I am restoring a 35mm film projectors from the 1930s, which at some point had a Kelmar optical sound head added onto it. The solar cell outputs a balanced signal through an XLR jack that is very weak (around 60-70 mV). If I run the XLR straight into a mic input of a mixer I can get some recognizable audio if I crank the gain way up but it is extremely muddy. Here is a image of an oscilloscope reading of the output signal:

reading of signal

I would like to build a preamp circuit to get this signal up to line level. I am very new to electronics (more comfortable with the mechanical aspects of this project) and would like to design a circuit based around the SSM2019 IC. I tried to replicate a mic preamp schematic with poor results and ultimately I somehow fried the IC experimenting with it.

enter image description here

Before I order another one I want to see if I am on the right track at all.

Specific questions:

  1. Is the SSM2019 an appropriate IC for this project?
  2. Do I need to correct for the DC bias of my input signal or is that insignificant?
  3. Can I/How can I design this circuit to run on DC power?
  4. How does the impedance of the solar cell affect my circuit design?
  5. If I measure across the outputs of the solar cell I read its 34k ohm resistance. Is that the same as the solar cell's output impedance?
  • \$\begingroup\$ Welcome to EE.SE. The SSM series of op-amps are the best on the market, and the SSM 2019 is an excellent choice. But to recover a signal buried in noise goes far beyond an amplifier. You may need to use a DSP engine to extract in-band noise, like sound studios use to re-master old analog recordings. Think about this before spending money. Right now any amplifier is going to amplify the good with the bad. \$\endgroup\$
    – user105652
    Jan 29, 2018 at 3:04
  • \$\begingroup\$ @Sean Can you expand on the signal description? It looks like a modulated carrier. \$\endgroup\$ Jan 29, 2018 at 3:12
  • \$\begingroup\$ That IC seems to be overkill. Is it calibrated? The preamp with a calibrated system and a Buzz track film from what I read ought to be 0.5Vpp output. Is it microphonic to motor noise? www.rsem.com/pdf_manuals/Strong_International/Soundheads/cntymr3.pdf \$\endgroup\$ Jan 29, 2018 at 3:35
  • 1
    \$\begingroup\$ @jre Actually a photovoltalic cell was the standard for this for many, many years, they are somewhat specialised parts but work well. \$\endgroup\$
    – Dan Mills
    Jan 29, 2018 at 13:46
  • 1
    \$\begingroup\$ @Sean: just find your post here and I wonder if you succeeded in build your preamplifier... I would be very interested because I need to do the same. Thanks \$\endgroup\$ Dec 1, 2018 at 4:55

1 Answer 1


For a variable area or variable density (OLD Movies!) track you will have audio riding on a fairly large DC bias, which will need removing before amplification.

I would suggest a series cap of maybe 47uF into a simple 5532 based amp is usually going to be at least as good as whatever was used originally, but actually a standard mic amp should do just fine, with a few tens of mV out of the cell you should only need ~30 - 40dB of gain to get to the usual sort of volt or so of line level.

Have you considered the need for de-emphasis in the playback processing (Either the "Academy" curve for mono sound tracks or the "X Curve" for Stereo variable area? It will likely sound severely odd without these, and that is before we consider Dolby.

One other gotcha is the existence of modern 'Cyan dye' prints which have replaced the silver halide with a dye which is transmissive to IR from the exciter lamp, to get optical sound to work with these films you need to replace the exciter with a laser source.

Sounds like a fun project.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.