Author Topic: Why so few low-end "manual" pick & place machines on the market?  (Read 8925 times)

0 Members and 1 Guest are viewing this topic.

Offline Dave8266Topic starter

  • Contributor
  • Posts: 35
  • Country: us
I've recently been researching pick & place machines, and it struck me as odd how few basic manual pick & place machines were on the market.    Does anyone know why?

The mechanics are trivial, and the electronics (if any) are absolute commodity stuff - In this world of $100 3D printers and DIY CNC everywhere, I would have expected basic X/Y/Z carriages with an aquarium pump vacuum pen to be a product that at least one vendor would have latched onto, if not dozens of options from overseas.   But it seems to be an abandoned market.

There are of course plenty of high-end "request for quote" options (Fritsch, Manncorp, etc), and mid/high-tier machines like the "SMT Caddy", Fortex MPP1, Gold-Place MPP-11, CIF V900121, those from Madell, etc.  But at prices from $1,700-$3,500 USD, you're already above the cost of some hobbyist automatic Pick & Place machines (Liteplacer, OpenPNP, SimplePNP (if it ever delivers), etc), and quite close to the low-end commercial automatic units with feeders (Charm High CHMT36VA, etc).   The market is also littered with references to nice looking mid-range machines that no longer exist - such as the Eurocircuits eC-placer, VEGATEC V900022, LPKF ProtoPlace E, etc, all seemingly discontinued in favor of costlier machines with more features.

Aside from a bunch of clever homemade machines (vpapanik, etc), really the only low-end manual machine I could find for purchase was the PickSoEasy PSE-20... but even that starts at $580 without cameras or feeders, and goes up from there.  The next closest was the Abacom ezPick at $700, which looks like it uses drawer slides for motion?


I'll probably end up making something myself - pricing out extrusions, linear motion components, etc, it's looking to be <$100 in parts.  I'm just really surprised that the market doesn't already have these.   :-//

« Last Edit: June 20, 2021, 05:20:42 pm by Dave8266 »
 
The following users thanked this post: electrolust, liteyear

Offline fcb

  • Super Contributor
  • ***
  • Posts: 2117
  • Country: gb
  • Test instrument designer/G1YWC
    • Electron Plus
Re: Why so few low-end "manual" pick & place machines on the market?
« Reply #1 on: June 20, 2021, 06:06:02 pm »
It's all about the FEEDERS!

Used machines without feeders are virtually worthless, and a decent compliment of feeders will often be worth 2-3x what a machine is worth.

Before you cut a length of rail, build a nozzle or write a line of code, sort the feeders out.
https://electron.plus Power Analysers, VI Signature Testers, Voltage References, Picoammeters, Curve Tracers.
 
The following users thanked this post: ANTALIFE, Reckless, wincyj

Offline jduncan

  • Contributor
  • Posts: 36
  • Country: us
Re: Why so few low-end "manual" pick & place machines on the market?
« Reply #2 on: June 20, 2021, 06:07:54 pm »
It doesn't offer anything to a hobbyist with good eyes and a steady hand, and it's completely worthless for commercial purposes. No market.
 

Online Bud

  • Super Contributor
  • ***
  • Posts: 6860
  • Country: ca
Re: Why so few low-end "manual" pick & place machines on the market?
« Reply #3 on: June 20, 2021, 06:26:36 pm »
Since the question was about "manual" pnp, it is quicker to assemble a board with a tweezers. Also no need to prepair parts, just take the cut tape out of the bag, take what is needed and put it back in the bag (or reel). With my hand and tweezers i easily align parts with pads. Not sure what it may take to rotate a part just a smitten on a "manual" pnp.
Facebook-free life and Rigol-free shack.
 

Offline jduncan

  • Contributor
  • Posts: 36
  • Country: us
Re: Why so few low-end "manual" pick & place machines on the market?
« Reply #4 on: June 20, 2021, 06:46:38 pm »
Exactly. It is not hard to do 500CPH by hand with tweezer, and for a familiar design or simple panel you could go quite a bit faster.

Yes, a microscope helps a lot for small stuff but that's just a couple hundred. I am faster without magnifcation down to 0603, when I was younger probably 0402 would have been fine.
« Last Edit: June 20, 2021, 06:48:58 pm by jduncan »
 

Offline Jackster

  • Frequent Contributor
  • **
  • Posts: 460
  • Country: gb
    • PCBA.UK
Re: Why so few low-end "manual" pick & place machines on the market?
« Reply #5 on: June 20, 2021, 07:22:55 pm »
I looked at a few. Came to the decision that once you set up manual pnp, you would have done most of the boards you wanted to do anyway.

There is a very thin margin (quantity of boards) between doing it by hand and sending panels off to assembly services..

Offline Skashkash

  • Regular Contributor
  • *
  • Posts: 118
  • Country: us
Re: Why so few low-end "manual" pick & place machines on the market?
« Reply #6 on: June 20, 2021, 07:51:32 pm »
I agree with the OP.

   I personally do feel the need for an "assisted" manual placement system with some vision assist. 

   I too am probably faster with tweezers and any two pin parts down to 0603,  but QFNs and other similar fine pitch parts give me a lot of trouble to align properly.   Not looking for something with automatic tape feeders. Just something that can grab loose parts from a work area.

   Hard to believe there is nothing reasonably priced available between tweezers / binocular microscope and a full blown PnP machine.  It may be a niche, but a niche that could be filled. 
   
   I think there is room in the market for that exactly that kind of machine,  the mechanical and vision bits and pieces appear to be commonly available.     

  I will probably wind up making my own version of this system eventually, hopefully before my eyesight and motor skills cease keeping up with the reduction in smd part sizes.   
 

Offline Doctorandus_P

  • Super Contributor
  • ***
  • Posts: 3313
  • Country: nl
Re: Why so few low-end "manual" pick & place machines on the market?
« Reply #7 on: June 20, 2021, 11:10:04 pm »

There are some articles such as the ezPick manual SMT P&P machines:
https://abacom-tech.com/product/manual-smt-pick-and-place-machine-ezpick/

But don't ask me why a few drawer sliders, an aquarium pump and the mechanism from a shaving mirror would be worth USD700.



There are similar products on the market for somewhat similar price, or even more expensiver, but with pneumatic brake systems installed.
Here are a few other ones (Oops, you already mentioned these):
https://www.fritsch-smt.de/en/manual-pick-place/
https://smtcaddy.com/

Or if you like pictures more better:
https://duckduckgo.com/?q=manual+pick+and+place+SMT&t=h_&iax=images&ia=images

There are a few cheaper options though. I remember seeing something similar (but simpler) on either kickstarter, indiegoo, or similar.


The chinese table top P&P machines start at around EUR3000 such as the TVM802A and TM220A. There are also some people who sell kits for products like litePlacer or OpenPNP.
For a small business that has production runs between a handful and 50 or so PCB's there seems to be a niche market for such products.
For 10 PCB's it's becoming tedious to put the stuff on manually, but the setup costs are problematic for out sourcing.

I have a moderate interest in products such as litePlacer & OpenPnP.
I would not buy one though, but design and build something myself, as it's more just for fun then for actually using it.
I've made a design for a SCARA arm that may have enough resolution for decent P&P work, but have not build it yet.
Feeders for this kind of machines is not a completely solved problem yet.
Hackaday has collected quite a bunch of DIY feeders over the last few years: https://hackaday.com/tag/pick-and-place/

For small runs and light use, the setup time of an P&P machine is an important factor, and so are the costs of the feeder itself.
Speed is not a mayor issue for such machines, and therefore I think a feeder that is operated by the P&P head itself is probably the best option.
Stripping the top tape reliable is also messy. Most feeders need long leader tapes and have a separate winding mechanism, which requires a lot of manual setup time.

I find the "Ploopy" quite charming. https://hackaday.com/2020/01/23/a-ploopy-pick-and-place/ It pulls the top tape loose on only one side, and then folds it over. This way no leader tape at all is needed. I do find Ploopy much to big though, and it does not have a built in "magazine" for parts.

Ploopy lifts the tape on one side and then folds it over. Maybe it's a better option to cut the top tape with a knife in the middle, and then fold both flaps outward.

This is a project for small SMT magazines: https://hackaday.com/2019/11/07/3d-printed-magazines-tame-the-smd-tape-beast/

The optimal P&P feeders for a DIY machine and fast setup would probably be a sort of cross between these projects.

* Magazine for 250 or so parts ( Which is  a meter or tape)
* Small form factor, so you can store a bunch of them in a small space.
* 3D printable, or other easy manufacturing method. (Maybe partly made out of FR4).
* Mechanism to advance the tape by a single component (multiple of two or 4mm).
* Operated by the P&P head to keep cost of feeders low.
* Self contained mechanism. Just take it out of a drawer and place it on the P&P. No other setup should be needed.
* Mounting on fixed positions on the P&P machine. Maybe with some lego like system.
* Label with bar code, or which can be read by OCR. (Idea is to have a camera on the P&P head, and let it scan the positions of all feeders itself.)
 

Offline Dave8266Topic starter

  • Contributor
  • Posts: 35
  • Country: us
Re: Why so few low-end "manual" pick & place machines on the market?
« Reply #8 on: June 22, 2021, 04:50:55 pm »
Thanks guys, good discussion.

Yes, I agree- in many cases, it makes sense to stick with tweezers for low-volume/prototype, and move to automatic machines or outsourced production beyond that.

But I can still see a market for a low-end manual machine - After all, I wouldn't have done the research above if it wasn't worth it to us, for example.   It could be hand shake issues, or worsening eyesight.  In my case, the breaking point was a physically large PCB, where the extended tweezer reach (and inability to brace the rear of my hand without disturbing other components) made fine pitch placements problematic, and we were spending too much time reworking assembled boards.   A vacuum pen also nicely solves the "tweezers sending the tapered part flying across the room never to be found" problem.  So why not address everything all at once for a couple hundred bucks or less?

We're not assembling enough boards for the setup fees for contract assembly to be worth it.  (That prototyping intersection of too many unique boards, with too many unique components, but not needing enough quantity of any one board.)   And we couldn't do that anyway, because too many of our parts are zeroed out by the chip shortages other than random scraps of cut tapes, tubes, and loose bags that we already have on hand- and most all contract assembly places require continuous cut tapes with ample spares.   If we bought an automatic machine ourselves, I'm not sure the feeder setup time and re-taping would be worth it- we could have had the board assembled already if we had just done it by hand.

So throw everything together, and a manual pick & place, literally the most simple type, really starts to make a lot of sense.


 
The following users thanked this post: liteyear

Offline Doctorandus_P

  • Super Contributor
  • ***
  • Posts: 3313
  • Country: nl
Re: Why so few low-end "manual" pick & place machines on the market?
« Reply #9 on: June 22, 2021, 09:41:29 pm »
I've thought a bit about the "gap" between these machines, and come to the conclusion that it's quite logical.

First it is a niche market.
Then there are the hobbyists. They do not want to spend any money if it can be helped, and much rather cobble something together themselves.

Then, for the commercial market it does not matter if such a machine costs EUR300 or EUR3000. Cost of the machine is small to the salary of the operator.

------------------
I don't like the PSE-20 much. It looks very un-ergonomical.

I do like the simplicity of the VPApanik with the big arm rest.
 

I do not like the flimsy aluminum guide rails though.
Rubbing the sharp edge of a ball bearing into soft aluminimum is a recipe for a very short lifetime.

This one is similar in appearance, but better thought out:
 
 http://trains.socha.com/2014/07/building-manual-pick-and-place-machine.html

And this one:
 

But I would not bother with home-made rails, and buy a few MGN12 rails.

The next step up would be to add some electromagnets (for example operated by a foot pedal) to lock the X&Y axis in combination with a (for example 4:1) pantograph to do the final placement.
 https://en.wikipedia.org/wiki/Pantograph

I am a bit surprised that none of the commercial machines use a pantograph to reduce hand movement, and make accurate positioning easier.


 

Offline jesuscf

  • Frequent Contributor
  • **
  • Posts: 499
  • Country: ca
Re: Why so few low-end "manual" pick & place machines on the market?
« Reply #10 on: June 28, 2021, 09:55:38 pm »
Coincidentally I have been working in a very simple P&P machine using the Silhouette Curio hobby machine.  The total cost of the setup is around $300 (US$). I just uploaded a video of an early "test" to YouTube I made about three weeks ago. I called it "test" because I ended up assembling a total of 300 boards.  In retrospective, that was a mistake, because now I don't have any boards to test "for real" many of the improvements I made since.  The smallest components I tried are 0805, but it may be possible to do 0603.



Roughly the workflow of using the Curio to do a P&P job includes:

1) Configure the job using a graphic editor (written in Python/tkInter).  No much trouble here as the editor should be easy to use.  Since the setup doesn't use a camera, it requires accurate measurements of the cut carrier tape to achieve good results.  The editor allows for copy/paste, so that is helpful to speed up the job.

2) Cut a craft foam gig to place the PCBs and cut carrier tape.  The Curio is used to cut this gig.  This craft foam gig can be used many times.  I used one 10 times and it is still good for more uses.  This is the hardest part of the whole process and may take some practice to master.

3) Load the PCBs and cut tape in the foam gig.

4) Place the gig in the Curio tray.

5) Remove the transparent cover from the cut tape.  This, as you probably know, is a pain!

6) Start the pick and place process.  Wait.  Remove the PCBs with the placed components.  Reflow. Done.

My setup uses, out of the top of my mind, the following:

1) 1 x Silhouette Curio cutter.  About 200$.

2) 1 x 12V vacuum pump. About 30$.

3) 1 x 12V solenoid valve. About 15$.

4) 1 x microcontroller to control both the pump and valve via a USB to serial adapter.  I used a custom made board, but an Arduino or similar should do fine as well. 20$?  Bot the pump and valve are turned on/off using an N-MOSFET.

5) 1 x Silhouette pen holder.  Actually, it is better to have two.  One pen holder for the pickup tool, and one pen holder for the sharpie marker. About 12$ each at Michael's.

6) 1 x 0.5 mm mechanical pencil with non-retractable tip.  The one with metal tip works better.  I am using the Pentel Twist-Erase Express which I got from Staples in a package of 4 for about 10$.

7) about 2 meters of vinyl tubing for the vacuum pump.  I think I got it from Home Depot. ??$

8) 3 x 'T's for drip irrigation or similar.  I 'inherited' a bunch of these, but you can find them cheaply at Amazon.

9) 1 x roll of double sided Duck tape or similar.  Amazon carries it.  About 10$.

10) 1 x roll of double sided Scotch 3M tape (I wouldn't compromise on this one as I tried other options and they were awful).  About 6$ from Home Depot; they call it "indoor carpet tape".

11) a cutting mat from the dollar store OR a piece of 30 gauge aluminum sheet or similar that is 256mm x 186mm OR a blank PCB that is 256mm x 186mm.  They all work fine.  The cutting mat is easier to work with, but the aluminum sheet and blank PCB are more robust.  I paid 1.5$ for the cutting mat

12) a 12V dc adapter. 15$ perhaps?  I am using a recycled one.  It powers the microcontroller, pump, and solenoid valve.

13) a Silhouette deep cut blade to cut the foam. About 30$.

14) one or two fine point Sharpie markers.  Red and black worked for me.  They are used to write the values of the components in the foam. Cheap;  I don't remember.

15) Lots of 2 mm craft foam. I got 40 sheets for about 15$ from Michael's, but also used the one from the dollar store.

16) A computer running Windows 10 or Linux.  If running on Windows 10, a driver must be installed to access the Curio.

I'll leave it here for now.  If anybody is interested on setting this up, I can continue with more details in a separate thread.

Homer: Kids, there's three ways to do things; the right way, the wrong way and the Max Power way!
Bart: Isn't that the wrong way?
Homer: Yeah, but faster!
 
The following users thanked this post: edavid, wincyj, JesseLackey

Offline RetireMeNow

  • Newbie
  • Posts: 3
  • Country: nz
Re: Why so few low-end "manual" pick & place machines on the market?
« Reply #11 on: October 20, 2021, 04:32:50 am »
FWIW We purchased a manual SMT PnP machine from Madell Technology, the SMT-PLC-3. It turned up broken due to poor packaging. It also was an incredibly poor quality build and looked like it had been made by a 5yr old. They offered to send us the 3D print file to replace the broken part ourselves but the whole unit was so mind bogglingly terrible we didn't bother for fear of other parts breaking. Quality issues were things as simple as cross threaded screws, very cheap/low quality 3D printed parts and an aquarium pump for the vacuum. Not what we expected for $2K USD. Feels like it was bought in from China for $200 and sold for $2000.
We eventually returned at our cost and got a partial (85%) refund. I would not go anywhere near this company again!  >:( >:( >:(
 
The following users thanked this post: electrolust, Reckless

Offline jmelson

  • Super Contributor
  • ***
  • Posts: 2752
  • Country: us
Re: Why so few low-end "manual" pick & place machines on the market?
« Reply #12 on: October 20, 2021, 03:28:37 pm »
I just don't see the point.  What does the manual machine actually DO, that you would not be doing by hand?
I just did a batch of 20 boards on my 20+ year old Quad QSA30A P&P machine.  That was 67 parts / board and it completed each board in 100 seconds, including nozzle change time and fiducial recognition.  This machine has a conveyor that fixtures the board, so no need for cut-out plastic fixtures to place the boards.  All the parts are checked and aligned by flying vision cameras.  Also, I take the P&P file from my PCB CAD software and run it through a little converter program and import it into the P&P machine controller, no manual teaching or laborious data entry.  I did spend some serious money and time to get this machine up and running (about $7500).  But, I am sure glad I got a real P&P machine and not a manual toy.

Jon
 

Offline Kjelt

  • Super Contributor
  • ***
  • Posts: 6458
  • Country: nl
Re: Why so few low-end "manual" pick & place machines on the market?
« Reply #13 on: October 20, 2021, 06:19:28 pm »
I just don't see the point.  What does the manual machine actually DO, that you would not be doing by hand?

First of all a manual pnp is a supportive tool, not a fast production machine like a full pnp, so comparison between the two is invalid. It is not for business and large quantity pcb production.

The whole point of a manual pnp is to aid a user placing parts that are either too small or have small pin distance too small to place with a tweezer or that the user is unable to place with tweezer or vacuum pincet due to any reason, being sight problems or shaking hands or too small parts to be placed by a human.

Besides placing on boardsfor hobbieists, I have seen small repair companies that uses manual BGA rework station to (re)place BGA’s for repair and another manual pnp vacuum tool placer to remove/replace 01005 sized components.
Very valid applications IMO.
 

Offline Rat_Patrol

  • Frequent Contributor
  • **
  • Posts: 301
  • Country: us
Re: Why so few low-end "manual" pick & place machines on the market?
« Reply #14 on: October 20, 2021, 08:33:20 pm »
Feeders.

I actually designed up some 3D printed feeders, I have a thread a few down about it.

I have them setup with around 80 unique components. Sure, you can dump a few components in a try and screw around with tweezers, but this keeps stuff organized and smooth.
 

Offline Harjit

  • Regular Contributor
  • *
  • Posts: 141
  • Country: us
Re: Why so few low-end "manual" pick & place machines on the market?
« Reply #15 on: October 23, 2021, 06:45:10 pm »
@Rat_Patrol - can you post a link to your feeder, please?
 

Offline Rat_Patrol

  • Frequent Contributor
  • **
  • Posts: 301
  • Country: us
Re: Why so few low-end "manual" pick & place machines on the market?
« Reply #16 on: October 24, 2021, 07:13:30 pm »
@Rat_Patrol - can you post a link to your feeder, please?

Here is a link to the discussion.

I haven't decided if I'm going to open source them or not yet.
https://www.eevblog.com/forum/manufacture/opinions-is-this-a-marketable-product-manual-smd-feeders/
 

Offline Doctorandus_P

  • Super Contributor
  • ***
  • Posts: 3313
  • Country: nl
Re: Why so few low-end "manual" pick & place machines on the market?
« Reply #17 on: October 30, 2021, 08:17:54 am »
Hackaday has collected some DIY projects around SMD feeders:

https://hackaday.com/blog/?s=smd+feeder

And there are plenty of youtub video's about those things. From simple DIY projects to probably most or all of the professional brands.
https://www.youtube.com/results?search_query=smd+feeder

One of the bigger problems for a DIY feeder is the cover tape. This always needs some extra mechanism which drives up complexity of the feeder and it usually takes a big part of the time to load ans setup the feeders.

Some time ago I saw a feeder that did not lift the cover tape at all.
Instead it lifted the center of the tape a bit, then cut it with a knife over the length of the tape and then folded both flaps outwards. I like this method a lot. You don't need any leader tape, nor do you have to put the leader tape into some spool. You can just push in a piece of cut tape and start using it.


 
The following users thanked this post: exe

Offline Topazio

  • Newbie
  • Posts: 3
  • Country: it
Re: Why so few low-end "manual" pick & place machines on the market?
« Reply #18 on: June 02, 2022, 03:44:23 am »
Hi

I own a PickSoEasy PS20 with 2 USB microscopes.

All the PSE 20 sold have at list 1 USB microscope, there is no model without microscope.

I don't know where you got that price...

on the website (where I bought) the entry level model with only 1 microscope is sold at 410 euros, I bought the double microscope for 460 euros. Honestly, it took out all my stress I had when I was using the tweezers.

The guy now is working on an upgrade kit for all the PSE20 to add stepper motors, he says you will control the machine with a gamepad... and you will be able if you want to use the centroid file of your eagle to send the machine automatically right to the pads position....

He even says you can connect a laser and cut MYLAR to make easily your own stencil....

It seems to be good to be true...

I will believe when I see it

However this is the machine I have ,  with double microscope...

 


It is sold with all these accessories




I have since 3 months now... it took me some days to calibrate and  learn how to use it, fortunately there are videos on YouTube

There are a lot of automatic expensive openPNP machines on the market but few manual for quick prototyping or small productions...

Before to buy I made an extensive research, and this was the cheapest considering the level... I was almost buying one from a German company for 2000 euros.... More refined, but with only 1 USB microscope... It had electromagnetic brakes, that it is interesting... It helps to stop in position the smd component before to put it down. I learned to use my fingers to stop the wheels on the aluminum bar, so it doesn't move...

However I'm waiting for these upgrade kit... let's see if it is true...



 

Offline Topazio

  • Newbie
  • Posts: 3
  • Country: it
Re: Why so few low-end "manual" pick & place machines on the market?
« Reply #19 on: June 02, 2022, 03:55:21 am »
I've thought a bit about the "gap" between these machines, and come to the conclusion that it's quite logical.

First it is a niche market.
Then there are the hobbyists. They do not want to spend any money if it can be helped, and much rather cobble something together themselves.

Then, for the commercial market it does not matter if such a machine costs EUR300 or EUR3000. Cost of the machine is small to the salary of the operator.

------------------
I don't like the PSE-20 much. It looks very un-ergonomical.

I do like the simplicity of the VPApanik with the big arm rest.
 

I do not like the flimsy aluminum guide rails though.
Rubbing the sharp edge of a ball bearing into soft aluminimum is a recipe for a very short lifetime.

This one is similar in appearance, but better thought out:
 
 http://trains.socha.com/2014/07/building-manual-pick-and-place-machine.html

And this one:
 

But I would not bother with home-made rails, and buy a few MGN12 rails.

The next step up would be to add some electromagnets (for example operated by a foot pedal) to lock the X&Y axis in combination with a (for example 4:1) pantograph to do the final placement.
 https://en.wikipedia.org/wiki/Pantograph

I am a bit surprised that none of the commercial machines use a pantograph to reduce hand movement, and make accurate positioning easier.

The PSE20 has got long aluminum bars where you rest the arm on...  maybe it doesn't look ergonomical, but if you do like in the video for sure you will not get tired...  I own one, I can say I've never had to hold the weight of my arms...  Plus has microscopes. I use it even for 144 pin 0.5 pitch microcontrollers... try without microscopes...

This video shows a method for precision with 1 microscope.... with 2 microscopes you don't need to use the reference method, you just go to the pad



« Last Edit: June 02, 2022, 04:13:00 am by Topazio »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf