Daniel,\
due to variation in the manufacture the voltage required for the DN2540 will be somewhere between -1.5 and -3.5V (depletion mode). The stage wants 12mA of current. So lets take a guess at your device and assume for the moment that it needs 2.8V. V=IR so 2.8 = 0.012 * R. R = about 233R.
Install a 330R resistor, carefully measure the voltage drop across it in operation, and then recalculate with the known voltage to get your desired current. You can adjust it by either piggy backing a resistor to the 330R or replaceing it with the correct value.
May I strongly suggest that you cascode two DN2540's. A single isn't much better, if at all, to a 10M45. You could easily deadbug it on the pcboard.
due to variation in the manufacture the voltage required for the DN2540 will be somewhere between -1.5 and -3.5V (depletion mode). The stage wants 12mA of current. So lets take a guess at your device and assume for the moment that it needs 2.8V. V=IR so 2.8 = 0.012 * R. R = about 233R.
Install a 330R resistor, carefully measure the voltage drop across it in operation, and then recalculate with the known voltage to get your desired current. You can adjust it by either piggy backing a resistor to the 330R or replaceing it with the correct value.
May I strongly suggest that you cascode two DN2540's. A single isn't much better, if at all, to a 10M45. You could easily deadbug it on the pcboard.
Last edited:
Gary Pimm has some measurements of the different CCS variants.
CCS performance measurments
(look at the graph at the bottom of page)
--
CCS performance measurments
(look at the graph at the bottom of page)
--
hey-Hey!!!,
In my next CCS I will use a 10M45 on top of DN3545N3. The 10M45 has nearly twice the negative bias of the 2540/3545 and so would lower the gate capacitance of the lower FET substantially( a useful thing I think ).
cheers,
Douglas
In my next CCS I will use a 10M45 on top of DN3545N3. The 10M45 has nearly twice the negative bias of the 2540/3545 and so would lower the gate capacitance of the lower FET substantially( a useful thing I think ).
cheers,
Douglas
- Status
- Not open for further replies.