floating current sorce

I'd try and find out exactly what you meant by a "floating current source".

so you could probably use it to make a Howland current source as described in Horowitz and Hill's "The Art of Electronics", if your idea of "floating" allows a common ground.

If you only wanted either only postive or only negative current, then a P- or N-channel FET would do the same job with fewer components.

--------- Bill Sloman, Nijmegen

How are you defining THD? Is that with respect to the current or the Voltage?

If the latter, then why does the subject of your post say "floating current source?" The reason I ask is that normally I see THD used in conjunction with Voltages, not currents.

Anyway, if you had to design a non-floating current source, could you?

So just design a non-floating current source, and feed it into a 1:1 transformer. 45 Hz is not very far from the 50 Hz which many small power transformers are specified to work at, so you'll probably be OK using one. If you can't find one with a 1:1 ratio, just use two identical transformers with their secondaries in parallel.

Or maybe you could find a small power transformer with dual isolated secondaries and use them, leaving the primary open.

Note that I've never done this, but I don't see why it wouldn't work.

Good luck!

--Mac

Common mode voltages will not be more than 20 volts. Current output would be about 10mA and load not more than 500 ohms. Frequency must be a

Me too. Harry

I read in sci.electronics.design that Mook Johnson wrote (in ) about 'floating current sorce', on Sat, 27 Aug 2005:

1. Look in my Little Black Book

1. Look in Art of Electronics

2. Look in data and applications handbooks

1. Google for 'current source circuits' - I get 2.4 million hits.

View in a fixed-width font such as Courier.

. . . +--[R2]--+ . | | . | | . | |\\ | . >--[R1]-+--|-\\ | . | >--+ . Vin |+/ | . >--[R1]-+--|/ | . | | . | | . | | . [R2] /| | . | /-|-+ . +---< | . | \\+|-+ . [Rc] \\| | . | | . +--------+ . | . | . V . . I to load .

Current source drive into a transformer is shunted by magnetizing current. Now you could use a "current transformer"- but the 1% accuracy requirement would require the primary current to be >100x the load current- say 1A.

Hmm. Guess I'll have to actually go open a textbook now to make sure I understand this. ;-)

--Mac

We would correct your data first: floating current source=? probably you meant a bipolar current source. Is the load grounded or referenced to a voltage rail? A current source is always "floating" with respect to voltage output.

Frequency must be a

The best way to do this is with a single-ended circuit powered from a completely floating isolated power source. There are plenty of ways to create the signal in the floating current source circuit, such as analog isolation amplifiers, etc. These are offered by TI (Burr-Brown) and Analog Devices.

By floating I mean that the output must not have a DC path between the current source terminals. It must be galvanicly isolated but the common mode voltage will not vary much beyond +/-20V from system ground.

By ,1% THD I am impling a clean sine wave if looked at on the scope. No major distortion products.

Idealy it would look like a voltage controlled current source (as seen in spice) where the input and output are isolated.

Hi ,

google around about electro impedance tomographie. These guys needs the same thing.

Marte