Comparator dying on CMOS

Thread Starter

BiscuitBarrel

Joined Aug 6, 2021
21
This is the first time I've ever used an op amp as a comparator and it hasn't gone smoothly!
As you see this is powered by 13V. It is meant to be used in a car. It detects when the car engine is turning by the spikes appearing on the 13V line.
The second voltage source simulates the actual voltage measured on the car, which is included.
On my bench, obviously, I just wanted to get the circuit operating in steady state.
I destroyed 3 op amps until I realised that my old Thurlby and Thandar PSU was ringing for a very short duration with a peak at 27V when set to 13V!!!
Please be aware I had this problem with an old T & T before so please check if you are having problems with dead components.
Only it wasn't, surprisingly, what was responsible for their demise. It was only one op amp in the quad pack, configured as a comparator, that was dying.
So I was l flummoxed!!
My only option was to turn the voltage up manually and leaving the output switch on.
That solved that problem, but yet another, number 4 blew.
Again the comparator.

The symptoms? Drawing approx 30mA at 13V (50uA average). The output is at near half rail in steady state.
So I left the damaged device in and did the following.
I applied a sine wave, to simulate the engine voltage spikes, directly to pin 3 then I get a square wave out at pin 7 of the comparator.
The current consumption drops to 15mA and the output toggles between 0V and the half rail steady state voltage.

In steady state this comparator has an approx 500mV input differential voltage. The common mode voltage is, obviously, about 1/2 rail.
In transient, the voltage at pin 6 is approx 12V, which is about 1/2V above the operational common mode voltage.(Vcc-1.5V to 0V)
So in theory you cant drive the input from the output of a neighbouring op amp if the output goes high!!!!
The maximum differential input voltage for this device is +/- 18V. The maximum voltage on any input is Vcc and minimum 0V. The maximum supply voltage is 18V.
So there's no violation of max values.
I didn't apply a hysteresis resistor to this circuit due to its nature, but I think I'm going to regret not using a couple of resistors to do so, but this should not kill this circuit? Surely?
I have checked all the protection diodes on every pin and they are ALL intact!!!
There are no components in this circuit that would put any node above Vcc.
The floating node, from the dual Schottky diode, just goes to a charging capacitor that charges between 2.5V and 8.5V and it works fine.
Any ideas??????
 

Attachments

Thread Starter

BiscuitBarrel

Joined Aug 6, 2021
21
Where is the DC path to GND on X9?
There is a capacitor that is charging. Its basically a cheap old RC circuit.
The idea was for the comparator to slowly discharge the capacitor when there are "spikes" on the Vcc line as X9s output toggles hight and low. So no dc path when the output is high, but only when its low.
Have I been a dope?
All the other op amps are in use and are fully functional.
Nothing floating

Welcome to AAC!
Were the inputs of the two unused op-amps in the pack left floating? That is a no-no with CMOS circuits.
All the other op amps are in use and are fully functional.
Nothing floating.

All the other op amps are in use and are fully functional.
Nothing floating.
And thank you for your warm welcome!!!!
 

Papabravo

Joined Feb 24, 2006
21,159
No I don't think that, it would be harsh. I'd just like to point out that IMHO the opamp inputs, regardless of what they are doing need a DC path to GND to function correctly. You have deprived them of this path, unintentionally of course.
 

eetech00

Joined Jun 8, 2013
3,859
This is the first time I've ever used an op amp as a comparator and it hasn't gone smoothly!
As you see this is powered by 13V. It is meant to be used in a car. It detects when the car engine is turning by the spikes appearing on the 13V line.
The second voltage source simulates the actual voltage measured on the car, which is included.
On my bench, obviously, I just wanted to get the circuit operating in steady state.
I destroyed 3 op amps until I realised that my old Thurlby and Thandar PSU was ringing for a very short duration with a peak at 27V when set to 13V!!!
Please be aware I had this problem with an old T & T before so please check if you are having problems with dead components.
Only it wasn't, surprisingly, what was responsible for their demise. It was only one op amp in the quad pack, configured as a comparator, that was dying.
So I was l flummoxed!!
My only option was to turn the voltage up manually and leaving the output switch on.
That solved that problem, but yet another, number 4 blew.
Again the comparator.

The symptoms? Drawing approx 30mA at 13V (50uA average). The output is at near half rail in steady state.
So I left the damaged device in and did the following.
I applied a sine wave, to simulate the engine voltage spikes, directly to pin 3 then I get a square wave out at pin 7 of the comparator.
The current consumption drops to 15mA and the output toggles between 0V and the half rail steady state voltage.

In steady state this comparator has an approx 500mV input differential voltage. The common mode voltage is, obviously, about 1/2 rail.
In transient, the voltage at pin 6 is approx 12V, which is about 1/2V above the operational common mode voltage.(Vcc-1.5V to 0V)
So in theory you cant drive the input from the output of a neighbouring op amp if the output goes high!!!!
The maximum differential input voltage for this device is +/- 18V. The maximum voltage on any input is Vcc and minimum 0V. The maximum supply voltage is 18V.
So there's no violation of max values.
I didn't apply a hysteresis resistor to this circuit due to its nature, but I think I'm going to regret not using a couple of resistors to do so, but this should not kill this circuit? Surely?
I have checked all the protection diodes on every pin and they are ALL intact!!!
There are no components in this circuit that would put any node above Vcc.
The floating node, from the dual Schottky diode, just goes to a charging capacitor that charges between 2.5V and 8.5V and it works fine.
Any ideas??????
Why not use comparators? LM339 Quad comparator perhaps?
 

LowQCab

Joined Nov 6, 2012
4,023
Cheap Op-Amps don't like having their Inputs near the Rails.

Trying to detect "Electrical-Noise" is not a very reliable method.
You need to detect the Voltage increase from the Alternator,
like this ...................
.
.
.
Engine Run Detector 1 FLAT .png
 

dcbingaman

Joined Jun 30, 2021
1,065
This is the first time I've ever used an op amp as a comparator and it hasn't gone smoothly!
As you see this is powered by 13V. It is meant to be used in a car. It detects when the car engine is turning by the spikes appearing on the 13V line.
The second voltage source simulates the actual voltage measured on the car, which is included.
On my bench, obviously, I just wanted to get the circuit operating in steady state.
I destroyed 3 op amps until I realised that my old Thurlby and Thandar PSU was ringing for a very short duration with a peak at 27V when set to 13V!!!
Please be aware I had this problem with an old T & T before so please check if you are having problems with dead components.
Only it wasn't, surprisingly, what was responsible for their demise. It was only one op amp in the quad pack, configured as a comparator, that was dying.
So I was l flummoxed!!
My only option was to turn the voltage up manually and leaving the output switch on.
That solved that problem, but yet another, number 4 blew.
Again the comparator.

The symptoms? Drawing approx 30mA at 13V (50uA average). The output is at near half rail in steady state.
So I left the damaged device in and did the following.
I applied a sine wave, to simulate the engine voltage spikes, directly to pin 3 then I get a square wave out at pin 7 of the comparator.
The current consumption drops to 15mA and the output toggles between 0V and the half rail steady state voltage.

In steady state this comparator has an approx 500mV input differential voltage. The common mode voltage is, obviously, about 1/2 rail.
In transient, the voltage at pin 6 is approx 12V, which is about 1/2V above the operational common mode voltage.(Vcc-1.5V to 0V)
So in theory you cant drive the input from the output of a neighbouring op amp if the output goes high!!!!
The maximum differential input voltage for this device is +/- 18V. The maximum voltage on any input is Vcc and minimum 0V. The maximum supply voltage is 18V.
So there's no violation of max values.
I didn't apply a hysteresis resistor to this circuit due to its nature, but I think I'm going to regret not using a couple of resistors to do so, but this should not kill this circuit? Surely?
I have checked all the protection diodes on every pin and they are ALL intact!!!
There are no components in this circuit that would put any node above Vcc.
The floating node, from the dual Schottky diode, just goes to a charging capacitor that charges between 2.5V and 8.5V and it works fine.
Any ideas??????
If the power supply was ringing at up to 27V when turned on, are you sure you are not having other issues with the power supply output while turning it up slowly? I would recommend trying the circuit with a fixed 12V DC source like a wall-wort to eliminate the possibility of it being the power supply. If this is your standard bench power supply, I would recommend replacing it with a more reliable model. It should not be misbehaving like that.
I looked at the datasheet for the OpAmp and agree you are using it within its specified safe limits. The datasheet is not really clear on what happens if an input exceeds the common mode input voltage range of Vcc-1.5V. It sounds like the device may not operate properly, but according to absolute maximum ratings, it should not be damaging it.
 
Last edited:

dcbingaman

Joined Jun 30, 2021
1,065
This is the first time I've ever used an op amp as a comparator and it hasn't gone smoothly!
As you see this is powered by 13V. It is meant to be used in a car. It detects when the car engine is turning by the spikes appearing on the 13V line.
The second voltage source simulates the actual voltage measured on the car, which is included.
On my bench, obviously, I just wanted to get the circuit operating in steady state.
I destroyed 3 op amps until I realised that my old Thurlby and Thandar PSU was ringing for a very short duration with a peak at 27V when set to 13V!!!
Please be aware I had this problem with an old T & T before so please check if you are having problems with dead components.
Only it wasn't, surprisingly, what was responsible for their demise. It was only one op amp in the quad pack, configured as a comparator, that was dying.
So I was l flummoxed!!
My only option was to turn the voltage up manually and leaving the output switch on.
That solved that problem, but yet another, number 4 blew.
Again the comparator.

The symptoms? Drawing approx 30mA at 13V (50uA average). The output is at near half rail in steady state.
So I left the damaged device in and did the following.
I applied a sine wave, to simulate the engine voltage spikes, directly to pin 3 then I get a square wave out at pin 7 of the comparator.
The current consumption drops to 15mA and the output toggles between 0V and the half rail steady state voltage.

In steady state this comparator has an approx 500mV input differential voltage. The common mode voltage is, obviously, about 1/2 rail.
In transient, the voltage at pin 6 is approx 12V, which is about 1/2V above the operational common mode voltage.(Vcc-1.5V to 0V)
So in theory you cant drive the input from the output of a neighbouring op amp if the output goes high!!!!
The maximum differential input voltage for this device is +/- 18V. The maximum voltage on any input is Vcc and minimum 0V. The maximum supply voltage is 18V.
So there's no violation of max values.
I didn't apply a hysteresis resistor to this circuit due to its nature, but I think I'm going to regret not using a couple of resistors to do so, but this should not kill this circuit? Surely?
I have checked all the protection diodes on every pin and they are ALL intact!!!
There are no components in this circuit that would put any node above Vcc.
The floating node, from the dual Schottky diode, just goes to a charging capacitor that charges between 2.5V and 8.5V and it works fine.
Any ideas??????
Occam's Razor: If we already know the power supply has damaged the opamp due to 27V ringing. It is likely still causing spike issues even when turning it up slowly. Try a reliable power supply next to eliminate that possibility.
 

dcbingaman

Joined Jun 30, 2021
1,065
This is the first time I've ever used an op amp as a comparator and it hasn't gone smoothly!
As you see this is powered by 13V. It is meant to be used in a car. It detects when the car engine is turning by the spikes appearing on the 13V line.
The second voltage source simulates the actual voltage measured on the car, which is included.
On my bench, obviously, I just wanted to get the circuit operating in steady state.
I destroyed 3 op amps until I realised that my old Thurlby and Thandar PSU was ringing for a very short duration with a peak at 27V when set to 13V!!!
Please be aware I had this problem with an old T & T before so please check if you are having problems with dead components.
Only it wasn't, surprisingly, what was responsible for their demise. It was only one op amp in the quad pack, configured as a comparator, that was dying.
So I was l flummoxed!!
My only option was to turn the voltage up manually and leaving the output switch on.
That solved that problem, but yet another, number 4 blew.
Again the comparator.

The symptoms? Drawing approx 30mA at 13V (50uA average). The output is at near half rail in steady state.
So I left the damaged device in and did the following.
I applied a sine wave, to simulate the engine voltage spikes, directly to pin 3 then I get a square wave out at pin 7 of the comparator.
The current consumption drops to 15mA and the output toggles between 0V and the half rail steady state voltage.

In steady state this comparator has an approx 500mV input differential voltage. The common mode voltage is, obviously, about 1/2 rail.
In transient, the voltage at pin 6 is approx 12V, which is about 1/2V above the operational common mode voltage.(Vcc-1.5V to 0V)
So in theory you cant drive the input from the output of a neighbouring op amp if the output goes high!!!!
The maximum differential input voltage for this device is +/- 18V. The maximum voltage on any input is Vcc and minimum 0V. The maximum supply voltage is 18V.
So there's no violation of max values.
I didn't apply a hysteresis resistor to this circuit due to its nature, but I think I'm going to regret not using a couple of resistors to do so, but this should not kill this circuit? Surely?
I have checked all the protection diodes on every pin and they are ALL intact!!!
There are no components in this circuit that would put any node above Vcc.
The floating node, from the dual Schottky diode, just goes to a charging capacitor that charges between 2.5V and 8.5V and it works fine.
Any ideas??????
Another thing I noticed from the datasheet, this device is very ESD sensitive:

This is the first time I've ever used an op amp as a comparator and it hasn't gone smoothly!
As you see this is powered by 13V. It is meant to be used in a car. It detects when the car engine is turning by the spikes appearing on the 13V line.
The second voltage source simulates the actual voltage measured on the car, which is included.
On my bench, obviously, I just wanted to get the circuit operating in steady state.
I destroyed 3 op amps until I realised that my old Thurlby and Thandar PSU was ringing for a very short duration with a peak at 27V when set to 13V!!!
Please be aware I had this problem with an old T & T before so please check if you are having problems with dead components.
Only it wasn't, surprisingly, what was responsible for their demise. It was only one op amp in the quad pack, configured as a comparator, that was dying.
So I was l flummoxed!!
My only option was to turn the voltage up manually and leaving the output switch on.
That solved that problem, but yet another, number 4 blew.
Again the comparator.

The symptoms? Drawing approx 30mA at 13V (50uA average). The output is at near half rail in steady state.
So I left the damaged device in and did the following.
I applied a sine wave, to simulate the engine voltage spikes, directly to pin 3 then I get a square wave out at pin 7 of the comparator.
The current consumption drops to 15mA and the output toggles between 0V and the half rail steady state voltage.

In steady state this comparator has an approx 500mV input differential voltage. The common mode voltage is, obviously, about 1/2 rail.
In transient, the voltage at pin 6 is approx 12V, which is about 1/2V above the operational common mode voltage.(Vcc-1.5V to 0V)
So in theory you cant drive the input from the output of a neighbouring op amp if the output goes high!!!!
The maximum differential input voltage for this device is +/- 18V. The maximum voltage on any input is Vcc and minimum 0V. The maximum supply voltage is 18V.
So there's no violation of max values.
I didn't apply a hysteresis resistor to this circuit due to its nature, but I think I'm going to regret not using a couple of resistors to do so, but this should not kill this circuit? Surely?
I have checked all the protection diodes on every pin and they are ALL intact!!!
There are no components in this circuit that would put any node above Vcc.
The floating node, from the dual Schottky diode, just goes to a charging capacitor that charges between 2.5V and 8.5V and it works fine.
Any ideas??????
ESD.PNG
1kV is not hard to generate. Are you using an ESD bench with ESD protection? Also, being this device is so sensitive to ESD I am not sure it is a good choice for automotive applications.
 

Thread Starter

BiscuitBarrel

Joined Aug 6, 2021
21
Another thing I noticed from the datasheet, this device is very ESD sensitive:


View attachment 245136
1kV is not hard to generate. Are you using an ESD bench with ESD protection? Also, being this device is so sensitive to ESD I am not sure it is a good choice for automotive applications.
Thank you for your replies. Yes I have taken notice of the ESD frailties of the device. I now wear a ESD strap just in case.
Its just always the same op amp. If ESD was a problem then I would be experience problems across the board.
It is a part of the circuit I have not been modifying. I have changed resistor values on all the others with my soldering iron and have had no problems. My whole bench has a ESD dissipative mat.
As regards my PSU, yes I have scoped the output to determine if it is behaving when winding up the voltage and whether there is transients in steady state. Yes, I really should update my PSU, its just the cost and its a nice piece of kit!!!
I hope to determine what's going on with the regulation circuit in the PSU as I have a schematic for it!!!!
I am using a 9V battery from here on in so we shall see what happens. Thank you for your feedback.
 

Thread Starter

BiscuitBarrel

Joined Aug 6, 2021
21
No I don't think that, it would be harsh. I'd just like to point out that IMHO the opamp inputs, regardless of what they are doing need a DC path to GND to function correctly. You have deprived them of this path, unintentionally of course.
I am sorry that I misunderstood your question.
When you asked about DC path to ground I assumed you were talking about the output of X9. I was confused as to why that should be the case. Just me being dumb. The reason I went down that anomalous road is that inputs do have a path to ground through the output of X10? There is no resistor from the inverting input of X9 to the output of X10, which will introduce a dc offset in X9, but this is more than catered for by the biasing of the non-inverting input by R6. In steady state there is a 500mV difference at the inputs. If I'm wrong then I'm glad I have someone that will straight forwardly tell me!!
Its what I need at the moment. I understand what you are telling IS BEST practice and I'll take it on board.
Sometimes we become sloppy.
I'm still hedging my bets that I've done something dumb and if I was everyone else here, on this forum, I would probably think the same!!!
Thank you Papabravo.
from RomeoMike
 

dcbingaman

Joined Jun 30, 2021
1,065
Thank you for your replies. Yes I have taken notice of the ESD frailties of the device. I now wear a ESD strap just in case.
Its just always the same op amp. If ESD was a problem then I would be experience problems across the board.
It is a part of the circuit I have not been modifying. I have changed resistor values on all the others with my soldering iron and have had no problems. My whole bench has a ESD dissipative mat.
As regards my PSU, yes I have scoped the output to determine if it is behaving when winding up the voltage and whether there is transients in steady state. Yes, I really should update my PSU, its just the cost and its a nice piece of kit!!!
I hope to determine what's going on with the regulation circuit in the PSU as I have a schematic for it!!!!
I am using a 9V battery from here on in so we shall see what happens. Thank you for your feedback.
It will be interesting to see how it does with the 9V battery. It looks like we are running out of possibilities. The only thing left I can think of is the op amp does not like the one rail going above Vcc-1.5V. You may try an experiment with just one op amp. Have the other ones in the package inputs grounded. Place a pot from Vcc to Ground with the wiper on the + input and ground the negative input. Place a small load on the output to ground say 10K. Pay close attention to the current from the power supply (or battery) as you adjust the voltage on the pot towards Vcc and watch the output of the op amp. It should obviously stay high. Interested especially what happens as the pot gets within 1.5V of the Vcc line. The circuit looks like that should not be possible being the DC gain on the first stage is 1 after the cap charges and I see you are handling the turn on case by having the other op amp having another RC to deal with the first stage.
 
Last edited:

dcbingaman

Joined Jun 30, 2021
1,065
Thank you for your replies. Yes I have taken notice of the ESD frailties of the device. I now wear a ESD strap just in case.
Its just always the same op amp. If ESD was a problem then I would be experience problems across the board.
It is a part of the circuit I have not been modifying. I have changed resistor values on all the others with my soldering iron and have had no problems. My whole bench has a ESD dissipative mat.
As regards my PSU, yes I have scoped the output to determine if it is behaving when winding up the voltage and whether there is transients in steady state. Yes, I really should update my PSU, its just the cost and its a nice piece of kit!!!
I hope to determine what's going on with the regulation circuit in the PSU as I have a schematic for it!!!!
I am using a 9V battery from here on in so we shall see what happens. Thank you for your feedback.
After looking at this again I ran LTSpice on the circuit. Could not use the same op-amp but similar one. The negative input to the 'comparator' op amp is at the supply rail voltage for the first 150ms. Until the cap on the first stage charges up. The positive input starts at 0V thus a difference in input voltage of the power supply rails at first startup.
 

Thread Starter

BiscuitBarrel

Joined Aug 6, 2021
21
After looking at this again I ran LTSpice on the circuit. Could not use the same op-amp but similar one. The negative input to the 'comparator' op amp is at the supply rail voltage for the first 150ms. Until the cap on the first stage charges up. The positive input starts at 0V thus a difference in input voltage of the power supply rails at first startup.
This the voltage measured on my scope. The device is powered by 9V battery as promised!!!! Vcc was measured at 9.1V
 

Attachments

Thread Starter

BiscuitBarrel

Joined Aug 6, 2021
21
At the end of the day, it doesn't exceed max values and secondly the output is damaged, the inputs are just grand!!!??*^$
The output has a dirty great big resistor on it too.
I blame aliens or Soviet EMP experiments. The other explanation is that I've done something dumb. I have in the past wondered why I never got a reading on the scope and discovered that the power supply output does need to be switched on :0D
I will replace the device and run it from a 9V battery and see if I can kill it again.
Killing op amps is pretty new to me. At least my surface mount heat gun work is improving. Before you ask, I have soldered by hand just in case I've fried it too much!!!!
Thank you for your support.
 

dcbingaman

Joined Jun 30, 2021
1,065
In other words you are human. I am not sure what else to check at this point. So this is a schematic that has been converted to a PCB or just a protoboard?
 
Top