Quad 4c not placing components

#83379

Quad 4c not placing components | 4 September, 2019

Hello All,

I've just bought a Quad 4c pick & place machine. It's so close to actually building boards, but just doesn't quite get there. I was wondering if there's somebody out there with experience of these machines that might be able to help...?

The basic problem is that it'll pick up a component, and when asked to place it, will move over the correct position on the PCB, hesitate, and then move to the reject location and dump it. About every third failure it re-calibrates the nozzle, and sometimes gives me a "Run error #1" (Pickup assigned number greater than 127, or pickup entered in sequence was not taught). This error makes no sense as it's already made the pickup. Usually though, there's no activity or error at all.

Initially, I thought that this was due to the laser alignment failing. I could get parts to place with the alignment disabled, and they'd be rejected with LAE parameters set up. However, this isn't 100% - some aligned parts would place okay, some of the time.

The alignment works fine if done manually (function 31) with a component on the nozzle - size and offset are bang-on, although rotation appears to be a random number..

I've even taken the EPCU battery out to reset all the parameters to default, and then just changed the MOD codes that I know need updating.

Anyway, I messed around with it for a few weeks on-and-off, and made no progress. I've got a batch of boards to build, and the fallback plan was to build them with no alignment and fine-tune the positions with tweezers afterwards. Three boards done like this (about 600 components), and I adjusted one of the pickups (just changed the Y by 15 thou/mil on the PC), and now it rejects everything!

I changed this co-ordinate back again, and even tried the backup of the sequence. I've reset everything, and even turned the machine off for an hour, but nothing is making any difference - it won't place anything at all now.

What's particularly annoying is that the software is clearly making a decision not to place the component, but as far as I can tell there's nothing output and no error message to tell me why.

I thought maybe it's a vacuum issue? I've measured it at the nozzle and that seems fine. Vacuum verify is set low (100). I though maybe a problem with the transport? - if it doesn't think there's a PCB there... It loads and unloads boards fine, so I can't see this being the issue.

Other than this, everything seems to work fine. There's generally no error messages (unless I do something stupid!), and everything that should move, seems to move fine.

Any suggestions would be gratefully received. If you've seen anything like this before, I'd love to hear from you - even if you didn't fix it, you might be able to help me rule things out!

In the meantime, if anybody needs a machine to remove all the components from reels and dump them in a box - I can help!

Many thanks,

Steve.

Steve Thomas

#83381

Quad 4c not placing components | 4 September, 2019

My recollection of the 4c is fuzzy but I know that there's a ribbon cable on the head that gets some abuse when the head rotates. Seems like mine would eventually break at the crimps and I had to replace it a few times. Eventually I bought the parts to build them so always had a spart. I wish I could remember the errors it generated when it started to die.

I'm not at all sure if this could give you those specific errors but if it's never been replaced it's not a bad place to look until more expertise arrives.

#83383

Quad 4c not placing components | 4 September, 2019

Thanks for that Steve - I'll try anything at the moment, but that sounds like a reasonable suggestion.

I've already replaced a few cables that looked dodgy, and the LEDs on the LED rings, but I've not done a ribbon cable on the head yet. I shall have a look and see what I can do with it...

Kind regards,

Steve.

DWL

#83384

Quad 4c not placing components | 4 September, 2019

Sounds like its just rejecting parts. Either the length and width of the component is wrong or the tolerance is set too tight. I don't remember the function codes to edit the tolerance but I left them at 30% (the max) and never had any problems with placement accuracy.

#83385

Quad 4c not placing components | 4 September, 2019

Try these things: 1. Do a function 30....your nozzle height should consistently be around 200 +/-20....but whatever number it does find it shouldn't change. (with nozzle on) 2. Most times its the z-bellows....an old z bellows that has a worn out spring will reject many parts due to height changing with and without vacuum on a part... and if the runout is too big it causes issues. 3. It could be dirt in the quadalign that causes problems. The quadalign is probably 20 years old and many times dirt finds its way in through worn out gaskets on the glass lens or in other ways. Regardless...clean the lens best you can. I have also had to remove the lens on the camera side and blow out the quadalign to get it working correctly.

good luck

#83386

Quad 4c not placing components | 4 September, 2019

Hello DWL,

Thanks for the response.

I've measured the components with a micrometer and set the sizes. Maximum tolerance on my machine is 20%, and that's what I'm using. If I do a manual scan, the measurements are exactly right (and what I've set up). Besides, it's rejecting components even with the align disabled.

Kind regards,

Steve.

#83387

Quad 4c not placing components | 4 September, 2019

Hello Bobpan,

Thanks for the response.

1. Function 30 gives me a consistent 182-185, even after nozzle changes. I can vary this by changing mod code 15 and I have been advised to aim for 90-110, but it doesn't seem to make any difference so I've gone back to the default (50) which gives the function 30 readings above.

2a. Hadn't considered the bellows getting shorter with the vacuum pulling it. It doesn't feel like this would be possible when I push it in manually, but it's something I can try to investigate - However, I'm getting the same failure with alignment disabled, so surely this wouldn't even be a factor...?

2b. I have no idea what a permissible runout is. On a full rotation of the nozzle I'm seeing maybe 0.3mm (12thou/mil) variation. I also have a new z-rod, but with this fitted the runout is actually worse, probably double. In theory the laser align could compensate for this, or detect it and cause a failure - but I don't know which (if either) it's doing. Again though: still fails with the laser align off.

3. It's a CyberOptics laser if that makes any difference? I'll have a go at cleaning it.

If the alignment is disabled though (extent is "empty" for all pickups), shouldn't it just pick & place blind? I can't see what error can occur - it's just following movement patterns isn't it?

I know that even without laser component alignment, the laser is used to check the nozzle position (function 30, and after a nozzle swap). Is the laser doing other things too that I don't know about? If I disconnect it, it's not long before the machine stops working.

Many thanks,

Steve.

#83388

Quad 4c not placing components | 4 September, 2019

It sounds like the vac-verify is not working. There is a command on the hht to turn it off. When turning off your hht will show either 'idle' or enter command' without the V showing anymore. Maybe try the cntrl and vacuum or shift and vacuum.

If its the vac-verify....you may have a clogged transducer on the headboard. Check the tubing running to it. Also there is a function code to check the vacuum.

good luck

#83389

Quad 4c not placing components | 5 September, 2019

Hello Bobpan,

Thanks for the suggestions. Sorry I didn't reply yesterday as I'd gone home - I'm in the UK, so I'm guessing there's a time difference.

Okay, I found the control to turn off the vacuum verify - it's SHIFT - SWITCH/VERIFY. Makes no difference though. Still hesitates over the PCB, then rejects the component every time.

Function 12 gives me the vacuum level. With a small nozzle, it's around 2200 with no component, and 3100 with a component on the nozzle. It drops to zero (over several seconds) if I switch the vacuum off.

If I measure the vacuum pump directly I get about 29in-hg, and the gauge on the Quad agrees with this. At the nozzle, it's about 26in-hg, which suggests a minor leak somewhere, although it has no trouble picking components up and holding onto them.

I've tried the single-step mode to see if this helps. Pickup is fine with the component (an SMA diode in this case) appearing centrally on the nozzle. The head then moves over the correct location on the board - the diode is still as picked up, 90 degrees out from its placement angle. Next step rotates the diode 180 degrees (so now 90 degrees out the other way) - I assume this is doing a scan, even though I've told it not to do alignment. The next step takes it to the reject location.

I don't think the wrong angles are a problem. When it previously did place this component (exactly the same program running), it placed it correctly. I think it's just rotating it 180 degrees to do a scan.

I'll take the CyberOptics unit off this afternoon and try to strip it down and clean it out. I still think that with alignment disabled this shouldn't be relevant, but it does appear to scan every component regardless (I don't know what settings - align type/height - it's using though as I've not told it any parameters).

Many thanks,

Steve.

#83390

Quad 4c not placing components | 5 September, 2019

Make sure your placement rotation co-ordinates are correct. If you do a function 3 it will show you. I think they should be in multiples of 6000....ex...6000 = 90 deg.

Also check the z axis number.

I dont think its cyber or vac related. It is something in the program...i believe.

#83393

Quad 4c not placing components | 5 September, 2019

Also check the lae on the pickup that is being dumped. I think its function 16....make sure that the alignment type is 0.