Current source circuit fail.

Zout should be theoretically infinite! It's a current source.

The '544 should have worked, with maybe a little frequency compensation.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

Right, head slap, Vout compliance is limited by opamp supply and Rset. (It wasn't very happy driving a capacitor... it'd go to one one rail or the other.)

I tried some C (100 pF) from feed back (pin 5) to ground, I probably put the C in the wrong branch.

The input and output of the 544 were backwards, not sure why. (I had no R's in the 544 circuit, just a 5 pin to-220 in space,

0.1uF C's on the power leads... word on the street is that the second amp needs to be faster. (I've never quite understood that.)

George H.

Wait...wait..... it's a current source!

If one is fast and one is slow, it's likely to be stable. Either one. If they are near the same speed, some theory may be required.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

Grin. Well I was hitting it with square wave and getting a triangle out. (but it drifted up or down.. as I tweaked the dc offset. Right that's expected.) Hey hitting it with a triangle wave gives an output that looks a little like a sine wave but isn't. (A parabola?)

OK. Or maybe I'll try the BUF634 from the spec. sheet. ~100mA would be enough... +/-1 amp would be cool though.

George H.

You could use a discrete follower pair, nfet+pfet or NPN+PNP.

TCA0372 is a dual 1-amp opamp for about 50 cents.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

Hey it worked OK when I had the power opamp inside only one loop.

gnd---RR---+--RR--+ ||\ | +|-\ | Ina134 | >--+-(A)-Rset-+---output to gnd. +|+/ | ||/ /| | Vin--RR--+----RR---< +-----+ \| Buffer opa134

I moved my (A) power opamp. Though the opa544 doesn't seem to like having it's output driven near the negative rail.

Chan. 1 is voltage across 200 ohm resistor, chan 2. is input. (100 ohm 'set' resistor. 1 V_in= 10mA

George H.

What is your load? Is the output connected to ground? If you leave the output open the current has nowhere to go and the circuit won't work correctly maxing the output because of the positive feedback.

If you ground the output the output of the buffer amp is ground and the op amp is a unity gain buffer into Rset (assuming RR is the same value everywhere). What are your power rails? Your scope traces pretty clearly show the amp is being maxed out.

Is your load 200 ohms? Then with -5 volts in, the current in the load is 50 mA and the voltage on a 200 ohm load is 10 volts. The voltage at (A) is -15 volts which likely is your rail. As the input exceeds -5 volts input the output can't drive harder and a large difference develops across the inputs of the op amp.

I guess you are just over-driving the circuit.

--

Rick C 

Viewed the eclipse at Wintercrest Farms, 
on the centerline of totality since 1998

[snip]

Why all that complexity? Plus the load behavior is inside your loop. A dual R-R OpAmp plus either a PMOS or PNP for the high current will easily solve your problem... and be stable. ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| STV, Queen Creek, AZ 85142    Skype: skypeanalog |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 

         To save time, Let's just assume I know everything.

Right 200 ohm load, +/- 15V rails. (I might have bumped 'em up a bit.)

Exactly. I wanted to see how it behaves when over driven.

George H.

Hmm well doesn't look all that complex to me. It started as a current source, bipolar with grounded load. Without the power opamp. The int-amp->R-> opamp circuit is cute I thought. Then I was wondering about more current. (It sometimes would be nice to drive something like magnet coils at ~1A on each side of zero... I don't need a grounded load for that, and could just use the power opamp in the typical V to I converter. Sense R to ground.... but I've done variations on that circuit a million times. (OK maybe ten times. :^) Anyway mostly playing around.

George H.

How would that work?

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

[snip]

See "CurrentSourceConceptual.pdf" on the S.E.D/Schematics Page of my website for a bidirectional source/sink version. Simplification to single-sided is left as an exercise for the student ;-) ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| STV, Queen Creek, AZ 85142    Skype: skypeanalog |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 

         To save time, Let's just assume I know everything.

Here's an old one, from 2002:

I wouldn't do it that way now. I'd replace it all with one part.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

Jim, Do you need that first pair of Fets? (Assuming the opamp can handle the supply voltage.) You are just using current to transfer the voltage from low side to high side... which I like.

George H.

The first set of FET's could just as well be bipolar's. Or, if you trust resistor ratios, just make the first stage a level translator to transform a 0 to +1V input into a +VCC to (+VCC-1V) signal (or whatever scale factor you want. ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| STV, Queen Creek, AZ 85142    Skype: skypeanalog |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
     It's what you learn, after you know it all, that counts.