Author Topic: Big Day at the Meshtron "Fake It Till We Make It" Home PCB Assembly Shop!  (Read 2715 times)

0 Members and 1 Guest are viewing this topic.

Offline meshtronTopic starter

  • Regular Contributor
  • *
  • Posts: 132
  • Country: us
I purchased, on Alibaba, a QIHE TVM802BX Pick-and-Place on 5-SEP-2023 (as in - I paid for it then).  The listing said and the agent confirmed it's roughly 30 calendar days from order to delivery and the unit was IN STOCK...  This morning, it's supposed to actually be delivered!

Tracking has been essentially non-existent the entire time.  I've requested an order cancellation and refund from the seller, but been reassured "it's definitely coming" multiple times, but without any real tracking data.  Alibaba has been helpful, but basically said "let's wait and see if it arrives."  Out of the blue yesterday I got a call from a live human who scheduled a drop-off this morning.  So, either the farce is truly end-to-end or a piece of equipment is REALLY coming!!

In a strange twist of fate, the ZB3040H stencil printer I ordered on 4-OCT-2023 is ALSO arriving today, but via UPS (which takes all the fun out of things because you can just track the package all along).

As it turns out, I've just started on the 3rd revision of my board design (first didn't really work well, 2nd was quite close, 3rd should work great AND have no PTH parts) and just landed a (non-PCB) design contract that will keep me busy with money-making things for a few months.  So I'll be tinkering and fiddling trying to understand the PnP machine and software through the winter and hopefully assembling my first panelized boards in-house.  But, I'm excited to have all the equipment I intend to buy in-house.

Anyway, it's hard to believe it's actually going to happen - we shall see.  In theory, my equipment list should be:

- ZB3040H Stencil Printer
- TVM802BX Pick and Place
- T937M Reflow Oven (here, working great)
- 2x 4k Overhead Digital Microscopes
- Hakko FX888D Soldering Iron
- YIUHA 8786D Soldering Station

I'll update with pics later - fingers crossed!
 

Offline meshtronTopic starter

  • Regular Contributor
  • *
  • Posts: 132
  • Country: us
Re: Big Day at the Meshtron "Fake It Till We Make It" Home PCB Assembly Shop!
« Reply #1 on: November 15, 2023, 07:26:32 pm »
Aaaaaaannnnnd - it's actually here!  Crate was in rough shape, but at least on preliminary looking - machine appears fully intact and all the parts are there.

Fingers crossed it fires up and makes machine noises - probably won't get to try it out until the weekend.
 

Offline Styno

  • Regular Contributor
  • *
  • Posts: 154
  • Country: nl
  • TÜV-geprüft
Re: Big Day at the Meshtron "Fake It Till We Make It" Home PCB Assembly Shop!
« Reply #2 on: November 21, 2023, 09:29:31 am »
Congrats on your purchase! I hope it works well for you. Please share experiences with using it, support and how the machine holds up over the coming months, years.

What kind of parts are you using (pitch/size, # of bom-lines), are there light components in plastic tapes/tray, components with odd shapes (e.g. antenna or USB connectors)?

I'm especially interested in how fiddly the feeders are. In my experience these are the parts that take the most time and are the biggest source of frustration (e.g. tape peel-back on reels with varying peel-resistance, advancement not being consistent). I don't care if the machine is slow, as long as it doesn't need my constant attention. The other interest is in the workflow: how does the machine handle mis-picks, empty feeders and setup and change a board or panel.
 
The following users thanked this post: meshtron

Offline Psi

  • Super Contributor
  • ***
  • Posts: 9953
  • Country: nz
Re: Big Day at the Meshtron "Fake It Till We Make It" Home PCB Assembly Shop!
« Reply #3 on: November 21, 2023, 10:03:11 am »
My advice for the stencil printer is to buy 4 different brands of solder paste and try them all.
See which one works best for you on your machine.

People spend way to much time trying to optimize and debug their solder paste application system to get the best transfer possible, when most of the issues are just their solder paste brands compatibility with what they are doing with it.

Also consider the ball size of the paste. T1-T5 etc..
« Last Edit: November 21, 2023, 10:04:45 am by Psi »
Greek letter 'Psi' (not Pounds per Square Inch)
 
The following users thanked this post: meshtron

Offline meshtronTopic starter

  • Regular Contributor
  • *
  • Posts: 132
  • Country: us
Re: Big Day at the Meshtron "Fake It Till We Make It" Home PCB Assembly Shop!
« Reply #4 on: November 21, 2023, 08:50:15 pm »
My advice for the stencil printer is to buy 4 different brands of solder paste and try them all.
See which one works best for you on your machine.

People spend way to much time trying to optimize and debug their solder paste application system to get the best transfer possible, when most of the issues are just their solder paste brands compatibility with what they are doing with it.

Also consider the ball size of the paste. T1-T5 etc..

Yeah thanks for the tip!  Size of solderball is definitely something I should've researched a bit more before I bought my first little jar of HC10.  But, I've had no issues with paste thus far (admittedly one or two at a time).  I have made some changes to the next board (switching from HASL to ENIG for example) to try to get a bit tighter control over the solder process, we shall see how it goes!
 

Offline meshtronTopic starter

  • Regular Contributor
  • *
  • Posts: 132
  • Country: us
Re: Big Day at the Meshtron "Fake It Till We Make It" Home PCB Assembly Shop!
« Reply #5 on: November 21, 2023, 09:03:32 pm »
Congrats on your purchase! I hope it works well for you. Please share experiences with using it, support and how the machine holds up over the coming months, years.

What kind of parts are you using (pitch/size, # of bom-lines), are there light components in plastic tapes/tray, components with odd shapes (e.g. antenna or USB connectors)?

I'm especially interested in how fiddly the feeders are. In my experience these are the parts that take the most time and are the biggest source of frustration (e.g. tape peel-back on reels with varying peel-resistance, advancement not being consistent). I don't care if the machine is slow, as long as it doesn't need my constant attention. The other interest is in the workflow: how does the machine handle mis-picks, empty feeders and setup and change a board or panel.

Yeah I will definitely share my (mis)adventures with this machine. :)

EDIT: Added a picture of 3 generations of my first board.  The one on the right is just plastic cosplaying as a PCB, but it's the final size and mostly final component layout :)

Right now I only have a single product, but there are 4 more related products in the pipeline.  Board size for this product is 72mm X 36mm.  This is the first version that has 100% SMT components, so we'll see how that goes.  I have a couple big-ish (4-gang) DIP switches and 2 bulk caps that I will be curious to see if the PnP can tolerate at all.  For all my passives, I've standardized on 0603 sizing since that's small enough to fit in the space I have but big enough to be reasonably easy to hand-assemble.  If all the PnP could do was place all the passives, that would still be a win of sorts, but I'm really hoping to get it to run them 100%.

Current board has 34 SMT components on it so pretty simple by all accounts.  My next one will have roughly double that, and the part after that probably closer to 100 with an array of connector headers too.

Yes, I think these feeders will be finicky.  Even my first spin of the tape reel wheels showed some differences in tension on them, and they're brand new!  I also have concerns about the TVM802 not having a tool-change holder whatsoever, so hopefully I can get away with just using 2 different sized heads.  Wouldn't be the end of the world to manually change during a run, but that's definitely not the goal!

Sadly I haven't even fired it up yet, but I'm getting all the rest of my "to-do" stuff off my plate so I can spend some quality time over the Thanksgiving weekend with it.
« Last Edit: November 21, 2023, 09:15:01 pm by meshtron »
 

Offline SMTech

  • Frequent Contributor
  • **
  • Posts: 846
  • Country: gb
Re: Big Day at the Meshtron "Fake It Till We Make It" Home PCB Assembly Shop!
« Reply #6 on: November 22, 2023, 01:26:51 pm »
With what I assume your Oven to be (IR heaters)  two large electrolytics like that could be something it struggles with, likewise direct harsh IR onto a plastic cased dip switch might make them less reliable than they might otherwise be, even if they don't look melty. It wouldn't surprise me if the placer wasn't capable of placing the caps either, that's a common weakness although there may be workarounds in some cases.

We used to run a conveyor based 5 zone IR oven; the oven zone (i.e the emitters themselves) temps to achieve a board profile vaguely resembling spec were set around 330C in the liquidous zone. This worked absolutely fine for passives with proper pads and balanced tracks all the way from 0402 to Case E Tantalum and around Case E Electrolytic, (as long as another one wasn't nearby). QFN and QFP up to about 20mm were also easily handled but a chunky 30mm 240pin QFP was really on the edge. FPC connectors seemed to work fine but white did come out very beige, I think this was flux staining as IR ovens have low to no extraction.

What absolutely did not work were:
10mm tactile push buttons despite these being smaller, shorter and lower mass than some things that did.
12*12*5mm inductors
Elecs Case F or bigger
Boards with heavy copper and thermally unbalanced tracking (tombstones galore)
Ceramic/Rogers Boards

On the whole IR does a fairly OK job if your design looks like surface mount assemblies from 2002, a drawer based one might be slightly more variable than a conveyor based one.
 
The following users thanked this post: meshtron

Offline meshtronTopic starter

  • Regular Contributor
  • *
  • Posts: 132
  • Country: us
Re: Big Day at the Meshtron "Fake It Till We Make It" Home PCB Assembly Shop!
« Reply #7 on: November 22, 2023, 02:20:42 pm »
Yeah @SMTech I definitely have some concerns about the bigger SMT parts.  You are correct that my oven is IR, and being of the "cash register" form factor, it only has elements on the top side.

I have been thinking about building a profile that ramps very slowly from 30C to just below flux activation, then follow the recommendations from Henkel about time and temp from there.  But, I expect it's just going to take some experimentation to see what is and what isn't workable.  If I need a better oven, I will do that too.

I am definitely mentally prepared for the PnP to not handle any of the 5 big parts (2 caps 2 dips, 1 pot).  Worst case I hand place them, but I will put in a fair amount of time trying to make it work.  Theoretically "it should."
 

Offline meshtronTopic starter

  • Regular Contributor
  • *
  • Posts: 132
  • Country: us
Re: Big Day at the Meshtron "Fake It Till We Make It" Home PCB Assembly Shop!
« Reply #8 on: November 22, 2023, 03:36:23 pm »
It also occurs to me with respect to the caps, am I at more risk of just getting poor/cold solder joints, or am I at risk of overheating and destroying the cap, or both?

And will it be obvious that I've hurt the cap, or can they be "thermally degraded" but still seem functional?  How would I test for this condition?  Measuring capacitance was not part of my test plan, maybe it should be?
 

Offline MR

  • Regular Contributor
  • *
  • Posts: 215
  • Country: tw
Re: Big Day at the Meshtron "Fake It Till We Make It" Home PCB Assembly Shop!
« Reply #9 on: November 22, 2023, 04:20:58 pm »
just for fun I went to an electronic shop nearby today they offer IR and resistive oven .. very neat small 8-9L ovens which are absolutely suitable for reflowing.
In general forget IR modules, if you have no choice - you could possibly go with low wattage and shielded IR modules.

In the electronic store I clearly saw:
1. resistive heating elements
2. protected IR heating elements

All the mini ovens were priced at around 30-40$ (incl. 5% Vat).

My conclusion is:
* IR works okay when shielded and low wattage - direct exposure of IR to the PCB is an absolute no go whatever settings you apply it will never be okay for PCBs with various inductors or large elements. Inductors won't scream but they will increase the power consumption when overheated. I have nice plastic inductors when the colour changes you know things went bad (and it can be confirmed by measuring the power consumption).
* resistive heating elements - and k-type probes - it's a better way to go.

We have a manufacturer nearby which manufactures all kinds of resistive heating elements (usual stuff stainless + magnesium oxide powder), we replaced our IR modules with those ones and they're pretty much okay.
 
The following users thanked this post: meshtron

Offline meshtronTopic starter

  • Regular Contributor
  • *
  • Posts: 132
  • Country: us
Re: Big Day at the Meshtron "Fake It Till We Make It" Home PCB Assembly Shop!
« Reply #10 on: November 22, 2023, 05:21:09 pm »
just for fun I went to an electronic shop nearby today...

I remember when we had electronics shops in the US!  Harder to find anymore, at least near me!

Thanks for the thoughts.  I'm hoping I don't have to give up on my 937M right out of the gate, but maybe that ends up the case.  Confirmed my IR elements are not shielded whatsoever so just blasting down on top of the board at the moment.

Could a person - to get started anyway - add something like an aluminum foil "shield" over the top of some of the parts?  Not sure how IR works with that, if there would be damage reflecting too much back at the element or if it would be okay.  Also not sure if the convection would make it impossible to keep it more-or-less in place during reflow.

Hoping to order boards today, I'll get plenty to play with.  Part of me wants to hedge my bet and add a couple PTH pins for the caps just in case I need to bail on SMD, but another part of me doesn't want an easy out that makes me give up on solving it too fast. :)
 

Offline MR

  • Regular Contributor
  • *
  • Posts: 215
  • Country: tw
Re: Big Day at the Meshtron "Fake It Till We Make It" Home PCB Assembly Shop!
« Reply #11 on: November 22, 2023, 05:30:54 pm »
https://m.media-amazon.com/images/I/51KrDIbEsJL._AC_UF350,350_QL50_.jpg

with shielded I mean the slotted shield which you can see here.

Having multiple IR modules next to each other (unshielded) will result in a disaster, and it doesn't matter if they have put a k-type temperature sensor in between of them they won't pick up the real temperature from the PCB.

If I would be you I'd immediately get rid of the chinese IR reflow oven, it's not worth the electricity you burn with it.
Easy PCBs will work but it's far from reliable and repeatable.
I know some people say they have a working version, my answer to this would be they are probably not using parts which are sensitive to this - so they're lucky to not run into those problems. However sooner or later it will also hit them. IR is much harder to control than resistive heating elements
« Last Edit: November 22, 2023, 05:35:19 pm by MR »
 
The following users thanked this post: meshtron

Offline meshtronTopic starter

  • Regular Contributor
  • *
  • Posts: 132
  • Country: us
Re: Big Day at the Meshtron "Fake It Till We Make It" Home PCB Assembly Shop!
« Reply #12 on: November 22, 2023, 07:10:05 pm »
https://m.media-amazon.com/images/I/51KrDIbEsJL._AC_UF350,350_QL50_.jpg

with shielded I mean the slotted shield which you can see here.

Having multiple IR modules next to each other (unshielded) will result in a disaster, and it doesn't matter if they have put a k-type temperature sensor in between of them they won't pick up the real temperature from the PCB.

If I would be you I'd immediately get rid of the chinese IR reflow oven, it's not worth the electricity you burn with it.
Easy PCBs will work but it's far from reliable and repeatable.
I know some people say they have a working version, my answer to this would be they are probably not using parts which are sensitive to this - so they're lucky to not run into those problems. However sooner or later it will also hit them. IR is much harder to control than resistive heating elements

Fair enough.  I will be putting in a bit of effort to see if I can get the oven I already have to work with the parts I'm making.  But, I'm old enough to recognize - if I get there - when I'm putting a square peg in a round hole too.  If that happens, it will be off to the resistive element reflow oven store with a bit more experience and knowledge than I have today.
 

Offline SMTech

  • Frequent Contributor
  • **
  • Posts: 846
  • Country: gb
Re: Big Day at the Meshtron "Fake It Till We Make It" Home PCB Assembly Shop!
« Reply #13 on: November 23, 2023, 11:51:21 am »
With small ovens these are moot points, if there is a direct line of sight between element and PCB, it is heating your board with IR to a significant degree, that means there is shadowing if you deploy some kind of shield or something tall is on the board. So what you will find with difficult parts is the solder not meting at all, your only weapons are time and temperature but as you fiddle with those; smaller parts can end up getting too hot for too long and. You really want to stick quite religiously to the TAL (time at liquidous) time and max temp ratings of your parts, and that windows is pretty narrow. You really want circulating hot air and as little radiated heat as possible, that gets heat transfer all around difficult areas and prevents weird cold zones, this is why phase reflow is also popular and efficient.

Don't confuse an Oven profile with a board profile, the oven profile is the temp cycle it needs to run to transfer heat at an appropriate rate for the PCB/solder to match its desired profile . An oven will run at far higher temps than the PCB inside it, element or air temp is typically what an oven controller is measuring but for a drawer oven a probe mounted on the board might work better as you are trying to control a single area where the oven and the PCB both lag your setpoints. There are a lot of DIY efforts out there, some may be worth copying, the commercial single zone reflow ovens out there are either crazy expensive or crap. A small single zone oven may also change its performance board to board if used continuously as well as become a bottleneck in your process, there are small conveyor oven out there such as the IN6 (OK performance possibly goes through heating elements) or hybrid Air/Reflow like a T980.

A conveyor based oven will have multiple zones running at preset temps, typically a manufacturer will create a sacrificial board and solder thermocouples in multiple locations with high temp solder, including difficult areas e.g directly under something high mass. They then run this board through the oven and adjust the various zone temps and conveyor speed until all those points meet the criteria set by the solder paste over the reflow process. Our old IR one had zones from ~195 to 330C to achieve what they call a ramp to peak profile. That isn't an ideal profile but it was what that oven could manage with a ~6min process. By contrast our new oven is full forced air convection, with 7 heated zones and I can design profiles where even the last zone is under 260C, this leaves little chance of anything getting overheated as 260C is a pretty typically max temp for a lot of components, it also lets me achieve a ramp-soak-spike profile. Ovens and reflow are more complex than they first appear, at the commercial end of things it can take a lot of money and space to get the performance and throughput one needs.
 

Offline ttt

  • Regular Contributor
  • *
  • Posts: 87
  • Country: us
Re: Big Day at the Meshtron "Fake It Till We Make It" Home PCB Assembly Shop!
« Reply #14 on: November 29, 2023, 08:06:05 pm »
Congrats on your purchases!

I also have a TVM802BX. Operating one of these can be daunting at first. Expect to spend a lot of time to fine tune the machine.

One of things which quickly becomes a PITA is changing reels. You need to remove and reinsert the center shaft through all reels each time (which I had to buy separately originally). Reel tape needs to have as little and consistent pull resistance as possible. If the design did not change in the last few years try one of these caddies I designed which remove the requirement for a single center shaft/rod: https://www.printables.com/model/298453-tvm802ab-reel-caddy

If you have issues with plastic tape jumping around this could help: https://www.printables.com/model/226361-tvm802ab-8mm-rail-guide
 
The following users thanked this post: meshtron

Offline meshtronTopic starter

  • Regular Contributor
  • *
  • Posts: 132
  • Country: us
Re: Big Day at the Meshtron "Fake It Till We Make It" Home PCB Assembly Shop!
« Reply #15 on: November 29, 2023, 11:15:24 pm »
Congrats on your purchases!

I also have a TVM802BX. Operating one of these can be daunting at first. Expect to spend a lot of time to fine tune the machine.

One of things which quickly becomes a PITA is changing reels. You need to remove and reinsert the center shaft through all reels each time (which I had to buy separately originally). Reel tape needs to have as little and consistent pull resistance as possible. If the design did not change in the last few years try one of these caddies I designed which remove the requirement for a single center shaft/rod: https://www.printables.com/model/298453-tvm802ab-reel-caddy

If you have issues with plastic tape jumping around this could help: https://www.printables.com/model/226361-tvm802ab-8mm-rail-guide

Thank you, I dig the caddy design!!  I am still in the "get everything calibrated" mode but hopefully will have that finished this weekend.  I have been surprised how far off some of the factory settings were, but I am making good headway now that I am straight on all the various offsets.

I appreciate the insights and will report back soon with more info and pictures.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf