LM321MF heating up at 24V

Go To Last Post
22 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hello,

 

i made a RC debounce filter using LM321MF (Edited JS)  as buffer for output. At 12V the circuit works PERFECT, but at 24V, the OP amp gets kinda hot (50 to  70°C). The op amp is made for max 32V. The current draw at 12V is 7 to 12mA, but at 24V, it is 35mA. Any idea how to prevent heating up, or if this heating is normal. It says that it work for up to 125°C, but still, i prefer to keep my electronics at or near ambient temperature. In worst case scenario i can still run the filters on 12V.

 

Here is the circuit.RC

This topic has a solution.

Last Edited: Fri. Nov 17, 2017 - 12:31 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Supply current should be less than 3mA worst case at 30V.

What is the value of R3 and what is the max voltage input for your MCU_IN?

 

This circuit does not look reasonable for typical uCs (3 - 5V supplies).

 

David (aka frog_jr)

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

R3 = 510 Ohm,  output goes to an constant current driven optocoupler on a off the shelf motherboard, which has 24V inputs. But even when i have nothing conencted to output, the OP amp is stil lgetting hot. 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Please post a link to the data sheet for the op-amp.

 

Add the values to the schematic.

 

JC

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

R1, R2 = 1kOhm

R3, = 510 Ohm

C1, C3 = 0.1uF

C2 = 22uF

 

Time constant for switch press = aprox. 100 us

Time constant for switch release = aprox. 200 us

 

Datasheet:

http://www.ti.com/lit/ds/symlink/lm321.pdf

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

if you disconnect the output of the opamp (mpu_in) does it cool down?

You said something about mpu_in being an opto-coupler?  Can you show the circuit for that.

 

Jim

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I believe you need to add hysteresis to the opamp circuit so that it is operating as a schmitt trigger rather than a buffer.

 

As a buffer, the 321 will be following the charge/discharge curve of C1. Depending on its characteristics, the "constant current driven optocoupler on a off the shelf motherboard" may be operating in its linear region during the charge/discharge of C1, which may not provide a clean signal out of the optocoupler.

 

Regarding the high current, make sure don't have any shorts on the pins of the 321 (e.g. between pin 2 & 3).

David (aka frog_jr)

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

why not just debounce in software ... ?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

because i am working as hardware subtractor along whit some of friends for some other company and they said they cant do debounce in software(dont as kme why) and they want hardware solution. I was planning to go whit schmitt triger but they dont make them for 24V(required spec) and i had to keep PCB as smal las possible so i can just wrap it in heatshrink and join it semesly into cable (pcb is 6x10mm). Strange thing is that the op amp heats up even when there is nothing connected to output. I also checked and there are no shorts. But after some more testing i noticed that on 12V it heats up to 40°C, at 24V to 75°C and at 30V up to 95°C. 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

You can wire the opamp as a schmitt trigger:

 

schmitt4

 

(See http://www.electronics-tutorial....)

 

or:

 

An op amp Schmitt trigger circuit diagram showing the three resistors that give the hysteresis on the positive feedback

 

(See http://www.radio-electronics.com...)

 

Something is very definitely wrong with no load temps like that...

Can you photo the board and post here?

David (aka frog_jr)

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Google doesn't bring up useful anything for MC321MF but you show the LM321 in your diagram, so we need to know the EXACT chip you are using ie the LMV321 only works up to 5V!!

John Samperi

Ampertronics Pty. Ltd.

www.ampertronics.com.au

* Electronic Design * Custom Products * Contract Assembly

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

anyway ... your R3 resistor is very wrong

CPU will work with 5V and you have 24V out of the oamp! 

it is 35mA to the cpu pin... normal your cpu must be burn! 

Thierry

 

Thierry Pottier

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

TPE wrote:
CPU will work with 5V and you have 24V out of the oamp! it is 35mA to the cpu pin... normal your cpu must be burn!

 

He is driving an opto (led, not shown on schematic) with  output of opamp.

 

Jim

 

Last Edited: Tue. Nov 7, 2017 - 09:12 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Output and the rest is fine, the whole device is running for couple of days so far, but the engineers that designed sw and hw on motherboard forget about debounce for mechanical switches and now they dont want or dont have time to mod SW, so i am adding hardware debounce filter. The MOBO has been working whit 24V switches for 2 weeks no, so no problem there.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I am sorry, i had typo in the titile, it is LM321, not MC321. There is link to datasheet 6 posts higher ;)

Last Edited: Tue. Nov 7, 2017 - 11:15 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thanks, resistor R3 is in there only in case i have to limit output current. I tried and changed the 510 Ohm R3 for a 0 Ohm jumper and stil lsame problem. But now after putting R3 a 20k resistor the heating problem is almost gone. It is possible that i damaged 3 op amps when soldering, i will try again whit 3 pcb but i will put them in overflow to remove human error factur when soldering.

Last Edited: Tue. Nov 7, 2017 - 11:14 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Sounds to me like the circuit is oscillating. The data sheet shows the ideal layout with a ground plane.

Does your PCB have such a ground plane? How clean is your 24vdc?

Jim

If you want a career with a known path - become an undertaker. Dead people don't sue! - Kartman

Please Read: Code-of-Conduct

Atmel Studio6.2/AS7, DipTrace, Quartus, MPLAB user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

it is LM321, not MC321.

I have edited the top post with the correct number.

 

R3 = 510 Ohm,  output goes to an constant current driven optocoupler

So are you saying that the load after the resistor is just the LED of the opto coupler? For 12mA across the 510R resistor the chip's output need to be at about ~7V so at 12V the chip is dissipating 12-7=5V * 0.012=60mW at 24V the chip need to drop 24-7=17V * 0.012=204mW (getting warm now) at 30V 30-7=23V * 0.012=276mW warmer still....get the picture?

 

And is the input to the op amp changing when the supply changes? It seems not because the current is increasing instead of being a "constant current"!

John Samperi

Ampertronics Pty. Ltd.

www.ampertronics.com.au

* Electronic Design * Custom Products * Contract Assembly

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I think we need to see / know the detailed specifications on what the op-amp is driving.

 

JC

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

DocJC wrote:

I think we need to see / know the detailed specifications on what the op-amp is driving.

 

JC

 

I think we need a full schematic AND a real picture of the circuit board it lives on.

 

\JIm

If you want a career with a known path - become an undertaker. Dead people don't sue! - Kartman

Please Read: Code-of-Conduct

Atmel Studio6.2/AS7, DipTrace, Quartus, MPLAB user

This reply has been marked as the solution. 
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

js wrote:

 

So are you saying that the load after the resistor is just the LED of the opto coupler? For 12mA across the 510R resistor the chip's output need to be at about ~7V so at 12V the chip is dissipating 12-7=5V * 0.012=60mW at 24V the chip need to drop 24-7=17V * 0.012=204mW (getting warm now) at 30V 30-7=23V * 0.012=276mW warmer still....get the picture?

 

And is the input to the op amp changing when the supply changes? It seems not because the current is increasing instead of being a "constant current"!

 

adding to this that the package does 265C/W this will give a temperature increase of 265 * 0,204 = 54C. So yes it is getting nice and hot what ever you do.

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thanks, whit what "js" and "meslomp" said, the warming of the op amp is kinda logical. I actually dont know what exactly will be put on the output of the op amp, thats why i put it in as buffer config in the first place. I wil lsend the PCB and schematics to the company that develops main board and they will simply have to conenct it in a way they would deam good enough.

 

 

Here is also the PCB:

 

PCB