Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.

SMTnet
»
Electronics Forum
»
Glue Witness Pattern Verifiable by FUJI???

Glue Witness Pattern Verifiable by FUJI???

Subscribe to Thread

Hello all, at the present I have some product onto which I ... - Dec 05, 2000 by NickMata

Nick, In siemens terms: Fuji has the ability to check "in... - Dec 05, 2000 by JAX

Nick, what does your witness look like? Let's see if I can... - Dec 05, 2000 by davef

Plaaaatttt!! [you can see it if you make like you're respo... - Dec 05, 2000 by davef

The witness pattern is one column formed by individual dots... - Dec 06, 2000 by NickMata

Nick and Dave, Did I miss the question? What does it ma... - Dec 06, 2000 by JAX

The pattern does mater with respect to whether they are sin... - Dec 07, 2000 by NickMata

It doesn't matter because on a dual dot both dots should be... - Dec 07, 2000 by JAX

You're correct, it does not matter on dual dots because the... - Dec 07, 2000 by NickMata

NickMata

#2306

Glue Witness Pattern Verifiable by FUJI??? | 5 December, 2000

Hello all, at the present I have some product onto which I glue backside components. I dispense a pattern of "witness" to use as a visual cue as to the quality of the dispensed glue dots. A problem which I have encountered is that operators will often forget or simply not inspect the witness pattern. This results in additional prior to the wave or in repair process.

I have at the moment Siemens tools which I am using to "verify" the witness pattern. It works rather well and keeps us from building product with missing glue dots.

My concern however is that we will be soon be going to the FUJI platform and not know much about these tools, I would like to know if anyone has used the FUJI tools in the same manner as I have the Siemens.

The manner in which I have gotten the Siemens to examine the glue dots is to teach them as local (component) fiducials. If the glue dot is oversized, undersized, deformed or missing the Siemens tool will treat it as a missing fiducial and not populate the card.

Any suggestions / inputs on whether FUJI can match this would be greatly appreciated.

reply »

JAX

#2307

Re: Glue Witness Pattern Verifiable by FUJI??? | 5 December, 2000

Nick, In siemens terms: Fuji has the ability to check "ink spot" which I assume is what you are doing now. The only difference is the vision capabilities between the two machines. You do not have the freedom of siemens type gray scale with Fuji. You should still be able to make everything work but it will probably be easier on lighter color boards. The darker the board is the more confusion you will have from telling the glue dot from the board itself. If you are changing your placement machine you might want to think about buying a new glue machine. Some of the newer machines on the market will probably give you a better deviation tolerance than you have now. Some of the machines can also be set up to do the same "ink spot" verifications that you are doing now.

That's just my opinion, I could be wrong!!!

reply »

JAX

#2308

Re: Glue Witness Pattern Verifiable by FUJI??? | 5 December, 2000

That's just my opinion, I could be wrong!!!

reply »

davef

#2309

Re: Glue Witness Pattern Verifiable by FUJI??? | 5 December, 2000

Nick, what does your witness look like?

Let's see if I can make me show correctly through this HTML stuff ...

o o oo oo oo o o o o

reply »

davef

#2310

Re: Glue Witness Pattern Verifiable by FUJI??? | 5 December, 2000

Plaaaatttt!! [you can see it if you make like you're responding to it]

So much for that ...

In words, its a star with five arms. * At 9 - two dots side by side * At 12 - two dots, one under another * At 3 - two dots side by side * At 5 - two dots on a 45 pointing to the center of the star * At 7 - two dots on a 45 pointing to the center of the star * At center - two dots centered in line with the dot of the 3 and 9 arms

reply »

NickMata

#2311

Re: Glue Witness Pattern Verifiable by FUJI??? | 6 December, 2000

The witness pattern is one column formed by individual dots as well as dot pairs. The two dots which make up a pair are both side by side as well one under the other.

Thanks

reply »

JAX

#2312

Re: Glue Witness Pattern Verifiable by FUJI??? | 6 December, 2000

Nick and Dave, Did I miss the question? What does it matter what pattern the glue dots are in, whatever vision you use will have to read the individual dots anyway. If your glue machine is any good you should be able to get away with simply placing your witness/test dots first to purge the needle, once it has been purged you shouldn't have any problems until you start to run out of glue.

reply »

NickMata

#2313

Re: Glue Witness Pattern Verifiable by FUJI??? | 7 December, 2000

The pattern does mater with respect to whether they are single or dual dots. From info which I have collected, dual dots would be an issue for the FUJI since they are capable of viewing one or the other as a fiducial but not both at the same time.

What planet are you from? Please tell me so that I can go buy some of their glue machines.

reply »

JAX

#2314

Re: Glue Witness Pattern Verifiable by FUJI??? | 7 December, 2000

It doesn't matter because on a dual dot both dots should be the same size. If one of the dots is to small and is not read by the vision system the dot is bad and the board will be skipped. You don't have to read the dual dot as a single item. With the extent of my experiance with Seimens I don't think it will read the dual dot as a single mark either. If this is not true feel free to let me know. As far as fuji being able to read only one size fid at a time you are correct. What you might want to try is having your glue machine set to read fiducials after your board is complete. It can read one size while your Fuji machine reads the other.

Thanks for the remarks, I was starting to miss the old MoonMan!

reply »

NickMata

#2315

Re: Glue Witness Pattern Verifiable by FUJI??? | 7 December, 2000

You're correct, it does not matter on dual dots because they should be the same size. My concern is given that the FUJI will not see dual dots as a single item, in order to verify both dots I would have to program as them as seperate fids in which case FUJI may pick up one of the dots and treat it as the two it is looking for, yes?

You'd be surprised what the Siemens vision system can do. Yes, it can in fact view the double dots as a single item. There has been instances in which I have used unique trace patterns as fids. Raw card variations can be dealt with a combination of proper location for the glue dots as well as the transform feature within the Siemens SW.

reply »