Photon Pickup Phototransistor

I bought a Photon MIDI guitar system (serial number 2605) several years ago on EBay and mounted it out an 1974 Les Paul I had. The arched top of the Les Paul prevented the pickup PCB from aligning ideally with the strings so I removed it and the system sat unused until the past year or so.

I bought a 500 Variax by Line 6, a modeling guitar which provides a wide range of guitar sounds, which I thought would be a perfect compliment to the Photon. Now I could add MIDI to any number of different guitar sounds, including acoustic, 12 string or even a barely passable sitar guitar sound. And the top was flat so I thought it would work well.

I drilled the holes for the optical pickup and mounted it on the Variax and quickly discovered that the G string was not functioning. This started a long journey to repair the system and led to this discussion of the ins and outs of what I learned about the Photon system.

The first thing I discovered is that there is very little technical information out there on this. I found one guy (I'll respect his privacy and not name him or his Gibson contact) who said in an email:

"All trace of old Photon stuff has disappeared along the road of life. All I could find was an old photon and busted up cartridge. There is one person who might be able to help you. "M. D." at Gibson. Not sure how if he is still there, but he built the pickups for John McLaughlin's rig using the Photon on an acoustic. I think he also did a magnetic hex pickup conversion as well.

Hope this helps."

It didn't of course. I could not find any information on M. D. at Gibson or anywhere else though I didn't search too deeply to find him.

As an electrical engineer with a background in optics I figured it wouldn't be too hard to find a replacement for whatever part had gone bad. I quickly determined that the photodetector on the G string was bad. By measuring the forward voltage on another string I determined it had to be a phototransistor versus a photodiode, though later measurements seemed to indicate that for the three low strings, they did use a photodiode.

I located a source for side looking phototransistors (Optek Technology ) and purchased a few OP550A phototransistors. The OP550 comes in four flavors, A through D which are selected for on-state collector current with the A version spec'd at 2.55mA minimum. It has a 30 degree (one side) half power point or 60 degrees overall. See the graph below reproduced from the spec sheet of the 550A.

Normalized Current vs. Angle

After installing the new phototransistor I found the G string worked...a little. The signal level was very low and was unworkable. I (quite wrongly) thought that maybe the right part would be a photodarlington instead of a phototransistor and tried one. It didn't work at all. At the time I didn't have an oscilloscope and didn't want to take the whole rig into work to troubleshoot it, so it sat again.

I recently obtained a scope and set aside the time to get things working. I found that the on state of my replacement phototransistor was not as low as the other strings so figured alignment was part of the problem. The new device is a little higher than the old device and I had installed it slightly rotated. I tweaked the alignment and raised the LED slightly and got a much better on-state signal. It was now time to mount the pickup and start measuring. The signal strength was still not good as reported by the Photon interface and I quickly realized that this system isn't about light, but "dark". The string interrupts the light path between the LED and detector and while I was getting appropriate on-state and off-state (full occlusion) signals, the AC signal generated by the vibrating string was weak. Unlike most photo-coupling system where getting sufficient light striking the photodetector is the challenge and therefore the common use of photdarlingtons to maximize photon to collector current transfer, with the LED just fractions of an inch from the detector this was no problem. The design challenge is getting the maximum AC signal from the vibrating string. Obvious once I thought about it but this wouldn't be the first time I dove into a problem without thinking (unencumbered by the thought process as the Car Talk guys say).

Therefore the right photodetector would have a very small acceptance angle so that with the string aligned just under the light path (as suggested by the Photon manual) vibration would cause it to obstruct or occlude the path as much as possible to create a strong AC signal. The 60 degree angle of the Optek 550A was simply too wide and was "seeing" too much of the LED and not enough string. A narrower angle would be a better choice although might create more LED to detector alignment problems.

As someone who has spent a lot of time working with LEDs, I know they are pretty awful optical sources. An LED is a (typically) square chip with a bonding pad on the top which is commonly centered.

LED Chip Top View

Typical LED Chip - Top View

So right from the start the center of chip does not emit light. Newer designs move the bonding pad or pads to the corner of the die but most inexpensive LEDs don't. There is often a dip in the on-axis response due to the bonding pad. Also, the sides or edges of most LED chips emit nearly as much light as the top. To collect and transmit this might in the forward direction the LED is mounted in a cup which improves the visible brightness but from an optical standpoint causes more aberrations in the radiation pattern.

In the case of the side looking package used on the Photon pickup the LED is mounted on a lead frame (no focusing cup) with a lens molded into the package. I don't have the equipment to do an accurate radiation pattern measurement but it is likely there's a bit of a hole in the center of the response so a bit of offset might not be a bad thing.

I searched the web for smaller angle photodetectors and found a couple possibilities. The SDP8406 from Honeywell has a slightly tighter 50 degree angle. I found some with an even tighter angle, such as the Sharp PR480E00000F which has about a 26 degree angle, but is in a thicker 3.2mm package versus the 2.54mm thickness of the OP550A. (see below) .

I installed the SDP8406 and it worked acceptably. The signal is weaker than the other strings but with gain adjustments can be brought into balance with the other strings. Based on this I would estimate that the original part has a 40 degree or so angle.