Raoul Duke wrote: ↑Tue Jun 04, 2024 2:28 am
So it is likely that the PI tubes I’ve been using in this build aren’t that well balanced then?
Probably.
Raoul Duke wrote: ↑Tue Jun 04, 2024 2:28 am
Is higher plate voltage desirable in a PI tube? I ask because I know I have a couple of ‘60s GE 5751s around that I’ve tried in the other amps where the plate voltage on “A” and “B” were within 2-3 volts, but they seemed a little on the low side - I.e. where a new JJ would average around 190-200 on each plate, the GEs would average around 170 on each plate. Better balance, but lower plate voltages.
5751 is drawing more current, probably because it will bias relatively hotter with the same Rbias.
Even with perfectly matched triodes, the PI tube will not have equal plate voltages. The LTP PI circuit is naturally unbalanced, and the offset plate loads are intended to fix that.
Interesting.
So HADs plate “recipes” for a given circuit is really just a starting point relative to other factors specific to the particular amp then?
Obviously, he was very careful when choosing components and I don’t doubt that the “ starting point” I refer to was more concrete for him because he controlled the variables better; but for those of us that lack the knowledge of both theory and component selection as well as the second and third order effects and interactions of those points (I’m referring to me - not the Martins, lans, Tonys etc) - it really kind of is a starting point maybe?
HAD tended to use higher plate loads than Fender, with less percentage difference to get closer to balance than the typical Fender 100k and 82k. This was his personal touch I suppose, and then he added the trimer to make it easy to adjust for even more precise balance and to compensate for variation in PI tube section matching.
Ok, thanks. That info helps me understand the relationships in the power amp better.
So in my case - the EH power tubes aren’t really the problem, they just highlight the imbalance in the PI tube triodes more - thus making it harder to adjust out?
bepone wrote: ↑Tue Jun 04, 2024 4:46 am
2mV really is very small negligible, so it is ok.. if is wanted zero (but really no needed), change - swap output tubes socket1-socket2 and retry
Thanks for the info! After doing some more reading, it seems that many Marshall users actually prefer slight PI imbalance and shoot for that on purpose. I’m sure all my other amps are imbalanced various amounts - but I feel like getting Dumble circuits to balance the PI is part of the exercise; whether practical/necessary or not.
Raoul Duke wrote: ↑Tue Jun 04, 2024 4:40 pm
So in my case - the EH power tubes aren’t really the problem, they just highlight the imbalance in the PI tube triodes more - thus making it harder to adjust out?
If all is perfect in PP stage, it is cancelling 2th harmonic, which are actually pleasant .. so in musical instrument what guitar amplifier is, it is not good to go this way. I see people still pushing and pushing on this forum to do so, but IMO is wrong way .
20k-47k trimmer instead of 10k in PI anodes solving all the problems, if someone call this a problem.
bepone wrote: ↑Tue Jun 04, 2024 7:27 pm
If all is perfect in PP stage, it is cancelling 2th harmonic, which are actually pleasant .. so in musical instrument what guitar amplifier is, it is not good to go this way. I see people still pushing and pushing on this forum to do so, but IMO is wrong way .
But in the thread on rack systems:
bepone wrote: ↑Tue Jun 04, 2024 7:31 pm
So i think also everywhere hi distortion preamp nicely matches clean output stage , we can see that topology in hi gain tube amps too :wink:
Dumble ODS creates its distortion in the preamp. The power amp is like a Hi-Fi.
think better, what is 2H and 3th and 5th H.., all are under "distortion" but 2H in small amount is good which you cannot say for 3H and 5H.. and you can cancell only second with perfect matching
bepone wrote: ↑Wed Jun 05, 2024 5:36 am
think better, what is 2H and 3th and 5th H.., all are under "distortion" but 2H in small amount is good which you cannot say for 3H and 5H.. and you can cancell only second with perfect matching
Perfect matching is only possible for a chosen output level. If you're "under" the chosen output level the balance is going to be off, same for "over" the chosen output level.
Fe. at pick attack, perfect match, when fading, no perfect match = bloom
You can easily try it Different levels are prone to the outcome of the matching.
erwin_ve wrote: ↑Wed Jun 05, 2024 6:37 am
Perfect matching is only possible for a chosen output level. If you're "under" the chosen output level the balance is going to be off, same for "over" the chosen output level. Fe. at pick attack, perfect match, when fading, no perfect match = bloom
You can easily try it Different levels are prone to the outcome of the matching.
Not sure what you are saying here (bold).
References to the using the Ampeg method wrt setting the ODS trimmer suggests balancing the power amp for minimum distortion at half-power, similar to tube Hi-Fi power amp practice. This does not eliminate any harmonics coming out of the preamp as long as a clean sine wave from a generator is used for the input to the power amp.
Martin: at pick attack there is a short maximum peak, let's say the balance procedure is done with the max peak of a playing situation in mind.
When the signal gradually fades, the balancing gets off hence return of 2nd order harmonics.
The balance also depends on the frequency used as the test tone when balancing the power amp. Try 100Hz then switch to 1KHz. It won't be balanced for both.
We don't know what frequency HAD used. The SVT method was obviously designed with bass guitar in mind. But it may make sense to still use 100Hz to keep the root notes cleaner in guitar. Or not, lol. I have also picked a note in the middle of the guitar register and used that.
In my limited experience I think the main thing is to balance the PI at the output volume you will play at most. For those of us that do not gig, that probably means a lower output level than what is referenced in the sticky. It does seem to me that the bloom is much easier to find at louder playing volumes though.
Ampeg specifies 40 Hz, approximately the low E on a 4-string bass. For guitar, that would be ~80 Hz. I have seen Hi-Fi amp procedures that specify 1 kHz.
That’s an interesting point as well. From what I’ve read here and other resources regarding PI balancing - there really is no RoT for which signal to use (other than sine wave). Seems to range from the SVT 40 Hz all the way up to 1 kHz. I’ve always followed Martin’s suggestion for 400 Hz thinking that the guitar is generally a mid range instrument.
I guess “to each their own” regarding the variations in the PI balance process. I suppose as long as the builder or “buildee” is satisfied with the outcome - it’s all good…
I appreciate everyone’s input. I learn a lot when this type of discussion happens
