Documentation, documentation, documentation.

Hello all!

Terence here, with another update on the Kossel Pro project.

Administrative:

In our mad rush to get units shipped out, we managed to corrupt the shared Google Spreadsheet that we've been using to track units going out and delivery.  Things have finally gotten to the point where I am no longer fighting fires left and right to allow me to take a half day of searching through all the email records, matching up tracking numbers, etc, to fix that hot mess.  After the audit, we found that we still owe about 18 printer kits.  We will be reaching out shortly to these 18 people to update them. 

After these last sets of printers, we still owe pre-assembled machines and we still owe plastic parts only pledges from Kickstarter.  Fortunately, we have plenty of extra parts.  I expect that I'll be getting to the plastic-only pledges in the next week or two (survey still needs to be sent).  With regards to the pre-assembled machine, our tech is in the process of moving after accepting a new job.  We are obviously waiting for these machines' assembly process to be completed, and we are also waiting for the new all-metal hot end.  We'll update in a week or two with the delivery status of both of these items.  

Manufacturing Update:

To make the project viable in the long term, it was imperative that we move as much of the kitting and handling to our overseas contract manufacturer.  (We simply do not have the space and human resources available to kit parts in the US, and at this stage of the business, entering into a long term commercial lease for a shop space seems a bit risky).  We were also able to get much better service from packaging vendors overseas:  We paid for cutting dies for all new packaging boxes, custom cut foam inserts, and the MOQ (up to a thousand piece for the foam inserts) and all that together, along with the NREs for the cutting dies, were still cheaper than buying and compromising with boxes from Uline.  

An example of our new kitting document, bilingual in Chinese and English.  My father's been helping with the translations, and in the few instances when I have to do the translation myself, I use Google Translate and Notepad to translate, copy and paste together my sentences.  My handwriting is still too poor for most handwriting recognition software to recognize (and learning Cangjie seems like an impossibly high bar to reach).  I do double check all the translations before they go out to our vendors.

An example of our new kitting document, bilingual in Chinese and English.  My father's been helping with the translations, and in the few instances when I have to do the translation myself, I use Google Translate and Notepad to translate, copy and paste together my sentences.  My handwriting is still too poor for most handwriting recognition software to recognize (and learning Cangjie seems like an impossibly high bar to reach).  I do double check all the translations before they go out to our vendors.

For the past few weeks, Mike and I have been rushing to update the kitting instructions to support our overseas vendor.  My Dad is also in Hong Kong right now, which really helps with the transfer; he's been handling the first article inspection of all the goods received, and also helping with communicating with our vendors.   We now have reached the point where our CM can start assembling the printer kits for us, which was a lot of hard work - but we expect work to go fairly quickly at this point on, and - more importantly - we can expect a relatively uninterrupted supply of kits, as long as we do our forecasting right.  This is a far cry from having to count screws, seal bags, and package and label parts, and will soon free us up to do engineering again, or at least maintain the flow of goods when Mike and I are bogged down by day job projects.

Customer Facing Documentation:

Studio OpenBeam... one day.  Photo gear does take up a LOT of room, especially if you are shooting large objects.

Studio OpenBeam... one day.  Photo gear does take up a LOT of room, especially if you are shooting large objects.

My wife have been threatening to evict me from my photo studio for the last month (I temporarily took over our guest bed room, after we got rid of the guest bed, before she could paint it and turn it into our nursery).  Now that ZT-KID-00101 is well on his way (in about 2 months), in between working on the kitting documentation (above) we've been banging away at the documentation pretty hard.  Tonight I'm happy to say that Rev 1 of the assembly documentation, covering mechanical assembly of the printer, is complete.  We'll be updating the links from KosselPro.com shortly.  We will update this with instructions covering systems power up test, firmware update, and best practices on the printer in the next month.

For those asking for more pictures, esp of the new hot end, please see below:

For those curious, these were shot with a Strobist setup.  On my last trip to Hong Kong, I picked up a bunch of YongNuo 560 IIs.  For less than the cost of a single 430 EX II from Canon, I bought 5 YongNuo speed lights, which are optically slaved, and had enough money left over for the collapsible background and the photo box.

Other Technical Updates:

Some days, it feels like our mad engineering rushes occur  between major Chinese holidays.  While our overseas partners are celebrating Qing Ming (or tomb sweeping) festival, we'll be writing and releasing the final build instructions for our printer kits.  Here are a few things that we've worked on in this period.

All Metal Hot End

We had one report of one of our early all-metal hot end jamming on ABS printing.  (Because of my pregnant wife, we are only using PLA for our testing until a better fume extraction system is setup, to minimize VOC exposure).  At first we though it was a thermal issue; turns out it was a roll of bad filament.  

And this is what happens when you read your print wrong, and do the calculations wrong for where to turn the dial to.

And this is what happens when you read your print wrong, and do the calculations wrong for where to turn the dial to.

The lathe we use does not have any sort of digital readout, so all our prototypes are machined by counting wheel rotation on the dial - much like how someone in Victorian era England would have made these parts. (Incidentally, when I started my machine shop training in college in 2002, I was the last class to be taught on fully manual machines.  Digital readouts and later on CNC were introduced by the time I graduated).   I started the project as an "OK" machinist and after a few frustrating setbacks, have gotten pretty proficient at making our prototype heat sinks.  I can now cut one in about 45 minutes.  

A not quite final all metal end effector prototype.  The one that's built to spec is currently being tested at MatterHackers - they've successfully printed Nylon and ABS on it, as well as quite a bit of PLA.

A not quite final all metal end effector prototype.  The one that's built to spec is currently being tested at MatterHackers - they've successfully printed Nylon and ABS on it, as well as quite a bit of PLA.

We now have the geometry of the heat sink finalized.  We were going to award the work to the shop I found in the LA area, but they quoted me 2 week lead time for my heat breaks when we awarded them the work 3.5 weeks ago - and not only have they been now pretty late, they've also been pretty uncommunicative.  Having been the small business owner trying to make it, I can certainly sympathize with what they may be going through, but we also have customers to take care of, and I'll be searching for an alternate supplier in parallel.

The good news is we've found a good supplier for heater blocks and nozzles, and we've also found a supplier for the fiberglass tubing used for insulating the thermistor.  We were really hoping to have a picture of our heat break to show, but unfortunately, we don't have an ETA on this part at the current time. 

New Heater Cartridges, power supplies and other small details:

For the next batch of Kossel Pros, we have commissioned custom heater cartridges with 22AWG wires pre-crimped, instead of the normal stuff with the high temperature insulation wires.  This will allow the crimp at the heater cartridge power terminal to meet pull test specs, as the current insulation is to wide to be properly crimped to the Molex SL crimp pins.  (Long term, we are going to switch to using ring terminals that anyone can crimp at home, but that requires retooling the plastic and the PCB and if we are going to do that, we might as well release the multi-material end effector that's been rattling inside my head for the last  year plus...).

Of course, leave it to the vendor to screw up on the cables from our first batch.  So my Dad is currently using it for a dummy load to test our power supplies.  Speaking of power supplies, we've found that the old vendor have been over driving the fan inside the PSU:  It was a 12V fan and they were running it on 24VDC. We commissioned a new batch of PSUs with 24V fans and they should run a lot quieter and last a lot longer.  

Finally, we've been making small improvements to the kit.   We've found a vendor that does good pre-printed labels (for cable labelling).  We also found laser cutting vendors and a new cable vendor overseas.  So far, pretty happy with the quality.

That's about it for this update.  We are pretty anxious to get the new kits in.  We'll be back in about two weeks with another update on the work we're doing to evaluate the auto-levelling routine.