Feeders, new hardware (split from Head wiping tapes)

[Gopher]

I'd be interested in hearing everyone's experience of good/bad points of the Versatronics feeders (all versions), perhaps what they would change/enhance? We have some of the early feeders (like the first posting's picture) we have used as mules for our prototype VRplacer.

Obviously there is one person on here who has heard every gripe and whine big or small regarding the feeders and their quirks, but given he now sells a new machine that of course addresses most or all of them and you are possibly developing a competing machine.....

The RV has been an awesome machine and done what we need of it for years, aside from a few quirks its pretty much what you expect of it. However if there is a part of it that can make me start swearing profusely when it starts being annoying, it's the feeders as solving issues when they start being awkward is time consuming and fiddly.

So the ones in Mikes picture:

Come in banks - a pain

Threading in tape - a time consuming pain (exacerbated by the fact they are in banks)

The tape doubles back on itself under the feeder meaning it can snag on other tapes (especially if plastic) in the same bank (or hit the deck on stop moving if it's short) and stop feeding through. If it's a plastic tape it might continue feeding but create a massive loop popping up out of the feeder in the pick position

The indexing pin pushes the tape rather than pulling it, creates an inconsistent pick height when coupled with the previously mentioned doubling back because at the pick point, its not under tension.

Its an indexing pin, not a gear, so it only engages in the index holes to move forwards, a heavy or slightly reluctant reel can then pull the tape back a tiny bit after the index.

Its an indexing pin, not a gear, so its only got hold of the tape by one hole, this makes plastic tapes vulnerable to ripping if the the tension if not perfect, they are spliced or perhaps not pristine thru' poor storage or shipping.

If the plastic tape rips, bends or is otherwise not perfect it will jam, then you can say goodbye to the parts between that rip and the pick position  because it will keep jamming until you cut that bit off.

Cover tape take off reels can't hold the cover tape for a full 5k component reel.

Cover tape take off reels always rotate a fixed angular rotation meaning the tension is not constant, nothing you can do about this the big boys have the clever alternatives patented.

The indexing is short and sharp with hard acceleration and deceleration, parts like sot23 in plastic tape are wont to leap around or out of the their pockets, it also doesn't help with the whole plastic tape rip problem

They are dumb feeders, intelligent feeders with suitable supporting software can do wonders for keeping track of what is fitted where, what got fitted to what, when it was changed how many were wasted and all sorts of things, this level of data logging is very important for high mix if you need good traceability especially for ISO or other approvals. Intelligent feeders also means you can move them with impunity, no addressing nonsense.

The reel holder bits only take 7" reels.

The reel holder bit has slots that are too narrow for the newer 7" reels made from thinner plastic that is then reinforced with a crimped edge, these jam solid.

When a reel gets near to empty, rather than turning it gets pulled off the reel holder and into the tensioning assembly and jams.

See Phono's posts on the things he fixed with a custom replacement control card.

If you look at pictures of other machines feeders they are mostly quite similar, often the brochures include pictures of how they work to brag about the issues they have fixed (from the 90's and before) that used to plague operators. (Who are now senior people buying new ones I guess).  How many of the things that bug me you could change on a machine like an RV is another question entirely, mounting system, arm reach, cost, influences from similar machines (A TWS Quadra for example), must all have conspired to make the feeders what they are. And when they are setup just right they are perfectly ok, it's just they can suddenly decide they are not any more.

[Mike]

Comments below on how you might make a new design better, not fix the existing ones...

I'd be interested in hearing everyone's experience of good/bad points of the Versatronics feeders (all versions), perhaps what they would change/enhance? We have some of the early feeders (like the first posting's picture) we have used as mules for our prototype VRplacer.

Obviously there is one person on here who has heard every gripe and whine big or small regarding the feeders and their quirks, but given he now sells a new machine that of course addresses most or all of them and you are possibly developing a competing machine.....

The RV has been an awesome machine and done what we need of it for years, aside from a few quirks its pretty much what you expect of it. However if there is a part of it that can make me start swearing profusely when it starts being annoying, it's the feeders as solving issues when they start being awkward is time consuming and fiddly.

So the ones in Mikes picture:

Come in banks - a pain

But probably inevitable on a low-cost machine. A banked feeder with removable lanes would be nice, but would still be more expensive

Threading in tape - a time consuming pain (exacerbated by the fact they are in banks)

agreed, but hard to see how you could improve this with a banked feeder. Maybe a per-lane stripper plate would help.

The tape doubles back on itself under the feeder meaning it can snag on other tapes (especially if plastic) in the same bank (or hit the deck on stop moving if it's short) and stop feeding through. If it's a plastic tape it might continue feeding but create a massive loop popping up out of the feeder in the pick position

This could pretty easily be fixed with some holes in the baseboard so tapes go down instead of backwards

The indexing pin pushes the tape rather than pulling it, creates an inconsistent pick height when coupled with the previously mentioned doubling back because at the pick point, its not under tension.

and  push-feeding is less tolerant of creased plastic tapes than pulling would be

The indexing is short and sharp with hard acceleration and deceleration, parts like sot23 in plastic tape are wont to leap around or out of the their pockets, it also doesn't help with the whole plastic tape rip problem

Easily fixable - just software

They are dumb feeders, intelligent feeders with suitable supporting software can do wonders for keeping track of what is fitted where, what got fitted to what, when it was changed how many were wasted and all sorts of things, this level of data logging is very important for high mix if you need good traceability especially for ISO or other approvals. Intelligent feeders also means you can move them with impunity, no addressing nonsense.

This would be trivially easy and cheap to implement nowadays, using an RFID chip or NFC type link or opto coupled data link.

The reel holder bits only take 7" reels.
The reel holder bit has slots that are too narrow for the newer 7" reels made from thinner plastic that is then reinforced with a crimped edge, these jam solid.
When a reel gets near to empty, rather than turning it gets pulled off the reel holder and into the tensioning assembly and jams.

All the above should be fixable by an adjustable-length arm arrangement that holds the reel by the centre hole.

[Gopher]

Some but not all of the current generation are now using single feeders not banks on their low end machines, but obviously the goal posts have moved in the last decade, Surface mount is less a proportion of a typical portfolio and instead is its mainstay. This means we want more out of our machines even at entry level so the budget is perhaps a little larger. That said RV feeder banks were not cheap, £2000 (iirc) for 10 8mm lanes gives you a reasonable budget to strap an indexing mechanism to something (as demonstrated by the cost of vivo feeders) , and flexibility can save you a lot of money very quickly. Our manual pick and place that must be 20 years old has single feeders too

The importance and low cost of making feeders smart is demonstrated by the fact that it is now pretty  much standard on any current machine, I can only think of one that makes you pay extra for this option.Its not even a cheap one..... There are of course different interpretations of smart, some implementations look like they could be as irritating as they are useful.

I have been tempted to try cutting a big hole in the base board

One of the most popular ways of holding reels seems to be just dropping them in slots in a curved trough, leaving the "feeder" to index and nothing else. If you held them by the centre on an arm system like an Rv you need some kind of brake or it will just empty itself over the floor by gravity. Well my attempt did lol. The worst option seems to be feeders that are designed for one size reel only.

[fcb]

Obviously there is one person on here who has heard every gripe and whine big or small regarding the feeders and their quirks, but given he now sells a new machine that of course addresses most or all of them and you are possibly developing a competing machine.....

I waited along time to see the new machine (still haven't), and during that time I had to keep my old RV going, hence the all the improvements. I finally made the decision to produce this machine when I gave up waiting. I like the look of the new machine, but feel that it doesn't address the 'bottom' of the market and that there is still a decent (if not latent) market for something like the RV.

The RV has been an awesome machine and done what we need of it for years, aside from a few quirks its pretty much what you expect of it. However if there is a part of it that can make me start swearing profusely when it starts being annoying, it's the feeders as solving issues when they start being awkward is time consuming and fiddly.

Agree, I think the problem mainly is that development stopped many years ago.

Come in banks - a pain

I agree that in the ideal world the feeders would be individuals, but this does distort the cost model of a lower priced machine. I have quotes for 2nd user machines where over 2/3rds of the quote value is the feeders.

Threading in tape - a time consuming pain (exacerbated by the fact they are in banks)
The tape doubles back on itself under the feeder meaning it can snag on other tapes (especially if plastic) in the same bank (or hit the deck on stop moving if it's short) and stop feeding through. If it's a plastic tape it might continue feeding but create a massive loop popping up out of the feeder in the pick position.

This could be solved in the basic mechanical design, I haven't had the snagging problem yet - but I tend to use alot of paper tapes.

The indexing pin pushes the tape rather than pulling it, creates an inconsistent pick height when coupled with the previously mentioned doubling back because at the pick point, its not under tension.
Its an indexing pin, not a gear, so it only engages in the index holes to move forwards, a heavy or slightly reluctant reel can then pull the tape back a tiny bit after the index.

Pulling the tape and doing this just prior to the pick (leaving the pin in the tape, during the pick) would solve this?

Its an indexing pin, not a gear, so its only got hold of the tape by one hole, this makes plastic tapes vulnerable to ripping if the the tension if not perfect, they are spliced or perhaps not pristine thru' poor storage or shipping.

Changing to a gear certainly solves the problem, but it does mean a total redesign, and the production cost of a sprocket system may be too high. Multiple pins might be feasible to spread the load.

If the plastic tape rips, bends or is otherwise not perfect it will jam, then you can say goodbye to the parts between that rip and the pick position  because it will keep jamming until you cut that bit off.

Perhaps a problem inherent in other feeder designs?

Cover tape take off reels can't hold the cover tape for a full 5k component reel.

Hopefully trivial to solve.

Cover tape take off reels always rotate a fixed angular rotation meaning the tension is not constant, nothing you can do about this the big boys have the clever alternatives patented.

Hmm - do you mean the Agilis system?

The indexing is short and sharp with hard acceleration and deceleration, parts like sot23 in plastic tape are wont to leap around or out of the their pockets, it also doesn't help with the whole plastic tape rip problem

They are dumb feeders, intelligent feeders with suitable supporting software can do wonders for keeping track of what is fitted where, what got fitted to what, when it was changed how many were wasted and all sorts of things, this level of data logging is very important for high mix if you need good traceability especially for ISO or other approvals. Intelligent feeders also means you can move them with impunity, no addressing nonsense.

I would defiantly make future feeders 'intelligent', I also plan on removing the AC mains from them. If the feeder banks could detect which position they where in (simple 'coded' sockets on the base unit, that also supplied DC and serial data) this would allow for modern 'intelligent' functionality.

The reel holder bits only take 7" reels.

The reel holder bit has slots that are too narrow for the newer 7" reels made from thinner plastic that is then reinforced with a crimped edge, these jam solid.

When a reel gets near to empty, rather than turning it gets pulled off the reel holder and into the tensioning assembly and jams.

Good feedback. Will lookout for crimped edge jamming.

See Phono's posts on the things he fixed with a custom replacement control card.

I'm aware of these cards - I think he is a hero for solving what should have been done sometime ago.

If you look at pictures of other machines feeders they are mostly quite similar, often the brochures include pictures of how they work to brag about the issues they have fixed (from the 90's and before) that used to plague operators. (Who are now senior people buying new ones I guess).  How many of the things that bug me you could change on a machine like an RV is another question entirely, mounting system, arm reach, cost, influences from similar machines (A TWS Quadra for example), must all have conspired to make the feeders what they are. And when they are setup just right they are perfectly ok, it's just they can suddenly decide they are not any more.

I think the bottom end of the market lacks innovation (don't get me going on those fishing lures...), my aim really is not to build a fast machine, just something that works well and is straight-forward to setup and maintain. Basically something you look forward to using and not get stressed about as you hit the on-switch. I think nowadays the mechanical complexity can be lowered dramatically and the quality of the software improved also.

[fcb]

Some but not all of the current generation are now using single feeders not banks on their low end machines, but obviously the goal posts have moved in the last decade, Surface mount is less a proportion of a typical portfolio and instead is its mainstay. This means we want more out of our machines even at entry level so the budget is perhaps a little larger. That said RV feeder banks were not cheap, £2000 (iirc) for 10 8mm lanes gives you a reasonable budget to strap an indexing mechanism to something (as demonstrated by the cost of vivo feeders) , and flexibility can save you a lot of money very quickly. Our manual pick and place that must be 20 years old has single feeders too

I have a target goal of between £50-100 per lane, this isn't really arbitary but relates to our overall model. We can certainly hit the upper end of this span.

The importance and low cost of making feeders smart is demonstrated by the fact that it is now pretty  much standard on any current machine, I can only think of one that makes you pay extra for this option.Its not even a cheap one..... There are of course different interpretations of smart, some implementations look like they could be as irritating as they are useful.

Agree, smart and intelligent are terms that seem to cover the inclusion of $0.20 worth of NVRAM.

I have been tempted to try cutting a big hole in the base board

I like this idea, it might make removal of a bank more difficult though?

One of the most popular ways of holding reels seems to be just dropping them in slots in a curved trough, leaving the "feeder" to index and nothing else. If you held them by the centre on an arm system like an Rv you need some kind of brake or it will just empty itself over the floor by gravity. Well my attempt did lol. The worst option seems to be feeders that are designed for one size reel only.

I like the simple designs the best.

[Mike]

The RV has been an awesome machine and done what we need of it for years, aside from a few quirks its pretty much what you expect of it. However if there is a part of it that can make me start swearing profusely when it starts being annoying, it's the feeders as solving issues when they start being awkward is time consuming and fiddly.

Agree, I think the problem mainly is that development stopped many years ago.

Yes - mechanical things can be fixed, but the biggest frustration for me is that there are so many things that could be improved if only the software was tweakable - many annoyances would be trivial to fix. This is why Open Source works so well - anyone can fix the things that annoy them.

I also plan on removing the AC mains from them.

Yes - if nothing else it would make the connectors & cabling cheaper. The silly thing is that as only one feeder will be drawing power at a time, you don't even need a much bigger PSU to run lots of feeders. OK when these feeders were designed, feeder electronics draw a fair amount but that's a complete non-issue these days

If the feeder banks could detect which position they where in (simple 'coded' sockets on the base unit, that also supplied DC and serial data) this would allow for modern 'intelligent' functionality.

I always feel uncomfortable at the thought of connectors mounted to something that vibrates a lot - unless they are in a floating mount, flying leads are probably a cheaper option than suitably rugged fixed connectors.

I think the bottom end of the market lacks innovation (don't get me going on those fishing lures...), my aim really is not to build a fast machine, just something that works well and is straight-forward to setup and maintain. Basically something you look forward to using and not get stressed about as you hit the on-switch. I think nowadays the mechanical complexity can be lowered dramatically and the quality of the software improved also.

I think that there is an increasing market for low-end (<<10K inc feeders) machines that is crying out to be filled. Look at what's been happening in the Open Source hardware world in the last few years - Products like Makerbot are revolutionising previously unaffordable technologies. There is an explosion of small outfits making small numbers of electronic products, and the few companies bringing assembly in-house like Adafruit are struggling with Chinese-built machines with flaky software and minimal support. If you ask me, a machine with a low entry cost is probably more marketable now than it has ever been, as well as more feasible due to the cost of computing reaching near-negligible levels. 
More so if you make all the software & hardware open source - this will not only harness a lot of potential work from some very clever developers out there, but create a lot of market goodwill and user-to-user support. Open Source is becoming a major selling point, and will only become more so.
I have no doubt that you could implement a good imaging system around the OpenCV vision library and some cheap webcams with no more than a little lens hacking.

[Gopher]

Hmmm sounds like an interesting project. I am also in the market for something to takeover from the RV but perhaps unlike you I do need it to be faster, easier (ideally a temp should be able to run the machine if not set it up) and more accurate. This means I have tossed out things from my potential list that might well go head to head with yours.
I am told that machines like APS novastor can come to about £35k which about 1/2 the next echelon up. That in real terms is perhaps cheaper than the rv4 was and it certainly looks like they have been actively developing the platform. Then there are weird MDC/Manncorp or Mechatronika tabletop ones and the assisted manual pick and place like the fritsch 902.
What is the bottom of the market? If our company for instance bought something like an RV again we would be standing still, and actively turning down work. People need and expect us to have abilities these machines don't provide. A low end machine that is standalone and of static configuration could end up costing you more than buying the more expensive modular option that lets you grow or even just handle that one off big job that happened along.
Feeders are certainly expensive both new and old, but if I account for the cost of changeover using an RV, that is expensive too.
What would be interesting is what killed Versatronics? Did it for instance saturate the low end market (which is perhaps smaller than one might think) and then not have the revenues to develop the something new that was needed. Or did they just lose the brains behind the next project? While it may be possible to build something good on quite a low budget (unlike HRT it seems) can you sell it & support it and still hold the entry level price you want.
The other issue is sub contract, as a small sub contractor ourselves there's no question our sector is after your potential market. Being small makes us flexible, with a machine with low changeover times and spare capacity after regular runs that are our main staple; companies like the one I work for are quite happy to do short notice, quick turnaround, low volume jobs, kitted or not. Companies like us do try to sell each other their services now and again but overall what we are not good at is getting out there and selling those services or people would not be buying decrepit Rv1's and a toaster oven and trying to do it themselves.

[Mike]

I think the newest potential market for low-end may be not so much  subcontractors but people wanting to bring things in-house for various reasons.
A large proportion of subcontract assembly has moved overseas, so there are fewer local choices, and trying to find the right fit for your requirements can be hard, especially if your volumes are low, more so if you need fast turnarounds.
A small suncontractor is vulnerable to disappearing if they lose a customer, and a big place will either not be interested or will charge high get-out-of-bed costs. Even when you find a good place, the lower-volume customer will get slotted in between the larger jobs, and the option to pay more for fast service is rarely there due to the overall job value.

Getting it right will require a lot of clever design work, so in terms of development cost and support, the only way I can see a good, useable low-cost machine ever happening is as an open source collaboration. 

[Mike]

Companies like us do try to sell each other their services now and again but overall what we are not good at is getting out there and selling those services or people would not be buying decrepit Rv1's and a toaster oven and trying to do it themselves.

BTW members here are more than welcome to include links to their business website in their forum sig.

[Gopher]

<£10k ? that's not even an operator! at this price point it's is surely not something you can sell for profit, it's collaboration for mutual benefit or nothing. I have wondered why it has not yet being attempted either for AOI or p&p, certainly there are closed source lab/homebrew projects (they look rubbish).
W.r.t your comments on subcontract providers, I have read some of the appeals made on other forums for help going the DIY route because they can't find a subcontractor who can do what they want. But in the appeal is the reason why, they are asking for the earth! In house assembly cannot provide what some of them think they can do, 24 busy pcb's with no notice, no kitting and no understanding of the process into the mix, oh and over the weekend please! If ones budget is as tight as some of these people claim, your pcb's are coming from outside Europe/USA, that means you can give your sub-con 3 weeks notice.With 3 weeks notice it's plain sailing even there is a big job to slot it round. And in our case, while we do have a core OEM product line, 28 years and counting, I don't think we're that risky.
Once you get a machine you start designing more things to use it, before you know it your original estimates -bunkum.

Being open source brings advantages and disadvantages

Pro's

It can keep going if its main backer fails or drops out
Anyone can contribute
Tweaks can be made for one off requirements
Interesting ideas can come in and you have a nice array of beta testers
Buying one would be less fraught with doubt than a new entrant who might go titsup 2 years later
Documentation through wiki, so all those questions get answered only once.

Con's

If its main backer fails or drops out they might well take the expertise needed to keep going with them.
The target market might not have the skills or time to help with development.
Most or indeed all open source projects have a few core contributors and a whole bunch of leeches.
The physical machine can be knocked off cheap by the Chinese who then undercut you by ignoring the original ethics and aims of the project. I would assume such a project would require at least some income form the sales of hardware and support if only to fund development


no doubt many edits for more...........

[Gopher]

Companies like us do try to sell each other their services now and again but overall what we are not good at is getting out there and selling those services or people would not be buying decrepit Rv1's and a toaster oven and trying to do it themselves.

BTW members here are more than welcome to include links to their business website in their forum sig.

To do that I'd have to delete half my posts, correct the typo's in the rest and go on a PR course . Not to mention update our website to properly include our sub-con work not just our OEM line that would interest none of you.

[SteveW]

I think the newest potential market for low-end may be not so much  subcontractors but people wanting to bring things in-house for various reasons.

That's me, in a nutshell. A little design consultancy that needs a faster turn than can be got from subbing the assembly out. We very rarely make more than 10 of anything, but I do need those 10 in a hurry. I also may need to build one, debug it, traipse it round to the customer, then build the rest. Manual pick & place onto solder paste has been the method in the past, but it's so tedious that babysitting an RV1 is a huge improvement. Not having to kit for external subbies (or wait the week while they kit) is compelling. Even knocking a day or two each way of transit time off saves me a working week...
While my RV1 is a bit frustrating, its pretty much the size, speed and capability I'm after. I don't need features that a full time board stuffer wants. Heck, the RV1 could be substantially slower and still be OK. I've got plenty to be getting on with while it chunters away. I just need it not to behead itself while I'm on the other side of the room.

The muttering about an open source, slow, flexible, machine appeals to me greatly. Knocking up some USB-powered feeders to sit round a makerbot, f'rinstance?

[SteveW]

If ones budget is as tight as some of these people claim, your pcb's are coming from outside Europe/USA, that means you can give your sub-con 3 weeks notice.

Err, no. My PCBs come from Cambridge Circuits, who are within cycling distance. 2-day turn PCBs, a day to assemble & test. Cycle round to customer to deliver. This is life in a prototyping shop...
This three weeks of which you speak - that's just a different world.

[fcb]

I have set a basic specification and spent alot of time thinking about the potential market.

Sub-contractors/faster machines
I'm not targeting sub-contractors specifically, mainly the 'little-guy' who wants to bring his low volume assembly in-house. The price point of the machine is going to be very important - we have a target of <£8-10K with a 'sensible' number of feeders and feeder mix.  I don't know what the market size is - I'm not sure anyone does really, as it's quite likely this machine will set a new price point/quality.

Open source?
Probably not. I'm quite sure that opening the code up to the wider world would have certain benefits - but it isn't a model I'm comfortable with. Mainly the fact that there is a risk of damage to warranteed hardware if the user can tinker or load a non-approved application. In fact the idea behind the code is to license it so that you would pay a small amount annually for the latest major revision.  This appears to be a model with a proven track record (I have used Easy-PC for many many years like this, and pay them about £60 per year for the latest version).  The incentive to keep supporting older machines in the market with new features then exists.

Sales/Support
The tough one.  How to support a low-cost machine? Well, for a start make it very reliable and very simple to setup.  Secondly make it intrinsically calibrated. Thirdly, make it fully user serviceable.
For example: how many people actually replace the RV vacuum pump themselves vs. getting perhaps a third party to do it? Same thing with belts? Perhaps on the present the only reason to involve a third party is calibration if/when it goes out - and why? Because it is designed to require a specialist tool/device which is (a) not supplied with the purchase (b) probably couldn't buy or afford to buy if you could.
So the result of the design (and I'm not attacking it - I understand the historic reasons for some of the machine) is that it creates servicing calls that just aren't necessary if designed differently.
I'm not suggesting at all that it won't be fully supported OR that the user won't be able to call up a service technician - just that it the requirement to do it will be minimised, and after all - if it takes 5 minutes to change a motor amplifier/camera/pump/etc... and it can be Fedex'd by 10AM next day, why would you want to schedule an engineer for later that week.

[Gopher]

Hehe, the difference in price between 3 weeks and 2 days tho' is terrifying. Most of our own stuff is done in batches of similar size to yours, sub-contract pays for the equipment to do it in house and allows to contemplate more ambitious things. If we ran would effectively be just in time or lean manufacturing on our own stuff and so bought pcb's on short notice in small numbers the cost would go through the roof, not feasible in our market.
3 days is an impressive turnaround , we do have customers who say, oh there's a board arriving 2morrow please build it so I can pick it up ASAP and while that can happen in 2/3 days what usually happens is they have designed something in that cannot be sourced in that timeframe  - usually because they picked their parts from digi-key or some ghastly internal approved parts list that needs updating. Without that stumbling block 5 working days or so low volume is a market I personally want my employers to look at even closer because it works well with everything else.
I would be very interested to know what kind of timeframes etc you ran into before going the DIY route