Problem with TL071 SMPS comparator - UK Vintage Radio Repair and Restoration Discussion Forum

mole42uk · 19th Dec 2014, 2:54 pm

Hi,

I'm working on a Marconi power supply which is a quite straight-forward SMPS. It accepts an input of anything between 12 and 36V DC and outputs ±5V and ±12V. It has it's internal clock synced to the 10MHz master clock in the instrument.

There's a problem that I have not resolved so far and I'm beginning to chew the bench so I thought I perhaps might get some help.

I attach a partial circuit of the offending part. It's an absolutely bog standard comparator - 6.2V zener-derived reference on one leg, and the monitored voltages on the other, adjusted with a preset. The output drives an opto-isolator the output of which adjusts the mark/space ratio of the SMPS. So far so good.

The problem I have is that the voltage output is too high and it cannot be altered with the preset control. Monitoring the voltages around the TL071, both inputs behave as expected, but when the TL071 output changes state it only goes from about 0v to about 75mV. Even if I disconnect the output pin and leave it floating. Or pull it either way. I have changed the TL071 and checked every other component in that circuit, but I cannot fathom why the comparator won't go high enough to turn the opto-isolator LED on hard enough the modulate the SMPS. I did prove the opto by pulling the LED on through a 4k7 resistor which quietened down the PSU, so that bit works!

Any guidance happily accepted. I'd prefer not to spend the next couple of weeks in a padded cell.....

cmjones01 · 19th Dec 2014, 2:59 pm

Has the TL071 got power and ground? I know it seems obvious, but when all other possibilities are eliminated...

Chris

mole42uk · 19th Dec 2014, 3:01 pm

Yes!

Alistair D · 19th Dec 2014, 3:03 pm

What are the voltages on pins 2 and 3 of the IC. For balance both of the pins should be at the same voltage.

Check C13 is not leaky.

Al

mole42uk · 19th Dec 2014, 3:17 pm

Pin 2 is 6.2 volts
Pin 3 can be varied between about 5.8v and 6.5v

I replaced C13 already.

Alistair D · 19th Dec 2014, 3:42 pm

When pin 3 is below 6.2V the op amp output should be high.

One possibility could be a problem with the input offset pins 1 and 5 of the chip. These need to be fully floating and not shorted to anywhere else.

It could also be worth scoping the output pin to ensure it is not oscillating.

Al

dave cox · 19th Dec 2014, 4:03 pm

A high resistance in pin 7 ?
The chip turns on but just drags the supply down with it.

dc

mole42uk · 19th Dec 2014, 5:57 pm

Quote:

Originally Posted by Alistair D

It could also be worth scoping the output pin to ensure it is not oscillating.

I did that. Nothing except a straight line (even at 5mV/cm on my 40MHz scope). Well, maybe a bit of noise, but then I'd expect that!

Dickie · 19th Dec 2014, 6:18 pm

Check R38 really is 2k2 ?

Lucien Nunes · 19th Dec 2014, 6:28 pm

See how it behaves with the feedback loop open.
Bypass integrator cap and disconnect ref input, see if IC works as op-amp.
Lift supply pin and measure supply current in both 'states'.
Check for stuck / shorted PCB nodes with power off.
Mock-up circuit on breadboard and test with IC that was in circuit.
Locally AC-bypass 12V supply or provide external supply.

And this is the one that I'd probably fall foul of: Make sure you didn't put dud 071 back in by mistake!

Radio Wrangler · 19th Dec 2014, 8:49 pm

At low frequency/DC the feedback loop is open.

The capacitor gives no feedback around the opamp at DC, so the DC condition is created using th rest of the PSU s the feedback path for the opamp.

If anything doesn't work in the PSU, there is no DC feedback, and the opamp is banged against the rail.

David

Tyso_Bl · 21st Dec 2014, 2:05 pm

Couple of questions, did this unit ever work for you? is there any possibility that it had been got at previously?

A couple of thoughts, using op amps as comparators isn't always a good idea, while different makes may all work as op-amps, they may not all work as comparators, I've had this with homebrewed circuits.

Is your replacement an exact replacement?

From what I've read so far your circuit may be latching up, some opamps can do that if the inputs are taken too close to the supply rails, usually the output will go in the expected direction then as an input aproaches the supply rail the output will reverse.

It may be possible that the inputs are being applied before the supply rail to the comparator has stabilised, it's such a simple circuit, the passives are easy to check, so it has to be something to do with how the chip is being fed, if the chip itself is the correct type and not faulty.

Finally are all the grounds grounded as they should be, ie nothing floating? and have you tried the opto isolator for any form of back leakage?

Alistair D · 21st Dec 2014, 2:29 pm

TL071 is a jfet input op-amp and to my mind is an unusual choice in this circuit. I would have preferred an LM321 or similar.

This range of devices can sense input voltages down to ground level which may help prevent latch up at power on.

Al

kellys_eye · 21st Dec 2014, 2:38 pm

Any 'crud' between pins 2 and 3? With a high impedance input the area needs to be clean.

mole42uk · 21st Dec 2014, 2:56 pm

Quote:

Originally Posted by Alistair D

TL071 is a jfet input op-amp and to my mind is an unusual choice in this circuit. I would have preferred an LM321 or similar.

My thoughts, exactly. Even a humble 741 would be a better choice, especially as we have ±12v available.

However, this is a Marconi product, so my view is that they probably used the TL071 for a reason, and I want to find out why this circuit doesn't work without re-designing it!

Eddy_Stone · 21st Dec 2014, 4:05 pm

I believe this is configured as a voltage error amplifier and not a comparator. The opto should work in its linear region. Can you show the rest of the circuit.

Eddy

Mooly · 21st Dec 2014, 4:06 pm

Have you scoped the opamp output ? It may be showing 75mv on a DVM but could be switching between the rails at high frequency and with some wide mark:space ratio.

mole42uk · 21st Dec 2014, 5:50 pm

Quote:

Originally Posted by Mooly01

Have you scoped the opamp output ? It may be showing 75mv on a DVM but could be switching between the rails at high frequency and with some wide mark:space ratio.

See post 8

mole42uk · 21st Dec 2014, 6:05 pm

Quote:

Originally Posted by Eddy_Stone

I believe this is configured as a voltage error amplifier and not a comparator. The opto should work in its linear region. Can you show the rest of the circuit.

I was expecting the opto to work in its linear region, I just pulled it hard on the prove that part of the circuit.

I don't understand the difference between a voltage error amplifier and a comparator in this case. I think the output of the PSU is compared against the zener controlled reference, and the op-amp, working linearly, adjusts the brightness of the opto LED to adjust the current flow through the opto taransistor. The change in current through the opto transistor changes the mark/space ratio of the SMPS chip. Or have I misundersood?

I've posted the whole circuit in two parts:

Mooly · 21st Dec 2014, 6:19 pm

Quote:

Originally Posted by mole42uk

See post 8

Should have spotted that, I got dragged away and posted in haste.

Reading it all and its very strange. The opamp output should switch between the rails as you alter the preset such that one input is either higher or lower than the other... standard comparator operation. Your range of voltage control by the preset is limited. Are you using a DVM ? (high ohms/v) so that we can be sure of those readings. Have you measured between the inputs to ensure you can get a real "difference" ? Heck, even a millivolt or two is enough to change states but something odd is going on.

Pin 2 is fixed at around 6 volts. What happens if you take pin 3 direct to 12 volt ? The output must swing high under those conditions. Although under these conditions it is working as a comparator, as Eddy_Stone mentions, when its all operating normally it isn't, the opamp + opto all become part of a "linear" system.

Something is fundamentally wrong here. Have you scoped all the rails. Check the 5 and 12 volt for hash ?

19th Dec 2014, 2:54 pm	#1
mole42uk Nonode Join Date: Aug 2010 Location: Resolven, Wales; and Bristol, England Posts: 2,614	Problem with TL071 SMPS comparator Hi, I'm working on a Marconi power supply which is a quite straight-forward SMPS. It accepts an input of anything between 12 and 36V DC and outputs ±5V and ±12V. It has it's internal clock synced to the 10MHz master clock in the instrument. There's a problem that I have not resolved so far and I'm beginning to chew the bench so I thought I perhaps might get some help. I attach a partial circuit of the offending part. It's an absolutely bog standard comparator - 6.2V zener-derived reference on one leg, and the monitored voltages on the other, adjusted with a preset. The output drives an opto-isolator the output of which adjusts the mark/space ratio of the SMPS. So far so good. The problem I have is that the voltage output is too high and it cannot be altered with the preset control. Monitoring the voltages around the TL071, both inputs behave as expected, but when the TL071 output changes state it only goes from about 0v to about 75mV. Even if I disconnect the output pin and leave it floating. Or pull it either way. I have changed the TL071 and checked every other component in that circuit, but I cannot fathom why the comparator won't go high enough to turn the opto-isolator LED on hard enough the modulate the SMPS. I did prove the opto by pulling the LED on through a 4k7 resistor which quietened down the PSU, so that bit works! Any guidance happily accepted. I'd prefer not to spend the next couple of weeks in a padded cell..... Attached Thumbnails __________________ Richard Index: recursive loop: see recursive loop

19th Dec 2014, 2:59 pm	#2
cmjones01 Nonode Join Date: Oct 2008 Location: Warsaw, Poland and Cambridge, UK Posts: 2,681	Re: Problem with TL071 SMPS comparator Has the TL071 got power and ground? I know it seems obvious, but when all other possibilities are eliminated... Chris __________________ What's going on in the workshop? http://martin-jones.com/

19th Dec 2014, 3:01 pm	#3
mole42uk Nonode Join Date: Aug 2010 Location: Resolven, Wales; and Bristol, England Posts: 2,614	Re: Problem with TL071 SMPS comparator Yes! __________________ Richard Index: recursive loop: see recursive loop

19th Dec 2014, 3:03 pm	#4
Alistair D Nonode Join Date: Jan 2010 Location: Edinburgh, UK. Posts: 2,008	Re: Problem with TL071 SMPS comparator What are the voltages on pins 2 and 3 of the IC. For balance both of the pins should be at the same voltage. Check C13 is not leaky. Al __________________ I won't tell you how I discovered that.

19th Dec 2014, 3:17 pm	#5
mole42uk Nonode Join Date: Aug 2010 Location: Resolven, Wales; and Bristol, England Posts: 2,614	Re: Problem with TL071 SMPS comparator Pin 2 is 6.2 volts Pin 3 can be varied between about 5.8v and 6.5v I replaced C13 already. __________________ Richard Index: recursive loop: see recursive loop

19th Dec 2014, 3:42 pm	#6
Alistair D Nonode Join Date: Jan 2010 Location: Edinburgh, UK. Posts: 2,008	Re: Problem with TL071 SMPS comparator When pin 3 is below 6.2V the op amp output should be high. One possibility could be a problem with the input offset pins 1 and 5 of the chip. These need to be fully floating and not shorted to anywhere else. It could also be worth scoping the output pin to ensure it is not oscillating. Al __________________ I won't tell you how I discovered that.

19th Dec 2014, 4:03 pm	#7
dave cox Nonode Join Date: Jan 2009 Location: Bristol, UK. Posts: 2,062	Re: Problem with TL071 SMPS comparator A high resistance in pin 7 ? The chip turns on but just drags the supply down with it. dc

19th Dec 2014, 6:18 pm	#9
Dickie Octode Join Date: Sep 2004 Location: St. Albans, Hertfordshire, UK. Posts: 1,478	Re: Problem with TL071 SMPS comparator Check R38 really is 2k2 ? __________________ Regards, Richard, BVWS member

19th Dec 2014, 6:28 pm	#10
Lucien Nunes Rest in Peace Join Date: Apr 2008 Location: London, UK. Posts: 2,508	Re: Problem with TL071 SMPS comparator See how it behaves with the feedback loop open. Bypass integrator cap and disconnect ref input, see if IC works as op-amp. Lift supply pin and measure supply current in both 'states'. Check for stuck / shorted PCB nodes with power off. Mock-up circuit on breadboard and test with IC that was in circuit. Locally AC-bypass 12V supply or provide external supply. And this is the one that I'd probably fall foul of: Make sure you didn't put dud 071 back in by mistake! Last edited by Lucien Nunes; 19th Dec 2014 at 6:50 pm. Reason: Got interrupted by nearby electronic malfunction!

19th Dec 2014, 8:49 pm	#11
Radio Wrangler Moderator Join Date: Mar 2012 Location: Fife, Scotland, UK. Posts: 22,902	Re: Problem with TL071 SMPS comparator At low frequency/DC the feedback loop is open. The capacitor gives no feedback around the opamp at DC, so the DC condition is created using th rest of the PSU s the feedback path for the opamp. If anything doesn't work in the PSU, there is no DC feedback, and the opamp is banged against the rail. David __________________ Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done

21st Dec 2014, 2:05 pm	#12
Tyso_Bl Hexode Join Date: Jul 2014 Location: Magor, Monmouthshire, Wales, UK. Posts: 436	Re: Problem with TL071 SMPS comparator Couple of questions, did this unit ever work for you? is there any possibility that it had been got at previously? A couple of thoughts, using op amps as comparators isn't always a good idea, while different makes may all work as op-amps, they may not all work as comparators, I've had this with homebrewed circuits. Is your replacement an exact replacement? From what I've read so far your circuit may be latching up, some opamps can do that if the inputs are taken too close to the supply rails, usually the output will go in the expected direction then as an input aproaches the supply rail the output will reverse. It may be possible that the inputs are being applied before the supply rail to the comparator has stabilised, it's such a simple circuit, the passives are easy to check, so it has to be something to do with how the chip is being fed, if the chip itself is the correct type and not faulty. Finally are all the grounds grounded as they should be, ie nothing floating? and have you tried the opto isolator for any form of back leakage? __________________ Adapt, Improvise, Oh Bother..... Last edited by Tyso_Bl; 21st Dec 2014 at 2:08 pm. Reason: Cat on keyboard

21st Dec 2014, 2:29 pm	#13
Alistair D Nonode Join Date: Jan 2010 Location: Edinburgh, UK. Posts: 2,008	Re: Problem with TL071 SMPS comparator TL071 is a jfet input op-amp and to my mind is an unusual choice in this circuit. I would have preferred an LM321 or similar. This range of devices can sense input voltages down to ground level which may help prevent latch up at power on. Al __________________ I won't tell you how I discovered that.

21st Dec 2014, 2:38 pm	#14
kellys_eye Octode Join Date: Jan 2014 Location: Oban, Scotland, UK. Posts: 1,129	Re: Problem with TL071 SMPS comparator Any 'crud' between pins 2 and 3? With a high impedance input the area needs to be clean.

21st Dec 2014, 4:05 pm	#16
Eddy_Stone Retired Dormant Member Join Date: Jun 2014 Location: Boston, Massachusetts, USA Posts: 32	Re: Problem with TL071 SMPS comparator I believe this is configured as a voltage error amplifier and not a comparator. The opto should work in its linear region. Can you show the rest of the circuit. Eddy

21st Dec 2014, 4:06 pm	#17
Mooly Octode Join Date: May 2011 Location: Lancashire, UK. Posts: 1,351	Re: Problem with TL071 SMPS comparator Have you scoped the opamp output ? It may be showing 75mv on a DVM but could be switching between the rails at high frequency and with some wide mark:space ratio.