Last night I finally got to test some of the Silonex NSL-5910 photocells and compare their responses to some other T4 units I had lying around. In summary, these cells have a little faster recovery time than optimal but I think they will make excellent sounding T4 units. Also, from the few that I had time to test, their responses are very consistent from unit to unit. Therefore, it is probably not necessary to spend a lot of time matching cells unless you're anal like me. Here are my preliminary test results:

Pulse test comparing a Teletronix T4B, Urei T4C and two NSL-5910 photocells

Sorry about the crappy quality of the screenshots. This weekend I'll get out the tripod and grab some better pics. Anyway, this screenshot shows my standard T4 pulse test. The X axis goes from 0 to 10 seconds and the Y axis shows resistance on a log scale ranging up to 10 megaohms. You can see that the NSL-5910's have the fastest recovery of any of the photocells in this test. Even so, I think they will make excellent sounding T4 units when coupled with an appropriate electroluminescent panel. They could also be used in an LA-4-type compressor when coupled with decent LEDs.

Pulse test comparing a Teletronix T4B, Urei T4C and two NSL-5910 photocells

The second screenshot compares the NSL-5910's with a vintage Teletronix T4B (a substandard one - look at the full dark resistances and poor matching of the photocells within the unit) as well as a Kenetek unit I built around seven years ago. Kenetek T4B #26 is one I keep around as a benchmark. It is a little slower than optimal but still sounds really good. An absolutely optimal T4 unit, from my experience would have a pulse response that fell right in between the NSL-5910 curves and Kenetek #26 curves in this screenshot.

Dynamic Range test comparing a Teletronix T4B, Urei T4C and two NSL-5910 photocells

This screenshot shows my standard T4 dynamic range test. The X axis is graduated from 0 to 100 volts and the Y axis displays resistance on a log scale ranging from 0 to 10 megaohms. You can see that the NSL-5910's track reasonably closely to the vintage Teletronix T4B. You can also see that the Urei T4C has a drastically different dynamic range characteristic to it. This is mainly due to the fact that I replaced the old, burned out EL panel with a new LSI unit. The LSI EL panel starts putting out some light with less than 20 volts applied to it, which is far lower than the vintage EL panels. My solution to this is to put a resistor in series with panel, with a value somewhere between 2 and 3.3 megs, depending on the panel.

I'll update this page over the Christmas holidays with more tests and better screen shots.