Sorry for the length of time away from this topic. I had another project that took a few months and now I'm happily getting back into this one.
The biggest difference between the original parts and the 3904/3906 combo (classic transistors) is the gain (hFE). Its not too different, but with small signals from a phototransistor its possible the circuit is not going to work.
Thanks for pointing out the hFE differences, which is not something I was aware of before. I'm also not yet understanding how to compare hFE values. Looking at the PUMX1 datasheet, it shows hFE of 120 at 6V and 1mA:
http://www.nxp.com/documents/data_sheet/PUMX1.pdfNow it gets confusing looking at the datasheet for the 3904 transistor, which shows a whole bunch of voltage and current values:
https://www.fairchildsemi.com/datasheets/2N/2N3904.pdfNot sure what to make of that. As you mentioned, the values are not too different--to me they look in the same ballpark, but they aren't using the same V and I values. What do you make of this?
Now looking at PUMT1 hFE values, it is also 120:
http://www.nxp.com/documents/data_sheet/PUMT1.pdfAnd the 3906 also looks similar:
https://www.fairchildsemi.com/datasheets/2N/2N3906.pdfBTW The 3904 and 3906 transistors are not Fairchild, they were bought from eBay from China but I assume they have similar specs.
Can you measure voltages at the output connector on the schematic, with and without light?
Ok, measuring both terminals of the phototransistor, the original circuit gives values of 4.1v when ambient room lighting is striking the phototransistor. When I take a DeWalt light and shine it directly on the phototransistor, it measures 0.17V. So there is a large range there which produces nice data readings (on my NXT brick it shows values from 94 in ambient to 245 in bright light).
Now when I'm using my circuit with the 3904 and 3906 transistors, it measures 4.18 volts with ambient room lighting, and 4.08 volts with direct light. So *some* change is measured, but I would like a greater range to register from this circuit (as it only produces values of 41 to 112 on the NXT brick).
Any ideas on some transistors with better hFE values?