tech: Voltage drop across ICs to check fi they are good
trebor
Reading in Randy Fromm's Blue Book he suggests using the diode test on a
multi-meter, red to ground and black on the input/ouput legs on the IC.
If the IC is ok you will get a voltage drop of .5 to .9, much lower or
higher and the chip is faulty.
Now this sound like a great way to find faulty chips on a board, however
does this apply to every leg on the chip?.
As in if some legs read bad (they might not even be input/output) but some
are okay then the chip is ok?.
So if I get a reading of .5 to .9 off any leg then the chip is ok?
Can anyone shed more light on this
Many thanks.
Rob
Adam J. Courchesne
1) The leg may have an ESD diode tied to it in which case all you may be
doing is verifying that this diode is okay
2) You may be verifying that one junction in the driving transistor is
okay but what about the logic that drives this transistor?
-Adam
Greg Menke
Its unlikely to be a particularly good test. I think its probably a
reasonable way to confirm a chip is bad because a chip with this kind
of problem on the vcc/gnd pins is certainly not going to work, but a
"good" reading won't confirm the chip is good because there are lots
of failures that don't involve shorting/opening the vcc/gnd pins.
Its likely he's only testing the vcc/gnd pins (upper right=vcc, lower
left=gnd), which are standard on DIP ic's- the other pins will vary
based on the ic itself.
This test might be of some help if you're doing triage on a massively
blown out board and you're trying to find the ic's you need to replace
straight away without further diagnostics.
Gregm
Randy Fromm
No, you have this IC test completely wrong.
This test works on input and output pins only, NOT on the Vcc and
ground.
This test takes advantage of the fact that integrated circuits belong
to a family of ICs known as TTL or "Transistor, Transistor Logic".
Simply stated, TTL means that the integrated circuits are made of
transistors. When transistors fail, they will usually short circuit
and we can detect the shorted transistor in the IC the same way we
test a regular transistor.
Set your digital multimeter to the "diode test" range. Connect the red
(positive) meter lead to the ground connection of the board. No...
This is not a misprint! In order to read the "junction drop" of the
pin of the IC under test you must ground the positive lead of the
meter.
Use the black meter lead to probe each of the input and output pins of
the IC. You don't even have to know which pins are inputs and which
are outputs. Just probe them all, one at a time. A good input or
output will generally display a normal junction drop reading
(generally between .4 and .9 volt). If the integrated circuit is bad,
the meter will generally display a much lower reading such as .05 or
.1 volt. Sometimes the bad IC will display a much higher reading
instead of the normal junction drop. If the reading is too low or too
high, the bad output pin of the IC will display a reading that is
obviously much different than all of the other pins of all the other
integrated circuit.
This test does not test the internal logic of the device, only the
inputs and outputs. But that is where the majority of the failures
occur. When a node is stuck "low" for instance, it's SOMETHING that's
holding it down. The SOMETHING is likely to be a shorted input or
output. Remember, a typical TTL device has a fan-out of 10, so any one
of the (up to) 10 inputs might be stuck low -OR- the ground switch
transistor (the current sink) in the totem-pole output might have an
emitter-to-collector short. Either failure will cause the node to be
stuck low.
There are a few exceptions where this test doesn't work. Open
collector or tri-state devices are among those that may test
differently (not a normal junction drop).
This IC test doesn't require any specialized knowledge of integrated
circuits to make it work. I have used this simple test to locate bad
integrated circuits in all kinds of equipment. For example, I was
brought a bad board from a skee-ball. I didn't have a schematic for
the board, nor did I have a game in which to test it. There weren't
that many ICs on the board so I used the meter test, probing each of
the pins in turn. I replaced all of the ICs that tested bad and it
worked just fine when it was installed in the game.
To Adam Courchesne:
In response to your criticisms:
I wouldn't trust this method to verify funtion as I see 2 problems
with it:
1) The leg may have an ESD diode tied to it in which case all you may
be doing is verifying that this diode is okay
-OR- you are locating a shorted PN junction on the input pin, showing
that the IC is BAD. INPUTS fail the most. I'm not verifying that the
IC is good, I am locating a BAD IC.
2) You may be verifying that one junction in the driving transistor is
okay but what about the logic that drives this transistor?
You are not testing the internal logic, this is absolutely true.
However, that is not a typical failure mode (except for shift
registers). Also, this is just a quick and dirty test to locate BAD
ICs, not to verify working ICs. If the PCB isn't working, we KNOW
there's a bad IC somewhere on it. If we can't find it easily with the
meter method, we can move on to other methods. No harm, no foul.
Note to all:
I write this stuff based on practical experience and research. I do
not write about things based on speculation, which is what you have
done, Greg. What the fuck is that all about? Stop trying to second
guess things when you haven't done the research and you don't know
what you're talking about. If you dispute something that you see in
print, why not contact the author and ask him/her to explain it to
you?
Heck, maybe you're correct or maybe you have a better way to do
things. If I am wrong about something, I TOTALLY want to know about
it. I would never take offense at having been shown to be incorrect
about something. It's the same as learning another language. How are
you going to learn if nobody ever corrects you? I have no problem with
being told I'm wrong about something.
But this sort of "non-advise" serves no purpose. You're just flapping
your gums, metaphorically speaking. The IC test works as advertised.
Try it and you'll see.
Randy Fromm
James Bright
Randy, man you need to lighten up. Neither one of these responses were
criticisms of you. Re-read the emails and you will see that In fact, I too
thought this was not a great test for ICs based on practical experience as
well. I use different tools to test circuits, as does Adam and probably Greg
as well. Your explanation validating your method was fine, and would be
greatly appreciated if you dropped "you have this IC test completely wrong",
"In response to your criticisms", and "What the fuck is that all about?
Stop trying to second guess things when you haven't done the research and
I'd really like to see more of on r.g.v.a.c. and for whatever reason you
took it personally and gave a good technical response that would have been a
lot better without all of the emotion.
(In case you don't remember, I attended your class the last time you taught
it in Weirs Beach and I am not just a hobbyist. I have also recommended your
books and your class in the past. I hope you're just having a bad day.)
--
JB
--James Bright
www.QuarterArcade.com
Restored Arcade Games for your Home
www.ggdb.com
Great Game Database. Pic, tech data, collectors & more
"Randy Fromm" <goo...@randyfromm.com> wrote in message
news:29990189.03081...@posting.google.com...
Randy Fromm
> greatly appreciated if you dropped "you have this IC test completely wrong",
Perphaps YOU didn't understand. He stated that the test was just on
the Vcc pins. HE WAS COMPLETELY WRONG. The test specifically EXCLUDES
the Vcc pin. THAT is what I meant.
My point is that this is a perfectly good test that even the most
inexperienced person can perform. When I see the test misrepresented
and "dissed" by someone that doesn't even know what the test actually
is, it does everyone a disservice.
That's all. Don't read any more into it than that.
rf
John Robertson
equipment can fix many common faults on pinball games. Most problems
(in Gottlieb for example) are caused by shorts from the solenoids to
the switch matrix - these show up very quickly with the diode test and
in many cases one can use the blown pins to identify the location of
the voltage surge...as just one strobe and one return line are damaged
then the common switch for these should be the first spot checked on
the game.
I try and identify this for customers so they know where the likely
cause of their blown MPU was...
Randy has been talking about this test for years as have others and it
is really useful trick to have under your belt!
John :-#)#
On 11 Aug 2003 19:46:09 -0700, goo...@randyfromm.com (Randy Fromm)
wrote:
(Please post followups or tech enquires to the newsgroup)
John's Jukes Ltd. 2343 Main St., Vancouver, BC, Canada V5T 3C9
Call (604)872-5757 or Fax 872-2010 (Pinballs, Jukes, Video Games)
www.flippers.com
"Old pinballers never die, they just flip out."
Craig Yarbrough
It doesn't sound like you're testing the input and output transistors
so much as you're verifying the clamp diodes.
Totem-Pole outputs (and complementary outputs) use two bipolar
transistors to drive the output, one sourcing from VCC and one sinking
to ground (push-pull). Each transistor consists of a sandwich of NPN
material, with N being negative and P being positive. Putting this
together, the output looks like:
VCC
|
N
L---P
N
|
---- Output
|
N
L---P
N
|
GND
(Hope that looks ok!) 'L' denotes the internal logic signals of the
device. The output is connected between the emitter of the push and
the collector of the pull. Note from the diagram that the path between
the output and GND and the path between the output and VCC both cross
over a full NPN junction. In the device's quiescent state the NPN
junction does not conduct and will not allow a voltmeter to read a
normal diode drop (0.6V). A better test of the transistors would be to
probe the voltmeter between the internal logic and the output, VCC,
and GND, which isn't really practical.
When you probe between GND and the output and see a diode drop of
~0.6V what you're seeing is the nominal drop across the clamp diodes.
Clamp diodes are placed between power (VCC or GND) and the output to
prevent transients from destroying the TTL structure. Take a look at
the following link for a decent schematic of clamp diodes: (sections
6.1.1 and 6.1.2)
http://www-ee.eng.hawaii.edu/~msmith/ASICs/HTML/Book2/CH06/CH06.1.htm
Likewise if you place the positive lead of your voltmeter on the
output and the negative lead at VCC you should also see a ~0.6V diode
drop. If you read significantly less than that it could mean that only
the clamp diode is bad, or only the collector-emitter junction of your
transistor is bad, or both. If you read significantly greater than
that it could mean that your clamp diode is bad, or that the device
simply doesn't have clamp diodes!
While your IC test might be ok for a cursory search for potential
troublespots, there's no substitute for reading the datasheets and
knowing fully what you're testing.
- Craig
goo...@randyfromm.com (Randy Fromm) wrote in message news:<29990189.03081...@posting.google.com>...
Greg Menke
I see your point about its applicability. I was responding to the
original question which implied a question about it being useful for
all IC's. For those circuits where its applicable, it seems like a
good first order test.
Second, my comments weren't "dissing" your test. I agree I
misunderstood it and thus stand corrected.
Gregm
John Robertson
ICs deep inside a circuit board, however most failures of boards are
caused by I/O damage due to voltage spikes and this test quickly
identifies those.
It is not a replacement for a logic probe etc, but it can make testing
boards easier if you are looking for beginner level tests. In 75% of
cases (pinball games) the problems are I/O related and can be found in
most cases then with the diode check.
John :-#)#
On 12 Aug 2003 04:03:21 -0700, hyar...@harris.com (Craig Yarbrough)
wrote:
(Please post followups or tech enquires to the newsgroup)
ra...@randyfromm.com
wrote:
While your IC test might be ok for a cursory search for potential
troublespots, there's no substitute for reading the datasheets and
knowing fully what you're testing.
- Craig
Of course not. I agree with you 100%. But just to put "The Test" in an
historical perspective, here's how the test came to be and what
happened (to me) while troubleshooting and repairing hundreds of
amusement machine PCBs while I worked for a game distributor in
London:
Using a logic probe and/or oscilloscope for digital repair, I notice a
common failure - a node that's stuck low. Often, you can see that an
output is trying to drive it high, but it never even reaches the gray
area.
Clipping and lifting pins, the output works perfectly when isolated
from all the inputs it is driving. When the offending input is clipped
and lifted, all other inputs on the node receive full amplitutde
pulses (good logic high).
I think: "Hmmm. If the input is stuck low, it should be obvious with a
passive test like a resistance check or by using the diode test. No
need to power up the PCB for that.
Sure enough, the bad input showed a low resistance.
That's ALL I am saying. Nothing more. Remember, I am NOT declaring
stuff to be good. I am only finding bad stuff.
If your car has a flat tire, you really don't need to start the engine
in order to determine that the car's not goin' anywhere.
Regards,
Randy Fromm
ra...@randyfromm.com
Oh, one other thing. I ground the red lead in order to see the
junction drop, NOT for the reason you think (checking clamping diodes)
but rather to indicate to the person making the test that he/she has
actually TOUCHED the pin under test and has made good contact. So
often I see where students, in the process of testing semiconductors,
blow the test because they are looking for an open reading (reverse
biased) and they achieve it by not making contact at all. They are
looking for open and they see it, regardless of the actual condition
of the component under test.
There's method in the madness, dude.
rf
On 12 Aug 2003 04:03:21 -0700, hyar...@harris.com (Craig Yarbrough)
wrote:
>While your IC test might be ok for a cursory search for potential
Dean Thompson
Learned a lot. Learned how to save myself a lot of time and money by
doing things his way. BUT, most important, learned how little I knew
before attending it.
Randy was a great teacher that taught from a stand point of "how to
fix the machines" and not trying to teach us to be electronic
engineers. Dean