high-side current sensing chips

- snipped-for-privacy@c65g2000hsa.googlegroups.com...

(snip)

Fred Bloggs turned me on to this circuit patented by Analog Devices:

. . -Is->

. . ----+------[Rs]------+----- . | | . [R] [R] . | | . .----+ | . | | | . | >| |< . | Q1 |-----+------| Q2 . | /| | |\ . | | |< | . | +-----| Q3 | . | | |\ | . >| | | | . Q4 |-------------------+ . /| | | | . | / \ | / \ . | | I | | | I | . Io \ / | \ / . | | | . | | | . '-------+--------' . | . --- . Yair, the ground currents gotta be fixed, not a mirror. If I use my circuit I'll replace the output transistor with a p- channel mosfet. gotta go.

Reply to
gearhead
Loading thread data ...

It is not 1 percent, it is 1 percent per degree C. if you have a 10 degree C temperature difference change you get a 10 percent unbalance.

Reply to
JosephKK

de quoted text -

At 300K and Vf=3D.6, a change of 0.1 degree would result in delta Vf =3D

200 uV. At a current decade for every 60 mV, that introduces an error (added to existing offset) of 10^.003333 =3D 1.0077 or about eight tenths of one percent. At Vf =3D .7, it comes out closer to 0.9 percent. So I get a round figure of about one percent error in the current mirror for each tenth of a degree temp mismatch centigrade in the transistors.
Reply to
gearhead

quoted text -

So you are paying attention after all. Good. Transistors not on the same die are difficult to keep within a few degrees of each other. In this case thermal design matters.

Reply to
JosephKK

Hide quoted text -

Pickings for current mirrors are slim. I found BCV62, but it is not what you would call well matched at all.

Reply to
gearhead

Hide quoted text -

Just use "matched transistor pair" with any search engine.

Reply to
JosephKK

The 100 ohm base-loading resistors are there to take care of gain and temperature match. You may still want to hand-select transistors, but that would be gilding the lily for your application.

Reply to
Tim Wescott

PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.