Hi zdr,
May I suggest you to measure the P type in similar manner? ... do not jump to conclusions so fast
Have fun,
Mihai
Hey, I like jumping to conclusions, saves time sometimes
Anyway, I have to listen to smarter then me, so here it goes. New N and P batch arrived today which derived a bit closer DMM Hfe, so I picked 8. This is the closest I could get from 100 (50+50) in total:
Tp Vbe Hfe Ic (Hfe DMM)
P1 614 121 100 118
P2 615 120 100 118
P3 615 120 100 118
--------------------------
N1 613 113 100 105
N2 614 113 100 106
N3 612 115 100 105
I ordered 40 more for next week (tell me when to stop).
Hi zdr,
What type of transistors did you ordered? Is seems to be NJL3281/1302, not the zero version which were more tight matched from the factory.
Stop ordering ... you are wasting money
Good luck!
Mihai
Mihai,
First set are 0302/0281, second are 1302/3281. I managed to match 1302/3281 from much smaller lot - last ten that arrived. Eight of them measured Hfe=104 straight in DMM.
Difficult to get in Brussels, you're kidding.
Vous is dingue ?
Btw, your numbers 10-20-50 suggest that you ordered power devices by the single digit.
For matching purposes, buying per rail (or tray in the old days) is smarter, higher chance that they're from the same batch, and consecutive dies from the same area on the wafer.
Downside is that if you do that for a couple of decades, you end up with a storage of parts that's the equivalent of an entry level BMW.
The way you're doing it will cost a 7 series in the same period.
With a 120mm high heatsink (3U internal height), the top row of output devices is located very near the top of the heatsink. This will make the top devices run hotter than the lower devices. 4U Will work better, 5u will be good.
It was to avoid using 5U and wide enough to accommodate the long board I chose to vertically orient the PCB and keep the output devices in the lower half of the tall heatsink.
It was to avoid using 5U and wide enough to accommodate the long board I chose to vertically orient the PCB and keep the output devices in the lower half of the tall heatsink.
One way around this is to use a nice thick spreader - something like 1cm+, cut to the size of the radiator. The spreader is not a bad idea even if the heatsink is larger, so that all the outputs are close to the same temperature.
Sheldon
As I am populating the board, I could not help noticing that having two schematics (spice and eagle) with totally different markings adds a lot to the confusion. One by one, I changed all eagle references to match spice ones. If someone thinks this might be at least a bit useful to the community and not a total waste of my time, I can post it here.
Well, my first channel is assembled and happily no worky. I have powered only front end with frontend ps and torroid with 2x30vac, giving 38vdc on the rails. When connected to frontend section, voltage drops to +-10v. Nothing really heats up in the process, all leds lit up. Fuses are not inserted. Curent drain from rails seems to be only 40mA, which points to a problem in ps maybe? Need some injection of intelligence please.