PCB in-circuit test: how do ATE machines measure resistors?

Hi folks,

does anybody know how ATE machine measures the value of a resistor
placed on the target PCB?

I know that ATE machines, through a bed of nail, inject a certain
voltage onto the test point.

Also, what happens if the resistor under measure is connected to other
resistors?

thanks for any help,
Enrico

On Thu, 2 Apr 2009 14:04:28 -0700 (PDT), zigbee@libero.it wrote:

>Hi folks,
>
>does anybody know how ATE machine measures the value of a resistor
>placed on the target PCB?
>
>I know that ATE machines, through a bed of nail, inject a certain
>voltage onto the test point.
>
>Also, what happens if the resistor under measure is connected to other
>resistors?

I don't know about all of them (obviously) but a general approach is to
sample a known good board (better, sample several) and compare the
values found on the unit under test to the expected values, +/- some
allowed sample variation.

-- 
Rich Webb     Norfolk, VA

zigbee@libero.it wrote:

> Hi folks,
> 
> does anybody know how ATE machine measures the value of a resistor
> placed on the target PCB?
> 
> I know that ATE machines, through a bed of nail, inject a certain
> voltage onto the test point.
> 
> Also, what happens if the resistor under measure is connected to other
> resistors?
> 
> thanks for any help,
> Enrico

Many years ago I programmed a incircuit tester. There was a solution like
this:
You have 3 resistors in a delta connection. The 3 nodes are numered 1, 2, 3
with its resistors R12, R13, R23. Connect a voltage between 1 an 2, connect
the same voltage between 1 and 3. --> U23 = 0 --> 
I(R23) = 0.
I(R12) = U/R12
I(R13) = U/R13

The current nail 1 = I(R12) + I(R13)
The current nail 2 = I(R12) + I(R23) = I(R12)
The current nail 3 = I(R13) + I(R23) = I(R13)

Hope this helps.
-- 
Freundliche Gr�sse -- Regards
F. Seuhs
Mailto: friedrich.seuhs@hasos.com

On 3 Apr, 13:21, Friedrich Seuhs <friedrich.se...@hasos.com> wrote:
> zig...@libero.it wrote:
> > Hi folks,
>
> > does anybody know how ATE machine measures the value of a resistor
> > placed on the target PCB?
>
> > I know that ATE machines, through a bed of nail, inject a certain
> > voltage onto the test point.
>
> > Also, what happens if the resistor under measure is connected to other
> > resistors?
>
> > thanks for any help,
> > Enrico
>
> Many years ago I programmed a incircuit tester. There was a solution like
> this:
> You have 3 resistors in a delta connection. The 3 nodes are numered 1, 2,=
 3
> with its resistors R12, R13, R23. Connect a voltage between 1 an 2, conne=
ct
> the same voltage between 1 and 3. --> U23 =3D 0 -->
> I(R23) =3D 0.
> I(R12) =3D U/R12
> I(R13) =3D U/R13
>
> The current nail 1 =3D I(R12) + I(R13)
> The current nail 2 =3D I(R12) + I(R23) =3D I(R12)
> The current nail 3 =3D I(R13) + I(R23) =3D I(R13)
>
> Hope this helps.
> --
> Freundliche Gr=FCsse -- Regards
> F. Seuhs
> Mailto: friedrich.se...@hasos.com

Hi, thanks for your answer.

First of all, I just discovered that PCB in-circuit test are carried
out keeping the board UNPOWERED.

I also found on this Google book, a couple of schematics:

"Building a successful board-test strategy" by Stephen F. Scheiber

They usually use 3 signals:

1. A voltage source (constant or sinusoidal)
2. An inverting Op.Amp. that shows Virtual Earth
3. Optionally a "Guard" GND that helps isolate the Part under test



As far as you know, is it possibile to distinguish 2 capacitance in
parallel?

Enrico

zigbee@libero.it wrote:

> On 3 Apr, 13:21, Friedrich Seuhs <friedrich.se...@hasos.com> wrote:
>> zig...@libero.it wrote:
>> > Hi folks,
>>
>> > does anybody know how ATE machine measures the value of a resistor
>> > placed on the target PCB?
>>
>> > I know that ATE machines, through a bed of nail, inject a certain
>> > voltage onto the test point.
>>
>> > Also, what happens if the resistor under measure is connected to other
>> > resistors?
>>
>> > thanks for any help,
>> > Enrico
>>
>> Many years ago I programmed a incircuit tester. There was a solution like
>> this:
>> You have 3 resistors in a delta connection. The 3 nodes are numered 1, 2,
>> 3 with its resistors R12, R13, R23. Connect a voltage between 1 an 2,
>> connect the same voltage between 1 and 3. --> U23 = 0 -->
>> I(R23) = 0.
>> I(R12) = U/R12
>> I(R13) = U/R13
>>
>> The current nail 1 = I(R12) + I(R13)
>> The current nail 2 = I(R12) + I(R23) = I(R12)
>> The current nail 3 = I(R13) + I(R23) = I(R13)
>>
>> Hope this helps.
>> --
>> Freundliche Gr&#4294967295;sse -- Regards
>> F. Seuhs
>> Mailto: friedrich.se...@hasos.com
> 
> Hi, thanks for your answer.
> 
> First of all, I just discovered that PCB in-circuit test are carried
> out keeping the board UNPOWERED.
> 
> I also found on this Google book, a couple of schematics:
> 
> "Building a successful board-test strategy" by Stephen F. Scheiber
> 
> They usually use 3 signals:
> 
> 1. A voltage source (constant or sinusoidal)
> 2. An inverting Op.Amp. that shows Virtual Earth
> 3. Optionally a "Guard" GND that helps isolate the Part under test
> 
That's exactly what I ment in my  posting.
> 
> 
> As far as you know, is it possibile to distinguish 2 capacitance in
> parallel?

I would say: NO !!

> 
> Enrico

-- 
Freundliche Gr&#4294967295;sse -- Regards
F. Seuhs
Mailto: friedrich.seuhs@hasos.com

In article <49da0c3b$0$2623$91cee783@newsreader01.highway.telekom.at>,=20
friedrich.seuhs@hasos.com says...
> zigbee@libero.it wrote:
>=20
> > On 3 Apr, 13:21, Friedrich Seuhs <friedrich.se...@hasos.com> wrote:
> >> zig...@libero.it wrote:
> >> > Hi folks,
> >>
> >> > does anybody know how ATE machine measures the value of a resistor
> >> > placed on the target PCB?
> >>
> >> > I know that ATE machines, through a bed of nail, inject a certain
> >> > voltage onto the test point.
> >>
> >> > Also, what happens if the resistor under measure is connected to oth=
er
> >> > resistors?
> >>
> >> > thanks for any help,
> >> > Enrico
> >>
> >> Many years ago I programmed a incircuit tester. There was a solution l=
ike
> >> this:
> >> You have 3 resistors in a delta connection. The 3 nodes are numered 1,=
 2,
> >> 3 with its resistors R12, R13, R23. Connect a voltage between 1 an 2,
> >> connect the same voltage between 1 and 3. --> U23 =3D 0 -->
> >> I(R23) =3D 0.
> >> I(R12) =3D U/R12
> >> I(R13) =3D U/R13
> >>
> >> The current nail 1 =3D I(R12) + I(R13)
> >> The current nail 2 =3D I(R12) + I(R23) =3D I(R12)
> >> The current nail 3 =3D I(R13) + I(R23) =3D I(R13)
> >>
> >> Hope this helps.
> >> --
> >> Freundliche Gr=FCsse -- Regards
> >> F. Seuhs
> >> Mailto: friedrich.se...@hasos.com
> >=20
> > Hi, thanks for your answer.
> >=20
> > First of all, I just discovered that PCB in-circuit test are carried
> > out keeping the board UNPOWERED.
> >=20
> > I also found on this Google book, a couple of schematics:
> >=20
> > "Building a successful board-test strategy" by Stephen F. Scheiber
> >=20
> > They usually use 3 signals:
> >=20
> > 1. A voltage source (constant or sinusoidal)
> > 2. An inverting Op.Amp. that shows Virtual Earth
> > 3. Optionally a "Guard" GND that helps isolate the Part under test
> >=20
> That's exactly what I ment in my  posting.
> >=20
> >=20
> > As far as you know, is it possibile to distinguish 2 capacitance in
> > parallel?
>=20
> I would say: NO !!
>=20

If they're the same type and value, the answer is probably no.
If they're of different types or values, you might be able to
distinguish them by their different self-resonant frequencies
or ESRs.


Mark Borgerson

Hi,

thanks for your answers.

I read many articles about in-circuit testing. The most important one
is the following by Teradyne:
www.teradyne.com/atd/resource/recording/general/ICTBasicsPart_2V4.ppt

Here are some important points:

1) It is impossible to distinguish 2 capacitors in parallel: the test
will measure the sum of the two.

2) In-circuit tests are carried out on UNPOWERED targets. In words,
ATE systems DON'T power up the boards during the test.

3) Probe signals have an amplitude of 0.2 Volt, in order not to power
the IC's that are loaded on the board.

4) Some components can be tested by DC signals, others require AC
signals (a sine wave at a certain frequency).

5) The polarity of Electrolytic Filter Capacitor is an issue. I'll
post a thread as soon as discover how to test it

6) A good ATE system may cost from 30.000 to 50.000 USD

ciao,
Enrico