MDC / Manncorp 77xx(FV) expert questions

[vincentp]

You might download the version again as 2.1.23 (updated yesterday) now has the LOW CAM button thing solved (*thumbs up* MDC).

Software upgrade : The upgrade is most of the time just the .EXE and a pair of text files with typo corrections of the constants help / description. I just overwrite the .EXE in the luna-win directory in the "Programs File" DIR on C:\ but I pay attention to keep a copy that I name with the version of the software so that I keep track of what has changed. This reminds me that I need to make a full machine backup now that I have several designs done and my constants fine tuned.

Part definition column / label : they changed along the years I think. There are still some issues in the handbook because I'm using old screen caps from MDC so it's not matching 100%. But most of the time it's because columns aren't placed exactly at the same place.
The part def as seen here
http://www.ladyada.net/wiki/mdcpickandplace/qfn
is very close to what you have in the current SW. Now you have a wt (wait ?) parameter that isn't used but that's it. You had versions for which 0° placement offsets were called dx0,dy0,da0, now they are dx,dy,da but overall it's the same. I was using 2.1.16 at the very beginning but I can't remember the exact changes.

Cassette feeders : Again, for 8mm feeders, the black spool should not touch the next feeder, even if you used some thick tape. Check the spool width or better post a picture of one of your feeders. When the spool is in place it should be much wider than the feeder. What tape do you use ?

cut tape feeders : I don't know much about those but what would make sense is that the location of where to pull on the tape is set in the feeder definition as it would depend on EACH feeder slot. It's probably NOT defined in the system constants, except if it's a constant offset that is defined for classic tape width (8-12-16-24 etc) but I don't recall seeing those in the system constants.
Please post a screen cap of your feeders definition, there might be some specifics for your machine as it has the cut tape pulling mechanism (which mine doesn't).
I'll ask MDC. I can eventually add a paragraph about those in the manual.

Machine speed : Like it's been posted on the wiki, I also experience the PSU problem when trying to use the machine at normal / default speed. It's my understanding that it's due to the PSU efficiency ratio when on 110 or 220V AC compared Japan where it's 100V.
Also, I verified that it's REALLY the Y axis speed that is the issue as I could upgrade all the other axis speed (X, Z, theta). My guess is that it's happening when the Y arm is slowing down before reaching the its destination. It's so heavy and going so fast that the current required to brake it goes above a threshold or simply goes back to the PSU because the windings generate current when slowing down (despite roll off diodes and yadayada). The previous owner of the machine was sent a replacement unit (Meanwell branded) so I proceeded with the replacement, I only had to drill and tap 3 x M4 holes and matching Allen screws and now the PSU replacement doesn't require to disassemble the whole electronic panel as the screws are accessible from the front. Don't forget the thread-locking fluid.
The Y axis is now back to 2000 speed, with a rate of only 150 compared to the X axis which rate remained 100 (rate = feed = slope = acceleration - higher number = reduced acceleration hence limited capacity of reaching the max speed over a short distance).
I experimented with placing a bunch of jelly beans on double sided tape (as I didn't want to test during an actual batch of boards) : no problem at all. With the stock TDK / Lambda PSU, the problem was almost immediate, after placing a few parts and checking the fiducials on each PCB.
PROBLEM SOLVED !

[vincentp]

@bmw : your machine has indeed some specific parameters for the cut tape feeders (my guess, confirmed by MDC). Also, some of the system constants differs to the 7722 as it has the additional actuator to pull on cut tapes.
The machine normally shipped with a manual addon dedicated to the cut tape feeders (printed copy or on the CD).

I asked MDC and they sent the latest version of the PDF document

Auto cut tape users manual.zip

auto cut tape user's manual from MDC

(489.92 KiB) Downloaded 62 times

it details quite well what you have to do. For your machine, the feeders definition have extended parameters such as the tape width. The pull pin offset that matches the indexing holes of the tape is defined in the system constants and the factory calibration should work as is it.
In you case, just believe the width of the feeder before the one you picture is incorrect and therefore you have an unwanted offset of pulling the tape. The distance between the part center and the 2 green lines suggest that this feeder is set as a 8 mm feeder, not a larger one. Check the Wid column in the feeders definition. As indicated in the manual you also need to tell the machine how many parts you have for a tape pulling operation. Auto cut tape are finally nothing than some dynamically refilled trays.

The cut tape manual explains also that if your part has a weird location on a larger tape (ie the part/pocket center is NOT the center of the tape width minus the hole width), centering the pickup location on the part will still result of an offset and wrong tape indexing. The quick fix (rather to play with system constants general offsets) is to slightly offset where you pick the part, and vision should be able to deal with centering the part anyway. That's for SMALL offsets. In you case the feeder definition is obviously wrong.

I won't be adding this in the manual as it's specific to the 77xx machines with auto cut tape feeders and the manual, while "perfectible" is easy to understand (once your figure the so many other aspects of the machine) as it's just a feeder definition adjustment.

hope this helps.

[vincentp]

found 2 more bugs, one is already fixed : checking the end of the piston / feeder indexing sequence at the end of the sequence to avoid moving the piston back to origin while it's still actuating the feeder lever.

[bmw330i]

Hello,

Thank you for contacting MDC and the new manual. I will read it today.

I am pretty sure my light is too bright as you say. It is now refusing to place a common 33R 0603 resistor.
I tried Lv=1, 126, 26 and with 1 it just rejects the part with an error seen in the first photo (ERROR 1 llv 233). To try and see more what's going on I enabled Lv=126 and it rotate the part around and rejected it. Same for 26.

I verified the parts definition for 0603. It works fine with 0603 caps. Well, it rejected a couple but kept going and succeeded.

So is it time to modify the pots on the machine for lighting ?

[vincentp]

thanks for posting those,

a few thoughts below :

Hello,
Thank you for contacting MDC and the new manual. I will read it today.

no problem. However, if I were you I'd check the CD rom you got with the machine, I checked mine, I had extra PDF in there. I'm pretty sure the auto cut-tape feeder manual is on your CD.

I am pretty sure my light is too bright as you say. It is now refusing to place a common 33R 0603 resistor.

it looks too bright indeed but I confirm you're too far. You need to check the bottom camera alignment and depth (what they call focus). See my other posts about that, your focus value is HALF of what I have.
Setting the lights now will be point less as you first need to look at the part at the right distance.

Second point : I suppose you can check a 0603 resistor with bottom vision but it's pointless IMO. That's something you should check and align with flying vision Lv=1. The 0603 is even smaller than a SSOP so you'll never be able to place it at that distance => fix your focus first.

I verified the parts definition for 0603. It works fine with 0603 caps. Well, it rejected a couple but kept going and succeeded.

no matter how many times you'll try, it's not a part definition problem, it's a focus definition (granted... I'm 99% sure).

So is it time to modify the pots on the machine for lighting ?

not yet, fix the focus first. See my other posts for the proper technique.
It's in the other topic about the 7722 alignment problems, read from there:
http://forums.adafruit.com/viewtopic.php?f=42&t=26353#p164665

From what I see of your potentiometers settings, their proportions are correct, VR1 quite maxed, VR2 quite low, VR4 in the middle. If you got a recent machine it's possible that the new leds are more efficient/brighter than the led lighting I have, therefore the stock settings are too high and you'll need to decrease all the 3 of them while keeping the proportions. I don't know if this is part of the calibration done by MDC. It should be, tested with a couple of standard ICs.

But again, that's for after you fixed your focus and bottom camera magnification scale. I provided some elements in my post about that and the manual will detail even more. I believe the lighting will LOOK different once the part is closer.

[vincentp]

just dropping a post about 0402 : I use them for the first time in a design and I applied the gained knowledge about half indexing (covered in the handbook I put online). It's working terrific ! Didn't have to touch anything about the calibration I did long ago now, placement is dead on, very accurate. *thumbs up MDC* !
to avoid pick up misses on the second pocket (the one the closest to the machine internals), don't forget to major slightly (2 to 5 Z axis steps) the stroke depth to compensate the fact the tape starts to bend down and is therefore further from the nozzle tip.

[bmw330i]

just dropping a post about 0402 : I use them for the first time in a design and I applied the gained knowledge about half indexing (covered in the handbook I put online). It's working terrific ! Didn't have to touch anything about the calibration I did long ago now, placement is dead on, very accurate. *thumbs up MDC* !
to avoid pick up misses on the second pocket (the one the closest to the machine internals), don't forget to major slightly (2 to 5 Z axis steps) the stroke depth to compensate the fact the tape starts to bend down and is therefore further from the nozzle tip.

HI Vincent, this is great news for me. I have a design that uses 0401. I'm only just now able to sit and focus this week on my machine. I am going to start with calibration. Wish me luck.

[vincentp]

calibration should be no problem, I really ran mine in less that 2 hours total while I still had to figure out what I was doing. Just make sure you have a few key parts, like one QFN and a bunch of 0603 (I prefer black resistors, I better see the edges when tuning / correcting offsets).

[bmw330i]

Hello all,

Thanks to your handbook Vincent I'm past many many issues that have come up. Just today I got a shottky diode to get recognized along with some other parts. I'm inputting my last part into the PCB parts menu and now have a really weird error/issue not found in the manual.

In the PCB parts menu I see parts 1-50. I have 36 parts on my board. The items are left over from copying another boards files and renaming them. I line by line went in and simply redefined 1-35 as new coordinates and parts on the PCB. It was all fine until I got to part 36. I added the part in the parts definition, then added the feeder 203 (3rd tray). I even in the feeder got the part to be recognized ok using LV1. However when I edited the PCB menu the line was not saved. I also noticed that when hitting <enter> the cursor did not drop to the next line. Usually enter goes to the next line. I tried then deleting and get an error. If I try deleting any line it says "Error" without a description of the Error. I tried add to add a line and it added line 51.

Is the file corrupt? Any ideas how to remove or get this to save? I tried exiting the application and re-entering. BTW the PCB Bias Point must be updated each time I start the application. But that's not as important is this issue ATM.

-DC

[vincentp]

interesting issue. Are you using the latest version of the software or an older one (if so : which version please ?)
You definitely should be able to delete PCB program lines from an other board, except the step copy lines.

I asked MDC to allow the removal of the step / repeat by clicking on it again and set 0 for X and Y repeats, that will leave only the first PCB. The other solution (I still use sometimes) is to select extend copies then delete lines.

is your PCB a single PCB or a panel with repeats ?

you might want to have a look to the files with an ASCII editor like notepad++. I can't remember if the PCB program is ASCII or hex but you might work on a copy of that file and remove some lines manually.

it could also be an error specific to the part you're trying to add to the model.
Suggestion : change the part / feeder of line 36 (put a 100nF capacitor instead)
try deleting lines / repeats

have you tried deleting a few lines after lines #36 but not everything up to 50 ? does it work ?

[isozee]

have you tried deleting a few lines after lines #36 but not everything up to 50 ? does it work ?

Tried for part type,tape feeder and PCB data deleting from 36 to 50 and did work.

[bmw330i]

Hello,

I tried:
1. in PCB parts list delete only line 36. Edit/Delete Input: 36, 36 <enter> (error)
2. in PCB parts list delete 36-50. Edit/Delete Input 36, 50 <enter> (error)
3. in PCB parts list add line. Edit/Add Input 1 <enter> accepted, now 51 lines.
4. in PCB parts list delete line 35. Edit/Delete 35, 35 <enter> accepted, now 50 lines.PCB shows now 34 parts instead of 35.

Is it related to me doing copy to create a panel and then later going back to edit the PCB lines?

Following the advice from the manual I first defined just the board. Then did a copy of the related files (in the OS doing copy command of the physical files). Then those are the files I am working with now.

-David

[bmw330i]

Vincent, I will power everything up and verify version.

Also, it could be a user introduced error. I possibly did not use the <enter> key for each line when I added the PCB part.

Yes, I have tried to make a panel. Have not tried looking at the file with an editor yet.

-David

[vincentp]

yes, it could be a file error, self corrected by deleting / adding more data.

normally, you can add new lines, the panel will automatically make the copies / replicates. But again, it might depend on the version. I'm not 100% sure as I always make the single PCB version, test on PCB #1 on the panel, then duplicate with step/repeat.

[bmw330i]

I solved it. The placement issue. Not the editing | deleting problem. Calibration. Vincent. I used your manual and went through the different calibrations. Turns out the K45 #1 value needed to be changed. Not sure if during moving the machine or what but it needed to go from 6 to 5. After everything is being recognized much better and placed very accurately.

The board is very small. I may just start over. I have another question for the experts. What is the best way to copy an existing board. Meaning I have a board that's very much like another board. I'd like to duplicate the one and just go change the placement values. The parts are almost the same between them.

-David