current driven diatones, open baffle questions

[enochRoot]

hey all_

i was curious if someone could shed some light on an idea i have. i have been interested reading all the talk of current drive (trasnconductance) amps, and wanted to see if it might work out w/ my diatones. i understand that the closer you get to the "perfect" current drive...the less that electrical Q plays a part in Qts. well...the diatones have a Qms of around 10. a Qes of around 1.9, and a Qts of around 1.5. since i am using these on open baffles...a high Qts is good, right? but how high do we want? is 1.5 about perfect (as it is), and i shouldn't mess with this, or will higher be better yet? i have run some numbers quickly (based on information from the other current-drive threads), and it seems that if i had a Zo of around 20...my new Qts would be (roughly) 3.5. if i had a Zo of around 40...the new Qts would be roughly (5.5). i'm quoting these numbers off the top of my head, but they are close enough for my questions. basically...since we need a high Qts...would these alignments bring a benefit to my setup, or would sticking w/ voltage drive (and a driver Qts of around 1.5) be better? obviously, if we keep inreasing Zo...we will start to approach the Qm of 10 for the Qts, right? is that *way* too high for open baffle, or well suited? sorry if this sounds ignorant. i've learned enough to be dangerous, but in no means come from an eletronics background.

as a side note on current drive...why would you just pick your favorite triode, and do away with the output transformer? wouldn't the really high plate resistance be good in this case? i'm not talking about the diatones, as it seems like my optimum value might be lower (ie: around 20ohms). or did i just answer my own question? do *most* speakers have a certain point they are looking for, and it is not always the theoretical "perfect" infinte impedence?

thanks for your help and time!

[Nelson Pass]

It all depends. Lowthers and Fostexes in a sealed box get pretty
close to liking a pure current source in a sealed box, and then only
at the lower frequencies unless you want to do a little eq on the
upper mid and top before the amplifier, otherwise a little loading
helps a lot to tame them up there.

While the Hawksford paper extolls the virtues of current source
drive, it also points out the probable need for motional feedback
to control the driver at resonance. I don't find this usually
necessary for the high-efficiency full-range cone types, but these
are a small minority.

Most other cases prefer some finite resistance, depending on
driver and enclosure and environment and taste, and some
would even prefer a negative source impedance. It's an area
where you can easily get some great results with a little
experimentation on your own.

[SY]

as a side note on current drive...why would you just pick your favorite triode, and do away with the output transformer? wouldn't the really high plate resistance be good in this case?

Current capability. You could possibly use a lot of high perveance beam-power tubes in parallel, which could work better (and more efficiently) than, say, a pile of 6080s.

[enochRoot]

Originally posted by Nelson Pass
It all depends. Lowthers and Fostexes in a sealed box get pretty
close to liking a pure current source in a sealed box, and then only
at the lower frequencies unless you want to do a little eq on the
upper mid and top before the amplifier, otherwise a little loading
helps a lot to tame them up there.

While the Hawksford paper extolls the virtues of current source
drive, it also points out the probable need for motional feedback
to control the driver at resonance. I don't find this usually
necessary for the high-efficiency full-range cone types, but these
are a small minority.

Most other cases prefer some finite resistance, depending on
driver and enclosure and environment and taste, and some
would even prefer a negative source impedance. It's an area
where you can easily get some great results with a little
experimentation on your own.

thanks a lot for the reply! i have read over your paper, but have not read the original hawksford paper. the main difference i was seeing (and hence, why i opened this topic) was the lack of a box in my scheme. i will definitely be experimenting with this, but specifically my question is about an open baffle (and this is due to my ignorance of what effect, if any, an open baffle has). seeing as boxes can further transform the overall system Q...would that mean that an open baffle sees no additional interaction, save what is going on between the amp and speaker? or put a better way...is the whole thing a bit easier to fiddle with, as in...if it doesn't sound good at any of the output Zs...it doesn't sound good...period? ie: there is no contribution to the crappy sound based on the "box" (or lack thereof). sorry this is kinda hard to put into words

anyhow...i plan on making these tests w/ ICs...as it is quick, easy, cheap, and within my ability. if i have favorable results...i'll wait until you have built all of the F1s, and release the schematic

when i finally get to play with this...i'll post the results. in a way similar to thorsten's thread...but different in that the Qms of the diatones is so high comparatively.

[enochRoot]

Originally posted by SY


Current capability. You could possibly use a lot of high perveance beam-power tubes in parallel, which could work better (and more efficiently) than, say, a pile of 6080s.

thanks for clearing that up! didn't even think of that! and also...based upon your answer...in doing this...we will get closer to a typical otl type of amplifier, and the output Z will go down significantly, right (ie: closer to 80 ohms than infinite ohms...the 80 being an arbitrarily lower number)? or would the number of tubes needed vs. plate resistance vs. feedback still leave us with a relatively high Z (depending on the "equilibrium" we reach)?

[SY]

You've got a lot of latitude on how to run that output stage. If the source Z is higher than you like, you can stick a resistor across the output- it will be big enough that it won't absorb too much current an be an oven element. Personally, I like the idea of CS drive using an output transformer, so much so that I gave one a shot today in the interest of science.

[enochRoot]

Originally posted by SY
You've got a lot of latitude on how to run that output stage. If the source Z is higher than you like, you can stick a resistor across the output- it will be big enough that it won't absorb too much current an be an oven element. Personally, I like the idea of CS drive using an output transformer, so much so that I gave one a shot today in the interest of science.

really?! cool. how is it working out for ya? is this via the resistor method w/ normal 4,8, or 16 ohm output taps...or did you wind (or have wound) a custom transformer w/ a higher output Z?

[SY]

enoch- No, all that is too hard for my old, feeble brain. I just used a pentode output stage and removed the feedback. The amp was stable and the source Z was high- that's as far as I got today before having to leave my lab and do the child-care thing.

If it measures out OK, I'll beg Pass to let me try it out on the KleinHorns. He's the evangelist who convinced me to give this notion a try.

[enochRoot]

Originally posted by SY
enoch- No, all that is too hard for my old, feeble brain.

yeah...somehow...i find that hard to believe. i *have* read posts of yours elsewhere, and "old" and "feeble" don't come to mind

Originally posted by SY
I just used a pentode output stage and removed the feedback. The amp was stable and the source Z was high

that sounds like a great way to do it! i might have to try that out. i traditionally come from a tube background. this solid state stuff is all new to me

Originally posted by SY
If it measures out OK, I'll beg Pass to let me try it out on the KleinHorns. He's the evangelist who convinced me to give this notion a try.

definitely report back, if that is the case!!