Cable and tube management on Pandaplacer

[Bob Amstadt]

I'm assembling a Pandaplacer that I bought over the summer, but I'm not liking the cable and tube management in the device.  I'd like to do some modifications in this area.  Anybody already do some work in this area?

Thx,

Bob

[Jarosław Karwik]

Hi,
I just assembled mine.
I am afraid there is not much you can do - their optimisation (smart, but questionable) makes it almost impossible to use different cabling . There is also no schematics so it would be hard to do modifications.
But what precisely you do not like ? I was able to arrange it quite neatly even if I preferred more classic approach.

I wanted to develop my own controller using Pandaplacer mechanics, but changes needed  for that would require e.g. conversion to umbilical cord head connection (like some 3D printers). And complete change of head boards.  So I plan to use it a bit and sell - I have already bought Bing machine ( which uses classical approach to tubing and cabling).

--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/openpnp/995f411f-ab20-4d2d-b5eb-67b309316fa3n%40googlegroups.com.

[Bob Amstadt]

I can live with the electrical cabling.  Seems a bit iffy to me but only if the head moves to the far extents.  The pneumatic tube wouldn't behave itself in the way that they specified.  It wanted to twist and wouldn't curve neatly on the side tray.  I have instead gone to trying to just take the tube straight from the head to a corner leg with a big arc in the air.  It seems to rigid enough to maintain the arc.  We'll see if that lasts.

To view this discussion visit https://groups.google.com/d/msgid/openpnp/CAC%2BEaogYV_Ze4zU%3DpV65Z2uj2tD-foK3CTapzH7qOy%2B%2BmU3wBA%40mail.gmail.com.

[bing luo]

3D printers are becoming increasingly lightweight, and I have a Bambu lab P1S with carbon fiber on both the X and Y axes. So it cannot use tank chains either, it also uses umbilical cords to provide pipes and cables to the head. I don't understand why PNP doesn't take this lightweight route. The finished PNP products you can buy in the market weigh over 50 kilograms  at least.   So I dare not change the traditional structure because I don't understand this.  

One more thing, because the shipping cost for international express delivery is very high,And it will take at least 2 weeks. I hope the accessories are easy to purchase and can be repaired or replaced by self.  This is different from in China, where sending a 1kg accessory from any city to any  another city only takes 1 day and costs 3 dollars.

[Jarosław Karwik]

Works better for me:

As I recall you need this - otherwise I had same issue:

Alternatively you can use something like this to guide the pipes -  almost umbilical cord :-)

@Bing - these days only cheapest packages from China need 1-2 weeks.

If you pay you have the parcel in 2-4 days, which is about same time I need for US orders.

About the umbilical cords - the solutions I have seen usually use supply + communication ( like FDCAN) and tubes.

I have considered such solution for my controller, but the price goes quite much up as you need to build dual controller - and the wires are just cheaper for simple machines.

[Ben]

I am not using pandaplacer, but bing's microsmt machine, and I do like neat wiring, so I printed a square PLA accessory and adding tie-wrap + nylon  accessory and it looks nice.

of course that it is from the bottom side and invisible from top view.

[mark maker]

> I don't understand why PNP doesn't take this lightweight route.

Me neither. It should be a central goal.

I also long suspect that a very lightweight single nozzle machine might even be faster than a multi nozzle machine, simply because its head (and drag chain solution) could be so much simpler and therefore lighter (assuming two designs of the same cost).

_Mark

To view this discussion visit https://groups.google.com/d/msgid/openpnp/7802dd17-dd28-4f9b-a033-6ea0306b9734n%40googlegroups.com.

[Jarosław Karwik]

Well,

I will give it a try - before going for chains I will see how it would work with Bing machine configured for cord.

I have already ordered electronics, so will be ready in ~2 weeks

[Ira Faberman]

I have no issues with most of the cabling and tubing on the Panda except for the long flat cable run on the X axis. It was hanging down and reliant only on the ribbon connectors to continue to hold the ends as gravity torqued the cable. This made me very uncomfortable so I made this support and it works well. The geometry is a bit tricky and it could use another tweak to get it perfected, but it is good enough as it is.

 

[bing luo]

> I don't understand why PNP doesn't take this lightweight route.

Me neither. It should be a central goal.

I also long suspect that a very lightweight single nozzle machine might even be faster than a multi nozzle machine, simply because its head (and drag chain solution) could be so much simpler and therefore lighter (assuming two designs of the same cost).

I found that my bambulab printer has super fast printing speed and it vibrates violently. 3D printers don't need to worry about the overall vibration, because PLA and ABS are semi-solid when they come out of the heating head, and they won't shift or jump out due to vibration. PNP cannot accept such overall vibration, as both 0402 and 0603    and  IC will  jump  out ,  so PNP cannot become lighter and move faster. PNP can only be heavier and then faster. 

[Bob Amstadt]

I can understand that.  In that case, can we get a solder paste nozzle
on the PNP?  Yes, I know that the tried and true method of using a
stencil does work well, but for a prototyper like myself, getting the
stencil is a waste if I won't use it many times.  Even if I do, it add
an extra complexity to my process.  My goal here is to make my life
easier and automate what I can.  Honestly, I don't need the second
placement nozzle and would gladly give it up for solder paste printing. 
Highest performance is not critical for me if it is completely automated.

Bob

[bing luo]

This is a meaningless attempt. If it could have been done, stencil would have been eliminated long ago.  PNP is not a new invention, it has a history of over 50 years.

[Bob Amstadt]

Ok, I'm not saying that you are wrong.  However, it is not a reasonable argument that just because something hasn't been done, that it can't be done.  For commercial PCB assembly, templates are very efficient.  They are fast to use if you intend to manufacture multiple boards.  I can easily imagine that nobody has gone down this path because it isn't commercially viable for standard PCB assembly houses.  As I said, you may be correct that it is not possible to do, but it may also just be possible that the incentive hasn't been there during those 50 years.

--

You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/openpnp/5ca61fda-340e-4bd7-b643-303d12dc2e57n%40googlegroups.com.

[bing luo]

like 0402 or 0603 can be achieved,  they are  just  not as perfect as stencil.   But I think the pins of the IC are difficult to implement.

[Bob Amstadt]

Admittedly, I may not by typical, but I do everything in 0805.  ICs are interesting and I wonder if less than perfect would be ok and then I just clean up by hand afterwards.  I do lots of audio circuits that are predominately 0805 parts with a very small number of ICs that I could even do by hand in a pinch.

To view this discussion visit https://groups.google.com/d/msgid/openpnp/1428ffaf-9b14-4b5e-8034-77724cebe87dn%40googlegroups.com.

[vespaman]

The professional paste shooters does 01005 and fine pitch stuff, I have used a subcontractor that had one of those (I think it was a MY700 Jet printer, or perhaps MY600). They seemed to prefer it over stencils for prototypes and very small batches.

Who would not like this feature in OpenPnP for prototypes..  But I have no idea how the dispenser would look like (and it would have do at least 0402). If  I understand correctly, there was something like this in OpenPnP earlier on.

 - Micael

[cborn]

I used the Voltera V-one to apply paste for a while, biggest problem for small runs was getting consistency from the paste.

It would tend to dry out in the nozzle, so you really needed to use a new nozzle each time, and they were quite expensive.

The paste is pushed from the top using a screw driven plunger which you calibrate with some test runs, that can be quite accurate for a start, but then the particles will bind and the flux comes out preferentially, and you need more force/time to push the solder particles through.

You can add software fixes for this, but it will vary with temperature, time, type of paste, etc.

In the end I went back to using stencils, they are quite inexpensive these days.

It probably comes down to finding the right paste for the job, one with a very fine particle size and long-term stability would probably work best.

A system to meter the paste amount at the nozzle end rather than at the top of the plunger would also help a lot. I looked into that, solutions exist but they are expensive.  

[vespaman]

LOL, What a coincidence lumen released this video not too long ago, which may give some paste dispensing. If it can be dialed in to at least 0402, I'll probably try it. It looks better than the pneumatic manual dispensers that was popular some 10 years ago. I really dislike cleaning stencils, so for a few prototypes, this might be it.

 -  Micael

[Alexander]

Pneumatic dispenser nozzle 0.2 mm - for components with a pitch of up to 0.4 mm / 0201 (0.6 x 0.3 mm) with paste type 6. The dispenser takes into account the map of the curved printed circuit board using PROBE.

суббота, 16 ноября 2024 г. в 23:44:52 UTC+2, micael....@gmail.com:

[bing luo]

his job does not require PNP, you can replace the extrusion head of the 3D printer with the clay extrusion head below. The pumping and extrusion of solder paste are both completed by driving the screw with a stepper motor.  The cost doesn't seem very high.

 

[vespaman]

To me, having the syringe on the head is much better - not as many part gets dirty to clean, less waste. 

But using a 3d printer would be ok, only add camera for fiducals, and some kind of board holder, and convert it to be able to talk to OpenPnP.

 - Micael

[bing luo]

If a syringe is used, a large amount of diluent is needed to dilute the solder paste to ensure its fluidity. Otherwise, it is difficult for the solder paste to flow out, and the solder paste contains excessive diluent, which can cause other problems. If using screws. Due to the high pressure, ordinary solder paste can be extruded directly without the need for thinner.

[bing luo]

Simplified version

[bing luo]

How to convert this drawing into a step file？

[vespaman]

>If a syringe is used, a large amount of diluent is needed to dilute the solder paste to ensure its fluidity. Otherwise, it is difficult for the solder paste to flow out, and the solder paste contains excessive diluent, which can cause other problems. If using screws. Due to the high pressure, ordinary solder paste can be extruded directly without the need for thinner.

Did you watch the video I linked to? He's using regular GC10. Not saying that solution is perfect, but looks promising enough to me. The air in the syringe is a problem that I recognize since the days I was using a manual (pneumatic) dispenser (with syringe), even though I bought ready made chip quick syringes.

 - Micael

[Alexander]

I think it would be right to create an editable library of paste cases, where you can specify where to apply the paste, and not in the center of the component output or several points on one output.

Example.

понедельник, 18 ноября 2024 г. в 12:32:43 UTC+2, micael....@gmail.com: