Grain of Rice (GoR) Compressor

Ok, I'm about to get weird.

I was thinking about taking some old tech and mixing it with new tech (just because it's less likely to have been explored).

A "Grain of Rice" (GoR) incandescent lamp is a tiny light bulb (literally the size of a grain of rice). These (or more commonly the bigger "Grain of Wheat" variant) can be used with an LDR as the variable resistor in a compressor but the bulb filament itself can also be used as a resistance element. The resistance of the filament depends on it's temperature. If it is illuminated / hot, it's resistance goes up. More specifically, if the GoR filament is cold, it's resistance is ~40 ohms but when it's hot it's ~400 ohms.

Here is a theoretical GoR compressor circuit:



So as signal level increases, more current is passed through the filament, it's resistance increases and thus you get attenuation. Very simple and elegant. But ...

Obviously this has some problems like the fact that the impedance of the element and shunt resistor is high enough that some significant gain is needed and that translates directly into noise. But the higher impedance is necessary for good attenuation. And, even if that can be overcome (drive differentially?), the timing is a little too slow for most things.

Ideas?

squarewave said:

Ok, I'm about to get weird.

I was thinking about taking some old tech and mixing it with new tech (just because it's less likely to have been explored).

The idea is to use a "grain of rice" bulb as the variable resistance element (more current == lower resistance) of a compressor. I've seen schematics that use them but I can't find one at the moment. So the new tech would be drive the lamp with TWO modern powerful high voltage op amps like OPA551. One supplies signal and the other applies DC. The DC could be used to both attenuate and set the threshold of the compressor yes? Then just difference the filament to recover the possibly attenuated signal.

Does anyone know anything about these "grain of rice" bulbs wrt audio? Are they microphonic / noisy?

Click to expand...

looked those up as I never heard of that bulb.  Seems to be nothing more than the size of a Christmas light used on the a string of light bulbs.

pucho812 said:

looked those up as I never heard of that bulb.  Seems to be nothing more than the size of a Christmas light used on the a string of light bulbs.

Click to expand...

It's way smaller than that. It's literally the size of a grain of rice. Model train crowd uses them.



12V / 30mA = 400 ohms

I am interested to see where this goes. 

"That was a great take, but we have to do it again after I replace the light bulb in my compressor"!

Light bulbs were widely used in old sine wave oscillator for gain/level control.

I have little attraction to old stuff just because it's old...

BTW I'm old.  ;D ;D

JR

Stuffing "DC" (control voltage) in the audio path will THUMP.

I can picture a balanced form but with likely filament variation it will still thump.

The classic light-bulb compressor would be a power amp driving a lamp and speaker (or resistor) in series. If the lamp gets hot, its R increases, voltage across load decreases. A 12V 30mA lamp might reasonably work with a opamp on +/-15V, lamp, and ~~100r resistor. Cold we may have 40r into 100r and low loss. Hot we have 400r into 100r and 14dB loss.

This is slow and will not handle a wide range.

E-V used a aircraft lamp in series with a tweeter in a well-promoted PA/Monitor speaker. It allowed more speech/music power input than the tweeter would safely deliver. Yes the system went "dull" at high input but they maybe didn't care; critical users could see the lamp glow and know they were over-driving and could not trust treble balance. I have not seen this used in any other system, which may be A Clue.

IIRC the Mastering Lab (Doug Sax) crossover used lamps in series with an Altec 604 speaker.  In later times, the UREI 81x monitors had the same design.  I seem to recall the lamp(s?) were automotive tail lamp bulbs.

Bri

Using lamps as protection for loudspeaker drivers is old and introduces issues.

A) when the lamp burns out, no HF driver

B) While operating as a temperature dependent resistance, it alters the crossover tuning.

C) the changing resistance can show up as a non-linearity, but generally slow moving so not audible. Fuses in series with LF drivers (pretty much same mechanism) have been blamed for measurable very LF distortion.

Last century Peavey developed a protection strategy that did not suffer from the obvious faults. A PTC fuse (low Z when cold, high Z when hot) in parallel with protection network.  Below protection thresholds, the PTC shunted across the dropping resistor and speaker behaved optimally. In protection mode the PTC fuse opened and the driver was protected. When it cools off the HF comes back.

A similar mechanism was taken advantage of for amp testing/repair load boxes. A 100W lamp in series with line cord would look like a low impedance when cold. If the amplifier being tested drew enough mains current the lamp would light up and protect the amp. When the amp was behaving on the test bench at low current the light bulb was bypassed for full power testing

JR

Peavey  used bulb limiters for tweeter protection also , visible light emanating from the ports was a sure sign of its operation  ,

The ducking effect on the tweeters is audible on smaller  peaks , but in the event of a major overload it tends to reduce any nasty harmonics hitting peoples ears , by comparrison with a properly set limiter light bulbs are very slow , but on the ocassional peak in program material I found it quite effective , thermally resetting solid state fuses seem to be the modern way to go , a lot quicker to act than a bulb ,but maybe slower to return to normal afterwards .

Tubetec said:

Peavey  used bulb limiters for tweeter protection also , visible light emanating from the ports was a sure sign of its operation  ,

The ducking effect on the tweeters is audible on smaller  peaks , but in the event of a major overload it tends to reduce any nasty harmonics hitting peoples ears , by comparrison with a properly set limiter light bulbs are very slow , but on the ocassional peak in program material I found it quite effective , thermally resetting solid state fuses seem to be the modern way to go , a lot quicker to act than a bulb ,but maybe slower to return to normal afterwards .

Click to expand...

The common light bulb protection at Peavey was upgraded to the PTC version last century. They came up with a fancy name for it, that I do not remember.

JR

PRR said:

If the lamp gets hot, its R increases, voltage across load decreases. A 12V 30mA lamp might reasonably work with a opamp on +/-15V, lamp, and ~~100r resistor. Cold we may have 40r into 100r and low loss. Hot we have 400r into 100r and 14dB loss.

This is slow and will not handle a wide range.

Click to expand...

Well, that's important! I had it all backwards it seems.

I have completely re-written the post to reflect reality.

Brian Roth said:

IIRC the Mastering Lab (Doug Sax) crossover used lamps in series with an Altec 604 speaker.  In later times, the UREI 81x monitors had the same design.  I seem to recall the lamp(s?) were automotive tail lamp bulbs.

Bri

Click to expand...

correct. When I  got mine, the issue of "insert speaker here" not working was related to bulbs that I acquired at the local auto parts store.