[stewi]

Intrigued by this forum, I like to share my knowledge about commercial high speed pick and place machine and component feeding in particular.
Unlike my previous attempt to market a feeder, which fits all machines, now, I leave this task to my perspective customers.
Although I’ve seen Universal feeders on Siemens machines and vice versa, each machine vendor has its own line of feeders together with protected firm and software operating these feeders.
Also, pick position, physical attachment to the feeder table and electrical or pneumatic connection are different from machine to machine.
In order to keep the tape advancing and cover tape peel mechanic generic and obtain a key advantage to have only one style feeder in the manufacturing environment, which can be swapped from machine to machine, each machine requires its own adapter and specific control electronic. .
I know, I’m an idealist and above-mentioned will likely not happen, but universal chargers on mobile phones give me some hope, the industry may consider my concept.

For most of the single nozzle pick and place machines feeder tape advancement doesn’t have to be much faster than 200 to 300 ms. Multiple nozzle heads could pick from the same feeder, by either advancing the rotary head or the X-axis for in-line nozzle configuration in much faster than 200 ms. In order to compete with Siplace feeders and the Hover Davis copy clones, my feeder should advance in 40 ms. At that speed the component should be covered during advancement and keep particular the small components in the pockets.
However, this is meant to be an open source project, my gear motor has the capacity of advancing the tape in max 40 ms, but I’ll leave this to my customers to regulate the motor speed and component cover feature to their liking.

New in this particular design is the method of tape advancing. As far I know this method is not patented or was not, when I did a quite similar patent in the year 1999. Patents don’t mean very much anyway in China and only if the copied product is exported from China, I could take some legal actions.

I mentioned this topic already in another thread. If you pull too hard on the cover tape, you’ll move the tape out of position. This led me to the idea the tape can be advanced by only pulling on the cover tape. Of course, I’ve tested it and it works very well.
The sprocket wheel is free spinning, but have slots or holes in order to detect 2 mm advancement.
The peel off is designed such, that the small plate with the groove wheel forces the tape down to the sprocket wheel also compensating different thickness of the tape material such as paper and plastic. Two pinch wheels transport the cover tape. The pinch wheels are forced together under spring tension and should allow for thicker splice tape, but will likely stall at cover tape knots.

I haven’t done a complete BOM yet, but I think I can easily beat all feeder prices out there by factor 3 to 5, plus I hope to get some feed back from the hobby and after market.

[scsi]

This is very interesting idea. Thanks for sharing! My only doubt is with gear motor. The cheap ones don't last long at all and in my experience fail somewhere between 20-50 hours of continuous operation. The good ones last hundreds of hours, but cost hundreds of dollars. Maxon and Escap are still in business because of that. The weakest point is the brush wear. Second problem is the gearing, which is hit and miss and can fail unexpectedly. Planetary gears last much longer, but are several times more expensive.

On the other hand, you probably don't need the motor to last a thousand hours. At 40ms per advancement, it can empty a 15,000 component reel in 10 minutes. But the gears can fail sooner because of the jerky stop-n-go operation.

I would look into using some kind of a small steper motor to skip the gearing and allow for more precise tape advancement. In that case your only wear item would be the ball bearings in the motor, which usually last well into thousands of hours. Nema8 is 20mm square. Is it going to be a problem? What's the maximum width of the 8mm feeder are you shooting for?

[stewi]

Thank you for your feedback and concern.
Besides Maxon, Faulhaber and Escap, I also used Canon motors in the past with good results and about half the price than the German/Swiss motors.
Feeder width should be 10.5 mm in order to compete with the triple feeder from Siemens and Hover Davis. The Hover Davis feeders are already significant less expensive, because they are using stepper instead of the Maxon planetary gear motors.
Using a tin can stepper or disc motor 10.5 mm high, I would be worried not delivering enough torque to advance the tape in the desired speed.
Some other ideas, I already discarded: I could tie 3 feeders together and arrange 3 larger motors in a 31.5 mm space to make it work.
Else, I don't use any motors in the feeder and place the motors in the feeder base/table. In this case one large motor could engage with each individual feeder via a solenoid activated gear (I've a patent on that).
Taking everything in consideration, I should get back to the task finding a better gear motor. By the way Canon offers their motors with longer lasting brushes.
BLDC out-runner motors are a nice option too. I could experiment with a tape drive motor and magnet replacements until I find a suitable size on the market.
However, for steppers and BL motors, I'll likely have to place the control electronic in the feeder, assuming most customers don't want to deal with additional circuits.
Bottom line is the price performance question. If the cheap gear motor survives a cycle (40 ms go, 40 ms stop) test under load of 40 hours, than it did 1,800,000 cycles. If I can make the entire feeder for $50, it could be considered as a consumable.
Again, your feedback is very valuable. It is just a concept drawing, I can change easily.

[scsi]

Right, I forgot about the can-stack steppers. They are dirt cheap and last much longer than those cheap gears even with the sleeve berings. Think of estimating the torque necessary to peel the tape in your feeder. I think it's going to work. You can order custom steppers from a manufacturer with really thin serrated shaft. Say, make it 2 or 3mm in diameter and in that case the motor should be able to pull the tape quite easily. For 16 and 24mm feeders you can use larger can-stack motors should you need more torque. Also, there should be a way of reducing the torque necessary by shaping the angle at which the cover tape leaves the component carrier tape. If you pull it at a right angle, it probably needs more torque to advance the tape. Make it more obtuse and it will advance the component tape much easier. Thats' what I found in my experiments with Madell's drag feeders. When I used a cross-bar with wery sharply cut edges, I had to use 6oz sinkers to be able to peel the cover reliably. Then I switched to a cross-bar made of a regular off-the-shelf aluminum strips, which had sligtly rounded edges, and now I use only 3oz weights for peeling. I think the edge has about 0.5mm radius to it and it makes a lot of difference.

I doubt that BLDC outrunner will be suitable mostly because of cogging. Torque is another concern there.

I have a bag full of small can-stack steppers from various floppy drives. I think they are right around 10mm OD. Let me know if you need a couple for experiments and I can mail them to you. They all have long worm shafts about 3mm OD.

[stewi]

Motor plus gear should not be thicker than 10.5 mm. I'll keep on searching, I only found this one http://www.pcbmotor.com/wlp1.php?rl=...FQXt7QodBF6zdw

I'll implement a small roll to reduce the friction on the peel off force.

[stewi]

I was actually looking into a 25 mm diameter stepper but no longer than 10.5 mm. In this case, I would locate the stepper the same way, than the sprocket wheel.
Assuming the pinch wheels have a 5 mm diameter and should pull with ~500g force, than the motor should have 125g.cm torque which convert to 1.7 ozf.in.
In a 1/4 revolution of the pinch wheels the tape advances about 4 mm. The motor should do 5-6 rev per second in order to get my desired speed of 40 ms. Looks like, I'll need to gear up for lower speed of the stepper and torque about 3 oz.in.

Do you have any technical info, source for your 10 mm dia steppers?

[scsi]

Nope, no technical info or source. I just bought a bag full of motors on ebay for my experiments with my auger dispenser.

Can-stack motors should give you 5 rev/sec easily without gearing, as long as the torque produced is sufficient.

[dandumit]

I was actually looking into a 25 mm diameter stepper but no longer than 10.5 mm.....
Do you have any technical info, source for your 10 mm dia steppers?

I have just broken apart my inkjet printer-scanner and scanner was containing exactly the motor that are you talking about. However I think that It had a high number of degrees per step since there where a lot of pulleys.

an example would be : http://www.sanyodenki.eu/60-2-phase-...tor-42-mm.html
But that one was thinner I think.
I will look for a source for such motors.
REgards,
Daniel

[dandumit]

However I still don't understand why do you want to have a motor for each feeder.
Under those wheels can be a linear stage that will move a motor where it's needed.

[stewi]

I found one of the gear motors at Kysan, which supposed to last 300 hours CW continuous operation: A12FT-5V400RPM I'll test and verify. Else, it does 400 rpm at 280 gram cm, which computes to 1.1 kg force on the 5 mm dia. pinch wheel. For a 1/4 revolution in order to advance 4 mm tape it will require 37ms and the feeder should be good for 40,000,000 cycles.
Unless I could be convinced otherwise, I'll be using this motor.
Thanks, Daniel, I found the Sanyo motor as well. Pretty impressive.

I'll finalize my drawings and will see, if I get anybody interested in quoting the job. Have not decided on basic material, sheet metal, milled or die cast alu or plastic (ESD safe) either.

[stewi]

However I still don't understand why do you want to have a motor for each feeder.
Under those wheels can be a linear stage that will move a motor where it's needed.

The disadvantage of this arrangement is, that I'll have to do the feeder base with the motor or linear stage as well and that for all type of machines out there.
I do have quite some knowledge about Siemens machines and could likely also figure out the communication protocol between the feeder and the feeder table in order to make my feeder act like a triple feeder. All physical data, like pick position in X,Y and Z have to be the same.
Again, I've done this before Apparatus for feeding electronic ... - Google Patents
but find, that a motor in each feeder, which cost only $4-7.- (dependent on quantity) is more universal.

[stewi]

Here is my feeder with one motor in the base.
feeder.wmv - YouTube

[Nails]

How do you plan on making the sprocket wheel cheaply? Is there a source for them where you can just buy it?

[stewi]

The only source I know is LaVezzi SK-Precision Sprockets & Rollers
They are definitely not cheap.
The sprocket wheel is a flat, 0.7 mm thick part with square teeth. Initially, I plan to have it laser cut and will have a stamping tool made, if I can make any money out of this project.

[jwolin]

Stewi, thanks so much for sharing this idea. I love open source thinking! If you get anywhere with this concept you might consider publishing your design at cncmentor.com (shameless plug). I don't have much personal experience with this, although I did tear apart a pick and place machine once that I picked up at auction and had quite a bit of fun trying to figure out how it worked.

[stewi]

How do you plan on making the sprocket wheel cheaply? Is there a source for them where you can just buy it?

I propose a wheel with 40 sprockets and 80 grooves in order to read 2 mm pitch.
Anybody here in this forum, who can build this to a decent price?
Wheel diameter is 50.93 mm, material 0.7 mm thick.

[jcdillin]

I'm in the process of getting my laser cutter set up in the next day or so, I can cut some out for you.

I would love to get set up with some feeders like this, i'm currently trying to implement drag feeders.

[stewi]

That would be great!
I'll make modifications to my hobby CNC and should be able to mill some of the other parts as well.
Last change to the sprocket wheel: I made the inner hole larger.
DXF added

[jcdillin]

So is the .7 mm thickness critical? I'm not finding that size easy to get, I can get .635 mm or .762 mm

[stewi]

I should finally convert to imperial!
I'll anticipate .635 and change the drawing.

I also found the pinion (pinch wheel) out of stock and will likely change to inch size as well https://sdp-si.com/eStore/PartDetail...51&GroupID=439