Messages

I had a machine from Mike which had had the camera removed... Had to recalibrate it completely.

Seeing as the camera was terminally dead there wasn't much option in this case!

If you use a hand held torch you might as well not bother. The Fid correction aligns everything. so make the dot you are looking at bigger or smaller or obround then the placement will be poor. its the same as using a hasl pad, poor and variable results every time.

As it's aligning to the centre, dot size shouldn't matter as long as it is symmetrical. Similarly any asymmetry should be fairly visible looking at the  onscreen image. Obviously you'd use a nicely diffused source, not a directional 'beamy' torch.

Applying some lateral thinking as to how you could realign the camera more easily, If before removing the camera, you were to move the head to a known position (e.g. centre of bed using feeder setup) , place a nice sharply drawn cross+box-shape at the right focus height on the bed, line it up with the camera centre and rotation and fix it down securely, it ought to be possible to realign the camera back to the same position to within a pixel or so (depending how easy it is to clamp the camera mount without it moving). 

Another pragmatic option may be to just make a new external LED ring to go round the camera body.

There as some things which aren't user serviceable on these machines. This is one of them.

Everything is user-serviceable, given sufficient time and determination 

Bear in mind that the alignment (especially rotation) of this camera is somewhat critical from the point of view of fiducial correction..
I have had one of these apart a few months ago but can't remember if the LED PCB is removable without taking the camera off the mount - looking at mine, there is a slot where the wire emerges, but I can't tell if there is any sort of internal retainer to prevent removal from the front. 
An alternative to replacement may be to use an external handheld torch when doing fiducial correction!

[Be very careful - there is a mixture of mains and low-voltage on this connector! Your PC will really not like having 240V shoved up its serial port]

There's a handy utility called feedertest.exe in the install directory of the RV software which can be used to test feeders on their own - very handy when doing repairs/maintainance.

You'll need to make up a cable to supply power & connect data from the PC COM port.

Be very careful - there is a mixture of mains and low-voltage on this connector! Your PC will really not like having 240V shoved up its serial port   

(RS sell the connectors, but you have to buy the pins in packs of 50 which makes it rather expensive to buy just one plug. I've just ordered some to make up some new feeder cables, and  will make the left-over pins available at cost+postage - I'll post another thread on this when available.)

Connect Live, neutral and Earth to mains, and the data lines to a D9 female for the PC COM port as follows :

0V : pin 5
Data : pin 3

Well according to the edit dialog in RV Setup, the total rotation is original rotation+CDF Prog rotation+board rotation*+any rotation in the edit dialog.

So the most plausible scenarios would seem to be :
1) It uses the rotation in the CDF at the time the job file is created by RV Setup.
2) It uses the rotation that was in the CDF at the time the CDF records were copied into the TFR file by RV gerber (or my PCAD import utility) - I've not yet looked into the details of what this copied CDF data is used for...
3) Something less obvious
4) It is buggy and doesn't work in some/all circumstances

So are you going to enlighten us Dave..?

* BTW I'm sure I've seen a board rotate option somewhere but can't find it - the RV Setup help references it, but where is it..?

Having played around some more, it looks like prog rotate is used by RV setup when creating the job file, so by the time you get to RV place and discover a wrong rotation, it's too late to change it the easy way by editing the CDF.

I'm having problems imaging SO14-Wide packages, due to the part not fitting in to the camera field-of view, and the software not trying to do the multi-section imaging it does for QFPs.
Regardless of specified package size it doesn't try to do the multi- segment imaging that it does with QFP packages - I can't figure out what parameter in the CDF controls whether it does the multi-image thing.

If I start with a CDF for a QFP package, it does do the multi-imaging, but still doesn't look far enough out to see the lead ends.

Is there a 'hidden' CDF parameter somewhere to determine the type of imaging - AFAIUI the 'shape' info isn't used for imaging
Is theer a setting somewhere to determine the part size at which it does multi-segment imaging?

Anyone figured this out?

Model G12/045 part no 50200 is listed in this catalogue :
http://www.gd-thomas.com/WorkArea/linkit.aspx?LinkIdentifier=id&ItemID=9738
Page 22

I suppose one way to test repetitive placement of a single part without wasting components would be to glue a part to a nozzle, or alternatively turn imaging off in the CDF.
Other than that I'm not sure what you could simulate realistically.

Can't help directly but these people make some small pumps that may be worth a look :
http://www.knf.co.uk/index.php?id=2&L=5

2. What do you use to clean the o-rings that tears the plastic off the tape, because I sometimes have a problem where the plastic doesn't  come off and then the part isn't picked up.

Not tried it on the RV yet but for rubber drive belts I use a product called platenclene, used for cleaning printer platens and restoring grip to rollers - it works very well on these so would espect it to be good for restoring grip on belts.

5. It sounds like the feeder is indexing but doesn't, and with the next index it it works again. Why?

Maybe the pin isn't always coming up - perhaps clogged with tape dust?
I've also had a feeder go erratic due to an intermitent connector from the end-limit switch.

Yes - I was going to do this at some point anyway - was there anything in particular you are interested in?

Apparently you have to be logged in to see the pictures. Odd.

This was a default option on the forum SW, presumably to limit bandwidth - changed now.

Hope those magnets are shielded - first time I made up a tray I used normal magnets and had some slightly magnetic components tombstoning in the magnetic field!

I also got one of these, which has slots with slight overhangs to hold strips of 8mm tape in place. Machining quality and finish is pretty awful , but still useable (throw away the reel holder!).
There is enough space between the lanes to stick odd-sized tapes down with double-sided sticky tape.

Yes, and to stick the leaders onto the pickup reels.
Wide thin-film double-sided sticky tape (e.g. as used for carpet) is also handy to put on PCBs to hold components down when doing dummy runs during setup - much less messy than solder paste!

Feel free to ask - use the SW/HW sections. Your questions and the answers will be useful to others!
You will find the first few jobs frustrating to set up, but it gets easier once you get the hang of things.
A big pain is making sure the orientation of parts in your library matches the 'natural' orientation of the machine (0 degrees is with index towards you in RHS feeders), and the rotate directions are the same (e.g. PCAD2006 rotations in P&P report are antioclockwise).
This is a particular pain for previous jobs that have old library defs - I've now got into the habit of changing libarary parts as I go for new designs, and combined with my export utility, the most tedious part is now threading tapes into feeders.
Hint - get some 4mm wide double-sided sticky tape from ebay!