I was playing around on the weekend with my dual-woofer speaker system. I had the bottom speaker disconnected while doing some frequency tests. I shorted the bottom speakers terminals together so it would not flap around so much and screw up some results I was testing for. That led me to think about changing the resistance between the terminals on my unconnected speaker. Which I did, and I found I could play around with the system frequency response at the low end by doing this. Of course, I started adding caps, etc. ... you get the picture.
I could not find any similar posts to this. Has anyone tried this or played around with this? No doubt it is common, but it was new for me.
Alvaius
I could not find any similar posts to this. Has anyone tried this or played around with this? No doubt it is common, but it was new for me.
Alvaius
Re: Using a second speaker as a passive radiator/controlling with circuit on terminals
I just want to understand one basic part of the picture. The unconnected speaker was reconnected to the amp before the resistors and caps you put on it had any effect, right?
I know it seems obvious that it would be reconnected, but if the speaker is acting as an unconnected passive radiator, I imagine it is sending current through it's wires as it moves, and who knows-passive components could possibly, concievably, maybe, perhaps have an effect on the frequency response.
Just want to make sure.
alvaius said:
I just want to understand one basic part of the picture. The unconnected speaker was reconnected to the amp before the resistors and caps you put on it had any effect, right?
I know it seems obvious that it would be reconnected, but if the speaker is acting as an unconnected passive radiator, I imagine it is sending current through it's wires as it moves, and who knows-passive components could possibly, concievably, maybe, perhaps have an effect on the frequency response.
Just want to make sure.
From the Shiva White Paper :
But the most important benefit for dual voice coils is flexibility for the T/S parameters. One can actually
“dial in” a desired Qts of the driver, by resistively loading one voice coil and actively driving the other.
This configuration, which we call Resistively Damped Operation (RDO), uses the second (undriven) voice
coil as an electromagnetic brake. In essence, the resistance across the second coil will determine how
strong the brake is. The smaller the resistance, the stronger the brake.
RDO affects the Qts of the driver by decreasing the Qms of the driver. The RDO brake acts to damp cone
motion, as if the suspension was considerably stiffer. However, as it’s an electromagnetic brake, the Fs is
NOT affected (as it would raise if the suspension components – the surround and spider – were stiffened).
But the most important benefit for dual voice coils is flexibility for the T/S parameters. One can actually
“dial in” a desired Qts of the driver, by resistively loading one voice coil and actively driving the other.
This configuration, which we call Resistively Damped Operation (RDO), uses the second (undriven) voice
coil as an electromagnetic brake. In essence, the resistance across the second coil will determine how
strong the brake is. The smaller the resistance, the stronger the brake.
RDO affects the Qts of the driver by decreasing the Qms of the driver. The RDO brake acts to damp cone
motion, as if the suspension was considerably stiffer. However, as it’s an electromagnetic brake, the Fs is
NOT affected (as it would raise if the suspension components – the surround and spider – were stiffened).
For clarification, the second speaker is NOT connected to the amplifier. It is disconnected, but between its terminals, I started putting in resistors and caps, and soon inductors..... sort of a tunable passive radiator.
On one post of motion feedback, I think the issue may be delay between the main speaker and the passive unit, but heh, you never know.
Alvaius
On one post of motion feedback, I think the issue may be delay between the main speaker and the passive unit, but heh, you never know.
Alvaius
This is a fairly old trick that gets rediscovered annually.
It works great. First, because the two drivers are identical,
it gets tuned perfectly no matter what you do.
From my experience, variable resistance is the way to go.
Values as high as 100 ohms or so have a fairly dramatic
effect on the damping of the "passive" radiator, and you
can tune it to taste.
It works great. First, because the two drivers are identical,
it gets tuned perfectly no matter what you do.
From my experience, variable resistance is the way to go.
Values as high as 100 ohms or so have a fairly dramatic
effect on the damping of the "passive" radiator, and you
can tune it to taste.
There's a good article in an old TAA titled, "The Not Quite Passive Radiator". The author experimented with several passive elements between the terminals.
Also, I believe one of Sonus Faber's speakers employed a similar technique and obtained very good low frequency response from a monitor-sized speaker.
Also, I believe one of Sonus Faber's speakers employed a similar technique and obtained very good low frequency response from a monitor-sized speaker.
Are you considering this passive radiator design or is this something you were just playing with? I noticed your other post asking about whether you can only get a 3-db baffle step. Passive radiator would be a way, but you wouldn't get response up to 400 Hz, I think. You almost done with that cross-over? I ain't gettin' any younger.
i dont remember where i learnt this but we used to do the same trick to DVC woofers. we would put a 100ohm pot between one set of terminals and dial the Qts. if i remember right the differende in Qts was about 10% between 0 ohms and open.
Mr. Pass, you seem as old as me although a lot more knowledgeable and a lot less senile.
Mr. Pass, you seem as old as me
although a lot more knowledgeable and a lot less senile.I thought of this a couple of days ago but then I found this post so it was not a new idea. I remember that someone told me many years ago that you should short out the terminals when transporting speakers to prevent damage. Then a couple of days ago I thought of this and using a variable resistance to tune a second speaker as a passive radiator. Oh well, I see this post is 2002 so it might be good to put some life into it.
The myth of speakers needing to be shorted out to prevent damage is indeed older than this zombie thread.
The myth is false, as can easily be observed by the fact that millions of raw speakers are shipped from manufacturers without being shorted with no damage, and millions of drivers in sound systems are transported daily without shorting or damage.
Sure one can do this - short out the passive radiator - but it is a very expensive way to do something of questionable usefulness. Normally a passive radiator does not want to be highly damped, and if it did, there are lots of cheaper ways to do this.
Fun to play with, but no real significant benefits that I have seen.
Fun to play with, but no real significant benefits that I have seen.
There is nothing terribly wrong with shorting a driver for transport, but it usually wouldn't be necessary unless it were very compliant and very fragile.
Adding resistance to a woofer used as a PR wouldn't tune the frequency of resonance, only the damping. You would need to add mass to loweer the resonance and vice versa. As Earl said, adding damping to a passive radiator doesn't accomplish much - other than reducing output and "control" on the active driver. The RDO thing I wrote about years ago is a valid way to play with parameters on a dual voice coil driver, but it is much better (more efficient) to just use both coils and equalize...
That is only useful if you have a suboptimal design to start with.
Driving a "passive" radiator is described here:
https://www.google.com/patents/US39...a=X&ei=OnClUZqcAs-64AOAvoHgDg&ved=0CDcQ6AEwAA
The version in the patent was used in some Speakerlab systems IIRC.
A minor variation is used in most Vandersteen systems.
Driving a "passive" radiator is described here:
https://www.google.com/patents/US39...a=X&ei=OnClUZqcAs-64AOAvoHgDg&ved=0CDcQ6AEwAA
The version in the patent was used in some Speakerlab systems IIRC.
A minor variation is used in most Vandersteen systems.
i disagree, i think it can be handy. For example visaton gf200. Using a single coil qts is 0.6 or slightly more, using both coils its a more usual 0.32. Large vas gives a large vented box F3 slightly above the Fs of 30hz. However, using a single coil it does very well in a moderate size sealed box. Qts is a tad high, but it could be dialled in to an optimal point for sealed use. The killer is the loss in efficiency, and in the other configuration it presents a 2ohm load...series wiring isnt much different to parallel IIRC. I still want to build with the GF200, and would probably go to the effort of tuning the other coil. Using reactors is something i thought to try, but wasnt aware exactly what it would do. Does 'compliance scaling' work on a dual VC driver like this? E.g. 1000uF + tunable R on the damping VC?
Been looking into experimenting with passive radiator tuning and am writing paper on it - I realise this thread is pretty damn old now but would love to do some experimentation of my own with this concept; what sort of source did you have on the inactive driver if not connected to an amplifier? Or are your caps and resistors literally preventing current that's been caused by the cone moving across the coil hence preventing the cone movement?
I've always sort of pictured in my head that a tiny tiny bit of DC could sort of hold a pseudo-passive radiator in place and act to damp its motion (although we all know this is probably a great way to fry your voice coil)
Interested to hear any thoughts, I'm relatively knew to loudspeaker theory, any academic paper recommendations anyone knows of related to this concept would be much appreciated!
Thanks
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.