Well the AM8 Build has been anything but smooth. Requiring multiple reprints, multiple assembly/disassembly cycles and a lot of ordering of extra bits, far more so than I thought would be necessary. Most of this was my own fault. However it is finally up and running. It still needs a few wires tidying up but as I write this it is currently printing a test benchy.
The ends of the pre-cut 2040 rails I bought were cut square with an accuracy of 0.1 degrees which over the length of a side rail produces a 0.65mm offset and if you connect one rail up and one rail down then you get a 1.3mm offset and a very warped base ! To get around this I simply rotated one of the side rails through 180 and accepted that the front and back rails would be 0.1 degree off vertical.
I ordered the optional longer side rails from RatRig - forgetting that this meant that I could not use the standard A8 Y rods. So I had to order longer 400mm Y rods (which for a couple of pounds more came with a complete kit of bearing/holders and Aluminium Rod Mounts).
Warning! Choosing to use the Aluminium Rod mounts over the lower profile printed, whilst more stable/accurate, did generate a significant amount of rework/redesign.
The RatRig kit comes with a good selection of 8mm/10mm/12mm bolts - However I ran out of the 8mm M5 screws from using the thinner Hafele brackets. So another postal delay whilst I ordered some more bits from my new favourite eBay supplier - Boltbase in the UK. For reference you have to match the bolt size to the depth of the plate they are securing, as too long a bolt will bottom on the 2040's internal profile before locking the plate/part in place.
8mm bolt will secure a 2mm or 3mm plate (I needed to pad out with a washer for 1mm thick plates/brackets)
10mm bolt will secure a 4mm or 5mm plate
12mm bolt will secure a 6mm or 7mm plate
About half of the cheap captive nuts provided with the 2040 rail kit were only partially threaded - so another trip to my garage to use the M5 Tap/Cutter on about 60 captive nuts.
On the subject of captive nuts you need to sit down and plan on the position of every 5mm captive nut and make sure that you have them in place before you start assembly as once the frame is assembled it is a complete faff to take it apart - just to add another captive nut.
As a partial work around if the bolt/nut is not that structural (e.g. used for Cable Clips/Endstops/Filament guides) then you can use these printable 3mm slot-in rotating captive nuts https://www.thingiverse.com/thing:3050607
A tip: if the nut is too loose in the captive holder it will slip out whilst you are fitting. You will need to locate and rotate these first using a bare bolt, then temporarily hold it in place by sliding the ramped end of a cable tie behind the captive nut, remove the locating bolt and then attach whatever you want and finally remove the cable tie.
I bought a new reel of SunLu PLA+ to print the AM8 parts. Only to find that I could not get it to print without warping. So after a long diversion, creating an enclosure, using brims, changing fan ducts, running multiple autoPIDs, trialling different temperature and cooling profiles; I gave in and printed most of the parts using the scrap ends of no name reels of standard PLA that I had in the cupboard. On the recommendation of some other cockpit builders I have ordered some eSun PLA+. So once the AM8 is up and running I will revisit this and I may reprint some of the parts.
The provided Z-Rod Holder design has no real adjustability. For both the Y and Z rods, I prefer to lock one rail/rod in the correct (square) position and then adjust the second rail/rod to be parallel to the first - so I have redesigned the top Z Rod holder to a) be adjustable and b) dispose of the anti wobble bearing for the drive screws - which I am vehemently against.
The Z-Endstop bracket design would never have worked with my X Carriage - so that also needed a quick redesign to push it further away from the Profile and under the X-Carriage.
As an aside: While debugging the AM8 I discovered a problem printing with the Z-home position moving 0.2/0.3mm between prints. That I have now put down to a worn/eccentric roller on the Z-Endstop microswitch which I have subsequently replaced.
Warning! Make sure you are looking at the right image before drilling ! The AM8 - horizontal Y Belt Upgrade assembly drawing shows the front and back bars as being connected inside the side rails and not to the end of the side rails as per the AM8 Drill Version which I am building. So I now have some spare holes on the bars!
The drill through design works well for all joints except the vertical rail to side rails which needed significantly beefing up to stop the frame going out of square at the slightest touch. I ordered some 70mm M5 through bolts and a pair of 180mm cupboard braces at 45 degrees.
The provision of stl files for build jigs and drill positions was excellent and did speed up drilling the frame. So much so that I added a couple more jig files for a) Y Rail position and b) the Vertical frame position.
Being a belt and braces sort of person I added Hafele Gusset Plates to some corners to ensure that they were held at 90 degrees to each other.
I would not recommend doing this change at the same time as the AM8 Switch. My usual mantra is one change at a time that way you can more easily debug the problems and also go backwards. The decision to go with this also created some problems
The design of the Y Belt motor bracket must have been done for a smaller motor than mine as the motor sticks out of the bottom of the frame by several milli-meters. This caused the first of multiple redesigns. It may also be that they assumed the frame would be mounted on rubber feet. I prefer to screw the frame down to a work top (IKEA Lack Coffee table ) and the feet to the bottom of the Lack table as this dramatically simplifies levelling the X Carriage and print bed.
As the new Y rods came with Aluminium mounts I decided to use these for better accuracy. But this had the effect of raising the Y rods 10mm's and so required reworking the Vertical Y Belt parts that I had already printed.
The Y motor was raised a second time
The tensioner block had to be re-designed, and of course using a different belt pulley (as I already had it in stock) meant that I had to rework the belt tensioner a second time !
The print bed Y-Belt attachment had to be redesigned to make it taller and this provided the impetus for me to alter it a second time to use my preferred tear drop clamp.
The bearing plate fouled on the Y rod holders, limiting the amount of motion.
I needed to add a 3mm washer to raise the plate above the Y rod Bearings -
With the plate raised 3mm higher the belt is now 3mm too low - another washer needed !
As a workaround I used some 3mm aluminium plate to get the printer working. I then printed the new Spacers and swapped these lighter PLA parts in for the Aluminium.
The AM8 X carriage design is 5mm narrower than the A8 shorter - So the rods needed adjusting within the Z Carriers.
After some initial safety checks, it powered up OK. Homed X and Z OK - I then went for Y and it ran the hot bed into the front of the printer - mad dash to kill the power. My choice of side (the right hand belt loop) for attaching the Y belt to the Y Carriage effectively meant the motor was now running in reverse. I could have taken the Y carriage apart again and swapped belt sides, or recompiled Marlin, but to keep life simple I just swapped the pins around on the motor connector and all is well.
First time bed levelling is always a pain in the proverbial, as I prefer to level to Z=0 ie no offset on the end stop. I needed to cut myself a new measuring stick to level the X Carriage above the base, Lack Table, and the higher print bed.
The test Benchy in progress - the robotty looking thing to the left is a freshly printed replacement Z End Stop holder. I use the back of the build plate (and orient prints long axis to left/right) to reduce the amount of motion on the bed heater cables.
The new Silver PLA filament, that I bought as a stop gap earlier this week, is acting a bit like PLA+ with very poor bed adhesion - it makes me wonder if the vendor is only stocking the one type and selling it as both. I have had to use a glue stick for the first time to get a filament to stick to painters tape.
The SunLu Filament that was on back order has now arrived and it initially seemed to stick a lot better than the 2 no name filaments. That was true for smaller parts. However Larger parts all suffer from Elephants Footing in the corners, requiring both brims and glue.