Contact an Expert

Thread: Preserving a test fixture


Permlink Replies: 2 - Pages: 1 - Last Post: Sep 15, 2011 1:29 AM Last Post By: adrianc Threads: [ Previous | Next ]
visanic

Posts: 1
Registered: 07/13/10
Preserving a test fixture
Posted: Jul 13, 2010 7:29 AM
Click to report abuse...   Click to reply to this thread Reply
Hello-

I have a simple question.

If I already have a fixture built and want to re-use it when the board spins, which 3070 file do I send to the layout engineer so that they do not move any existing testpoint locations?

Is it board_xy?

Thanks-
Ikon

Posts: 1
Registered: 10/08/10
Re: Preserving a test fixture
Posted: Oct 8, 2010 9:29 AM   in response to: visanic in response to: visanic
Click to report abuse...   Click to reply to this thread Reply
Hi,

normally I find it better to work with the CAD. The physical designer will have testpoint attribute assigned for the CAD. This is normal for all CAD packages.

The only problem then is how the fixture nails have been allocated for the test fixture. Common mistake is that testpoint are not marked as mandatory and vias are still left in as possibles. Then what happens is that to make the probe allocation use bigger probes VIAS are used instead of testpoints. :(

What ever CAD processing tool you used to process the CAD to BOARD_XY should be able to provide list of testpoints with known reference that you could provide designer.

But I always check no matter what the designer says when there has been a re-spin of CAD and check test access, component moves/changes and topology changes.

Hope this helps,
adrianc


Posts: 69
Registered: 04/08/08
Re: Preserving a test fixture
Posted: Sep 15, 2011 1:29 AM   in response to: visanic in response to: visanic
Click to report abuse...   Click to reply to this thread Reply
Attachment getprobexy.txt (1.6 KB)
Hi,
I think one way is to get the coordinates from the fixture file.

NODE "$2N509"
PINS
215152;
21552;
PROBES
P186 32500, 17550;
ALTERNATES
29250, 53935 l1.1 50MIL;
30950, 47600;
WIRES
215152 TO P186 Blue 28 5.00;
21552 TO P186 Blue 28 5.00;

you will see the the probe locations are specified (P186) with the coordinates with respect to the board origin (32500, 17550).

Sometimes, the system will do an offset of all the board coordinates (including the probe locations) when compiling the board_xy file. You can see this when you compile the board_xy with list option (or find in board_xy.l file).

compile "board_xy";list

scroll to almost the end of the board_xy.l file ... you may see something like this ...

All locations have been moved by 2500, 6500 (tenth-mils) so they are positive.

If so, then the x-y coordinates in the fixture file should minus the offset to get the actual x-y coordinates with respect to the board origin.

for P186 : (32500-2500), (17550-6500)
which makes it ...
P186: 30000, 11050

----------------------------------
I have a small script that can extract the probe locations from the fixture file and drop it into a text file delimitted by comma.

It is in PERL, so you would have to run in windows.

1) rename the file to "getprobexy.pl"
2) drop it into the "fixture" directory where the fixture file resides
- you can list it from the object file by executing the following command in BTBasic window
list object "fixture/fixture.o" to "fixture/fixture"
3) run the script from a korn shell window and read the resulting text file ... "probe_xy_fixt.txt"
$ getprobexy.pl
or
$ perl getprobext.pl
The script does not take into account the initial offset added when the board_xy file is compiled. Please use Excel to add it in.

Hope this helps.

getprobexy.txt

Point your RSS reader here for a feed of the latest messages in all forums