Op_Amp_Booster_Circuit

[321Mark]

I'm trouble shooting a circuit for an R/D converter. One circuit is an op-amp booster. It takes a sine wave in differentially and drives it single-ended. It uses a booster stage to drive the signal at 10V peak. I'm not familiar with booster circuits using transistors. Can anyone tell me if this circuit is sufficient and what to what to watch out for? The load is about 120 ohms. Is there a better solution?

 

[IRstuff]

"Boost" is probably a misnomer. The transistor stage is simply a unity-gain buffer used to increase the current drive capability beyond what the op amp can do.

The topology is similar to National's LH0033 buffer amp:

http://www.datasheetcatalog.com/datasheets_pdf/L/H/0/0/LH0033.shtml

TTFN

FAQ731-376

 

[sreid]

Also look at the Linear Technology LT1010 unity gain buffer.

http://www.linear.com/pc/productDetail.jsp?navId=H0,C1,C1154,C1009,C1028,P1219

 

[electricuwe]

The discrete circuit in your file should be also able to do the job.

What is the frequency of your signal ?

What is the problem ?

 

[Skogsgurra]

Agree. This is the standard way of doing it. Frequency probably in the few kHz range. If higher, you may have a distortion problem.

Have you looked at the output signal under load? What does it look like? What is the problem you are troubleshooting?

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[321Mark]

This circuit was designed for us for an Inductosyn system. It drives a high level sine wave (10v) to the "primary" coil. The signal looks ok on the scope, but on the analyzer it's showing distortion products. The drive frequency is 20KHz. This circuit has enough drive, but I was looking to get the distrotion down. This is a hi-rel application so I'm forced to use a military op-amp with a transistor output stage. No commercial parts.

Thanks

 

[IRstuff]

Almost all push-pull circuits will tend to have crossover distortion at the zero crossings. The two diodes are there to try and minimize the crossover distortion. But, if you're driving lots of current, Vbe and Vd may mismatch, thus causing a notch at crossover.

Did the circuit work at some point in time, and stopped working when some component was changed?

TTFN

FAQ731-376

 

[321Mark]

I think there is a cable drive issue associated with the circuit. The output drives a long flex cable and I suspect there may be a fair amount of capacitance in the cable. I'm going to try and measure the capacitance. The output looks OK when its not driving the cable, but distorted when its driving the cable. The capacitance my be adding pole to the response and reducing the phase (gain) margin?

Also, the inductosyn coil looks like an inductor at the end of the cable.

 

[Renovator1]

The 2222/2907 pair aren't exactly high current devices (Ic ~ 600mA) and Pd limits you to about 200mA, I'd guesstimate. There are a number of integrated, Mil-spec buffers and opamps that can deliver up to around 200mA; a buffer that fulfills your implied requirements is Intersil's (nee Harris) HA-5002. The performance of this part will greatly exceed the discrete one as drawn as well as "reduce" the number of components (satisfying that inane mil-spec rule that the fewer the components the more reliable the circuit).

Cable capacitance can definitely affect stability but your (implied) bandwidth requirements mean you can use pretty much any of the standard compensation techniques. A capacitor across the 10k feedback resistor (say, 330pF to 680pF) will curtail the high frequency response somewhere north of 20kHz. More elaborate schemes can help tailor the pulse response, but it doesn't sound like you need to worry about this. At any rate, if you are getting obviously visual distortion at 20kHz with the amount of capacitance from a few feet of cable then you are hard up on the slew limit of something (e.g. - U1, the discrete output buffer or both).

 

[Warpspeed]

I agree, I would try to find a power op amp with suitable specifications, and then build a simple differential amplifier with four resistors.

 

[Skogsgurra]

Inductosyn scales are rather inductive. Especially the circular ones, which I think you have. Sometimes there is a compensating capacitor that makes the combination purely resistive, if correctly sized, and thus reduces current.

Is it possible that there has been such a capacitor and that it is now missing? That could explain why you have problems with a device that seems to have worked once.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[321Mark]

Ok, I'll check for the compensating capacitor. There is not one in the system. Yes it is a circular inductosyn and the length of the cable is over 20 feet and its flex cable where one conductor is over top of the other. We measured these two conductors with no source or load connected and are reading between 3000 and 4000 pf. But I'm not sure this is a good method. I'll look at the HA-5002. I have not come across this one yet.

Thanks for all the help.