PI Balance/Unbalance for Harmonics

[erwin_ve]

This afternoon I did some tube rolling in my '70s 50W ODS. I should preface my remarks by saying that this amp uses a 12AT7 in the PI, not a 12AX7, and 47K/51K plate resistors with a (HAD not approved) 5K trimpot on the B+.

I ended up with an old Mazda in the PI, and proceeded to do the harmonics optimization procedure, brushing the strings and listening, moving the trimpot around, to find maximum harmonics. After I had found it, I measured the plate voltages and they were identical at 293.4V, ie this is perfect DC voltage balance.

Usually we find maximum harmonics several volts out of DC balance. This intrigued me, because other tubes I have tried always ended up that way at best harmonics content.

So I put a 400Hz sinewave into the amp and measured the AC balance on the plates, after the coupling caps so as to eliminate PS ripple. 5% spread between the triodes, and there are the harmonics.

Anyone else had this happen?

The DC balance of this tube was average, not very close like I usually try to get, and I only have one of them.

Seems like ALL my builds work when i set the trimmer outa balance, i can do it by ear NO problem, The opposite way is HIFI and cleaner, no brainer....imho!! Just set the way ya EARS like it..., no test equip needed imho..

jimbobampdudeinmoravianfallsnc

Taking measurements from the PI is not taking in account of the powertubes. If you use the Ampeg procedure(basically at the power tube cathodes) it is. And will explain why there may be a difference using your ear vs. measuring at the PI.

[talbany]

This afternoon I did some tube rolling in my '70s 50W ODS. I should preface my remarks by saying that this amp uses a 12AT7 in the PI, not a 12AX7, and 47K/51K plate resistors with a (HAD not approved) 5K trimpot on the B+.

I ended up with an old Mazda in the PI, and proceeded to do the harmonics optimization procedure, brushing the strings and listening, moving the trimpot around, to find maximum harmonics. After I had found it, I measured the plate voltages and they were identical at 293.4V, ie this is perfect DC voltage balance.

Usually we find maximum harmonics several volts out of DC balance. This intrigued me, because other tubes I have tried always ended up that way at best harmonics content.

So I put a 400Hz sinewave into the amp and measured the AC balance on the plates, after the coupling caps so as to eliminate PS ripple. 5% spread between the triodes, and there are the harmonics.

Anyone else had this happen?

The DC balance of this tube was average, not very close like I usually try to get, and I only have one of them.

Seems like ALL my builds work when i set the trimmer outa balance, i can do it by ear NO problem, The opposite way is HIFI and cleaner, no brainer....imho!! Just set the way ya EARS like it..., no test equip needed imho..

jimbobampdudeinmoravianfallsnc

Taking measurements from the PI is not taking in account of the powertubes. If you use the Ampeg procedure(basically at the power tube cathodes) it is. And will explain why there may be a difference using your ear vs. measuring at the PI.

Erwin
Balancing the signal at the cathodes still doesn't take into account output, transformer and amount of GNFB applied..If you are going for the perfect balance you have to scope and sweep it at the output after the OPT..However IMO perfect balancing of the PI may not be what we are doing when we balance by ear ..Push pull output sections generate predominately 3rd order harmonics so by purposely throwing off the balancing of the PI we shift the dominating 3rd harmonic to a 2nd..I believe the dominate 2nd order harmonic is what helps promote the bloom whatever you want to call it effect(although I cannot prove it)..Also the cathode follower driver in the output section in some Dumblelands and SSS generate 2nd order harmonics and IMO a big part of why those amps sing..

Tony

[vibratoking]

I am not sure whether the dominant second order harmonic approach or the lowest THD method is what I prefer. I can tell you that I have tweaked the PI while by looking at the spurs on a spectrum analyzer. At this time, I have not come to a 'real' conclusion, but I can tell you that I have tweaked the trimmer for the lowest THD and also unbalanced it so that the 2nd harmonic was about 20db above the 3rd. I liked what I heard with lowest THD. The amp was very clean and punchy in the clean mode. With a dominant 2nd harmonic, the clean side of the amp is bluesy dirty and there were no sparkling cleans to be found - it was not as Dumble-like as the lowest THD setting. I am planning to continue with some better test equipment and I will let you know what I find. I will also be investigating the adjustment based on Talbany's posts and reference to a paper discussed here: http://www.spiritone.com/~rob_369/audio ... 0Shape.htm

I'll let you know what I find.

[Max]

Also the cathode follower driver in the output section in some Dumblelands and SSS generate 2nd order harmonics and IMO a big part of why those amps sing..

Tony,

AFAIR all the power amps of the 150W(+) Dumble amps I know of that have been measured concerning their harmonic content with a spectrum analyer didn't show any special amount of 2nd order harmonics. Just very low overall THD as in one of the "best" old school class AB Hifi tube amps (McIntosh etc.).
The "2nd order generators" in Dumble 150W(+) amps I know of are in their preamp sections and not a feature of their power amps. And AFAIR the power amps of the Dumblelands, SSS, ODS 150W and Odyssey amps I know all are very "classic" class AB stuff in regard to their harmonic content measured by a spectrum analyser.

So IMO the "sing" of Dumblelands and SSSs has not much to do with 2nd order harmonics generated by their power amps.

BTW: adjusting the dynamic balance pot for lowest THD, as recommended by Alexander, does take into account any possible unbalance of the OT, too. That's IMO one of the the advantages of the "lowest overall THD method" compared to the "zero voltage difference between TP1 and TP2 (SVT: K1/K2) method".

All the best and have a great weekend,

Max

[talbany]

I am not sure whether the dominant second order harmonic approach or the lowest THD method is what I prefer. I can tell you that I have tweaked the PI while by looking at the spurs on a spectrum analyzer. At this time, I have not come to a 'real' conclusion, but I can tell you that I have tweaked the trimmer for the lowest THD and also unbalanced it so that the 2nd harmonic was about 20db above the 3rd. I liked what I heard with lowest THD. The amp was very clean and punchy in the clean mode. With a dominant 2nd harmonic, the clean side of the amp is bluesy dirty and there were no sparkling cleans to be found - it was not as Dumble-like as the lowest THD setting. I am planning to continue with some better test equipment and I will let you know what I find. I will also be investigating the adjustment based on Talbany's posts and reference to a paper discussed here: http://www.spiritone.com/~rob_369/audio ... 0Shape.htm

I'll let you know what I find.

Vibro
Thanks for the very informative post!!..
Although 20db sounds a bit extreme and may be the cause of the dirtier clean tones.. as the paper stipulates a 1 to 2 db imbalance..

These experiments showed that you only need to cause a 1 or 2 dB impbalance to get most of what you want in harmonic products. As tubes age, a given designed imbalance will drift. This is another issue to look at if you're a perfectionist.

Give that a shot and please let us know what you find..

All The Best!!

Tony

[glasman]

I am not sure whether the dominant second order harmonic approach or the lowest THD method is what I prefer. I can tell you that I have tweaked the PI while by looking at the spurs on a spectrum analyzer. At this time, I have not come to a 'real' conclusion, but I can tell you that I have tweaked the trimmer for the lowest THD and also unbalanced it so that the 2nd harmonic was about 20db above the 3rd. I liked what I heard with lowest THD. The amp was very clean and punchy in the clean mode. With a dominant 2nd harmonic, the clean side of the amp is bluesy dirty and there were no sparkling cleans to be found - it was not as Dumble-like as the lowest THD setting. I am planning to continue with some better test equipment and I will let you know what I find. I will also be investigating the adjustment based on Talbany's posts and reference to a paper discussed here: http://www.spiritone.com/~rob_369/audio ... 0Shape.htm

I'll let you know what I find.

Your abservations are exactly the same as mine. I tune the PI/Power amp for minimum 2nd harmonic (200hz) based on a 100hz fundamental as measure at the 4 ohm tap. Normally I can get to -55 to -60 dBV below the 100hz fundamental. Depends on the PI type. For example the 70's style AT7 has lower THD than say the Skyliner style.


I normally use either a 5K or 10k trimmer, but 5K are much easier to tweak. To be honest, I rarely look at the voltage diference on the plates as I have found it has no impact on the adjustment.

YMMV,
Gary

[glasman]

FWIW, I don't go as low on the fundamental as the Ampeg procedure as I feel the phase funnies (due to tubes, transformer, couplie caps etc) at the low end impact the adjustment.

[Max]

As tubes age, a given designed imbalance will drift. This is another issue to look at if you're a perfectionist.

My impression was always concerning the Dumble amps I had and have the fun to play, that a very good dynamic balance and low THD values over the complete output range help a lot to bring out the best of them.

So to have a set of power tubes that isn't matched only for idle current but for tranconductance, too, IMO is an important starting point.

And because tubes age it's of course necessary to readjust the idle current and the dynamic balance from time to time. That's why I think an external access to bias and dynamic balance test points and pots is rather practical.

BTW: IMO a Dumble amp is a perfectionistic approach anyway: "The best or nothing!" ("Oh Lord, won't you buy me..."). That's IMO one of the reasons why they are great musical instruments for players and such a most interesting topic for technophiles, too.

Have fun and a nice weekend,

Max

[talbany]

So IMO the "sing" of Dumblelands and SSSs has not much to do with 2nd order harmonics generated by their power amps.

Max..Here is a pretty good article
http://www.freewebs.com/valvewizard1/dccf.htm

So we now have a cathode follower that is permanently 'stealing' current from the previous stage; what is the result? If a down-going signal appears at the gain stage's anode the grid voltage of the follower is pushed down back into the area of "normal operation", and grid current stops flowing more or less immediately, for the duration of that cycle. But when the incoming signal is positive going the grid is pushed more positive, which induces even more grid current to flow into the cathode follower as it tries to maintain bias, which in turn 'drags down' the anode voltage of the gain stage which is trying to move positive! In other words, the up-going cycles are very heavily compressed, but the down going ones are not, which generates a lot of second harmonic distortion. This is only made possible by the DC coupling, and is why this circuit is so often used in high gain amps- it can warm up a signal that already contains too many high-order harmonics, and return a rather fuzzy sound to a rich, creamy distortion tone.

Last I checked that follower resides in the power amp section..

FWIW..To be clear I wouldn't consider that follower to be the main reason why DL & SSS/Pentode amp w/follower sing, only a contributing factor..IMO of coarse..
When I finish my SSS build I'll do some spectrum analysis and get back

Have a Great Weekend Back!!

Tony

[DimenBrun]

Here's a thought that's maybe > 10 yrs too late for this thread. For a SSS amp, if you have cathode followers in both sides driving the power tubes, even if they are generating 2nd harmonic, it should be cancelled out (depending on how well the two sides are balanced) since the two sides are opposite polarity. That's my two cents on the topic.

[martin manning]

...even if they are generating 2nd harmonic, it should be cancelled out (depending on how well the two sides are balanced) since the two sides are opposite polarity.

True, but unlike the direct-coupled CF Blencowe is talking about in the dccf link above (5F6-A Bassman, JTM, etc.), the SSS driver stages are ac coupled, and biased in the "normal" operating range. They are highly linear, and grid current does not play any significant role. The Valve Wizzard dccf link above is broken; here's the new one: http://www.valvewizard.co.uk/dccf.html, and see http://www.valvewizard.co.uk/accf.html for the ac coupled version which applies to the SSS driver stages.

[DimenBrun]

AC coupled, huh? I guess I looked at the wrong schem. for a SSS. What is the point of running AC coupled followers to drive the PT's since they can't supply much grid current? Is there some other sonic effect that one gets?

[DimenBrun]

Ok, I see the error of my ways. AC coupled follower input, not output.