I've read a couple of Thorstens posts mentioning how current drive would be a good thing to use with the "gainclone" circuit. I was wondering if anyone has experimented with this and what their results were, either way I think I'll give it a try and see how it sounds.
Only thing is, how would we apply this to the inverted LM3875 type circuit that everyone seems to be using?
One last thing, I read mentioned that it works well for midrange and treble, but why not bass exactly? Just curious, because I would be using it for 150Hz-20kHz duties, wondering if in theory it would work well for those frequencies...
Only thing is, how would we apply this to the inverted LM3875 type circuit that everyone seems to be using?
One last thing, I read mentioned that it works well for midrange and treble, but why not bass exactly? Just curious, because I would be using it for 150Hz-20kHz duties, wondering if in theory it would work well for those frequencies...
Konnichiwa,
Well, you read of my ones....
The usual way I suspect. I PERSONALLY would fit a pair of 100k Resistors in series as the conventional feedback resistor in the IGC circuit and place a 1uF Cap to ground from the junction between the resistors.
Then place a 1R resistor in the "ground" lead of the Speaker driver. From the speakers groundside connection or the resistors top side connection a 24k Resistor connects to the inverting input. This should increase the output resistance of the Amplifier markedly to at least something close to current drive.
By switching to current drive you do two fundamental things.
1) You ensure that the force applied to the radiation surface is now strictly proportional to the signal.
2) You remove all electrical damping on the Drivers Cone/Suspension oscillating system (it is basically a spring/mass system), meaning driver will show excessive output at the resonance.
Hence current drive should be used only for drivers with either very low Qm or considerably above or below the drivers fundamental resonance. Another option is to use various methodes of mechanical damping on the cone with suitable arrangements in the enclosure.
Typhcial cone drivers used for Midrange these days tend to have very low fundamental resonances compared to the oprtating range and thusly can be operated quite well using this system. Please note that a highpass should preceede the Amplifier to prevent the undamped fundamental resonance from being (unduly) excited.
Sayonara
JoeBob said:
Well, you read of my ones....
JoeBob said:
The usual way I suspect. I PERSONALLY would fit a pair of 100k Resistors in series as the conventional feedback resistor in the IGC circuit and place a 1uF Cap to ground from the junction between the resistors.
Then place a 1R resistor in the "ground" lead of the Speaker driver. From the speakers groundside connection or the resistors top side connection a 24k Resistor connects to the inverting input. This should increase the output resistance of the Amplifier markedly to at least something close to current drive.
JoeBob said:
By switching to current drive you do two fundamental things.
1) You ensure that the force applied to the radiation surface is now strictly proportional to the signal.
2) You remove all electrical damping on the Drivers Cone/Suspension oscillating system (it is basically a spring/mass system), meaning driver will show excessive output at the resonance.
Hence current drive should be used only for drivers with either very low Qm or considerably above or below the drivers fundamental resonance. Another option is to use various methodes of mechanical damping on the cone with suitable arrangements in the enclosure.
Typhcial cone drivers used for Midrange these days tend to have very low fundamental resonances compared to the oprtating range and thusly can be operated quite well using this system. Please note that a highpass should preceede the Amplifier to prevent the undamped fundamental resonance from being (unduly) excited.
Sayonara
To throw another question in there quickly...
You meantioned in the gainclones TNG thread that if it were to be used with a woofer then there would have to be substantial mechanical dampening. Couldn't this be avoided by the use of a motion feedback system, be it whatever method you choose?
Thanks again.
You meantioned in the gainclones TNG thread that if it were to be used with a woofer then there would have to be substantial mechanical dampening. Couldn't this be avoided by the use of a motion feedback system, be it whatever method you choose?
Thanks again.
Konnichiwa,
Actually, it is there to make sure that the feedback applied is only DC. And I forgot to mention that the resistor from the 1R resistor needs to connect via a capacitor of it's own, or you could simply mobe the DC Blocker to the inverting Input and connect the input Resistor (10k) and Current Feedback Resistor (24k) ahead of the DC blockeing capacitor.
Sure, but a Motional Feedback System that ACTUALLY WORKS is not a beginners design job nor for the fainthearted.
If you where to tell me which driver you intend to use and such things I might be able to be a little more specific.
BTW, by sizing the 1uF Capacitor apropriatly you can re-introduce Voltage Feedbac and thus make sure the Amplifer transits at low frequencies near the drivers resonance back into "voltage output" mode. This would be usefull with a Driver operated via a suitable highpass but not quite far enough from the Fundamentals for comfort without steep filters....
There are many factors here and they need to be understood. What you need to do is to design an Amplifier to suit a given specific Driver & Enclosure....
Sayonara
JoeBob said:
Actually, it is there to make sure that the feedback applied is only DC. And I forgot to mention that the resistor from the 1R resistor needs to connect via a capacitor of it's own, or you could simply mobe the DC Blocker to the inverting Input and connect the input Resistor (10k) and Current Feedback Resistor (24k) ahead of the DC blockeing capacitor.
JoeBob said:
Sure, but a Motional Feedback System that ACTUALLY WORKS is not a beginners design job nor for the fainthearted.
If you where to tell me which driver you intend to use and such things I might be able to be a little more specific.
BTW, by sizing the 1uF Capacitor apropriatly you can re-introduce Voltage Feedbac and thus make sure the Amplifer transits at low frequencies near the drivers resonance back into "voltage output" mode. This would be usefull with a Driver operated via a suitable highpass but not quite far enough from the Fundamentals for comfort without steep filters....
There are many factors here and they need to be understood. What you need to do is to design an Amplifier to suit a given specific Driver & Enclosure....
Sayonara
I was thinking of using it on a vifa p17wj and a vifa D25 setup. Only thing is I would be using the P17 down to nothing, but I guess I could use a 2nd P17 for the lower frequencies, or a bigger woofer even, it's just that I have these on hand. These (well the P17s) would be in a dipole, because I like dipoles.
Konnichiwa,
The P17WJ has a usefully low Qm. With pure current drive the Drivers Qt becomes Qm so that in an infinite baffle the driver would show around a 4db Peak at 40Hz. In a sufficient large baffle this would allow the Driver to be flat into the decent LF region, try 2' X 4'.
A single P17WJ on such a baffle should support a sub 50Hz in room response. Of course powerhandling at LF is very limited as is the maximum SPL. I personally would probably run 3pcs of P17WJ of which two are pure "woofers" with a suitable X-Over point and a first order Lowpass on a narrower baffle (folded back wings unless very deep give little improvement in LF response) and a 1st order highpass on the "woofer/midrange". A 200Hz X-Over seems suitable for a baffle of 10" width and a suitable hight.
Give each drive it's own Amplifier with suitable Equalisation of the woofer/midrange Response (due to the cancellation of the drivers inductance it will now be rising). Use a subtractive X-Over for Midrange/Tweeter X-Over with a 3rd Order HP on the Tweeter and the woofers drive formed from the subtracted signal of Tweeters Signal and the Input Signal and slope the baffle for time aligning the drivers. The X-Over can be formed directly around the Chip Amplifiers without recourse to external X-Over circuits.
The lowpass Filters for the Woofers can be placed in the feedback loop, so all you would need is a single buffer and four modified Gainclones.
Anyway, that's what I'd do IF I where building something like this.
Sayonara
JoeBob said:
The P17WJ has a usefully low Qm. With pure current drive the Drivers Qt becomes Qm so that in an infinite baffle the driver would show around a 4db Peak at 40Hz. In a sufficient large baffle this would allow the Driver to be flat into the decent LF region, try 2' X 4'.
A single P17WJ on such a baffle should support a sub 50Hz in room response. Of course powerhandling at LF is very limited as is the maximum SPL. I personally would probably run 3pcs of P17WJ of which two are pure "woofers" with a suitable X-Over point and a first order Lowpass on a narrower baffle (folded back wings unless very deep give little improvement in LF response) and a 1st order highpass on the "woofer/midrange". A 200Hz X-Over seems suitable for a baffle of 10" width and a suitable hight.
Give each drive it's own Amplifier with suitable Equalisation of the woofer/midrange Response (due to the cancellation of the drivers inductance it will now be rising). Use a subtractive X-Over for Midrange/Tweeter X-Over with a 3rd Order HP on the Tweeter and the woofers drive formed from the subtracted signal of Tweeters Signal and the Input Signal and slope the baffle for time aligning the drivers. The X-Over can be formed directly around the Chip Amplifiers without recourse to external X-Over circuits.
The lowpass Filters for the Woofers can be placed in the feedback loop, so all you would need is a single buffer and four modified Gainclones.
Anyway, that's what I'd do IF I where building something like this.
Sayonara
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.