Ok, sorry. I didn't think it could be a serious attempt since you started out, with Si sensor and hoping you didn't need an IR filter. Here's why you need one: human vision covers roughly from 400 to 700 nm wavelength, and Tri-X film falls off a bit sooner on the red end, at about 650 nm. So anything your sensor can see towards IR, from 650 or 700 on, MIGHT be problematic.Fun??!! I've resoldered and traced point-to-point 22ga solid core wire hundreds of times and written 700 lines of C. I'm doing this because I need a spot meter!
If every light source you used for photography had the same spectral signature, this would be ok. However, if you sometimes want to meter in daylight, and sometimes under tungsten light, big problems! Compared to a nominal 5000K daylight source, a tungsten source, at nominal 2800 is drasticially deficient in blue (yet still loaded with IR). So if your meter can see some IR, and is calibrated in daylight, it would barely notice the missing blue of tungsten light. Tri-X, on the other hand, would suffer greatly from this loss. So the result is properly exposed Tri-X under daylight conditions, but heavy underexposure under tungsten lighting. Of course, the issue is more complicated, but I hope this is a convincing argument.
For serious work, I think you need to know more about your combined spectral response. However, if you can match a commercial meter under both tungsten and daylight, you might be ok. Otherwise, various "heat absorbing glasses", such as Schott KG-5(Edmund Optics cat# K49-092), might be players.
Don't overlook the need for light baffles in your optical path; flare light can really distort the readings.