New revisions as of 05-26-05 with new schematic diagram
Way back when I was a high school student there was a device the peaked my interest. It was the photocell that was used to turn lights on after dusk, and off at dawn. I was kind of a hi-fi nut back then. The surface noise found on my record collection got worse with repeated playings. Scratch filters found on stereos of the time obliterated the highs. Somehow it came to me to try to use the photocell to turn off the surface noise between songs on a album, when it was most apparent. I devised a circuit that put the photocells in line with the tape monitor loop. I used the tape out audio as the control signal to drive a pair of green leds that shined on the photocells when the music was at normal level. Between songs the leds would not glow, and therefore caused the photocells to turn the audio down, therefore gating it. I did not know of gating or compressors, or any of that studio stuff at that time. This system worked reasonably well for me. I thought of the possibility of the photocells working to do the opposite by keeping loud passages from getting too loud for cassette tapes made for the car. I could get the overall audio level much louder so I could hear it over the road noise. Many years later I was shocked to discover that just such a device had been in service since the mid sixties, and used extensively on many of my favorite records. After wiping the egg from my face, I pondered the relative ignorance that existed in the 1970’s due to the lack of the fountain of information that is now available to all on the internet.
After reading about people that are trying to build their own opto compressors to get that renowned sound of that universally accepted standby that exists in just about every major studio in the world, I decided to try building one for myself. I searched the web to find countless examples of electro-optical implementations. Some used Vactrols such as those that are use in many Fender amps for the channel switching and tremelo circuitry. Many commercially available compressors also use these devices. They have a certain sound, but not the legendary sound original. Some have used electro-luminescent panels with varied results. The circuitry required to sufficiently drive the panel is tricky and usually requires transformers or somewhat high voltage power supplies.
While on the subject of these electro-luminescent panels, It is known by the designers that these panels have a rather short life, shockingly so. They lose half their brightness in around 2000 hours of use depending hard you push them. Buy two panel nite lites use one for a month or so, then plug them both in next to each other. You will see the difference. Imagine what is happening in your studio.
Then I spent alot of time trying many different photocells. There are really only four photocell formulations in use. Some compressor manufacturers have the vendor create a special part number for them based on their requirements. You can do this selection yourself, however. I designed a test rig to select and match the photocells. Once I settled on the correct photocell for this project, I went on to search out the led that would posess the right linearity charicteristics needed for the sound I was after.
Once I was satisfied with the results, I designed a totally op amp based circuit that used the leds I found to be perfect in replacing the electro-luminescent panel and it's associated tube circuitry. I then added some control over the actual decay time external to the led. This is primarily to duplicate the natural decay time of the phosphors used in the electro-luminescent panel. I have conducted tests on several photocells and led/photocell combinations. The Vactrol VTL5CL exhibits far too fast a natural decay time to be suitable without some degree of external decay time to make them sonically pleasing. Many other photocells exhibit this problem. The Silonex NSL 5910 does have the correct decay characteristic, but must be tested and matched. The external decay circuitry gives the user the option to "fine tune" the essential charicteristics.
If you intend to build this compressor please note:
The 5k trimpot and adjusted to 2.82k to ensure the best compression slope when using the Silonex photocell.
Clairex no longer manufactures photocells as of this writing. I like the results I have had with the silonex photocells. I was told that these photocells are the actual ones used in today's re-issue of a very famous optical compressor.
Controls:
R1....input gain.............adjusts the gain of the input circuit.
R11...compression............adjusts the gain of the light amp that drives the led.
R8....output level...........adjusts the amount of make up gain after the compression circuitry.
R18...decay time.............fine tunes the decay time to optimize for different instruments.
R4....Ratio slope............adjusts compression ratio.
The special green LED is catalog # 25-377 from www.hosfelt.com. The all important photocell is the Silonex NSL-5910 specially selected by matching for stereo pairs if desired. They are available from Allied electronics. The red front panel indicator leds are radio shack 276-307 or just about any general purpose led.
Notes for this circuit:
An all tube signal path can be implemented with this compressor. Observe all precautions regarding dangerous high voltages inherent in tube circuitry. (see disclaimers)
I have found that the use of tube circuitry for this compressor is not worth the effort because the change of sound is not really a factor in this implementation.
Power supply is a split supply of +/- 15 volts.
Any decent quality op-amps can be used.
All diodes are 1N914 except Led's
Parts list:
R1....100K AUDIO TAPER R2....1K R3....10K R4....5K TRIMPOT R5....SILONEX NSL-5910 PHOTOCELL R6....10K R7....10K R8....100K AUDIO TAPER R9....1K R10...22K R11...100K AUDIO TAPER R12...1K R13...10K R14...10K R15...10K R16...10K R17...10K R18...1MEG LINEAR TAPER R19...4.7K R20...4.7K R21...1MEG R21...1MEG C1....0.47UF C2....0.1UF C3....0.47UF C4....1UF FILM C5....0.47UF C6....2UF FILM
More notes:
R4 can be mounted on the front panel if desired to provide a ratio control. In my prototype I chose to set it to 2.82k internally.
C2 can be reduced in value to make the detector less responsive to low frequencies. Some find this desirable. Experiment with smaller values.
R10 can be increased to 56k if you need more output gain.