Hello,

I've bought a used SMT VP-2500DP. After reading all the useful informations here and in the manual I was able to get the machine working very quickly. With "easy" components like 0603 or 1206 I have no problem and I'm very happy with the placement results. At the moment I try to place a QNF48 and a TQFP64 and have no success to place it exactly. The following steps I have done:

- Go to the calibraton chapter of the manual and check all the settings
- Adjust the Nozzle1 to Mark Cam Distance (like described here in the forum)
- Adjust the Noozle1 to Noozle 2distance
- Check and Adjust the Noozle to Camera1 and Camera2 Position
- Check the Noozle to PCB-Holder Z distance

I've run the the machine in mid-speed setting and set the speed template to "take the chip speed" for the described parts. The parts come from a grid feeder. I adjusted the pick-up height in the feeder settings for the parts and set the right component parameters and an Auto-Magnify setting of 1.3 . For the PCB I used the mark point calibration as described. The feeder related settings are:
- visual collect time 300ms / alignment precision 1 for XY and A
- pick delay 200ms
- put delay 300ms
- I use a CN140 Noozle

For the QNF48 I use camera 1 and for the TQFP64 I use camera 2. The pickup works perfect and also the visual alignment tests give good results (picture show detection of the QNF48). But both parts are a little bit missplaced on the board (see pictures). For the test I use a thin double sided tape on the PCB.

Did I miss a setting? Or did someone have an idea what could be the reason for this?

Many thanks for your help and best regards

Sebastian|addpics|vgh-1-9c09.jpg,vgh-2-207f.jpg,vgh-3-c873.jpg|/addpics|

Hi Michael,

thanks for your reply - I will have a try :)

Hello again,

I tried to reduce the height settings for the parts. After this it was a little bit better. I also tried to use the camera correction settings, but at my last tests the missalignment was not the same everytime. The testjob contains a panel with 3 single PCBs. I use mark detection on every single PCB. The first board has only a small missalignment, that should not be a problem (attached pic.) The second PCB has a bigger missalignment (attached pic.) that is diffrent to the first board and that could make bridges after soldering. The third board again is better as second, but again diffrent. So for the moment I have no direct idea what settings I need to tweak to make the placement better. I also made two video of the process:

Edit: I'm unable to include external links in the post. I will attach this in the next post.

Does anybody has an idea?

Many thanks and best regards
Sebastian

Sorry - "placeholder-post" for external links ...

Placement-Video: youtu.be/fw4gW9heLuw
Software-Video (for the placement): youtu.be/lQObNXhC6Dc

Edit: Sorry forgot one point. I also installed some kind of support-structure under the middle PCB to avoid bending.

I have the same problem with LQFP-64 and LQFP-100 (0.5mm pitch) positioning on multiple board panels. On the first board, everything is almost always located correctly. On the second and third with an offset.
I recalibrated the upper chamber, second head and lower chambers. The result is practically the same. Placement speed is set to "Chip speed", collection time is 500 msec.
There are no problems with other components, for example 0402 or QFN.
My guess is that the analog camera cannot handle large components with small steps.

I ran out of ideas on how to defeat the LQFP chip.

Zitat

---

You all use belt machines?
Recalibration of X and Y axis on changing temperatures is important.

---

VP2800HP

Zitat

---

Second problem check Z axis distance on different PCB positions and adjust PCB holder height if different.
Posiible PCB got bent?

---

Perhaps the best way to do this? Use a nozzle on head?

Zitat

---

Too high speed used for placement?

---

Low Speed on "Chip speed", collection time is 500 msec.

Zitat

---

Right nozzle used?

---

Yes. How can this affect?
Visually, the chip presses down a bit when placed.