as it will in cathode bias making the amp very tube fussy.

So then is gm the only characteristic which makes two otherwise identical tubes require different bias settings in the same circuit?  I had just assumed that plate resistance was a major factor since it fits using ohms law.

Do LEDs have a constant voltage drop across them regardless of current?

BTW, after our last discussion of biasing a few years ago I stop using the term auto- and self-biasing b/c they are so misleading.

Thanks for the info!

[Roger A. Modjeski]

Hi Roger
actually
jfets similar to tubes
mosfets enhancement mode similar to bjt
mosfets depletion mode similar to jfets and tubes

kind regards

 

Yes, that's what my previous posts said,  however none of the above behave like a triode. They all behave like a pentode where the output is high impedance and the transfer function is basically current per volt. A triode is the only device I know that is volts per volt which results from its output impedance that changes with current in opposition to the transconductance. Thus a triode can have low impedance output and not need feedback which a pentode needs to achieve low impedance. It is interesting to note that the power triodes we use for output tubes have very high transconductance, low output impedance and thus low voltage gain (3 to 5).

I recently made (and sold) a pair of 12 watt monoblocs using 13EM7s in push pull. With no feedback the results are very good for both low distortion and good damping. What more could one want.

[G Georgopoulos]

 

Yes, that's what my previous posts said,  however none of the above behave like a triode. They all behave like a pentode where the output is high impedance and the transfer function is basically current per volt. A triode is the only device I know that is volts per volt which results from its output impedance that changes with current in opposition to the transconductance. Thus a triode can have low impedance output and not need feedback which a pentode needs to achieve low impedance. It is interesting to note that the power triodes we use for output tubes have very high transconductance, low output impedance and thus low voltage gain (3 to 5).

I recently made (and sold) a pair of 12 watt monoblocs using 13EM7s in push pull. With no feedback the results are very good for both low distortion and good damping. What more could one want.

Hi Roger

in ss in order to acheive good damping is design a big nfb factor

nfb factor = open loop gain / closed loop gain

then

damping factor = Rload / (Zout / nfb factor)

Can you throw in some equations,cause i cant follow you with what you said :

"With no feedback the results are very good for both low distortion and good damping. "

kind regards

George

[Roger A. Modjeski]

Hi Roger

in ss in order to acheive good damping is design a big nfb factor

nfb factor = open loop gain / closed loop gain

then

damping factor = Rload / (Zout / nfb factor)

Can you throw in some equations,cause i cant follow you with what you said :

"With no feedback the results are very good for both low distortion and good damping. "

kind regards

George

I am trying to avoid equations for my gentle readers.  Actually Solid State amps can have good damping without much feedback as most of them are emitter followers which have low output impedance by their nature.  I made several SS amps without feedback.

As to my triode amp: There are listeners who prefer amplifiers without any global feedback. However many of these amplifiers have a lot of coloration due to distortion and poor damping. What I have done is to make a no feedback amplifier where the distortion is below 1%. As to damping I prefer to quote the output regulation which is 2 dB from no load to rated load. I find this more useful as it directly tells us the maximum number of dB that a speaker can influence the flatness of the amplifier. In other words, if the load goes from 8 ohms to infinity the output will only go up 2 dB or about 20%. This converts to a damping factor of about 5. Changing the output transformer ratio can improve this with a little loss of power.

[corndog71]

Are individual bias pots per tube better or a waste compared to one bias pot per 2 tubes (1 channel)?  My understanding is that the individual pots would allow greater freedom when using unmatched tubes. 

The kits I've built using the Curcio driver mods use 1 pot per channel with the option to also balance the bias between the tubes using a 2nd pot. 

While I haven't looked extensively I haven't seen anyone claiming their bias controls equates to better sound.  Then again I tend to gloss over the marketing bs unless it's really entertaining.

[Roger A. Modjeski]

Are individual bias pots per tube better or a waste compared to one bias pot per 2 tubes (1 channel)?  My understanding is that the individual pots would allow greater freedom when using unmatched tubes. 

The kits I've built using the Curcio driver mods use 1 pot per channel with the option to also balance the bias between the tubes using a 2nd pot. 

While I haven't looked extensively I haven't seen anyone claiming their bias controls equates to better sound.  Then again I tend to gloss over the marketing bs unless it's really entertaining.

Its a good question and the answer depends on several things.  Although David Manley likes to have a pot for each tube this can be difficult for the user especially when installing a large set of new tubes. If you have 8 pots you should turn them down siginificantly (not necessarily all the way) to avoid overheating the new tubes which will likely be stronger as they are new or just because they are different. There is a terribly large range of bias voltages for output tubes from maker to maker, batch to batch and even within a batch. When I match up tubes I have at least 100 tested and measured with precise bias and gm values. The range from top to bottom forms a bell curve and the ends are usually 10 volts apart. At typical gm values every volt is about 6-10 milliamps of difference. This is why we match to 0.5 volts typically and we can provide matches to 0.1 volts which is a difference of say 4 mA out of 50 which is less than 10%. Read my paper on matching for more details.

http://tubeaudiostore.com/virofpowtubm.html


The main problem with idivudial bias pots is interaction caused by the fact that as you increase any one tube it reduces the current in the other tubes due to power supply sag. (Someone called me yesterday and said "tube sag" is all over the internet and he thinks they are talking about power supply sag}. Any links would be appreciated). This is quite normal and it forces the user to go around the circle of bias pots several times to get them all equal. What I did in my RM-9 SE was to provide one bias pot per channel and one DC balance pot so once the bias is set the balance, which is more important, can be nulled to zero. With individual bias pots this is very difficult to do as one has to learn to overbias a tube by the proper amount to compensate as the others are adjusted, just as a skilled piano tuner sets a string a little sharp to compensate for the torque of the tuning pin as he releases the tuning hammer, its a learned skill.

If you read the paper linked above you will also see other advantages of matched tubes like symmetrical clipping and reduced distortion. Bottom line, there is just no good reason to use unmatched tubes unless you can confirm that they are reasonably close. Two tubes from the ends of the spectrum would sure degrade the performance of an amplifier.

[rbwalt]

i have a RM9SE and setting the bias and balance is so easy that a 3rd grader can do it. thanks roger for making it very easy. i set my bias at about 27.5 for my EL34's.

rob.

PS. by the way the SE is to say the least a fantastic sounding amp!!!! Kudos's roger.

[jeffreyfranz]

There are the three. Any others?

4. Excuse Me Bias: The amp is developed as a learning computer with a power output stage. Its OS is preloaded with all the world's great music, and it anticipates problems in plenty of time. It then says, "excuse me" (like the female voice on my Mac), and tells the listener to rise, grab the appropriate tools and check the bias. Wait, I think you covered that, above (Fixed Bias).

I don't mean to be a punk. This is an interesting discussion. Thanks, Roger.

[Roger A. Modjeski]

According to Dick Olsher in the most recent TAS Bob Carver has a new EL34 amplifier with one cathode resistor  Although Dick Olsher states in the review that he previously thought individual cathode resistors are better he appears to have been won over by Bob's ability as a hawker to believe that now that one is better To make matters worse Bob has put a 14 volt zener diode across the cathode resistor which will likely short out the first time a tube shorts or flashes. Then there will be no bias resistor. After the tubes meltdown the amp can visit the shop to have its zener diode replaced, which is there only so the amp can do 20 watts sine wave (remember Bob likes specs). I have not seen a schematic, would love to, however its pretty clear to me what the circuit looks like from Olsher's description. Oh one other thing. It appears Bob did something with the feedback to raise the output impedance to 1.5 ohms making the damping about 5. I guess he wants the amp to sound real tubey.