Quad 4C part placement rejection

#65447

Quad 4C part placement rejection | 15 November, 2011

Well, everything WAS going OK until I started setting up for a board job (40 boards with about 450 components each - many image repeats). We setup 37 feeders with their parts and started defining the pickups and placements. Part #1 is an SOD123 zener diode and I had a bit of trouble until I moved the QA (Quad Align) Z ht to 5 mils, then it scanned and placed OK. Next I configured pickup #2, a 1206 resistor. Then I setup placement and did a Run Step placement to test it and, REJECT... I set the Alignment type to 4. If I change it to 0, it will place every time. I called PPM and talked to their 4C tech. I went through all of the detail with him and he has had me make a few changes in machine settings but none of them help the situation. Here's what he asked and told me to do.

1. He told me at the V Enter CMD to... I said wait, my hand held always only ever says Enter CMD - no "V". He said the vacuum cerify is off and should be turned on (had me do a Shift Switch Verify and it turned on). Next he said that an erronious run out reading after reject sometimes could be that it is trying to check run out with a part still on the nozzle at the reject point. I told him that every time I looked there was no part. He had me do a Function 12 to see what the vacuum readings were with a part attached and without. Without a part the reading was ~1700 and with ~3470. He said those were OK numbers. I mentioned that Mod 9 (vacuum setting) had been changed twice by the previous owners from the default setting of 1000. That setting was readjusted to 500 and then crossed out in pencil and changed to 200. He had me bump it up to 350.

2. He asked if the nozzles were fully seated on the z rod and i told him that they were (and I again checked using my Optivisor and a very bright light while pushing up on a nozzle) Yep - fully seated.

3. Then he said something that I find very puzzling. He said that Quad Align doesn't do anything during a Run Step operation. It ONLY functions during a full program run. Somehow I find this hard to accept. The only thing that causes failure during this placement is if I set the Alignment type to anything other than 0. Now the first part (the SOD123) is setup as a type 4 Alignment and it repeatably places perfectly. (just like components on the previous board I was running). When I pickup #2 and do a Function 31 it tells me the extent data and length, width are well within tolerance and X and Y and Theta all show up and look good. But... when I try to place the part - REJECT. I have tried every variation of the Z setting and the part orientation in the feeder and the feeder locations are the same for pickup 1 and 2. And... they both use nozzle 2 (XG) so there's no nozzle change. The customer was expecting me to have one of these boards reflowed by the end of today and I've got a boatload of programming to do before that's gonna happen. Any thoughts? (other than dynamite, that is).

Thank you in advance for any assistance you can provide.

Tim

#65456

Quad 4C part placement rejection | 16 November, 2011

A couple of things i would try:

1. Does function 30 give you a nozzle height that is repeatable. The number should be around 200.....+/- 20 or so. 2. I would change the z-rod because the spring could be worn out and when the vacuum turns on....it can collapse the z-rod causing rejects. Maybe the zener diode is a thicker part? 3. I would leave the vac verify off. If it was turned off.. there is a reason. Many of the machines i worked on had bad vacuum transducers that had parts stuffed in them and didnt work correctly. 4. Are you seeing the parts in the dump bin? Make sure it is picking them up and they are not being blown out of the feeder by too high of a puff off pressure setting.

Good luck,

aci

#65459

Quad 4C part placement rejection | 16 November, 2011

Tim, Bob is the C series master so try this.

#65460

Quad 4C part placement rejection | 16 November, 2011

Bobpan, thank you for your speedy response. As to the items you noted: 1. I did a test of Funct 30 for the first 5 nozzles, 2 times each as follows: noz1 Ht=215 & 219, center=1005 (all nozzles all tests), run out 2.2 & 2.7. Noz2 Ht=228 &228, runout= 2.4 & 2.5. Noz3 Ht=218 & 218, runout=2.2 & 2.7. Noz4 Ht=214 & 214, runout=4.8 & 5.0. Noz5 (with rubber gasket) Ht=153 & 151, runout 7.5 & 7.8 (probably looking at rubber). 2. The zener (part 1) is thicker than the 1206 resistor (part 2). I pulled the z-rod and compared it to a brand new one in terms of "springiness" to see if it felt "limp" as it reseats to the bottom position. It behaves just like the new one. (I do have one that was used that the previous owner marked and left with the spares and it seemed less springy). So should I change the z-rod anyway? 3. I've turned vac verify back to off as you sugggested. Although this reject problem occurs with vac verify on or off. 4. When rejects occur they are always in the dump bin. I've been watching each attempted placement and the parts are on the nozzel as it tries to place and are in the dumper after it rejects.

In addition to my earlier post, here is some more information: About 50-60% of the time when the machine rejects a Run Step Placement part now, it moves to the dump spot and after dumping the part the machine starts beeping and the HHT has the following error: Run Error 1 The manual defines this error as: “A pickup number greater than 127 was assigned, or the pickup number entered into the sequence steps program was not taught.” The first part of that error is just not true because as of today I only have 11 pickups taught. As for the second part, I got excited for a moment because when I went to the central controller and did a database download from the assembler to the central controller and looked at the pickups, placements and extent data, I then looked at the program sequence and the steps from the previous board were still there. I thought aha, it’s looking at the old sequence program where the pickups and placements were different. So… I deleted all of those program steps and then uploaded to the assembler to replace what was there. I retested doing another pickup and placement … and the placement is still rejected every time (unless I set Alignment type to 0). So, that was a bust. – worth trying though. So, any idea what’s with this error message and how it might relate to the placement reject issue?

One other thing... Does Quad Align do its thing during a Run Step Placement? (It seems like it must or this porblem wouldn't be occuring - right?) Again, thank you for all of your help.

Tim

#65462

Quad 4C part placement rejection | 16 November, 2011

Hi again Bobpan, One thing I noticed when I pulled the z-rod to inspect it was that it was "bearly" finger tight. When I replaced it, I snugged it a bit better (not overtightening). After resuming setting up additioanl placements (this board requires 38 feeders), just for fun I did a Run Step Placement and the part placed perfectly - WOW! So... I went back and tested placement #2 againg and it also placed just fine. Thinking it may have been a loose z-rod, I started testing other placements and...reject - damn... So I angain tested #2 and reject. My excitement was short-lived. Now, I'm currently working on pickup 20 but I thought (just for fun I'd test 19. It worked (once). The I tested 20 (both 19 and 20 are same part packages 1206). Pickup went fine but at Run Step Placement the gantry stopped and "QA error #0" showed on the HHT. I hit Enter and it rejected the part. I tried the same part 3 times and got the same message. I've never seen this one before. Do I need to call an exorcist?

#65468

Quad 4C part placement rejection | 17 November, 2011

wow.....lets try some more things.......

first....i think quad align error 0 has something to do with the theta number not being at 0,6000,12000,18000. I am not for sure about that but check it out by doing a function 3 on your pickups.

second..... You should be scanning as high up on the resistor as you can in your extents. Set your scan height at 3 mils or so.....this would scan at the top of the part and if the nozzle is not seated good then it will still pass the part. What is your tolerance set to on your part....make it at least 20% for the x and y scan size.

check to make sure your extent information is attached to your pickup by doing a function 16. As for the quad-align...it really does only work the way its supposed to in the run mode. But you should be able to test parts with function 31 or 32.

Good luck

#65471

Quad 4C part placement rejection | 17 November, 2011

Thanks again Bobpan for all of your help. I hope some of my questions are dumb ones.

Regarding "first", the QA error 0, you were right. In checking the setup, for some reason theta was 65227 instead of the 0 that I thought I had set.

Regarding "second", I will try what you recommend and report. I do Function 16 regularly to make sure all of the extent data is there (and correct).

About Quad Align not working in the Run Step mode, then how does it know to reject the part if z or tolerances are wrong? Just curious.

In you earlier post you seemed "unimpressed" with Quad C's Vacuum Verify capability and I have turned it back to the off as mentioned. However, one frustration I had with an earlier board project (a panel of 32 6-component image repeats) was that a DPAK (fed by a belt feeder) would miss pickups about 30% of the time and just continue placements without a part on the nozzle. As a result I had to physically place (with tweezers) to populate the board. With the vacuum readings (with and without part on nozzle), PPM seemed to think it was working great. Vac Verify would have caught the missed pickups and (I think) doen retrys to get it right (or am I not understanding?)

I will report on the above later today, again thank you.

Tim

#65473

Quad 4C part placement rejection | 17 November, 2011

Hi Bobpan,

Regarding "second", the results of the test are no difference - still rejects the part. I tried 2 mils, 3 mils, 5 mils, 7 & 9. I changed the tolerances fron about 15% to 20%+. By the way, I did put on the new nozzle and it is still the same. I'm curious about the Run Error 1 that I get about 1/2 the time after a Run Step Placement (earlier post), any thoughts on that?

Something somewhere is flaky because placement 2 actually worked 2 or 3 times in between not working (earlier post).

I apologize for the following somewhat convoluted question, but when setting up a pickup, and setting the Alignment type to non-zero, then setting up a placement, what actually goes on in the placement process with regard to this Alignment type value? If the values are 0 then it will place all day long with missing a beat. However, the placements will not be accurate (no adjustments for feeder pocket differences or part skewing on the nozzle). But, if when I do set an alignment value to, say type 4 (slowest non-leaded)and it doesn't reject, the placement is perfectly aligned every time. I've even tried setiing the machine to slow operating mode (wow is that sloooow) and it has no effect on the problem. I'm actually happy about that in that I think it means the encoders, motors, bearings, etc are in reasonably good shape. Could something in the head assembly be loose or require lubrication?

Ever-indebted, Tim

#65475

Quad 4C part placement rejection | 17 November, 2011

Tim, 1. You dont need the vacuum verify if the quadalign is working properly. If you have the part quad-aligned and dont have alignment type at 0 it will never 'not' place a part. More or less...the vac verify was needed for the chucking machine (3c). It does have a purpose but like i said....its not really needed.

2. Truthfully.....the whole deal with the quad align and not being in the run screen and just doing a runstep place....i am not sure about. I just know that it really only works the way it is supposed to when running the machine in run mode. 3. Have you cleared the memory and reloaded in your mod codes...this will probably not fix the problem but its one thing that i always eliminate. Just make sure you have all the original codes written down. 4. It does sound to me like its a quad align problem. Although you are getting a consistent reading with function 30...it is possible that there is dirt inside or on the outside causing the image of the part to look larger than it is. I have dealt with this before. If the windown is totally obstructed...you will get a 1 for nozzle height with function 30. 5. Depending on the firmware version....there is a mod code up in the 60-70 range that is a quad align illuminator value....make sure this is set at 255. 6. Make sure the z-belt that drives the z up and down isnt slipping....that is about the only thing that could be loose.

Your straining my memory with this stuff....haha....havent worked on but 2 c-series machines in the last 8 years.

Good Luck, Bob

#65476

Quad 4C part placement rejection | 17 November, 2011

Hi again Bob, I'll go by your item numbers:

1. OK, I'll not worry about it.

2. Somehow, it seems to me that it must be doing something or in Run Step Placements there wouldn't be different behavior for Alignment type 0 vs. other settings. I know I'm hanging on to this like dog to a steak, but being an engineer myself, I just can't (mentally) let go of it.

3.I didn't clear memory (not that brave yet) but I did check every mod code against the settings in the log book and they're the same.

4. I agree about it being a quad align problem and yes, Funct 30 is repeatable. Regarding dirt, I cleaned the exterior of the two lens (glasses) with a lint-free isopropyl / DI water cloth and made sure it was spot free dry. If dirt gets on the inside, how do I clean that. I assume that if the quad align assembly is removed from the machine, there would be some special realignment procedure required and I have no clue (or special tools from Quad) for that. I've never seen a 1 for nozzle height. It's always around 216-228.

5. Since the previous owner was still using the machine within a couple of months of my purchase (January 2011), I assume firmware is the latest rev. I can give you the number if you want. The align illuminator value is 255 (the highest setting).

6. Since the Z belt is a toothed belt i don't see how it could slip, but it is nice and tight (no slop).

I'm sorry for drawing on old info, but I don't know where to turn and since I'm a one-man show, I can't really afford to fly someone out here from PPM (I'll stop whining now).

If something comes to mind, don't hesitate to let me know. Also, mosborne has my phone number, should you need it.

Again, thank you for all of your help. I hope I can resolve this and keep my customer smiling.

Best regards,

Tim

#65477

Quad 4C part placement rejection | 18 November, 2011

try this.....

I think there is a coupling for the theta motor....make sure both set screws are tight.....maybe the theta is slipping. Make sure there are no bright lights shining into the machine....maybe even throw a dark cover over the top and try to run it. Bright light into the quad align will cause rejects. If all else fails....write down all your mod codes by stepping through them on the hht and do a battery flush and re-enter the mod codes.

Bob

#65489

Quad 4C part placement rejection | 18 November, 2011

Hi again Bob,

I checked the theta motor coupling and the set screws are definately tight. I noticed that the theta motor drives an snti-backlash gear (nice design) so there is no play there. It appears that the gear drives a slotted shaft which is definately well greased. I presume this shaft is the z-rod shaft and the reason for the slot is to allow vertical travel. If the slot in the shaft or the mating "tooth" on the gear set is worn there could be some play there. There may be another anti-backlash arrangement at that point but I couldn't see well enough.

Regarding lighting, the shop does have a bunch of 4-foot double fluorescents, but they were there when the last board project ran fine. Just the same I removed the tubes from the two directly above the front of the machine and it didn't make a difference.

Regarding the mod codes, since occasionally a Run Step Placement actually works, it should't be the codes right? How do you do a battery flush for wiping the codes? As always, thank you for everything. Tim

aci

#65503

Quad 4C part placement rejection | 21 November, 2011

Tim,

The main controller card has a battery on it. Power down the machine and disconnect power. Stick a credit card between the battery and the clip. Hold for at least 60 seconds. Then power back up the machine. Make sure to have yourmod codes written down. The machne will go through a homing sequence then you need to re enter your mod codes. You will need to upload your mod codes to the central controller as well.