Troubleshooting differential amplifier fault. Advice needed.

Have you monitored the power rails during start-up with a fast scope? There is a possibility you may have a switch-on spike that is well beyond the limitations of the amplifier to withstand (even during the 5ms ramp-up).

--
******************************************************************** 
Paul E. Bennett IEng MIET..... 
Forth based HIDECS Consultancy............. 
Mob: +44 (0)7811-639972 
Tel: +44 (0)1235-510979 
Going Forth Safely ..... EBA. www.electric-boat-association.org.uk.. 
********************************************************************

"Damaged"? Or malfunctioning? I've seen situations (aka I've been burned :) where a slowly ramping power supply allowed a circuit to latch up and draw destructive currents.

...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142     Skype: skypeanalog  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.

Can you measure the raise time of the linear supply?

Something tells me that maybe the SMPS could be coming up to fast for an error in design of the circuit.. Are both rails coming up the same time?

Did you put your scope in glitch capture mode? it's possible the supply is generating a short over voltage pulse. This is common where the supply is over rated , where the load can not suppress the pulse enough to allow the supply to get in operating conditions, putting a burden R on the supply sometimes helps this.

Maybe you can put rail diodes in for the inputs to keep it from getting over driven?

Jamie

ge is sensed by a differential amplifier input. The differential voltage is amplified linearly. The full scale output is +/-10V.

** That a LEM Hall sensor ?

I'm familiar with the LTA50P ( which has exactly that spec) and a few othe rs.

/-15V supply, the sensor functions normally. But when I use a switched mode supply, the op amp is damaged on switch-on and the output is either +13.5V or -13.5V irrespective of the input.

period of about 5ms. I wonder if this might somehow be causing some common mode input problem for the op amp.

** Are the +/-15 volt rails coming from the SAME switching transformer?

Cos if there are two SMPSs involved, one may come up before the other and t hat can cause the op-amp to latch up. Also, the first supply to come up wil l often momentarily drive the other rail with reverse voltage producing the same outcome.

FYI

Many linear PSUs used with audio mixing consoles have twin, indentical sup plies for the op-amps where the +17V output of one links to the common or g round terminal of the other to give -17V. If only one supply comes on due t o a fault in the other, this can result in damage to a lot of ICs.

The fix is to link the supplies outputs with a simple circuit that inhibits both supplies unless they come up simultaneously.

Eg:

See TR1 and TR2.

.... Phil

On a sunny day (Fri, 31 Oct 2014 15:00:11 -0700 (PDT)) it happened John Aston wrote in :

Are you sure that SMPS has no overshoot at switch on?

V/-15V supply, the sensor functions normally. But when I use a switched mod e supply, the op amp is damaged on switch-on and the output is either +13.5 V or -13.5V irrespective of the input.

I can't get the same results consistently. Sometimes the op amp survives sw itch on. Sometimes it will latch up but then recover when I switch on a sec ond time. Sometimes it is destroyed. Maybe the fault is at the margin of wh at is safe for the op amp.

I've tried two different SMPS with the same results. Both have a ramped out put at switch on.

One problem may be that the circuit draws less than 10mA per supply rail an d the SMPS is rated at up to 1A.

I haven't checked the rise time of the rails from the linear power supply. I'll do this.

The SMPS rails go from 0 to full scale over a period of 5-10ms so I was assuming that was slow enough not to be the reason.

As I mentioned in another post, the circuit draws less than 10mA per supply rail and the SMPS is rated at up to 1A. Maybe adding some extra burden as you suggest might help.

By "rail diodes", did you mean diodes between the rail and the input? I wonder if the leakage current will affect accuracy.

tage is sensed by a differential amplifier input. The differential voltage is amplified linearly. The full scale output is +/-10V.

hers.

5V/-15V supply, the sensor functions normally. But when I use a switched mo de supply, the op amp is damaged on switch-on and the output is either +13. 5V or -13.5V irrespective of the input.

a period of about 5ms. I wonder if this might somehow be causing some commo n mode input problem for the op amp.

that can cause the op-amp to latch up. Also, the first supply to come up w ill often momentarily drive the other rail with reverse voltage producing t he same outcome.

upplies for the op-amps where the +17V output of one links to the common or ground terminal of the other to give -17V. If only one supply comes on due to a fault in the other, this can result in damage to a lot of ICs.

ts both supplies unless they come up simultaneously.

Are the +/-15 volt rails coming from the SAME switching transformer?

Yes they are. The positive and negative supplies look like a mirror image o f each other on the scope. I can't see any short term spikes.

oltage is sensed by a differential amplifier input. The differential voltag e is amplified linearly. The full scale output is +/-10V.

others.

+15V/-15V supply, the sensor functions normally. But when I use a switched mode supply, the op amp is damaged on switch-on and the output is either +1 3.5V or -13.5V irrespective of the input.

r a period of about 5ms. I wonder if this might somehow be causing some com mon mode input problem for the op amp.

nd that can cause the op-amp to latch up. Also, the first supply to come up will often momentarily drive the other rail with reverse voltage producing the same outcome.

supplies for the op-amps where the +17V output of one links to the common or ground terminal of the other to give -17V. If only one supply comes on d ue to a fault in the other, this can result in damage to a lot of ICs.

bits both supplies unless they come up simultaneously.

of each other on the scope. I can't see any short term spikes.

Maybe try a different opamp? How about a little resistance in the supply rails? (Can you share a circuit diagram?) George H.

[snip]

Yep. Schematic worth thousand words ;-) ...Jim Thompson

-- | James E.Thompson | mens | | Analog Innovations | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | San Tan Valley, AZ 85142 Skype: skypeanalog | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at

| 1962 | I love to cook with wine. Sometimes I even put it in the food.

besides the +15V and -15V connection, do you also have a proper 0V or ground connection between your circuit and the power supplies?

Are the bypass caps in your circuit from the +15 to the -15 or from each to ground?

The problem is likely either a ramp up, transient or common mode issue.

Do you have a scope?

Mark

Yes I have a storage oscilloscope. I agree with you about the likely source of the problem.

I'm going to build up the circuit from scratch and stress it with different power supply levels. There does seem to be a common mode voltage developed at the input to the differential amp so I will examine the behaviour of this first.

There are 1uF tantalum capacitors between each rail and ground.

I'll report back after testing. Thanks.

You might post a schematic. ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142     Skype: skypeanalog  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.