I think it's time to do something different. I want to try Current drive instead of voltage drive. I also would like to combine this with a full range solution avoiding cross over networks. And why not try to combine this with open baffle. All three items have their pro's of their own, so why not combine them?
I am fully aware of the problem with current drive; we will have to consider the driver impedance, and especially the resonance in the low end. But since the problem with OB is the lack of descent bass reproduction, with current drive we will have a bass boost for free. Suitable drivers could be Tang Band W4-1320SJ for a small room/office and W8-1772 for a living room.
Unfortunately I only have some experience with amplifiers and filtering, but not with speakers. So I would like to know if someone have tried the combination of full range open baffel with current Drive. Or has someone raised the output impedance of their amplifiers to extend the low end with OB?
I am fully aware of the problem with current drive; we will have to consider the driver impedance, and especially the resonance in the low end. But since the problem with OB is the lack of descent bass reproduction, with current drive we will have a bass boost for free. Suitable drivers could be Tang Band W4-1320SJ for a small room/office and W8-1772 for a living room.
Unfortunately I only have some experience with amplifiers and filtering, but not with speakers. So I would like to know if someone have tried the combination of full range open baffel with current Drive. Or has someone raised the output impedance of their amplifiers to extend the low end with OB?
For an OB, W8-1808 is probably better suited. There is another active thread discussing driver selection.
The idea of using the OB roll-off to naturally EQ the bump at resonance of a driver with Qm>1 is a very good one (but not new). Careful matching of baffle with driver will be needed.
dave
The idea of using the OB roll-off to naturally EQ the bump at resonance of a driver with Qm>1 is a very good one (but not new). Careful matching of baffle with driver will be needed.
dave
The 8" Tangband drivers have very strong machanical damping, I don't know whether influencing electrical damping by current driver will change so much. Supravox 215-2000 would be nice, but expensive. And current drive won't liberate you from building big baffles, like these U-shaped bending plywood constructions.
Attachments
Why is that? With current drive, a low Qm is preferable. W8-1772 has Qm = 1.02 and W8-1808 has 1.33. Are there other parameters that are more important?
You may already be aware of my article on current driving
full range speakers. If not, check it out at either
FIRST WATT HOME or Pass Laboratories
It works pretty well with OB's, and I demonstrated this with
the new SEAS full range in an OB at BAF two weeks ago.
It does generally need some compensation - some resistance
in parallel with the driver (or in series with a voltage source
amplifier - you don't necessarily have to use a current source
amplifier) to set the bass character and a high frequency
RC network (zobel) to adjust the upper mid and top.
full range speakers. If not, check it out at either
FIRST WATT HOME or Pass Laboratories
It works pretty well with OB's, and I demonstrated this with
the new SEAS full range in an OB at BAF two weeks ago.
It does generally need some compensation - some resistance
in parallel with the driver (or in series with a voltage source
amplifier - you don't necessarily have to use a current source
amplifier) to set the bass character and a high frequency
RC network (zobel) to adjust the upper mid and top.
Measured Qms a bit higher than that.
It is a juggling act. Qms>1 causes a lift in the bass with a current amp, which you try to counter with the appropriate baffle width (then filter as per Nelson).
From my listening, the 1808 is the better sounding driver (and has less ripples in the impedance curve) But i personally wouldn't use either.
dave
Q and compensation
So If I understand you right, it is not a first goal to look for drivers with low mechanical Q. It is, as Nelson says, maybe necessary to compensate for the high resonant impedance somehow. But to my knowledge it is easyer to compensate for a low than a high Q. I would think that for a DIY project, an active compensation (here I am talking about a partly compensation with a series resonant circuit, realized with e.g. a JFET as the active component) would be the way to go. Since I don't have the bank account as packed as Nelson, I would not afford a full range SEAS driver. My choice of TB drivers is also a question of economy.
So If I understand you right, it is not a first goal to look for drivers with low mechanical Q. It is, as Nelson says, maybe necessary to compensate for the high resonant impedance somehow. But to my knowledge it is easyer to compensate for a low than a high Q. I would think that for a DIY project, an active compensation (here I am talking about a partly compensation with a series resonant circuit, realized with e.g. a JFET as the active component) would be the way to go. Since I don't have the bank account as packed as Nelson
, I would not afford a full range SEAS driver. My choice of TB drivers is also a question of economy.
OB and Current Drive
Having read the excellant article of Nelson, http://www.firstwatt.com/pdf/art_cs_amps.pdf, it seems to me that for sealed enclosures, we would anticipate a bass lift that can be taylored by a passive or active compensation network. In addition it may be necessary to soften the high end of the frequency spectrum. But if I understand speaker theory right, the OB should need less compensation at the resonant frequency - but current drive would have a great impact on amplifier overloading (due to the high resonant impedance), or?
Having read the excellant article of Nelson, http://www.firstwatt.com/pdf/art_cs_amps.pdf, it seems to me that for sealed enclosures, we would anticipate a bass lift that can be taylored by a passive or active compensation network. In addition it may be necessary to soften the high end of the frequency spectrum. But if I understand speaker theory right, the OB should need less compensation at the resonant frequency - but current drive would have a great impact on amplifier overloading (due to the high resonant impedance), or?
I'm playing right now with TB 1772's running OB and driven by the First Watt F-1. Very pleasing!! Several guys were in tonight and just going wild over how it sounds.
In my case I made some smaller baffles that acoustically make it down to a measured 200hz. Fooling around right now with bass coming from Avantgarde woofers, which nicely make up high enough (being more woofers than subwoofers!). Adding in a HP crossover doesn't seem to change performance as much as I'd have though at the levels I usually listen at. Higher SPL would certainly differ.
The F-1 gives a very nice big picture with great sound stage all the way across. Super nice - Very dynamic!
Likely the 1808 is a better idea OB. These are just what I happen to have. I've gotten lots of complements on the result!
Mark
In my case I made some smaller baffles that acoustically make it down to a measured 200hz. Fooling around right now with bass coming from Avantgarde woofers, which nicely make up high enough (being more woofers than subwoofers!). Adding in a HP crossover doesn't seem to change performance as much as I'd have though at the levels I usually listen at. Higher SPL would certainly differ.
The F-1 gives a very nice big picture with great sound stage all the way across. Super nice - Very dynamic!
Likely the 1808 is a better idea OB. These are just what I happen to have. I've gotten lots of complements on the result!
Mark
First & foremost you have to keep in mind that dipole rolloff can be considered compensation for the lift in frequency response from the interaction of current drive and rising response. If you can get one to perfectly counter the other you have an ideal situation.
In a voltage drive OB we look to use high Qt to compensate for the dipole rolloff... in a current drive OB you look at the Qm.
Also if one surveys the Qm of available drivers, <2 is actually a low Q.
dave
I doubt that a fullrange small-size OB-solution (except for desktops) is possible because the displacement limits compensation of the dipole rolloff. But you could try what can be done with drivers that are already OB-suitable with voltage drive driven by a current drive amp. Such drivers (e.g. from Sica) are not expensive, could be fun to try.
The best compensation is no compensation...in terms of OB. In other words, the bigger the baffle, the less compensation is needed, which makes it easier to get the correct balance. On the other hand, the bigger the baffle, the lower the WAF , (so you have to compensate elsewhere)
Most of us here would have a hard time to accommodate baffle sizes that Nelson utilizes, so we must turn to compensation. It is interesting to see how Mr. Linkwitz does it via shelving filter, which compensates at the rate of 6 db/oct, from about 10 Hz up to 300 Hz (this would depend on the width of the baffle). But then it makes it necessary to use biamping, going active..more complicated...so..the quest continues
, (so you have to compensate elsewhere) Most of us here would have a hard time to accommodate baffle sizes that Nelson utilizes, so we must turn to compensation. It is interesting to see how Mr. Linkwitz does it via shelving filter, which compensates at the rate of 6 db/oct, from about 10 Hz up to 300 Hz (this would depend on the width of the baffle). But then it makes it necessary to use biamping, going active..more complicated...so..the quest continues
Last edited:
I find that the issue of compensation appears after the
baffle is built and you take a listen. Often it is independent
of baffle size because the compensation we are talking about
is up in the upper frequencies - 3 KHz and beyond. Your
ordinary baffle step is in a much lower range, and I don't
seem to run into that much. It looks to me like it has more
to do with the shape of the baffle than the size - the edge
effect when you have two identical waves cancelling is
different than having the baffle surface turn at a right angle.
baffle is built and you take a listen. Often it is independent
of baffle size because the compensation we are talking about
is up in the upper frequencies - 3 KHz and beyond. Your
ordinary baffle step is in a much lower range, and I don't
seem to run into that much. It looks to me like it has more
to do with the shape of the baffle than the size - the edge
effect when you have two identical waves cancelling is
different than having the baffle surface turn at a right angle.
Interesting. I was mostly concerned with the compensation in the bass range 300 hz and downwards. I don't have an idea about the >3Khz compensation, other than to notice that midrange gets more pronounced with on the Open baffle. Then, I scratch my head and wonder if Linkwitz 3 Khz notch is something we should look for (or, other means to shape a frequency response). By the way, sorry for going almost off topic...
Compensation
I would think that a possible compensation would be easier to do actively; it is a much cheaper solution, and it is easier to modify than a passive solution. As Pass says, this can be done after the baffle is built. I would guess that a rate of 6 db/oct or maybe less should be sufficient. I don't think a notch filter would be the way to go.
I would think that a possible compensation would be easier to do actively; it is a much cheaper solution, and it is easier to modify than a passive solution. As Pass says, this can be done after the baffle is built. I would guess that a rate of 6 db/oct or maybe less should be sufficient. I don't think a notch filter would be the way to go.
Many years ago I built open baffle design with 3x 10” seas driver Qts around 0,28 if I don’t remember wrong. With 15 Ohm in series on each driver, crazy thing to do but it worked very well. Fastest open baffle design I have heard. No active compensation, decent SPL/W quick bas with good low frequency performance and you could play rely loud. We play with a 200W amplifier and also 50W and both worked well.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.