exxos's Atari, Amiga & retro forum

PhilC wrote: 26 Oct 2020 19:53 @exxos nicely done. Is that using the built in heated bed as well?

No, I cannot really use the heated bed as board I am doing now has parts on the underside of the PCB.. I had to put some small metal washers under the PCB clamps to lift it up a bit because the PCb won't lay flat on the bed... The built-in heat bed does actually work well it seems, but I've only used it the once so far... In the above I just used my heat-gun... I think that could be why my first one ends up with solder balls because I was heating it up too quickly... So I went from 420c air temp to 350c air temp and seems a lot better now.

@exxos I use around 350 when soldering with hot air

Did not last long before I ended up with a blocked nozzle :roll: I tried increasing the pressure on the cartridge no end only just would not unblock.. So I scooped out as much paste from it as possible and put it in the ultrasonic cleaner for about 20 minutes at 60° C.. Now it is good again :)

Incredibly boring video below...



I also realise now they have a in chat box in their software itself way you can talk to manufacturer directly from within the software :o I was having a bit of problems calibrating the printer in respect to alignment.. Long story short, it seems other people have requested a fine adjustment from the printscreen itself but there was a workaround via a command console.

Problem being with their calibration probe, relating to the difficulty of getting the magnifier close to the nozzle to do... Regardless of the number attempts I made, I really think the calibration probe is ever so slightly bent or something.. But it is rather difficult to visually see and align the probe on a pad you can not see without magnifier anyway... I mean we are talking down to like 0.2mm pads here anyway..

But at least with this command console I can nudge the head over a bit will save a lot of hassle when the calibration is not 100% spot on.. I really hope they will include this as a feature in the software eventually..


In any case, the support in their software with the in chat module, even though I have only talk to them directly once so far, has been extremely productive. I don't think I ever heard of in app support like this in any other software. So I'm actually really impressed with the level of support I have received so far. I literally only had to wait a few minutes for a reply, and the reply was nothing short of excellent and got my problem solved within a few minutes. And myself saying I actually are impressed with something... Well that pretty much never happens.. :lol:

Getting the hang of it all now :) The pads are deliberately "short" of paste :)

I have been looking into alternative pastes.. Ironically the current paste composition actually matches the paste I am currently buying in the UK anyway... After talking to the manufacturer, their smallest syringe is still marginally too big to fit the printer :cry: Bizarrely the printer manufacturer actually sells empty cartridges so I purchased a pack of those and some other bits...


All I did was transfer the paste out of one syringe into the printer size syringes... I was a bit worried about airballs and such but it did not seem to be a problem...


Next I realise you cannot fill these cartridges more than about halfway :roll:


So had to take some paste back out again...


I also purchase some disposable nozzles.. But I don't know why, but the paste just seems to get stuck in these incredibly easily
:roll:

It seems to come out perfectly well for a short time and then just blocks up to the point where suitable forces needed to get the paste flowing again..


Later I used the metal nozzle which came with the printer, which again seems to get damaged when putting the protective cap on the thing :roll: after straightening it out it just simply snapped off... But in this case I actually wanted a slightly larger hole anyway, so I think it actually worked out as a good thing :lol:


I was doing some experiments with my 30pin simms... If you look closely from left to right the paste seems to get less and less..

But amazingly, and thankfully, the printer software has a compensation feature which over time you can apply a percentage of extra pressure to compensate for such problems.


This is after reflowing on the printer itself... Overall I will probably need more solder on the pads, but I was really just trying to get the whole thing working to some degree before anything else..


Mostly I was working on the philosophy of a larger nozzle will mean faster prints.. And after talking to the manufacturer, there is not actually a setting for nozzle sizes or anything like that... BUT.. There are ways of getting around this problem... For example a larger nozzle will require less pressure otherwise it will come out too fast, and I can easily adjust the pressure and flow rates etc you can also control how many "passes" of paste goes on each pad in the software.. And another fiddle is the actual paste on the PCB file can be changed.

The problem overall is that while I have the printer maxed out speed wise, it still takes over five minutes to bring the pace on a single simm :( It would basically take 10 minutes to paste 2 up.. Whereas if I paste them by hand manually with a stencil and my rig, it does not really take more than a minute.. But again the problem overall is that even just 10 sets of simms is 40 simms which have to be pasted.. This is one reason why my Simms of been out of stock for so long because it really takes its toll on my fingers :( so the printer is the only real alternative to doing this work anyway..

But it does not necessarily mean the overall job will take longer.. As pasting stuff by hand is subject to errors and smudges and basically all sorts of failures and shorts which I then have to spend time fixing... Theoretically the printer should do a consistent job across all of them and reduce the amount of fixing time I will need to spend later on. Of course other than placing the PCB on the printer bed itself, those five or 10 minutes I can actually spend doing something else anyway.. So I'm not overly bothered about the slow printing.. But considering it takes to force an entire day to assemble 40 simms.. Then another couple of days testing and fixing... It would unfortunately mean pasting an assembly may well end up taking a lot longer.

After some research I have found some nozzles in the UK which I think will fit the syringes in the printer.. These are massively cheaper than like £20 a pop on the proper ones.. though finding some low profile ones has been a bit of a challenge.. Anyway, I have ordered some 0.25mm and 0.5mm tips.. 0.25mm is the default size for the printer and the ones I am currently using.. I went with 0.5mm then multiplied by pie and chips or something will increase the flow rate more than 3 times. I'm hoping the larger nozzle will help prevent the flow problems with the paste, and I also hope I can fiddle the printer to do less passes on each pad and put more paste down every single pass.. So I really hope I can get the paste time down to for like two or three minutes...

I think the only possible problem I may end up with, is I need more paste on the simm contacts than on the DRAM pads.. Putting based on the DRAM pads may well end up to much.. But saying that the stencils I was using down way too much paste anyway... Some hoping I can just do a single quick pass on the DRAM pads, and maybe 3 passes on the simm pads... I guess it is a bit like setting up a 3-D printer, that everything has to be set up perfectly to get the results you want..

I tried the disposable nozzles a couple more times, I don't know why, but they just clog up extremely easily compared to the metal ones :shrug: Considering they were not exactly cheap either it is slightly annoying :roll:

The syringe tips I ordered in the UK before Christmas seem to have vanished :roll: Though while trying to tidy up yesterday a bit, I found a bag full of various sizes of syringe tips :) they were all too long, but I managed to score them with some sharp cutters until it broke cleanly..

I had to do some more calibration which is a bugger like 3-D printers are :lol: :roll: The problem is there is settings like the initial "kick" To get the paste flowing, then the " while drawing kick" followed by retraction at the end of the line.. And then height the paste is put down at, and the height the actual head is retracted to prevent stringing :stars:

In any case, I seem to be getting good results now, because I only have to print half the amount of lines, the pasting time has now halved. While this is marginally more messy, does not really matter for this particular job.

Are you using leaded or lead free soldering paste. From my (small) experience leaded on is much easier to work with.

tzok wrote: 01 Jan 2021 19:41 Are you using leaded or lead free soldering paste. From my (small) experience leaded on is much easier to work with.

Been using lead free paste for years. Seems to be fine for me.

You are selling these, so I guess you have to use lead-free, to be RoHS compliant. But for prototyping, try leaded. It flows much better, and in lower temperatures. If something goes wrong, you can reflow it, even without adding flux. It also requires much less precision while applying it on the PCB. As an extreme example, if you have a solder mask between the QFP IC pads, you can put a line of leaded paste across the pads, and it will still work. Ironically leaded solder is also safer to use, as it usually contains less active fluxes, and flows in lower temperatures, thus produces less toxic fumes.

If you can suggest a paste I can try it. Though all you mention I can do with lead free anyway..