Neoden 4 pick and place

[AaiRIz]

Any one can tell that when N5 will be Available ??

[pnp_teltow]

Hi all,

we just bought one Neoden4 and it runs well so far but I have to asked some questions from time to time and this blog seems to be the best way with a lot of skills people.

So here comes the first one:

Is there any way to change the crosshairs colour because ours is black and it`s pretty hard to see espacially when the camera is showing a IC. In most of the vedios and manuals it`s blue and much better to see.

Thanks for you help

[bluetinc]

All,

After much research (this thread was great!) I am almost ready to order a N4/N5.  I have been talking with NeoDen customer support with my questions, and I have been looking into how to slow down the Z access during a pick and place.  They responded that this is already built into the software using the following settings.  Since I don't have a NeoDen machine yet, could anyone with the machine confirm the "pick delay" and "place delay" setting do what I want? 

Thanks, Peter

[TheSteve]

I believe those delays determine how long the nozzle stays down to pickup the part. It would be nice if they added variable speed for the nozzle travel itself.

[Eclipze]

They adjust the delay, but not the speed.  I don't think slowing down the placement head is that important.  It has been my experience that getting the pickup and placement height as accurately as possible is more beneficial.  Particularly with small parts that tend to bounce in the tape.  In some difficult cases, like a SOD323, I have to set those with respect to a specific nozzle, as the nozzle all function at slightly different heights and there is no independent calibration for them.

[mikeselectricstuff]

The reason to delay after placing is to allow time for the vacuum to fully dissipate - this can be an issue for large nozzles and lightweight parts. Also for LEDs with silicone surfaces, which can stick to the nozzle as they seal well.

Obviously these settings should be settable per part type, not globally, though from what I've seen of the N4 software, I wouldn't be surprised if it was the latter.

[TheSteve]

The placement delay is per part.

[rx8pilot]

Is it possible to delay the pickup so the vac has time to build on heavier parts?

[mikeselectricstuff]

Is it possible to delay the pickup so the vac has time to build on heavier parts?

I'd assume that's the "pick delay" value

[Smallsmt]

Pick delay is used to stabilize vacuum on part pickup used for big parts.

Put delay starts after closing vacuum can be  used for light parts to prevent not blown away on vacuum close. I don't know if Neoden 4 aplly a pressure push when vacuum closed.

The pressure push needed to separate parts having silicone surface.

I don't understand why there is no nozzle calibration for each nozzle ask Neoden if this is right maybe function covered in a dialog ?

[Eclipze]

I went back an forth with Neoden regarding the high offset between different nozzles.  I provided pictures and descriptions.  The nozzle are the same length.  I can swap the nozzles around and the difference in pick height always remains the same.  The machine has the error.  I spoke with Haimi last September about the issue.  He advised to check the head height, which I did as best I can.  Rather difficult to measure this.  The 14th email we were back to square one, being asked to try a different nozzle fitted.  Gave up.

I think the picture below is evident enough there is a significant height error.  I didn't resolve how to adjust this myself.  Nozzles 2 and 3 are the closest match and therefore the ones I use the most for volume placed parts.  Otherwise, I'm limited to using nozzles 1 and 4 where they are only used on that reel.  Bigger parts it's not so much of a problem, as I can overshoot the height without risk of bouncing the tape.

[bluetinc]

All,

Just to follow up, NeoDen did confirm after working through a communication hiccup that as Steve stated, those delays are only milliseconds and there is no way to slow only the Z velocity.  I explained how useful having independent control over the Z velocity would be, and they took it down as a possible future improvement.  Hopefully, how useful this would be to users translated well.

Peter

[TheSteve]

We have suggested more control of the Z axis as well.

[rx8pilot]

We have suggested more control of the Z axis as well.

It is so odd that would not be a very high priority from the very beginning.

[mikeselectricstuff]

We have suggested more control of the Z axis as well.

It is so odd that would not be a very high priority from the very beginning.

As well as more fundamental things like that when set to move slowly, it appears to do so even when not carrying parts. Could be that they only have very crude, global control over motor speeds.

[genobpgh]

Just got my Neoden4 and here is my review:

I ordered -

Neoden4 Machine w/ Rails
Solder Paste Stencil "Machine"
Some Extra Feeders
Stand
Monitor/Keyboard/Mouse

After placing the order, it took about 2 weeks to receive the packages.  There were 4 packages, DHL brought 3 of them in a van, and the last in a box truck (it was in a big crate).

Everything was packed VERY WELL.  The machine was bolted to rubber pads that were bolted to the crate floor.  It looked like they put a lot of care in shipping.

Since this thread focuses on the machine, I won't go into the Solder Paste Stencil Thingy-ma-bob.

The Stand -

It is made of thicker steel (like a sheet steel), and bolts together with a zillion allen head bolts.  It came with a thick rubber mat that lays on top of the finished stand.  Assembly time was pretty quick and went without a hitch.  It's not super heavy, but has some weight to it.  I can understand when people say you should fill the inside of it with sand bags or concrete  It's slightly shaky when you have that machine on top of it swinging the head around...

The Monitor/Mouse/Keyboard -

Typical old school VGA dell monitor (was it 1280x1024 or something like that?).  They give you a super long VGA cable which came in handy.  The mouse and keyboard were actually really nice.  Both wireless and uses a little dongle that you can plug into one of the USB ports around the machine.


The Machine (Neoden4) -

It was well packed, very clean, and very heavy duty.  I was very surprised at the quality of the machine.  It looks like it has a lot of Steel and Aluminum parts.  It fit on the stand with the help of a friend lifting it up and putting the bolts through the tabs.  I had to drill out 2 of the holes in the tabs on the stand, they were a bit too small for the bolts that were shipped (no big deal, stepper bit to the rescue).

Before I started up the machine, I went through and removed all of the clips that held the X,Y head in place.  I also double checked the feeders to make sure they were tight and plugged in. 

Everything looked good, so I fired it up.

The software boots up pretty fast.  Before you know it, you're at the main screen which allows you to pick a file.  Everything on my machine was in English.  It did have a few Chinese PNP files which I assume are for testing?

The software is pretty simple, it doesn't have a ton of menus or config - it gets right to the point.

Since I had a bunch of boards ready to go, I skipped messing around with their test board and went right to my stuff.

The machine came with a little plastic case with a few heads in it.  I put the smaller ones in (I don't have the sizes in front of me).  I then loaded up a test reel of 10k resistors (nice and cheap!!  can waste them all day!).  Now I had to add the part to the software - this is pretty easy to do.  I will admit I watched the youtube videos about 10 times and had to decipher some of what was going on looking at the videos then back and the screen. 

The software can definitely use an "English" overhaul - but if you can get past that, it's pretty straight forward.

Once I had the part in, I fiddled with a few things and tried to do a test pick ONLY.  Well, the head just went down and picked up - NOTHING.

ok?  so I went over to the system menus and saw that when I put my finger under the head and pressed "SUCK" there was no vacuum. 

I realized (after about 30 minutes of futzing around) that the vacuum pumps weren't on.  I hit up my representative at NeoDen and filled her in with the problem.  In the mean time, I said the hell with it, I'm going to see what's going on...  It's in our nature!

I took the gamble of fiddling with the machine - well aware of the consequences...  I figured if I break something, I ain't shipping this thing back and I'm sure they're not going to send a repair man from china.

I lugged the machine back off of the stand, and took the panel off of the bottom (thanks to this thread, I know exactly where the guts of the machine were and how to get to them).  I looked around, and noticed that the molex plug from the transformer was plugged into the 220v slot.  Right above it was a 110v slot.  I also measured the voltage on the vacuum pumps.  They are 12v pumps (written on the side of the pump) and the voltage going to them was 6v.  Well since I'm in the US, we need 110v   So I moved the connector to 110v and crossed my fingers, and turned it on. 

The vacuum pumps kicked on and the machine booted up. 

In about an hour (8am Chinese time), Neoden got back to me.  I sent them pictures of the change, and they confirmed that the machine was shipped for 220v and my fix was appropriate.  I was surprised at how responsive and helpful they were. 

After that minor hiccup, I got the machine back together and started testing some feed, picks, etc.  Everything worked great.

I then wrote a program to take my data (from pcbartist) and put it in the proper .csv format for the machine, and got it all imported to do a dry run of the board. 

Loading the reels...  This was a pain until I got the hang of it.  The hardest part is getting the plastic peel through the peel feed slot...  If the plastic doesn't have tape on it to help guide it through, it is a f%$#ing pain to maneuver it through that slot.

After getting 12 reels loaded, I started working with the software to set everything up.  Like I said earlier, it's not super difficult once you get through the wacky translation issues. 

At this point I'm going to jump to the process of board recognition and "Mark Points".

The first thing I did is setup the edge recognition.  I put the board in the feeder, and clicked FEED.  It feeds the board and attempts to find the edge.  It found the edge every time on my boards.  At this point you can do some adjustments on the alignment if needed.

The next part was dealing with the "LEFT BOTTOM".  I am not using panelized boards, so my rows/cols parameters were 1x1.  This is definitely a part where the software could be more clear.  The "LEFT BOTTOM" refers to the first part.  (I got this from watching the videos).  If someone else understands this better than me, please chime in.

What I did -

I created a part in the parts list called "FID1".  I picked a fiducial and entered in the coords from the pcb software.  I then went to the "LEFT BOTTOM" param and manually aligned in to the FID1 part.  It seems like at that point the software can get a relative coord calculation in place.

After doing this I added the 3 "Mark Points" in which I select 3 fiducials.

Now I was ready to go. 

I ran the board through - it found the fiducials, and started placing parts...  It placed them all over the office    Pretty much everywhere except where they were supposed to go!

Not even the double sided tape on the board helped.

I quickly learned that I had to spend some time adjusting the pnp parameters.

After going reel to reel (in the software) setting up "Pick Height, Place Height, and Delays", I got it tuned up and was able to get parts down with no issues.

I then switched of to a pasted board, and had a few problems but did a bit more fine tuning and was able to get a board completely finished.

*NOTES:

I was using 1206 passives, and a few SOT23 parts.  I am in the process of making a tray that I'm going to cut out of plexi on the laser cutter to try some IC's.  Once I get a tray I will try some of the fine pitch stuff.

I did about 50 components and they laid down very accurately.

Also, I had 3 SOT23 transistors that were going down about .5mm off.  After blaming the machine, I realized the part center in the PCB software was off.... 

Overall - I would say the software is very decent once you get past the translation learning curve, and some of the minor quirks.  I think the machine is great for the money spent. 

Thanks for reading my long post!

[penguinman]

I have a Neoden-3 and a TM220A

The Neoden-3 was no faster than the TM220A, because it is larger and heavier.
However I needed the extra tape reel spaces and large-IC placing ability.

These machines have saved me a huge amount of time and money, I've used the TM220A for 3 years, and the Neoden-3 for 1 year.
However I always need to visually check the boards for misplaced or missing components.

A vision system would be awesome and and the Neoden-4 looks great.

The TM220A is for sale.
It works perfectly, however as mentioned I am using the Neoden-3 because it has more reels.

[rom67]

Hi Guys!
Could anyone share the last stable firmware for NeoDen?
Is the firware ver. 4.1.1.B7 the last one?

[TheSteve]

Hi Guys!
Could anyone share the last stable firmware for NeoDen?
Is the firware ver. 4.1.1.B7 the last one?

Neoden is always happy to send the latest version, just email them.

[rom67]

Thank you TheSteve!

Unfortunately, the NeoDen is forwarding  me to the local distributor.
I would like to have free opportunity to change to a new firmware by myself.

For example, I know that Neoden has recently released a new firmware.
My local distributor tells me that it is raw, and that it should not be used.
I would like to be able to decide on my own whether to install or not.

I think (hope) that the new firmware should be free and open without any links with the distributors.
So I hope anyone give me the information about the last version of the firmware and share.

Thank you.

[LitlePCB]

Currently I am really pissed off by this device!

Support no longer responds... probably a lawyer will have to take over.
Positioning error on X and Y axis (about >=0.2mm/100mm)...
Feeder positioning error on 0402 components so 0201 is absolutely impossible!!!! (do not believe the described possibilities of their advertising)
Not described Software setting in the operating instructions (size correct???)

[Eclipze]

I don't use the size correct option.  It seemed related to vision checking, but I found it just dumps components unnecessarily.

I use 3 fidicuals for alignment, as wide spaced as possible.  Also, make sure parts placement is based on the copper.  Never use the silkscreen or mask as a reference.

I've place a lot of 0402.  You really need to set the pickup height with care, otherwise you'll just bounce them around.  I use different height offsets, as the pickups aren't precise enough across the entire machine.

Hope that helps!

[rom67]

Yes, I also have misalign of mounting the SMD components.
Here is the pic.
Unfortunately the camera is not good image and the nozzle has NOT symmetrical hole, so I can't get the stable calibration position of the nozzles.

Also I can't achieve a stable capturing of the SDM 0402 by the nozzle XN03. The XN07 is too much.

Eclipze, thank you for share.
I wonder to know what the nozzle type are you using? (XN03, XN07)

[Eclipze]

I'm not sure on the nozzle size names.  I use the smallest I have.  The next size up is just too big for them.

Sounds like you haven't calibrated the nozzles to the cameras.  It's in the manual.  Go into the System Configuration.  The Config pwd is nothing, just hit enter.  You need to do the lens rotation and then adjust the crosshairs on each so the upward camera is aligned with the nozzle.  Then there is another process where you align the down camera to the nozzles, where you move location to a PCB and mark it.  Need to have something like carbon paper so you can see where it pushes the nozzle.  Then you line up where it's marked with the downward camera.

[LitlePCB]

I don't use the size correct option.  It seemed related to vision checking, but I found it just dumps components unnecessarily.

I also do not use this option because it is not described anywhere.

I've place a lot of 0402.  You really need to set the pickup height with care, otherwise you'll just bounce them around.  I use different height offsets, as the pickups aren't precise enough across the entire machine.

The pickup heigt has nothing to do with the xy coordinate System. I have placed graph paper on the workbench and did some measurements with the camera. In the X axis I have an accumulating error of >= 0.2mm per 100mm. In the Y axis round about 0.15mm/100mm.