Connect WECO Trace 2 (3DFT+) for remote edging

[OptoMan99]

Hello,
I am trying to connect an old Weco Trace 2 to my remote edging system. Reading online I found out that this machine does not use the standard VCA-OMA protocol, but it uses some old Weco communication.

Our software receives the data correctly from the tracer, but it can't convert it in VCA-OMA for our edgers.

We receive the data from the tracer in this way:

02
DEV:WECO-FormTracer V 04.15
CON:010707072E0C068400000000000000009D3B00009D4B00007CC90000 0603

NMBR:000001 SIZE:0000 DBL:0134 OFFS:0000

RIGHT: STRT:0A62
FDF6EDE5DEDADADBD9D7DADBE2E6E5E9ECEEF0F3F5F5F5F3F1EFEFE9E8E6 E5E2
E0E2E5EBEFF5FB01070D141B21262C32363B3B3C3D3C3936312B251F1810 08FF
F6EDE4DCD3CBC8C4C2C0C2C1C5CACCD3DAE0E5ECF3FAFF040B0F1312120F 0A05
02FEFCF7F5F4F5F7FA00020810151A21262B30353A3F41403F3B342E261C 1208
INF:040060AB423F110000000000AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA 0000
CIRC:1539 BOXX:0543 BOXY:0351

LEFT: STRT:1A6E
02FAF0E8E3DDDAD8D7D6D9DADEE0E4E7EBEDF0F2F5F5F6F5F2EFECEAE8E4 E3E1
E0E3E6EAEFF6FB01090E16191D232A3032373D3B3D3C3A34322E27201A12 09FF
F5EAE2DAD2C9C4C1C1C2C2C2C3C8CED5D9DFE6ECF3F8FE050C1115151512 0F08
03FEFAF7F5F5F6F6FAFE03070E131A20262B3034393A40423F39372F271E 180D
INF:040B6BA9443B110000000000AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA 0000
CIRC:1536 BOXX:0541 BOXY:0350
ERR:0000
CHCK:597A


We have searched on internet for some manual, or contaced weco, but with no success.

Has anyone else ever had this problem? How can I convert this data in OMA?

[MakeOptics]

Hello,
I am trying to connect an old Weco Trace 2 to my remote edging system. Reading online I found out that this machine does not use the standard VCA-OMA protocol, but it uses some old Weco communication.

Our software receives the data correctly from the tracer, but it can't convert it in VCA-OMA for our edgers.

We receive the data from the tracer in this way:

02
DEV:WECO-FormTracer V 04.15
CON:010707072E0C068400000000000000009D3B00009D4B00007CC90000 0603

NMBR:000001 SIZE:0000 DBL:0134 OFFS:0000

RIGHT: STRT:0A62
FDF6EDE5DEDADADBD9D7DADBE2E6E5E9ECEEF0F3F5F5F5F3F1EFEFE9E8E6 E5E2
E0E2E5EBEFF5FB01070D141B21262C32363B3B3C3D3C3936312B251F1810 08FF
F6EDE4DCD3CBC8C4C2C0C2C1C5CACCD3DAE0E5ECF3FAFF040B0F1312120F 0A05
02FEFCF7F5F4F5F7FA00020810151A21262B30353A3F41403F3B342E261C 1208
INF:040060AB423F110000000000AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA 0000
CIRC:1539 BOXX:0543 BOXY:0351

LEFT: STRT:1A6E
02FAF0E8E3DDDAD8D7D6D9DADEE0E4E7EBEDF0F2F5F5F6F5F2EFECEAE8E4 E3E1
E0E3E6EAEFF6FB01090E16191D232A3032373D3B3D3C3A34322E27201A12 09FF
F5EAE2DAD2C9C4C1C1C2C2C2C3C8CED5D9DFE6ECF3F8FE050C1115151512 0F08
03FEFAF7F5F5F6F6FAFE03070E131A20262B3034393A40423F39372F271E 180D
INF:040B6BA9443B110000000000AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA 0000
CIRC:1536 BOXX:0541 BOXY:0350
ERR:0000
CHCK:597A


We have searched on internet for some manual, or contaced weco, but with no success.

Has anyone else ever had this problem? How can I convert this data in OMA?

FIRST STEP MAKE SURE IT'S NOT A SETTING THAT CAN BE SET TO OMA!

If you can provide some information about the frame?

It looks like:

CIRC = Circumference = 153.6mm
BOXX = A Size = 54.1mm
BOXY = B Size = 35.0mm

If you can confirm that data, the next step would be in deciphering how the right and left data is encoded. My first guess is hexidecimal since no letter's above an "F" exists in the code.

Decimal = 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Hexidecimal = 0 1 2 3 4 5 6 7 8 9 A B C D E F

I would start by taking a look at how many data points the tracer traces, this will give you a package size to divide your data into and then we can analyze further, your data has 260 characters for the right eye (removing spaces).

[OptoMan99]

FIRST STEP MAKE SURE IT'S NOT A SETTING THAT CAN BE SET TO OMA!

If you can provide some information about the frame?

It looks like:

CIRC = Circumference = 153.6mm
BOXX = A Size = 54.1mm
BOXY = B Size = 35.0mm

If you can confirm that data, the next step would be in deciphering how the right and left data is encoded. My first guess is hexidecimal since no letter's above an "F" exists in the code.

Decimal = 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Hexidecimal = 0 1 2 3 4 5 6 7 8 9 A B C D E F

I would start by taking a look at how many data points the tracer traces, this will give you a package size to divide your data into and then we can analyze further, your data has 260 characters for the right eye (removing spaces).

Yes the circumference and the Boxx and boxy are all correct, and the dbl 13.4 is also correct. Only the shape is what i'm missing now...

Time ago we read on an article about the weco trace 2 and it said the tracer reads 128 points. Now if I remove spaces and newlines it means according to what you say that I have 256 Hexidecimal characters that express in some way my 128 points. I think that the STRT:0A62 (0A62 hex => 2658 dec) could mean the starting point, and that the next characters are the differential points from the starting point (similar to the VCA differential format).

But if I simply convert the hex values to decimals, how do i then use these numbers to obtain the real radius points?

Thank you very much

[MakeOptics]

Yes the circumference and the Boxx and boxy are all correct, and the dbl 13.4 is also correct. Only the shape is what i'm missing now...

Time ago we read on an article about the weco trace 2 and it said the tracer reads 128 points. Now if I remove spaces and newlines it means according to what you say that I have 256 Hexidecimal characters that express in some way my 128 points. I think that the STRT:0A62 (0A62 hex => 2658 dec) could mean the starting point, and that the next characters are the differential points from the starting point (similar to the VCA differential format).

But if I simply convert the hex values to decimals, how do i then use these numbers to obtain the real radius points?

Thank you very much

Yes so you start at: 2658, with 256 characters and 128 points then every 2 characters is the differential for the next point so you add that to the start measure and you get the radius. Just to be sure when I run this I get a weird pattern not an actual frame if this is correct I will post my methods.

[OptoMan99]

Yes so you start at: 2658, with 256 characters and 128 points then every 2 characters is the differential for the next point so you add that to the start measure and you get the radius. Just to be sure when I run this I get a weird pattern not an actual frame if this is correct I will post my methods.

Yes exactly. I think we are on the right way, just have to understand how to manage the hex characters for the conversion of the differential points. Of course in the shape, sometimes the differential can be positive and sometimes it can be negative values. So I think in the method their has to be a sort of "rule" that indicates if the differential point is negative or positive....

[MakeOptics]

Here is an implementation in JavaScript: http://jsfiddle.net/harrychiling/bdf4wa8v/

I would guess next, that the data in INF maybe a negative value indicator, however their are only 60 characters. Unless it is a position of the point in the array to be negated?

[OptoMan99]

Here is an implementation in JavaScript: http://jsfiddle.net/harrychiling/bdf4wa8v/

I would guess next, that the data in INF maybe a negative value indicator, however their are only 60 characters. Unless it is a position of the point in the array to be negated?

Your code seems interesting, but I don't think the negative values are at a fixed position in the array, because every shape could have different angles where it increases or decreases.
But about one thing I am sure, for experience with OMA, I know that the first point has to be the same or really near the last point, in that case it means the drawing of the shape is correct and the last point connects to the first one (like a circle). With this method the shape starts at 2911 and ends at 2666 (they are too far from each other I think...)

[MakeOptics]

Yes the values are not adding up, I am looking over the VCA comm standard again just to refresh.

The start value for the left eye seems high?

[MakeOptics]

Can you force it in your software to only use format 1? (All OMA devices can use 1)

Here is what I have so far

[OptoMan99]

Can you force it in your software to only use format 1? (All OMA devices can use 1)

Here is what I have so far

Apparently these old machines don't have OMA in their software because the OMA was still not around. In fact I tried to change all the settings on the machine but it never shows me OMA anywhere....

[jspayneii]

Does this look like the shape you traced? I have attached the excel file I used to interpret the points. We can make the program work off of this. Perhaps you want to post a few more shapes to run through to verify it.

[MakeOptics]

Wow, what's your interpretation of the data?

[jspayneii]

Wow, what's your interpretation of the data?

I believe the data in the tracing is essentially offsets. I also believe that there is some sort of negative value in here. For my formula, I assumed values over 128 were negative, so I subtract 256. I applied each offset to the starting point.

That being said, after taking a second look, I don't have the v_box correct as it is totalling 60. So I will look at that. I think perhaps I need to look at the offsets as an accumulation.

It would also be nice to have a picture of the lens or tracing, so we know we are on the right track.

[PRECISIONLAB]

Started with WECO line (tracer, blocker, and edger) as a stand alone set up in 2000. When we upgraded LMS and additional Optronics edgers, communication with Innovations and older WECO equipment required a Current Loop Card and WECO current loop software running.

[OptoMan99]

I believe the data in the tracing is essentially offsets. I also believe that there is some sort of negative value in here. For my formula, I assumed values over 128 were negative, so I subtract 256. I applied each offset to the starting point.


That being said, after taking a second look, I don't have the v_box correct as it is totalling 60. So I will look at that. I think perhaps I need to look at the offsets as an accumulation.


It would also be nice to have a picture of the lens or tracing, so we know we are on the right track.

EXAMPLE.zip


Here is attached another example:
- the "000001.dat" file is the original weco trace 2 data received from the tracer (open it with notepad)
- the "000001.txt" file is the conversion in OMA (done by Weco software)
- the "000001.png" is the image of the traced shape


Your calculation seems very interesting, could you try to applicate it on this "000001.dat" shape? So we can see if the result is correct.....

[MakeOptics]

I believe the data in the tracing is essentially offsets. I also believe that there is some sort of negative value in here. For my formula, I assumed values over 128 were negative, so I subtract 256. I applied each offset to the starting point.

That being said, after taking a second look, I don't have the v_box correct as it is totalling 60. So I will look at that. I think perhaps I need to look at the offsets as an accumulation.

It would also be nice to have a picture of the lens or tracing, so we know we are on the right track.

Nice, I like the way you think.

[jspayneii]

It was a good theory, but I haven't been able to get the new shape to come out. I plotted the changes against the oma file. I feel like there is something in the CON or INF content that we need.

[MakeOptics]

It was a good theory, but I haven't been able to get the new shape to come out. I plotted the changes against the oma file. I feel like there is something in the CON or INF content that we need.

I'm thinking black hat, I have a dll that I can reverse engineer just to get the algorithm. I will give it a shot.