Is it possible? Testing an IC with a multimeter?!

[ali6x944]

Hi everyone,
I was making a inverting converter with an MC34063A and I did it with SOIC-8 package and in a small space, I was very happy with it until I discovered that the chip is broken and dosen't work, so I ask is it possible to test ic's with a multimeter or a scope?
Or a method u know of to test small size smd packages and even DIP packages without soldering them ?
Help would be appreciated

[uncle_bob]

Hi

The data sheet on the 34063 gives you a very good idea about what is inside the part. Provided you have a way to attach to all the leads on the package, you can test it with fairly basic gear. The problem is the mechanical one of how to get at the leads. With a DIP package, you can use a socket and get the job done quite easily. With SOIC's things like probes are often used.

Bob

[ali6x944]

Hi

The data sheet on the 34063 gives you a very good idea about what is inside the part. Provided you have a way to attach to all the leads on the package, you can test it with fairly basic gear. The problem is the mechanical one of how to get at the leads. With a DIP package, you can use a socket and get the job done quite easily. With SOIC's things like probes are often used.

Bob

Thanx very much, but do u have any ideas about how to test more complex ic's that usually don't have internal structure diagram in the datasheet but instead offer in this case a less useful block diagram?

[uncle_bob]

Hi

The data sheet on the 34063 gives you a very good idea about what is inside the part. Provided you have a way to attach to all the leads on the package, you can test it with fairly basic gear. The problem is the mechanical one of how to get at the leads. With a DIP package, you can use a socket and get the job done quite easily. With SOIC's things like probes are often used.

Bob

Thanx very much, but do u have any ideas about how to test more complex ic's that usually don't have internal structure diagram in the datasheet but instead offer in this case a less useful block diagram?

Hi

It's the same thing. You set up the supplies and feed in test signals. You look at the responses you get to the signals. This goes up and down on an input and that happens on an output. Exactly what the signals and the responses need to be is highly varied. There are many different categories of IC's out there. A microwave L band LNA would be tested very differently than an audio op-amp or a Pentium CPU....

Bob

[Audioguru]

I designed an equalizer circuit for a new speaker. It used a TL074 quad opamp in a SOIC package. Tens of thousands of the product were built, tested and sold and all worked perfectly except two. One had the IC mounted backwards and the other had a shorted capacitor. The ICs were not tested before assembly of the circuits. After a few years none of the product were ever returned for replacement.

[uncle_bob]

I designed an equalizer circuit for a new speaker. It used a TL074 quad opamp in a SOIC package. Tens of thousands of the product were built, tested and sold and all worked perfectly except two. One had the IC mounted backwards and the other had a shorted capacitor. The ICs were not tested before assembly of the circuits. After a few years none of the product were ever returned for replacement.

Hi

We build many 10's of thousands of circuits every week. Absolutely none of the components that go into any of the are ever tested prior to assembly. That just isn't the way modern production does it. Instead you make very sure that the parts come in "right" before they ever go into stock.

Bob

[ali6x944]

I'm not trying to make an industrial produced product...
So only one... two... maybe 4 chips will be used....
Also I'm kinda cheap poor guy so it is much cheaper to salvage parts -for me- than to by them... thus I don't know if the chip is ok to use or not

[uncle_bob]

I'm not trying to make an industrial produced product...
So only one... two... maybe 4 chips will be used....
Also I'm kinda cheap poor guy so it is much cheaper to salvage parts -for me- than to by them... thus I don't know if the chip is ok to use or not

Hi

Wire it up in the circuit and see if it works. Probing surface mount parts ... not easy at all.

Bob

[ali6x944]

I can't do that without soldering the part...

[uncle_bob]

I can't do that without soldering the part...

Hi

That's correct. You are saving money by using scrap parts. The work you do to solder and un-solder the parts is a direct consequence of your saving that money. If you can get a brand new part for <3 cents, that work probably does not make sense. If the new part costs > $20 then it may be time well spent. Getting new parts cheaply does involve a bit of planning and buying a large group of parts at one time. The effort put in there also counts as "work".

Scrounging parts is a random process. We all started out that way. I still have parts I pulled out of gear ... err... 50 years ago. Somehow I doubt I will ever use them. I spent time pulling them and sorting them into groups. That process taught me a lot about parts and how stuff was put together, so time well spent. I have hauled them all over the place on many moves. The ones I have used out of the pile could have been bought new, just on the freight and storage costs. Sometimes you do things, to save money, that costs you money ....

Bob

[ali6x944]

You are right, but here in saudi unfortunately unlike the US  electronics as a hobby Is very rare thus hobbyists like me find it incredibly hard to find cheap gear or even parts an example here in the eastern province there is only a couple of shops selling row electronic parts like microcontrollers ,diodes, transistors, caps...est , the price of thous parts are two to four times thier original manufacture price and even parametric search engines like digikey, or jameco, so it is much more economical to desolider parts rather than buy them, the mc34063 it self sold for 10SR or around 3USD , in digikey it is 0.5USD! , it a real problem that hobbyists faces here and in a lot of of developing countries, maybe here in saudi we don't suffer from it as much as others...

[uncle_bob]

You are right, but here in saudi unfortunately unlike the US  electronics as a hobby Is very rare thus hobbyists like me find it incredibly hard to find cheap gear or even parts an example here in the eastern province there is only a couple of shops selling row electronic parts like microcontrollers ,diodes, transistors, caps...est , the price of thous parts are two to four times thier original manufacture price and even parametric search engines like digikey, or jameco, so it is much more economical to desolider parts rather than buy them, the mc34063 it self sold for 10SR or around 3USD , in digikey it is 0.5USD! , it a real problem that hobbyists faces here and in a lot of of developing countries, maybe here in saudi we don't suffer from it as much as others...

Hi

It has been maybe 40 years since I bought electronic stuff regularly in a shop. I probably spent $30 in that time on parts from a retail store. Some of what I have came from the local flea markets. That's a very hit or miss sort of thing and has gotten more difficult in the last 20 years. Most of my parts came from auction sites or various mail order vendors. In some cases the stuff was shipped from half way around the world for next to nothing. In other cases the fixed shipping costs were pretty significant. In those cases I spent time to come up with an order that made the shipping come out to a workable part of the total cost. Maybe 20% of what I use has come directly from electronic distributors.


Bob

[ali6x944]

Well lets stop doing economics, and talk electronics how to test these complex ic's now, or any ic in general?

[uncle_bob]

Well lets stop doing economics, and talk electronics how to test these complex ic's now, or any ic in general?

Hi

Testing an ic is simple. Build up a probe station, drop the IC in. Hook up the test bench. Run the software that puts in the signals and measures the responses. That's the *general* approach to testing any IC. With the possible addition of cooling, that is how all of them are tested. Simple enough in an industrial setting. Not going to work on a chip scale part in your basement lab.

If you can substitute  a socket for the probe station .. then it becomes basement lab compatible. That's not practical on most surface mount packages (no socket).

Bob

[ali6x944]

Last question:
Could the resistance power pins on the ic be used as an indicator of chip useability?

[gildasd]

The first thing to test is Ohms between the Vcc pins and the ground pins, it should not be a low value (3Ohms) except in documented special cases.

For the other pins, you need a "good" IC as a reference. If a value is significantly off (half an order of magnitude) or worse, a short not present on the "good" one, then it's probably dead.

[uncle_bob]

Last question:
Could the resistance power pins on the ic be used as an indicator of chip useability?

Hi

It is very dependent on the exact IC and what it was designed to do. Some chips are designed to go "open circuit" below a set input voltage. Equally important, many IC faults have nothing to do with the power pin. A single blown input (that takes out a wire bond) will simply read as open circuit. That's pretty much what some CMOS inputs will read ...

Next layer to the problem:

IC pins may have a very limited amount of current they will accept without damage. In a power down state, simply probing pins could inject more current than they are rated for. How much current is injected is very much a function of your meter and what you have it set to.

Next up:

You can have ICs that are made with a bipolar process, a CMOS process, a BI-CMOS process, ECL, or Analog. Each of them will have it's quirks when it comes to protection and testing. They each will give you different sneak paths on a powered down chip. There are sub variations in each process so keeping track of what it what can take a lot of work. It also can involve chip level information that the manufacturer does not publicize.

The list goes on and on ...The bottom line is still that you need to power up the device to start doing tests.

Bob

[rstofer]

Most often when you try to read the resistance on pins, you are just testing the ESD clamping diodes and this has little to do with the internals.  It's not even clear to me that it is a good idea to use an ohmmeter because on the high resistance scale, the open circuit voltage on the probes might be too high.  On some meters there is a 9V battery for some scales and a 1-1/2V battery for others.

I use a lot of SMD devices and I just don't think there is any good way to test them.  I will usually socket DIPs but that is a often not possible with SMD devices.

Short answer:  No, you can't test a chip for functionality short of a multi-bazillion dollar tester.  Just solder it in and hope for the best.  To be honest, I have never run across a defective chip unless I did something to let the magic smoke escape.  It's true that I don't buy on the secondary market but that's because I have ready access to primary suppliers.  I don't know how I would handle the problem of not having credible suppliers.

I guess a hot air rework station would be a requirement!

[uncle_bob]

Hi

One alternative would be to solder the SMD stuff to DIP conversion boards. Then you can plug the board into a socket and set up a test on the part. If it's bad, the only thing you might toss out is the conversion board. You still need a full bench to do the testing, but it gets you around the probe station.

Bob

[rstofer]

You could do as Bob suggested and solder the SMD to a DIP adapter board.  Then you could just plug the entire assembly into your project board.  The project board will be somewhat larger due to the size of the adapter board but at least the devices are interchangeable.

http://schmartboard.com/smt-to-dip-adapters/

I have no idea how difficult these may be to obtain outside the US.  I have used them for breadboarding and they work well.

[ali6x944]

Thanx for all the help