Like I said, one way or another, you've got to reduce the drive at resonance. Now if you're going the speaker-level network route, you'll divert current, but how much (in relation to the impedance minimum or any other reference point) will depend on how much you're trying to peak up the bass. Too little diversion and you're back at compliance problems and boom. Too much and the bass dries up and goes away. I don't think you can have a strict rule of thumb- it's going to depend on the driver and enclosure.
In a sense, this is a way of manipulating EQ.
In a sense, this is a way of manipulating EQ.
In that case then this article could be of interest:
http://www.pearl-hifi.com/06_Lit_Archive/07_Misc_Downloads/Critical_LS_Damping.pdf
It contains a discussion about speaker damping and describes a bit about amplifiers with variable damping factor, they mention damping factors of 0.125 which is the same as an output impedance of 64 ohm, not quite a transconductance amplifier but close.
Regards Hans
Steve Eddy said:
It's pretty simple- with a pure current drive, most speakers will peak at resonance. So the drive near resonance has to be decreased by one means or another. Less voltage, higher Z -> lower current, lower power. I'm not sure what else you're looking for.
Now, there's probably some bizarro driver out there with a smashingly low Qms and that could change the calculus here. I just haven't seen it yet.
SY said:
Yes, less voltage and higher impedance equates to less current and lower power, if wer're talking about a voltage source.
But with a curent source, which is the context here, higher impedance means higher voltage and greater power.
In order for Fs to be "the place where you'll least need power," the impedance at Fs would need to be reduced to below the driver's minimum impedance, which for a typical 8 ohm driver is around 6 or 7 ohms and occurs at some point above Fs.
So are you saying that the network you speak of to reduce the response peak at Fs will present an impedance to the current source which is lower than the driver's minimum impedance?
I'm looking for the answer as to exactly how Fs becomes "the place where you'll least need power."
se
I'm sort of running out of ways to explain it. If you don't want the bass to peak near resonance, you can reduce the drive there relative to other frequencies by EQ or velocity feedback. But the impedance is still unchanged. That's lower power.
If you use a speaker-level conjugate network, that may sometimes be the case. In those cases, you'll need to drop the impedance below that of the driver's minimum. And that's OK, since P is proportional to Z at a constant current. So power is reduced using that technique, too, both total power and (even more so) power to the speaker. CS amps are quite happy running into low impedances- that's why they're also (paradoxically) a good choice for high-power ribbon designs.
If you use a speaker-level conjugate network, that may sometimes be the case. In those cases, you'll need to drop the impedance below that of the driver's minimum. And that's OK, since P is proportional to Z at a constant current. So power is reduced using that technique, too, both total power and (even more so) power to the speaker. CS amps are quite happy running into low impedances- that's why they're also (paradoxically) a good choice for high-power ribbon designs.
SY said:
Well sure, if the load remains the same and you reduce the current, that's lower power than if the current remains the same as well.
But even at that I don't see that Fs will necessarily be the "place where you'll least need power."
Would you necessarily need to?
Oh, and while you're here, any idea where I might find an M9.5x1 hand tap?
se
Sure SY, 'tho I hope you don't mind if we first hook 'em to some old car stereo speaks I've got to see if we launch the cones....
I guess I'm not too sportin'
Nelson has done the research on my drivers, right?
Interestingly, I have been collecting the parts for a Son o Zen for a couple of years now- looks like I was on the right track!
But toooobs - now that would be cool!
I guess I'm not too sportin'
Nelson has done the research on my drivers, right?
Interestingly, I have been collecting the parts for a Son o Zen for a couple of years now- looks like I was on the right track!
But toooobs - now that would be cool!
Let me get this straight
The theory sugest that we used current source amp for single drivers like fostex
So what i should do is drive my fostex with a pentode output stage say el34 wire in pentode. Output trans of 3K with no form of feedback at all? With a very high output impedance around 30-50ohms?
Is this right. Any other requirements?
The theory sugest that we used current source amp for single drivers like fostex
So what i should do is drive my fostex with a pentode output stage say el34 wire in pentode. Output trans of 3K with no form of feedback at all? With a very high output impedance around 30-50ohms?
Is this right. Any other requirements?
I'd be interested in seeing your design SY.SY said:
I'm looking at building a new pair of speakers for when I move home, and I want to run a a pair of 15" JBL K140s in parallel in a large sealed enclosure ~ 40 - 500Hz, so a transconductance amp is on the list to try to get some bottom end (before I cross to subs and mid/high horns above).
NickC said:
My first proof-of-concept was exactly like this. I took an ST70 with EL34s and a SYclotron driver, converted to pentode (regulated screens at 400V), removed the loop feedback and... voila! With the stock OPTs, turns ratio to the 16 ohm winding is about 16.5:1. I measured about 125 ohms source Z at three spot frequencies (50, 1K, 10K). It gets pretty ugly in the deep bass, but that's the hazard of using tubes and OPTs.
The "real" version will be similar, though using different tubes and OPT. High perveance sweep tubes will probably yield a less-high output Z (though a lower OPT turns ratio may more-than-compensate), but if one is to believe Pass, 47 ohms is adequate. If I can't achieve that open-loop, the current feedback is an option.
Steve Eddy said:
I gave you an answer, but I didn't give you an understanding. My bad.
Would you necessarily need to?
Oh, and while you're here, any idea where I might find an M9.5x1 hand tap?
se
Well, if you read my earlier responses to you, you'll note that I believe that it wouldn't always be necessary. As for the hand tap, be careful or you'll go blind. Blind men with hairy palms have trouble selling pencils.
Doesn't McMaster-Carr have one?
Hi,
Steve,
I see what you're getting at but it would be easier for all if you'd care to give an example of what you mean.
I know I could but it's not a very common situation with tube amps in general, assuming I'm reading you correctly in first place.
Using an impedance correction network with a FR speaker is neither a sin nor a shame.
Quite to the contrary; it will make an already decent speaker even easier to drive and will even out an otherwise peaky response, especially when losing a dB or two of efficiency is not an issue.
If you know that tube amps tend to increase their output power with rising load impedances and most of the time the peaks occur in the bass region and exhibit a rising Z it should be clear that there's more than one direction in which one could control it IMO.
Idem ditto here.
Cheers,
Steve,
I see what you're getting at but it would be easier for all if you'd care to give an example of what you mean.
I know I could but it's not a very common situation with tube amps in general, assuming I'm reading you correctly in first place.
Using an impedance correction network with a FR speaker is neither a sin nor a shame.
Quite to the contrary; it will make an already decent speaker even easier to drive and will even out an otherwise peaky response, especially when losing a dB or two of efficiency is not an issue.
If you know that tube amps tend to increase their output power with rising load impedances and most of the time the peaks occur in the bass region and exhibit a rising Z it should be clear that there's more than one direction in which one could control it IMO.
Idem ditto here.
Cheers,
fdegrove said:
Well, just take a look at the impedance plot of a fullrange driver.
Typically you'll see that at Fs, the driver's impedance ranges from around 50 ohms to well over 100 ohms.
All else being equal, when driven from a current source, Fs represents the point at which you'll need power the most.
If you use a speaker level network to reduce the peak in response at Fs due to the high output impedance of the current source, you won't necessarily end up with Fs being the point of lowest impedance which subsequently means that Fs won't necessarily be the point at which you least need power.
No, not in any absolute terms.
I was simply saying that many find the simplicity of a direct connection between the amplifier's output and the voice coil attractive in a Zen sort of way.
Sure. Just that not everyone will like the result. Not everyone looks at things in purely technical terms. Some have more philisophical or asthetic bents, not to mention varied subjective tates and preferences.
se
Hi,
Yep....
Just that you won't be able to achieve that with a transconductance amp using penthodes and an OPT IMO unless perhaps in the mildest of cases or an extraordinary amount of power.
As an example:
The famous VOT speaker using an Altec driver (416 something IIRC) exhibits a nasty peak somewhere in the upper bass area that you won't be able to control with an ordinary amp, tube or transistor except for an OTL amp.
This is because the OTL can deliver an exceptional amount of current_ for a tube amp_ an amount that will increase with a rising impedance which so happens is just what the speaker needs to behave.
By speaker I mean the combination of the Altec unit and the loading of the VOT cabinet combined which is of course something that needs to be considered.
Mind you, the effect of the amp is not subtle in this particular case and certainly is very audible.
I could dig up the details on what causes what effect and so on but I think I have described the general idea correctly.
Certainly.
Then not everyone wants their Hi-Fi to be High in Fidelity either....
The question that keeps hanging in my mind is now: how will a xformer coupled penthode transconductance amp convey its power into a given load?
Will the OPT hamper the performance in any way?
I vaguely recall this could be problematic but I could well be wrong about the isue. It's just been too long ago since I looked into this...
Cheers,
Yep....
Just that you won't be able to achieve that with a transconductance amp using penthodes and an OPT IMO unless perhaps in the mildest of cases or an extraordinary amount of power.
As an example:
The famous VOT speaker using an Altec driver (416 something IIRC) exhibits a nasty peak somewhere in the upper bass area that you won't be able to control with an ordinary amp, tube or transistor except for an OTL amp.
This is because the OTL can deliver an exceptional amount of current_ for a tube amp_ an amount that will increase with a rising impedance which so happens is just what the speaker needs to behave.
By speaker I mean the combination of the Altec unit and the loading of the VOT cabinet combined which is of course something that needs to be considered.
Mind you, the effect of the amp is not subtle in this particular case and certainly is very audible.
I could dig up the details on what causes what effect and so on but I think I have described the general idea correctly.
Certainly.
Then not everyone wants their Hi-Fi to be High in Fidelity either....
The question that keeps hanging in my mind is now: how will a xformer coupled penthode transconductance amp convey its power into a given load?
Will the OPT hamper the performance in any way?
I vaguely recall this could be problematic but I could well be wrong about the isue. It's just been too long ago since I looked into this...
Cheers,
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.