jensenmann
Well-known member
Has anybody ever tried to built a compressor with VDRs? Is there anything to read about this subject out there?
VDRs for use in a compressor?
- Thread starter jensenmann
- Start date
Help Support GroupDIY Audio Forum:
Samuel Groner
Well-known member
You mean a varistor? That doesn't work, for compression you want a three terminal device where one terminal sees the control voltage (which comes from the side chain) and the other two provide the variable resistance--the most typical example for such a device is a JFET. A varistor has just two terminals and the resistance is just a function of the applied voltage, i.e. it distorts (and they don't care which frequency they distort either).
Samuel
Samuel
jensenmann
Well-known member
ok, thanks, now I see. simply limiting the voltage would mean distortion because the limiting would happen immediately. Does it happen immediately in real world? Does a varistor react fast?
Sorry, these may be idiot questions but if the varistor needs time to react then no sidechain would be necessary when you feed it with the right voltage. The attack and release control would be implemented in the varistor itself - like the timeconstant of an T4B is fixed through the EL panel. The compression ratio and threshold would be controlled through the amount of the incoming voltage and the varistors transferfunction (Kennlinie). I find the idea very interesting. The entire compresser would be a driverstage to drive the divider/varistor, followed by the gain/output stage.
Sorry, these may be idiot questions but if the varistor needs time to react then no sidechain would be necessary when you feed it with the right voltage. The attack and release control would be implemented in the varistor itself - like the timeconstant of an T4B is fixed through the EL panel. The compression ratio and threshold would be controlled through the amount of the incoming voltage and the varistors transferfunction (Kennlinie). I find the idea very interesting. The entire compresser would be a driverstage to drive the divider/varistor, followed by the gain/output stage.
G
Guest
Guest
I used a diode limiter sensing current through it that was amplified to drive LED in an opto pair. However, it was implemented to protect SS power amps from clipping, but that's the single point where I use compressors (except one more device, a Bi-El-Contour vocal processor, but that's not about limiting of a dynamic range). A first, diode limited peaks, then compressor reduced gain, then I saw a LED flashing and pulled down faders.
A steep or hard threshold voltage dependent resistance would be little different than clipping and not very useful for dynamics processing.
For 2 terminal devices with a V:R or I:R relationship like a pn (diode) junction where the impedance difference is more gradual, some degree of gain control can be realized by imposing a DC component on top of the AC (audio). For this to be useful the AC component must be small relative to the DC component so either linearity, or noise floor will suffer. Usually both. I experimented with these but dismissed them as too much baggage for audio use.
FETs are literally "field" effect, but that field is modulated by a gate "voltage" so could arguably be considered a VDR... These are more linear than diode limiters but less linear than VCA or other more neutral sounding approaches. FET gain elements are popular in several audio applications for their quick speed and low cost.
JR
For 2 terminal devices with a V:R or I:R relationship like a pn (diode) junction where the impedance difference is more gradual, some degree of gain control can be realized by imposing a DC component on top of the AC (audio). For this to be useful the AC component must be small relative to the DC component so either linearity, or noise floor will suffer. Usually both. I experimented with these but dismissed them as too much baggage for audio use.
FETs are literally "field" effect, but that field is modulated by a gate "voltage" so could arguably be considered a VDR... These are more linear than diode limiters but less linear than VCA or other more neutral sounding approaches. FET gain elements are popular in several audio applications for their quick speed and low cost.
JR
> Does a varistor react fast?
Yes. Often used for lightning-spike protection.
It is not absolutely necessary to have a 3-terminal device; indeed when we have 3 pins the control often goes in the same port as the audio.
There is a varistor (bridge) VCA in one of the old guitar amp temolo schemes. It seems to rely on some rather sloppy old varistors; modern sharp-knee ones sound bad.
> attack and release control would be implemented in the varistor itself
Thermistors and incandescent lamps can be worked this way. The time constants are never really what you want, tend to be more symmetrical than music wants, and not easily changed.
E-V used a car tail-lamp as a power-limiter on a tweeter. Just put it in series. Cold resistance was like 1 ohm, negligible effect. Many milliseconds of a dozen volts heated it up to 10 ohms, cut 6dB off the tweeter power. Try the same with a grain-of-wheat lamp and a 100 ohm resistor.
Yes. Often used for lightning-spike protection.
It is not absolutely necessary to have a 3-terminal device; indeed when we have 3 pins the control often goes in the same port as the audio.
There is a varistor (bridge) VCA in one of the old guitar amp temolo schemes. It seems to rely on some rather sloppy old varistors; modern sharp-knee ones sound bad.
> attack and release control would be implemented in the varistor itself
Thermistors and incandescent lamps can be worked this way. The time constants are never really what you want, tend to be more symmetrical than music wants, and not easily changed.
E-V used a car tail-lamp as a power-limiter on a tweeter. Just put it in series. Cold resistance was like 1 ohm, negligible effect. Many milliseconds of a dozen volts heated it up to 10 ohms, cut 6dB off the tweeter power. Try the same with a grain-of-wheat lamp and a 100 ohm resistor.
[quote author="PRR"]> Does a varistor react fast?
Yes. Often used for lightning-spike protection.
It is not absolutely necessary to have a 3-terminal device; indeed when we have 3 pins the control often goes in the same port as the audio.
There is a varistor (bridge) VCA in one of the old guitar amp temolo schemes. It seems to rely on some rather sloppy old varistors; modern sharp-knee ones sound bad.
> attack and release control would be implemented in the varistor itself
Thermistors and incandescent lamps can be worked this way. The time constants are never really what you want, tend to be more symmetrical than music wants, and not easily changed.
E-V used a car tail-lamp as a power-limiter on a tweeter. Just put it in series. Cold resistance was like 1 ohm, negligible effect. Many milliseconds of a dozen volts heated it up to 10 ohms, cut 6dB off the tweeter power. Try the same with a grain-of-wheat lamp and a 100 ohm resistor.[/quote]
Not to quibble, but incandescent lamps are "temperature" dependent resistances.
JR
Yes. Often used for lightning-spike protection.
It is not absolutely necessary to have a 3-terminal device; indeed when we have 3 pins the control often goes in the same port as the audio.
There is a varistor (bridge) VCA in one of the old guitar amp temolo schemes. It seems to rely on some rather sloppy old varistors; modern sharp-knee ones sound bad.
> attack and release control would be implemented in the varistor itself
Thermistors and incandescent lamps can be worked this way. The time constants are never really what you want, tend to be more symmetrical than music wants, and not easily changed.
E-V used a car tail-lamp as a power-limiter on a tweeter. Just put it in series. Cold resistance was like 1 ohm, negligible effect. Many milliseconds of a dozen volts heated it up to 10 ohms, cut 6dB off the tweeter power. Try the same with a grain-of-wheat lamp and a 100 ohm resistor.[/quote]
Not to quibble, but incandescent lamps are "temperature" dependent resistances.
JR
Yves
New member
Here is one S20K75 potted:
http://www.dissident-audio.com/Traceur/PDFs/S20K75.pdf
A nice soft knee :thumb:
Account for a 1nF range shunt capacitance.
Yves.
http://www.dissident-audio.com/Traceur/PDFs/S20K75.pdf
A nice soft knee :thumb:
Account for a 1nF range shunt capacitance.
Yves.
jensenmann
Well-known member
[quote author="PRR"]> Does a varistor react fast?
Yes. Often used for lightning-spike protection.
[/quote]
I should have better asked are there any slow ones available? :grin:
Yes. Often used for lightning-spike protection.
[/quote]
I should have better asked are there any slow ones available? :grin:
G
Guest
Guest
[quote author="jensenmann"][quote author="PRR"]> Does a varistor react fast?
Yes. Often used for lightning-spike protection.
[/quote]
I should have better asked are there any slow ones available? :grin:[/quote]
Use some external heater, like transistor glued to the varistor.
Yes. Often used for lightning-spike protection.
[/quote]
I should have better asked are there any slow ones available? :grin:[/quote]
Use some external heater, like transistor glued to the varistor.
I know it is hard to remember what all the different "___istors" do, but this information IS available, no need to recycle bad information.
A VARistor is pretty much a hundred teeny spark-gaps, made by pressing a high temperature powder together. It acts "instantly" for all audio purposes; spark-gap breakdown is very fast. It is NOT very temperature sensitive, heating would not be a good control method.
A THERMistor is a temperature sensitive resistor. It could be used for audio gain control. They are made in both positive and negative temperature coefficients. They can give a very wide change of resistance for a small change of temperature. They can be used "self-heating": this is a problem in temperature control applications (the measurement current changes the temperature you are trying to read) but can give a self-acting audio AGC with just two parts (and some spare gain). Disadvantage is that they can not be made to a precision value, you will usually need to trim a resistor, and of course any temperature effect is a problem for non-studio work, such as sound for Ice Festival or Desert Race.
The last "new" variable gain cells were chips, both transistor-bridge and analog switches.
Unless you go with the DBX/THAT chips, most of the best techniques are the oldest techniques. As Narma says in CJ's interview, most tubes have too MUCH gain and too stable. Photo-resistors are useful, but modern burner-safety LDRs do not have the long tail of decay that the early ones did, which made the LA2a so sweet.
A VARistor is pretty much a hundred teeny spark-gaps, made by pressing a high temperature powder together. It acts "instantly" for all audio purposes; spark-gap breakdown is very fast. It is NOT very temperature sensitive, heating would not be a good control method.
A THERMistor is a temperature sensitive resistor. It could be used for audio gain control. They are made in both positive and negative temperature coefficients. They can give a very wide change of resistance for a small change of temperature. They can be used "self-heating": this is a problem in temperature control applications (the measurement current changes the temperature you are trying to read) but can give a self-acting audio AGC with just two parts (and some spare gain). Disadvantage is that they can not be made to a precision value, you will usually need to trim a resistor, and of course any temperature effect is a problem for non-studio work, such as sound for Ice Festival or Desert Race.
The last "new" variable gain cells were chips, both transistor-bridge and analog switches.
Unless you go with the DBX/THAT chips, most of the best techniques are the oldest techniques. As Narma says in CJ's interview, most tubes have too MUCH gain and too stable. Photo-resistors are useful, but modern burner-safety LDRs do not have the long tail of decay that the early ones did, which made the LA2a so sweet.
G
Guest
Guest
[quote author="PRR"]
A VARistor is pretty much a hundred teeny spark-gaps, made by pressing a high temperature powder together. It acts "instantly" for all audio purposes' spark-gap breakdown is very fast. It is NOT very temperature sensitive, heating would not be a good control method.
[/quote]
Thanks PRR, I always knew you are a walking Encyclopedia!
PWM resistors as well.
A VARistor is pretty much a hundred teeny spark-gaps, made by pressing a high temperature powder together. It acts "instantly" for all audio purposes' spark-gap breakdown is very fast. It is NOT very temperature sensitive, heating would not be a good control method.
[/quote]
Thanks PRR, I always knew you are a walking Encyclopedia!
PWM resistors as well.
Tomba
Member
I think, at least in theory, it could be possible to make compressor with varistors. Last night I made some simulations and it worked quite well. If you use balanced topology, keep audio-signal under 1V and use high sc-voltage (0...+40V), distortion is under 0.1% ...well, in simulation. But you have to have perfectly matched varistors. Is that possible? I have no idea.
mikep
Well-known member
The idea of a self-heating thermistor as a compressor is very interesting. the long portion of a complex release time constant could be provided by external heating.
if this isnt "warm" sound, nothing is.
mike p
if this isnt "warm" sound, nothing is.
mike p
