In-circuit IC tester?

[Russ]

I have been searching for a long time for a device that will test
digital and analog ICs in circuit. I have seen some old ones but I
would like something newer. Something with a USB port that you can
hook up to a computer would be great and the ability to add devices to
the list is a must have. There has got to be something out there. I
would like the price to be under $1000.00. I have so many boards that
I look at and think "Is this chip working or not" Some boards I can't
apply power to. It would be nice to be able to have adaptors for the
newer SMD device and a 48 pin ziff socket also. I have a Huntron with
a HSR-410 but I don't always have duplicate boards or chips to compare to
But I would think there are other companies out there that
make something like this. Now I have searched google for hours, day
after day, for a long time. So I have seen pretty much anything that
comes up with a related word search using google. So unless I missed
something someone is going to have to have experience in the
electronic troubleshooting field and maybe worked with something similar.
If I was smart I would make one.
I have seen many out of circuit checkers. And like I mentioned some
old discontinued devices that check digital IC chips in circuit. I
also have the HP Comparator but that has so many limitations. I would
like this device to have most chips stored in memory and you punch the chip
number in and hook up a clip onto the chip that is in circuit and it
runs some tests on it, with and without power. For both digital and
analog and other types of devices. Has anyone seen anythink like
this?Thanks for the help.

 

[Robert Baer]

I have been searching for a long time for a device that will test
digital and analog ICs in circuit. I have seen some old ones but I
would like something newer. Something with a USB port that you can
hook up to a computer would be great and the ability to add devices to
the list is a must have. There has got to be something out there. I
would like the price to be under $1000.00. I have so many boards that
I look at and think "Is this chip working or not" Some boards I can't
apply power to. It would be nice to be able to have adaptors for the
newer SMD device and a 48 pin ziff socket also. I have a Huntron with
a HSR-410 but I don't always have duplicate boards or chips to compare to
But I would think there are other companies out there that
make something like this. Now I have searched google for hours, day
after day, for a long time. So I have seen pretty much anything that
comes up with a related word search using google. So unless I missed
something someone is going to have to have experience in the
electronic troubleshooting field and maybe worked with something similar.
If I was smart I would make one.
I have seen many out of circuit checkers. And like I mentioned some
old discontinued devices that check digital IC chips in circuit. I
also have the HP Comparator but that has so many limitations. I would
like this device to have most chips stored in memory and you punch the chip
number in and hook up a clip onto the chip that is in circuit and it
runs some tests on it, with and without power. For both digital and
analog and other types of devices. Has anyone seen anythink like
this?Thanks for the help.

There are too many *different* ICs of different types in a GIVEN
package for this to be possible.
Take something "simple" like the SOT23-3 package; 1) dual cathode
diodes; 2) dual anode diodes; 3) Jfets(P and N channel); 4) MOSfets(P
and N channel); bipolar transistors (NPN, PNP; single, darlington); 5)
single diodes (may be various lead configurations).
Now that may not be an exhaustive list, but it gives you an idea as
to variety.
The SOT23-5 allows even more variety; 8-pin DIP/SO-8 etc is worse yet.

And "in-circuit"??
Forgeddit.
Too much other things can be connected to the DUT that could be DUDs
to mess with analysis of any type; even if all is OK, other devices can
cause abnormal readings (where "abnormal" means same detection scheme on
an isolated known good part).

 

[Russ]

Thanks for the feedback. I almost ready to give up on troubleshooting. I am
in the Coin-op equipment business and there are many pcb's that I encounter.
But my success rate isn't very good. I have lots of books, and equipment.
But 80% of the time I have to send it in. My last hope was for this kind of
device. Something to give me a fighting chance. Your right pcb's are
getting to complex. I guess at one time it was simpler and that is why they
had the Fluke 900?, and other TTL and CMOS 20 pin device testers. Like the
Huntron HSR-410 and others. Then I guess they gave up because it just can't
be done. I am not sure it would sell if it could be done. With all of the
SMD's and double, triple layer boards you probable don't stand a chance.
Thanks for the input.

Russ

 

[Walter Harley]

I have been searching for a long time for a device that will test
digital and analog ICs in circuit. [...]

Think of the very simple case of a single logic gate, say a two-input OR
gate, in circuit. How would your hypothetical tester test it? There are
two inputs and one output; there's nothing else that can be measured, so you
have to base your decision on that. Okay, so suppose you measure zero at
both inputs and at the output. That's working, right? Well, it's correct,
but maybe the chip is completely dead; you'd get the same results.

So, you say, fine, I'll inject a signal into one of the inputs. Ah, but can
you do that? Suppose that the reason the inputs are zero is that something
else connected to the inputs is shorted to ground. So, you can't inject
enough signal to get the logic level to change. No way to test. Or,
suppose that the inputs are high, but the output is zero. It could be that
the gate is defective; or, it could be a short-circuit somewhere else,
pulling the gate down.

And so on. The bottom line is that there are very, very many circuits where
it is simply not possible to test a device in circuit, because the fact that
it is in circuit interferes.

Okay, that was assuming that you could get to all the pins. But modern
chips often have the pins on the bottom - Ball Grid Array. Once it's
soldered to the board, there's no way to access the pins, even if you could
do something useful with them.

And speaking of chips with lots of pins: they tend to be memories,
microcontrollers, and the like. That is, they have software and/or data
inside them. Suppose that a particular chip is a microcontroller, and one
of its functions is that a particular output is supposed to either be a
buffered or an inverted copy of one of its inputs, depending on a
user-configurable value in flash memory. How would you possibly test that?
It's not enough to know the datasheet for the chip: to know what its
behavior is supposed to be, you have to know the software that it is
currently running, as well as its entire history of input data.

So, that's why general-purpose IC testers in general, and in-circuit ones in
particular, are so hard to find. They are only useful on an extremely
limited set of cases.

 

[Russ]

Thanks for the well thought out reply. I am begining to see the light. I
am ready to call it quits on being able to service the equipment I operate.
It least it will save me money in buying more test equipment.
Thanks
russs

I have been searching for a long time for a device that will test
digital and analog ICs in circuit. [...]

Think of the very simple case of a single logic gate, say a two-input OR
gate, in circuit. How would your hypothetical tester test it? There are
two inputs and one output; there's nothing else that can be measured, so
you have to base your decision on that. Okay, so suppose you measure zero
at both inputs and at the output. That's working, right? Well, it's
correct, but maybe the chip is completely dead; you'd get the same
results.

So, you say, fine, I'll inject a signal into one of the inputs. Ah, but
can you do that? Suppose that the reason the inputs are zero is that
something else connected to the inputs is shorted to ground. So, you
can't inject enough signal to get the logic level to change. No way to
test. Or, suppose that the inputs are high, but the output is zero. It
could be that the gate is defective; or, it could be a short-circuit
somewhere else, pulling the gate down.

And so on. The bottom line is that there are very, very many circuits
where it is simply not possible to test a device in circuit, because the
fact that it is in circuit interferes.

Okay, that was assuming that you could get to all the pins. But modern
chips often have the pins on the bottom - Ball Grid Array. Once it's
soldered to the board, there's no way to access the pins, even if you
could do something useful with them.

And speaking of chips with lots of pins: they tend to be memories,
microcontrollers, and the like. That is, they have software and/or data
inside them. Suppose that a particular chip is a microcontroller, and one
of its functions is that a particular output is supposed to either be a
buffered or an inverted copy of one of its inputs, depending on a
user-configurable value in flash memory. How would you possibly test
that? It's not enough to know the datasheet for the chip: to know what its
behavior is supposed to be, you have to know the software that it is
currently running, as well as its entire history of input data.

So, that's why general-purpose IC testers in general, and in-circuit ones
in particular, are so hard to find. They are only useful on an extremely
limited set of cases.