Scientists don't usually worry about the sound of capacitors. If you want a DC coupled amplifier with no offset and drift you make a chopper this is VERY old.
Last edited:
They needed zero volts DC at the output on each of several amp cards/pcb's and not to drift away from zero. But servo would have had long TC response..... drift was very slow.... maybe weeks/months? IIRC it was part of scanning scheme. Not FET detector related.
[Choppers --- speed of level change is too slow and too noisy]
-RNM
Last edited:
At one time chopping was done with reed relays and called a 'contact modulator.' Used to have a little hardcover book describing it.
Remember, the overhung coil going in remains fully coupled to the magnetic circuit, but overhang leaving the gap does not.
Cost is their driver, but not ours..
Jn
Underhung voice coils with lots of travel need large, expensive, heavy magnetic structures. I saw a 12" driver with 1" of peak excursion (2" P-P) at a driver vendor targeting the autosound market. It was impressive but pretty ridiculous to address the issue and back then (25 years ago) it was like $250 OEM. Also lots of expensive flux is not in use for a long throw driver so you lose sensitivity fast that way.
When we were designing the Entec subwoofers all the issues of getting linearity over long travel were in front of us. We looked at optical solutions (behind the driver) but that would require interaction with the driver assembly process. For a small company already stressed by the minimums for custom drivers something that exotic would be far too risky even if simple (led to linear sensor with Voicecoil between).
The most cost effective position sensor used the coil sensing the projection of the overhung voice coil in the driver. We also had a bridge sensing the acceleration. Since its a narrow band I used a simple resistive bridge to compare the through the driver to the voltage from the amp. Its the .1 Ohm resistor in the schematic on the return lines from the driver.
Back then I was not a fan of adding an accelerometer to a driver. The mass was high, not symmetrical and it was another flex lead destined to fail. Today's accerometers are 1/10 the weight or less and cheap by comparison. Still you have 3+ flex leads to deal with. If the very special wire for the voice coil (tinsel covering cotton or nylon) can still fail making a coax that can flex that much for that long is a challenge.
I'll have an example of one of the woofers here soon. People are still getting them fixed after 30 years. I'll measure it and share. When we were building them the tools we had were pretty primitive. I'm much better instrumented today.
The less that great scan attached is all I have on hand for the circuitry but its all there.
Also neon lights and photocells. And in some cases rotating disks. Not much bandpass in those cases. The neon lights were in a number of HP instruments from the 1960's and 1970's. They did get us to the moon.
Hi Scott,
5 1/2 digit voltmeters made with tubes ! Man, would I ever love to see that meter. Years later the HP 3478A would have been an equivalent replacement being the same at 5 1/2 digits. Would you know if they were just voltmeters (amazing enough) or did they have other functions, like AC volts, current or resistance? The things we take for granted these days!
By chance would you have any pictures or remember a brand and model? This is something that was totally missed in my reading.
I even have a pair of Heathkit digital meters, an HD-1202 (?) and the other model # I can't remember, and a Fluke 8200 (I have three) 3 1/2 digit meter that used nixie tubes for the numeric display. One has all the options, the other two are DCV only.
The last time I calibrated the Fluke meters, they were right on the money. I think they were actually good to 4 1/2 digits but didn't display the LSD because they read bang on the money without the LSD flickering or being one count either way. I was amazed by this as I was use to the last digit being a little unstable on the other meters. I used the one with all the options while in University. The other two were picked up at a government auction years later. Still have them.
Just thought of another question Scott. Do you know what they would have used as a voltage reference? I'm dying to know, a standard Weston cell maybe? Wouldn't be a portable instrument in that case.
-Chris
5 1/2 digit voltmeters made with tubes ! Man, would I ever love to see that meter. Years later the HP 3478A would have been an equivalent replacement being the same at 5 1/2 digits. Would you know if they were just voltmeters (amazing enough) or did they have other functions, like AC volts, current or resistance? The things we take for granted these days!
By chance would you have any pictures or remember a brand and model? This is something that was totally missed in my reading.
I even have a pair of Heathkit digital meters, an HD-1202 (?) and the other model # I can't remember, and a Fluke 8200 (I have three) 3 1/2 digit meter that used nixie tubes for the numeric display. One has all the options, the other two are DCV only.
The last time I calibrated the Fluke meters, they were right on the money. I think they were actually good to 4 1/2 digits but didn't display the LSD because they read bang on the money without the LSD flickering or being one count either way. I was amazed by this as I was use to the last digit being a little unstable on the other meters. I used the one with all the options while in University. The other two were picked up at a government auction years later. Still have them.
Just thought of another question Scott. Do you know what they would have used as a voltage reference? I'm dying to know, a standard Weston cell maybe? Wouldn't be a portable instrument in that case.
-Chris
No just volts IIRC I mentioned it here years ago and someone had the model number. The readout used the same relays that the phone company used to dial numbers. The display was a deep array of plastic wafers that were side lit by incandescent lamps so the character etched in only the selected one stood out. Predated Nixie tubes.
My most missed item is probably one of a tiny handful of surviving Lumitron sampling oscilloscopes, 25pS in 1963. Packed with tunnel diodes, snap diodes, and gold doped transistors all long gone. The sampling head was a hybrid of a very early Schottky diode bridge and a 6CW4 where the gate capacitance was the sampling capacitor. The diodes were soldered directly to the inner conductor of a short hard line terminated with GR connectors on both ends. Probably the greatest example of bringing the past and future of discrete design together ever, unfortunately fueled by the cold war.
Last edited:
Hi Scott,
Thanks, this I have to find out about. I've seen pictures of frequency counters that used a vertical line of lamps to represent each decade. What you are describing is an amazing application of optics on top of even getting the meter to work.
The things that previous engineers and technicians created never ceases to amaze me. Well, at least I know that it exists now so I can go on a hunt.
-Chris
Thanks, this I have to find out about. I've seen pictures of frequency counters that used a vertical line of lamps to represent each decade. What you are describing is an amazing application of optics on top of even getting the meter to work.
The things that previous engineers and technicians created never ceases to amaze me. Well, at least I know that it exists now so I can go on a hunt.
-Chris
No.
A dual vc does not have one of them poised to cancel the inductance.
You do not appear to understand the concept.what part do you have issues with?
Jn
For those of you who are confused with the subject here, I would listen to 1audio (Demian) to get a clue. He appears to be the only one here who has some EXPERIENCE in designing exotic woofers. Also, he has worked along side Keith Johnson, who is exceptionally brilliant (known his work for 50 years) who also contributed to the design. His input makes sense.
Sorry. But I lost interest and havent followed real closely. If it can be used to lower distortion (MFB) then who can make them and lets do it. I'll order the first ones.
Now what Demian and his partner (RIP) did was good with an existing driver. Hopefully it is easier and less expensive to do now. I'd like to see an update or at least suggestions for an update.... note again... we want lower distortion from drivers not extended freq response here.
So far we limited ourselves to woofers/subs because of the desired use of MFB. But in general ,dynamic drivers are really bad in most any way you measure them. high distortion prevails.
How has the AES, Pro, consumer allowed this to go on and on and on like this? I went to ESL that had the required low distortion (0.1%) and ruler flat FR but limited in bass, significant room influence and limited dynamic range.
A few days ago, Mark (Markw4) was over to T&M his DAC project and to listen. I had torn down my system and was replacing everything... starting with the speakers.... M2's. But it is so good and extended clean deep bass, maybe I dont have to do anything. Just upgrade the rest of the system. One being replacement PA's.... Then, I would modify the amps with MFB also.
Just because I can.
Many others, I would assume, might also want low distortion bass to add to thier loudspeaker systems. If we can figure out something practical and do-able and not expensive, we can go over to a place where it is a DIY. And, maybe start a trend in the industry.... lower distortion. Right now the bar is fairly low.
Sincerely,
RNMarsh
Last edited:
I don't recall seeing posts from Demian discussing magnetic design, can you point them out for me?
All my peers certainly aspire to be exotic woofer designers, what we work with is so mundane and low tech..
Jn
- Status
- Not open for further replies.