It's not exactly a casual conversation. But I know what you mean. 😉
Yes, the idea of current-drive is that you have a transconductance amplifier. Voltage in and a relative current out. Double the input voltage and you get double the current output.
Some years ago I did a design what many should be able to do, and I have that amp here right now and anybody is welcome to come and have a listen to it.
Link:
Joe Rasmussen "Trans-Amp" - 40 Watt Transconductance "Current Amplifier"
You do realise that I am not promoting current-drive.
Discussing the role of current and how the amplifier delivers it, does not mean that the discussion is about current-drive.
What I am concerned about is current delivery into the driver. Hope you can see the difference?
I do see a role for current-drive in active speakers, where you can control all the difficult design parameters and the end user does not have to be concerned about it.
This is of course you quoting Purifi. And I have "experiemented" too and can attest that even 20dB improvement has been achieved. That is 10:1 in amplitude/magnitude terms. That is wow! and cannot be ignored.
My position, which is for some reason not been understood here (social media is broken, but not when talking fac2face with scientists, engineers and mathematicians actually works brilliantly) is that you don't need to go down the current-drive route because this is not going to take over the world. Current-drive is only going to be a niche and nothing more.
You can lower distortion by 15dB with voltage-drive by using techniques that works.
I have been trying to explain this to the others here, but it seems to go over their heads. Very odd.
One of them is to EQ the current of the amplifier (voltage) and also do what current-drive does, get better isolation between the amplifier and the speaker by designing the crossover to achieve that. The idea that an amplifier has any control is simply not backed up by the physics. When a voltage source is used and you still want much lower distortion, then the speaker designer must design control into his design and not reply on the amplifier. The amplifier should not be relied upon for control, but we do need some control over the current.
My line?
YOU DO NOT NEED CURRENT-DRIVE.
That is good news!
Yes, good drivers use a lot of copper. Have you seen the insides of a Purifi driver. No shortage (no pun intended) of copper. 😉
I want drivers to have the lowest possible inductance (copper rings) and that this be excursion stable, as the cone moves in and out. Even the standard EL-6 with MFC drivers is top drawer in this respects. Then you can take that to the next stage using Purifi ULD drivers. But you need deeper pockets.
Again, this is quoting Purifi.
We are incredibly sensitive to distortion in the mid-band. We can certainly hear down to -80dB which is 0.05% and even lower than that. At LF the tables are changed, we are not so sensitive to distortion, but we are sensitive to low frequency dynamics and that hardly ever gets mentioned.
Cheers, Joe
Yes, perhaps you have become so sensitized that you cannot recognize when someone is trying to offer information that might support your position. You focused on "current drive", when I saw "effect on the impedance".
Yes which is why in my post it was in quotation marks and italicised with a source attributed underneath, preceded by "Purifi actually say..."
Could I have made it anymore obvious?
I am surprised by your aggro, my reply was very measured. What makes you feel so insulted?
Hello,
You own the name, please do bring out the Mini Elsinores design.
I like the idea of a high end bench size speaker.
First blush I like a high pass into the mid-woofers in the range of 100 to 125Hz (you know in an effort to keep the large slow moving frequencies out of the voice coil). mid to tweeter about 2500 to 3000 ish, you know depending on what works.
Then hide a 12 inch Peerless XXLS woofer under the bench. The first effort will be active. This driver is discontinued, Digikey has a few still in stock. There are 6 on my shelf.
It is too late to run the test sweep frequencies tonight. That will be in the morning. I will be glad to share the results.
Thanks DT
Last edited:
hi Joe,
This is where I am not quite following. I have seen that you flatten the box impedance by adding LCR’s across the amp. Why would that linearise the current through the driver?
cheers,
Lars
That's the question alright. I am having this conversation right now with somebody else and I want to see how that goes. It is with somebody who is a known to be a presence here, but I would not like to ID him.
But I realise that I am sticking my neck out, oh boy!
But sometimes you have to be brave. The worst that can happen if you fail. lol
So many people have heard this simple trick where you have a typical 2-Way speaker system, 8 Ohm, usually a peak Z somewhere in the critical midrange, sometimes up to near 40 Ohm. I tell them, since it is rated at 8 Ohm, get a decent cheap 8R2 10W resistor and put it across the speaker terminals. That should make it into a 4 Ohm speaker, so a safe test. I suggested to a friend who has Revel M126Be speakers and he promptly at lunch time went down to the local Jaycar electronics store (I used to consult for them) and buy the resistors. Later in the afternoon he called me back and was very excited what he heard.
That 8R parallel resistor is what I call Poor Man's Current EQ.
This is the Z plot of the Revel speakers:
Note that midrange Z emphasis.
Let me tell you a story involving un unusual Hypex amplifier.
At ETF17 held at Tisvildeleje (I was there when ETF was founded in 2002 near Aarhus, with Kurt Steffensen). I remember Menno Vanderveen having this modified Hypex NCore amplifier and it had a switch where he could change the output impedance away from stock near 0 Ohm. The other two settings were 3 Ohm and 18 Ohm.
He then proceeded to connect these speakers, which I recall belongs to Jan Didden.
Looks like Audionote copies?
These will have that typical high Z in the midrange for sure. Anyway, he connected the amp up for us as we had a number of people in the room. Everybody preferred that 18 Ohm setting. I pointed out to Menno that it would have been just affective, and maybe even more so, if he had left the amp on zero Ohm and used an 8R resistor. It would not affect the bass and the mids would sound more balanced as the FR would not be modulated. But the improvement they heard in that room was far more that FR related.
Menno then asked the question, does less damping factor sound better? Personally, I don't think it was about damping factor.
I did a discussion paper and sent it to Menno and also to a guy called Hans van Maanen (of Temporal Coherence) because he employs current EQ in his active speakers, because he wanted the current phase angle to be flat with his Class AB amplifiers. This way, when you had the Class A-B transition, the voltage and current would line up, he had observed spurious behaviour, high order harmonics being fixed. He also rightly pointed out that feedback is a voltage correction, not current. Anyway, adding those passive EQ improved the sound, but I think there is more to it. But as I pointed out to Menno, he was using a Class D amplifier, not AB. So there has to be more to it.
It is clear that if the driver delivers a perfect load (a resistor) then a parallel resistor would do nothing. So whatever is going on, it must be driver related.
Anyway, Hans was very dismissive and even a bit blunt. So I remember walking through the door at ETF18 in Normandy, and Menno saw me from across the room and went for me to come and sit down with him. He was worried whether I was upset by Hans' reaction and that he wanted me to know that he was very much interested and that I had indeed made him take notice. I said I took Hans' reaction on the chin.
At this point I said to Menno, and I had prepared the exact words used:
"I am convinced that we are listening to the current of the amplifier and not its voltage."
After a short pause he told be that he agreed with me and added "the good news is that we should be able to prove it."
The series 18 Ohm is not so much of a surprise, So why does a parallel resistor?
I don't know if you know Richard "Dick" Marsh? We are still in touch, he has a daughter here in Sydney. He has also told me that he and many others have hard this parallel resistor trick. So what is going on?
I reckon I have worked it out. I believe the mystery can be explained. It took a while
Late here. I am going to see if I can convince Mr Mystery, who is very skeptical, via that PM conversation. If this very skeptical person can be convinced by the argument, that will be a big step forward. I must convince him that the answer is not invoking magic and that Ohm's Law remains inviolate.
But right now, I am off to bed.
Cheers, Joe
PS: Alan's amp should be arriving tomorrow. The pics I have seen shows he has done a nice job with your... 😉
Last edited:
Hello All,
Here are some quick measurements.
ScanSpeakl P17WJ-00-08 driver in 0.55 cubic foot sealed enclosure, same as previous measurements.
APx555 analyzer, APx1701 test amplifier with current sense resistors. APx500 software.
2.83volts output voltage.
Current is calculated across a 0.1R resistor. Measures current sensing voltages are small for useful dB values I added 130dB's to to equal dbrA. close to microphone calibrated SPL numbers. Better later.
The current sensing resistor measured harmonic distortion is not too bad. There is not too much room for improvement with added LCR's or controlled current amplifier. First blush.
Two Tone tests later
Today is a travel day, I need to run away.
Thanks DT
Here are some quick measurements.
ScanSpeakl P17WJ-00-08 driver in 0.55 cubic foot sealed enclosure, same as previous measurements.
APx555 analyzer, APx1701 test amplifier with current sense resistors. APx500 software.
2.83volts output voltage.
Current is calculated across a 0.1R resistor. Measures current sensing voltages are small for useful dB values I added 130dB's to to equal dbrA. close to microphone calibrated SPL numbers. Better later.
The current sensing resistor measured harmonic distortion is not too bad. There is not too much room for improvement with added LCR's or controlled current amplifier. First blush.
Two Tone tests later
Today is a travel day, I need to run away.
Thanks DT
Last edited:
This is the black "NBAC" aluminium driver and the ceramic coated aluminium tweeter. All the internal wiring done and six wires out of the back.
The Crossover will be developed outside the box. The next stage are measurement that can be done indoors. These are Z plots of the driver in situ and nearfield measurements of all drivers and the port. Then on a good day we set up the equipment so that all the farfield measurements, 0°, 15° and 30° off axis. Then all that calibrated data gets prepared in Clio V11 and the exported to SoundEasy (an Australian software from Melbourne) and this is where the Crossover takes place. It will then be built and the final testing, tweaking and listening -for some weeks the Crossover will be external. Then the very last step is to make it internal at the base. We are looking at a couple of months.
For those who constantly suggest changes in drivers, may I point out that is not so easy, there is a regimen that has to be followed and it is not something that can be done in a haphazard way.
There is another pair down in Melbourne. It has a temporary Crossover fitted. The difference in voltage sensitivity is expected to be -1.5dB. So it is a bit less than the "ULD" and "MFC" versions that have the same sensitivity. Still, it is still going to be above average sensitivity and easy to drive - all Elsinores are the amplifier's dream loads. Do I think that helps the sound? I do.
The Crossover will be developed outside the box. The next stage are measurement that can be done indoors. These are Z plots of the driver in situ and nearfield measurements of all drivers and the port. Then on a good day we set up the equipment so that all the farfield measurements, 0°, 15° and 30° off axis. Then all that calibrated data gets prepared in Clio V11 and the exported to SoundEasy (an Australian software from Melbourne) and this is where the Crossover takes place. It will then be built and the final testing, tweaking and listening -for some weeks the Crossover will be external. Then the very last step is to make it internal at the base. We are looking at a couple of months.
For those who constantly suggest changes in drivers, may I point out that is not so easy, there is a regimen that has to be followed and it is not something that can be done in a haphazard way.
There is another pair down in Melbourne. It has a temporary Crossover fitted. The difference in voltage sensitivity is expected to be -1.5dB. So it is a bit less than the "ULD" and "MFC" versions that have the same sensitivity. Still, it is still going to be above average sensitivity and easy to drive - all Elsinores are the amplifier's dream loads. Do I think that helps the sound? I do.
I could imagine that having the ULD version sitting in your listening room would make it hard to put so much effort into these lower-tier drivers. I am interested in what the results will yield.
What may be exciting is that this might even be possible with hard cone aluminium drivers. I never thought it would be possible. The drivers were paid for by the Monash University connection in Melbourne, the box and everything else is my part of the collaboration. Keep in mind, when it is all done, I can still pull out all the drivers and put new "MFC" drivers in and change to the "MFC" crossover. So will it stay "NBAC" or will it later become "MFC" - that we shall see.
Hello Joe,
Just a question, not anything else. In the heat of the above debates I lost track of what was intended.
You experimented with a parallel 8 ohm resistor to a nominal 8 Ohm loudspeaker system, i.e. parallel to the x/o input.
Q1 Is that correct or is that resistor placed parallel to the woofer, so within the x/o?
Q2 what in your view is the result of that parallel resistor?
Thanks.
Just a question, not anything else. In the heat of the above debates I lost track of what was intended.
You experimented with a parallel 8 ohm resistor to a nominal 8 Ohm loudspeaker system, i.e. parallel to the x/o input.
Q1 Is that correct or is that resistor placed parallel to the woofer, so within the x/o?
Q2 what in your view is the result of that parallel resistor?
Thanks.
I now see the R is placed parallel to the x/o terminals.
So far so good, but what is the result supposed to be?
So far so good, but what is the result supposed to be?
Hi Boden
Thanks for your enquiry, it is the big question alright.
That parallel resistor is seen directly by the amplifier, but it is what I call a Poor Man's Current EQ, a crude way and it suits a lot of typical 2-Way speakers.
But I have to say that the use of these resistors are far more common than most people realise and are used in commercial speaker systems, but the designers don't tell you as it is part of their secret sauce. Here they are more likely to put the resistor across the driver's terminals.
Why are they doing it and incorporating them silently in their designs?
Because it sounds better.
I don't do these resistors myself, I go much further, I EQ the current, so that the amplifier produces the same current at all frequencies. So if the load is 1 Amp at 100 Hertz, then the amplifier will also be delievering 1 Amp at 1KHz, 10KHz, 20KHz and 100KHz. I have done quite a number of these, but the flagship of these designs is the Elsinores, particularly the "ULD" version, but the "MFC" version is pretty good too. They are free DIY design.
Why does it sound better? Good question. I have been thinking long and hard about that myself.
There are three ways to do it.
1. That quick 8R resistor across the speaker terminals with global results and no need to do any design work. The results may vary, but in most cases you will end up with an 4 Ohm load and better sounding speakers.
2. Use the resistor across the driver terminals, this will have a more local and targeted effect. But this must be integrated into the overall design - and this is already done in some commercially available speakers and best to leave well alone.
3. The full discrete current EQ, this is also targeted but not much heat wasted. The other two above will produce will produce quite a lot of heat. My approach does not, not even close.
But there are heaps more benefits to EQ the current this way. Such as total insensitive to the amplifier's impedance (you can even use a current-drive). It also locks in the Crossover and makes it far more dynamically stable.
But all three just improves the sound and the question is why?
I have developed some ideas about that, that I believe can account for the better sound. It seems that there is an interaction between the speaker and no measurements yet. We need to measure what the amplifier does on the current side when the amplifier is a voltage source.
So I am working on a measurement that will. Wish me luck.
But people who have heard this are way too numerous and not deaf. Something is going on, even if those using these techniques themselves don't fully understand why it works this way. We have designers wanting their speakers to sound better and if using resistors that gets hot, then so be it.
Oddly, one designer who started using resistors secretly, he called the affect anti-jitter. That made me laugh, but anti-jitter crossovers has a nice commercial ring to it. Have you guessed the brand?
Thanks for your enquiry, it is the big question alright.
That parallel resistor is seen directly by the amplifier, but it is what I call a Poor Man's Current EQ, a crude way and it suits a lot of typical 2-Way speakers.
But I have to say that the use of these resistors are far more common than most people realise and are used in commercial speaker systems, but the designers don't tell you as it is part of their secret sauce. Here they are more likely to put the resistor across the driver's terminals.
Why are they doing it and incorporating them silently in their designs?
Because it sounds better.
I don't do these resistors myself, I go much further, I EQ the current, so that the amplifier produces the same current at all frequencies. So if the load is 1 Amp at 100 Hertz, then the amplifier will also be delievering 1 Amp at 1KHz, 10KHz, 20KHz and 100KHz. I have done quite a number of these, but the flagship of these designs is the Elsinores, particularly the "ULD" version, but the "MFC" version is pretty good too. They are free DIY design.
Why does it sound better? Good question. I have been thinking long and hard about that myself.
There are three ways to do it.
1. That quick 8R resistor across the speaker terminals with global results and no need to do any design work. The results may vary, but in most cases you will end up with an 4 Ohm load and better sounding speakers.
2. Use the resistor across the driver terminals, this will have a more local and targeted effect. But this must be integrated into the overall design - and this is already done in some commercially available speakers and best to leave well alone.
3. The full discrete current EQ, this is also targeted but not much heat wasted. The other two above will produce will produce quite a lot of heat. My approach does not, not even close.
But there are heaps more benefits to EQ the current this way. Such as total insensitive to the amplifier's impedance (you can even use a current-drive). It also locks in the Crossover and makes it far more dynamically stable.
But all three just improves the sound and the question is why?
I have developed some ideas about that, that I believe can account for the better sound. It seems that there is an interaction between the speaker and no measurements yet. We need to measure what the amplifier does on the current side when the amplifier is a voltage source.
So I am working on a measurement that will. Wish me luck.
But people who have heard this are way too numerous and not deaf. Something is going on, even if those using these techniques themselves don't fully understand why it works this way. We have designers wanting their speakers to sound better and if using resistors that gets hot, then so be it.
Oddly, one designer who started using resistors secretly, he called the affect anti-jitter. That made me laugh, but anti-jitter crossovers has a nice commercial ring to it. Have you guessed the brand?