current mirror

Jul 15, 2010 2 Replies
A

I was thinking of using discrete single or dual transistors to make a current mirror, for currents of about 1uA to 3mA. Then I found that Zetex makes current mirror devices.

formatting link
Zetex ZDS1009TA OK, fine, but what I don't get is the suggested application of high side sensing as shown in the datasheet. Why not just use a RR op amp? These current mirrors are 5$. That's crazy, no? Is it just the fact it can withstand a certain voltage differential and can be used to level translate?



Anyways my initial application was to translate the current output from an old 8 bit current output DAC in the range of 0 - 0.3mA to the needed 0-3mA.



I suppose I can use this Zetex part and try to get a gain of 10, but I have to admit I haven't looked at this stuff in ages. I can barely handle Ohms law lately. I guess it's time to go into management?! :)


Vote
Twitter Facebook LinkedIn
M

Watch out. If one transistor gets as little as a degree or two warmer than the one across from it, it will kill the accuracy. 3 mA could do that. There's an interesting thread about current mirrors you can find by going on google groups and searching for "accuracy of bcv62"

Vote
Twitter Facebook LinkedIn
B

..

than

on

As soon as your current mirror becomes asymmetrical, you also have to worry about intrinsic emitter reisitance - base resistance can also screw you up, but not usually as much.

Cascode transistors can decrease the differential self-heating - if you wanted top be really cute you could organise to have ten times the emitter-collector voltage drop across the 0- 0.3mA device that you had across the 0- 3mA device.

Bill Sloman, Nijmegen
Vote
Twitter Facebook LinkedIn

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required