Stuffed IC in backwards..now fried

[Rob Lo]

I am building an analog synth from a kit I received recently...I just learned a very valuable lesson of "measure twice cut once"..

Except I learned "measure once first"

The board basically called for a LM741 opamp @ the OP4 position and OP3 called for a TL702...

First issue was - I installed them in opposite sockets....second issue - I installed the TL702 backwards.

It smoked and crackled and got pretty hot.....at first I thought that it was the correct part...so I just reoriented it just to see it was fried even further.

What are the chances other components were taken out with it? I don't see any traces lifted but the OP4 socket smells burnt.

I've attached a schematic if it helps glean any information.

[helius]

Unlikely to have damaged any other components. When you connected the TL072 backwards, it received power with the opposite polarity, which will destroy most chips. The clamp diodes, that force signals to be less than V+ and greater than V-, are connected forwards across the rails and act as a near short circuit.

[Rob Lo]

Thanks a lot, Helius. I feel a bit better.

Some rare IC's on the board. 

The clamp diodes you mentioned are the integrated circuit or the ones on the board?

[helius]

They are integrated into the opamp, in this case as a part of the output JFET.

[Zero999]

I think you meant TL072, rather than TL702.

Fortunately both the TL072 and LM741 are fairly common, cheap and easy to get hold of. Just buy a few spare, just in case, replace them and it should work.

[Rob Lo]

Yes....sorry TL072

Replacing the TI opamps is no big deal. Was more concerned about the russian IC's and transistors on the board..since those are relatively difficult to source.

It's based off an early 80's design..I haven't seen any data sheets so I don't know the pinouts.

[VK5RC]

I done that before! If you haven't, you aren't trying hard enough! HiHi.
A respected senior colleague once said after you finish, check, then go and have a cup of tea or something , come back re-check, then switch it on. Amazing the times I have done that I have found something else wrong. if the bit of test gear that is connected is really expensive-get someone else to check as well!

[Audioguru]

You are consistent. First you got the opamps installed backwards then stuffed the TL072 dual opamp backwards then you wrote the TL702 also backwards. Do you ride a bicycle backwards?

[Ian.M]

You are consistent. First you got the opamps installed backwards then stuffed the TL072 dual opamp backwards then you wrote the TL702 also backwards. Do you ride a bicycle backwards?

[StillTrying]

 Just make sure the new ICs are in the correct sockets and the right way up!

[Ian.M]

..... and the sockets aren't fitted the wrong way round if you are relying on their pin 1 marking (a common problem with single sided and home etched boards that don't have any topside pin 1 indicators)! 

[janoc]

Yes....sorry TL072

Replacing the TI opamps is no big deal. Was more concerned about the russian IC's and transistors on the board..since those are relatively difficult to source.

It's based off an early 80's design..I haven't seen any data sheets so I don't know the pinouts.

Well, that's why one uses a current limited power supply to bring up a new board, with the current limit set to a low value. If it goes into current limit/constant current because it is drawing more than reasonable amount of juice (where reasonable depends on what is on the board) or things are starting to get hot, cut the power and check!

This will save you a lot of magic smoke. Everyone makes mistakes but the key is to anticipate that something could be wrong and prevent mistakes from turning into disasters.

[JS]

..... and the sockets aren't fitted the wrong way round if you are relying on their pin 1 marking (a common problem with single sided and home etched boards that don't have any topside pin 1 indicators)! 

 That's why I always edit my boards from the component side in the EDA, I'm familiar with it the way it goes, I've seen too many errors because of being familiar to see the board from the back. Even easier because I don't have to mirror the PCB image to do the transfer, as I worked it all the way mirrored... I know the standard way is the other way but I just can get it, it's much more prone to confusion!

Yes....sorry TL072

Replacing the TI opamps is no big deal. Was more concerned about the russian IC's and transistors on the board..since those are relatively difficult to source.

It's based off an early 80's design..I haven't seen any data sheets so I don't know the pinouts.

Well, that's why one uses a current limited power supply to bring up a new board, with the current limit set to a low value. If it goes into current limit/constant current because it is drawing more than reasonable amount of juice (where reasonable depends on what is on the board) or things are starting to get hot, cut the power and check!

This will save you a lot of magic smoke. Everyone makes mistakes but the key is to anticipate that something could be wrong and prevent mistakes from turning into disasters.

  That takes the fun of the first power on away! Today someone brought the simulator as a tool to avoid breathing magic smoke and I said that you learn so much more when a cap explodes over your face than when the PC shows you a red flag... When playing with TL072 they well worth blowing, just a few pennies and you will remember to pay more attention next time, for the time you are playing with $20+ chips you already learned the lesson and you also have the tools for when your brain farts.

  From my point of view, well done Rob, I don't need to tell you pay more attention next time! Few cents for that chip are well worth the lesson!

JS

[StillTrying]

Yep, if you have to use full power the first time, do it without the chips in, then add them 1 by 1, switching off each time.

[Ian.M]

Its not just power up without chips in, one should also check each power and ground pin of every socket is at the voltage you expect it to be, with the pinout for each chip printed out and annotated with the IC socket numbers right next to the board while you do so - its another chance for you to catch f---ed up pinouts before your board elects a pope!

[Rob Lo]

Yea I checked the voltages on the sockets before seating the IC's..

I did almost everything right. One big thing wrong lol.

[DimitriP]

Well, that's why one uses a current limited power supply to bring up a new board, with the current limit set to a low value. If it goes into current limit/constant current because it is drawing more than reasonable amount of juice (where reasonable depends on what is on the board) or things are starting to get hot, cut the power and check!

Adjustable power supply: $50-$600
The confidence to power up your project with full power:  Priceless. 

[Rob Lo]

Well, that's why one uses a current limited power supply to bring up a new board, with the current limit set to a low value. If it goes into current limit/constant current because it is drawing more than reasonable amount of juice (where reasonable depends on what is on the board) or things are starting to get hot, cut the power and check!

Adjustable power supply: $50-$600
The confidence to power up your project with full power:  Priceless.

Double speak for stupid..lol...I get it.

To add to my ignorance, I have a variable dc bench supply here. I was looking on how to connect it to the 10-pin header of the module which needs  -12v and +12v supply.

I

[helius]

To add to my ignorance, I have a variable dc bench supply here. I was looking on how to connect it to the 10-pin header of the module which needs  -12v and +12v supply.

There were special bench supplies with "tracking" that were for that purpose. You would dial in Channel 1 to +12V, and Channel 2 would follow it and output -12V.
If you need to power the device from a single channel, you can set it to 24V and make a virtual ground with two equal resistors. The value of the resistors isn't that critical, but if it's too high then the circuit may be distorted. Try 1K ohm.
Exercise: how much power will be wasted in the resistors, and what power rating do they need to have?

[Cerebus]

Yes....sorry TL072

Replacing the TI opamps is no big deal. Was more concerned about the russian IC's and transistors on the board..since those are relatively difficult to source.

It's based off an early 80's design..I haven't seen any data sheets so I don't know the pinouts.

Well, that's why one uses a current limited power supply to bring up a new board, with the current limit set to a low value. If it goes into current limit/constant current because it is drawing more than reasonable amount of juice (where reasonable depends on what is on the board) or things are starting to get hot, cut the power and check!

This will save you a lot of magic smoke. Everyone makes mistakes but the key is to anticipate that something could be wrong and prevent mistakes from turning into disasters.

Doesn't help if you connect your bench PSU backwards, which is my usual method of getting the magic smoke out. Consequently when I'm breadboarding things I put a couple of reverse biased diodes across the rails on the board so that trips the current limit.

[DimitriP]

Well, that's why one uses a current limited power supply to bring up a new board, with the current limit set to a low value. If it goes into current limit/constant current because it is drawing more than reasonable amount of juice (where reasonable depends on what is on the board) or things are starting to get hot, cut the power and check!

Adjustable power supply: $50-$600
The confidence to power up your project with full power:  Priceless.

Double speak for stupid..lol...I get it.

To add to my ignorance, I have a variable dc bench supply here. I was looking on how to connect it to the 10-pin header of the module which needs  -12v and +12v supply.

I

No. Not doublespeak. Quite the opposite. 
I beleve having the confidence to turn on a board at full power is indeed priceless. In the good sense of the word.
After something goes south it makes you pay more attention before turing the power on.
Life doesn't have a safety net.

[Rob Lo]

To add to my ignorance, I have a variable dc bench supply here. I was looking on how to connect it to the 10-pin header of the module which needs  -12v and +12v supply.

There were special bench supplies with "tracking" that were for that purpose. You would dial in Channel 1 to +12V, and Channel 2 would follow it and output -12V.
If you need to power the device from a single channel, you can set it to 24V and make a virtual ground with two equal resistors. The value of the resistors isn't that critical, but if it's too high then the circuit may be distorted. Try 1K ohm.
Exercise: how much power will be wasted in the resistors, and what power rating do they need to have?

Well my bench ps only goes up to 15v so I guess a purchase is in order...Never thought forward that I'd be getting involved enough to need that much power.

I'm unsure how to arrive at how much power is wasted in the resistors..