Feeder placement and filling

[Graham]

What are the thoughts on what order you have components loaded on the feeders?

I've got permision to completely revise the way comps are loaded in the reel feeders and do away with the ad-hoc arrangement used ATM.

Taking it as read that the most used comps should go near the fixed camera, how do others load them?

Do you try to have a set of common components that are permanently loaded, adding job-specific ones to just one/two feeders? Or are there other ways of looking at this?

FWIW the production manager is wedded to the idea of interleaving Rc and Cs on the feeders, saying something about being able to see where things go and where they've come from.

(Edit: Just realised I've stupidly posted in the wrong area. If you can move this thread across, Mike, then please do.)

[Mike]

On the basis that the feeders are quite fiddly to load, my policy is to leave anything that may be used in future where it is,and make space for new parts by unloading the least likely to be used again. However all my jobs are random and unpredictable - if you have a better idea of ongoing requirements than you may be able to be more organised.
The most-common nearest camera should probably override most other considerations.
Whatever scheme you use, I'd strongly advise physically labelling each lane with what is currently loaded.

[Gopher]

I have a 1.5 banks of std parts that pretty much lives on the machine, these are in positions A and B as I only need to reach one reel at a time when they run out,  I have an RV4 so these are imaged on the fly anyway, I keep my vib feeder closest to camera. Beyond the stuff that never changes anything else just goes in whatever empty slot is available, yes closest to board is nominally quicker but I'm not sure it ends up making a measurable difference to an actual run even when typically my job runs are 1.5 to 3 days long.
I can't subscribe to your pm's theory of interleaving, I don't see that helping at all. However I do this on my manual pick and place when I stick strips of tape to a board, I also highlight them in different colours .
As Mike has mentioned before it is quite useful to load the same packages in the same lanes so you don't need to do the feeder biases again.

[SteveW]

I also have preferred banks for plastic & paper tapes - probably doesn't matter for more modern feeders, though.
(two banks dedicated to the common stuff, in paper tapes, one bank for common stuff in plastic, and two multi-width for per-job parts, with other banks & vib feeders as needed. Although nothing seems to come in sticks any more. Well, that's the plan. Obviously some jobs don't play nice...)

Steve

[Graham]

Figured out what the PM was on about when we did my first run this week.

The pick position on the 8mm feeders were all set to the second position, away from the cover tape pull-back, and components were bouncing out when other lanes fed. The PM had been recovering the bounced-out comps for manual placement and to identify which MLCs were which they wanted gaps between their lanes.

Moving to the first pickup position sorted it completely.

A lot of things have now dropped into place. I didn't realise you could set up custom feeder definitions for example.

The main thing that's left to sort ATM is I can't get it to pick up SOT-23s properly - around 50% of the time they're picked up rotated by +/- 30 degrees and are placed thus. Any ideas?

[Gopher]

Sounds like  the SOT's are bouncing around in their tape possibly due to too much tension (makes it jerky) or the empty tape is snagging a little on the way out against some other part, depending on just how tolerant your imaging parameters are a rotated sot23 is the same size, straight on or not. The other possibility is the machine is not picking your Sot23 up dead centre  but instead by one of the leads or edges so it lands end on and any old way round. I don't know if anybody else has found this, but I find that getting the pick position just right and then locking X and Y pick positions in the CDF gives me very reliable placing, the self correcting pick position feature can't cope with the jiggly nature of plastic pockets and just makes things worse.

Only devices I pick one away from position one are 2512's because position one varies between just uncovered and not quite so fails all the time. In fact this is often true of parts loaded in the 3*12 5*8 feeder

[Ed]

check that the angular range is set to 130 or less. its prob at 300 which allows the devices to be place at the wrong angle. its cos the size of the triange fits the overall device size parameter even when twisted.

[Gopher]

Ooo thats what those values do, lol. I could swear our trainer said "leave them alone" and did not expand on that, mine are set at 125 with a step of 2.

[Graham]

Sorry about suddenly going quiet - things have got a bit tense with the firm I'm doing this for over remuneration.

Hopefully I'll be back but ATM I've lost interest.

  Graham

[Loftman]

I acquired an RV1s a while ago, still lots to learn.
Thanks for the tip on SOT23 pickup, it has helped a lot.
I wonder if anyone has any tips/warnings for the use of the 'merge moves' option.
What it does seems obvious, but I'm not sure of the consequences.

[Gopher]

Just what it says in the manual, avoid for finer pitch and stuff in plastic tapes, its another source off vibration and device/image instability I guess. As with "Fast" movement its probably also best avoided with large things too.

[SteveW]

I wonder if anyone has any tips/warnings for the use of the 'merge moves' option.
What it does seems obvious, but I'm not sure of the consequences.

I have an RV1S and (some) old (orange belt) feeders. If I engage 'merge moves', the machine destroys nozzles by moving sideways before it's cleared the dip in the stripper plate. This is somewhat annoying.
If you suffer from this, you may well also suffer from the machine not knowing that the cover tape slopes upwards, and wiping that out, too.
Old feeders can be a pain!

Steve