RAMPS 1.4.2 fork: RD3D v1.0 (6 steppers, 24v, Due)

Quote
obelisk79
That is a lot of holes. Wow! It's one thing to see in the files, another entirely too see them on a physical board.

tell me about it. it's why i want to split this board into upright daughtercards (like the old PCs from the 1970s / 80s), even for the MOSFETs. that way people can pick 2-MOSFET boards, 3 MOSFET boards, different power levels etc. so, i'll put 3 headers on the left side with 3 pins each (they have to come off of D3-D12 with D13 being power-enable as well as LED), with I2C (and possibly SPI) running to them as well. another header on the left side for DCin. then 4 14-pin expansion headers spaced near-equally top-to-bottom for steppers.

i worked out that it's just possible to fit 3 steppers onto one of those 14-pin expansion boards, if you use two sets of STEP/DIR and use I2C to communicate with an STM8 which controls the EN# of the stepper ICs. so i could do 2-stepper and 3-stepper boards, potentially for a total of *twelve* steppers, which actually starts to run into problems with RepRapFirmware as it uses a 32-bit number to store the stepper status in order to get as fast simultaneous control over all steppers as possible.

the nice thing about splitting all that out is, you now don't need heavy power-handlng on the main PCB *at all*, all the daughter cards take their own independent power, you could do a solid-state relay board as well (for 110/240v mains style silicon heaters and so on). and unlike BAM & DICE, right-angle 2.54mm pin-headers are a lot cheaper than 52-pin PCIe connectors.

ok so that's all the ICs (except the EEPROM, why the heck is that needed anyway??), resistors and capacitors, plus the inductors, MicroSD, power connector and fuses.

if i put on the 3-pin 5/3.3->VCC connector on i can test it powering up an arduino, then try some things like switch on MOSFETs (without MOSFETs) and confirm that the SN74LVC125 is working, and so on.

i've made absolute minimal usage of the hot-air gun, here. bit 100uF capacitors and the MC7808, that was all. the last experience i had using it on the 0805s was it really failed to do a decent job, i had to do more recovery than actual soldering.

hand-soldering of 0805s seems to be straightforward, with the exception of when one end is on a GND plane. those actually sink so much heat that the soldering iron (set to 400C) actually starts to stick to the PCB. amazing. but also a good sign, as far as connectivity is concerned. this is presumably primarily what "thermals" are for... but i entirely switched generating those off, right across the entire board.

I use hot air for reworking boards, but for initial assembly of an SMD board I find a soldering iron or a hotplate easier.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

ok so SN74LV125 for the MOSFETs is working. managed to smoke an 0805 capacitor.....

young kids of today, will smoke anything!

i really should have paid attention to the fact that one of the 0.1uF capacitors went up in smoke and produced a small flame. the one that i put on the main power rail also decided to go up, this time because it had the full power of the PSU (350 watts) behind it, and it was also right next to the fuse, the fuse plastic caught fire and it's badly damaged the copper tracks on *both* sides of the PCB.

thinking about it: i ordered just "0.1uF capacitors". they're quite likely to only be 6.3v or 10v, they almost certainly won't be ceramic, or X5R or X7R temperature-rated. so... putting 12v through them, with the full power of a 350 watt PSU behind them and no fricking wonder they got hot enough to catch fire.

so.... hum... yyehhh... i can't take *this* board to SZMF! i think i have some X7R 0.1uF 50V capacitors here.

[edit] damnit, taking out the charred remains of the 0805 capacitor, the fire was intense enough to burn several millimetres through the FR4 PCB, and it's also shorted out the 12v and GND copper. i'll have to recover some of the components and start again. ho hum. this time not putting on 6.3v capacitors where it's going to be supplied by 12 to 24v, eh?



Edited 1 time(s). Last edit at 11/02/2017 04:14AM by lkcl.

Can't make an omelette without cracking a few eggs right?

Quote
obelisk79
Can't make an omelette without cracking a few eggs right?

*sigh*... apparently not i may be able to recover some of the components from this board, but after the experience of messing up due to re-soldering some of the connectors i'm inclined to just chuck this board and start completely again. nggggh! 2 days gone! and i'm leaving for SZMF on the 5th!

okaay so the next one, much better (this is with an atmega2560) a few quirks, the capacitors and on-board (ATMEGA) pullups work well but previously i was used to no pullups, had to work out the endstop configuration: all good now. 4 of the steppers are working great, i had messed up with the names of the NETs so had X and E0 DIR and STEP reversed... doh. 4 large MOSFETs all good: reduced the indicator resistors to 4.7k as 1.8k was horrendously bright. heated bed works great, thermistors work great.

so as i'm short on time i've moved on to the TMC2660 expansion board, i've assembled one and am going to fire it up tomorrow (and take photos). i'm not connecting the SPI signals for the first test, i'm just going to rely on "default" settings and see what happens. if it doesn't go "pop" i'll be happy.

i'll need to take a closer look at marlin-rc source as apparently it does actually already have support for TMC2660 which would be fantastic, but using SPI so that involves a bit more investigating.

Hope you managed to get everything working in time! Let us know how it goes.

why not you use a Smart Ramps to hach,now you give me another choice update my printers to 32bits.

Quote
datesheet
why not you use a Smart Ramps to hach,now you give me another choice update my printers to 32bits.

i haven't time to do it myself, but check the video, i go through the process of identifying whether "thermals" are enabled or not on the power / gnd flood-fill / planes. you can do exactly the same thing, for yourself, and will be able to check if there is the same critical design fault. it would help other people a lot if, after doing so, you reported here but also made a note on the reprap.org wiki page of the findings [reprap.org]

without checking: if they've just put SMT components on, and based it off of RAMPS 1.4, chances are extremely high that yes, they didn't correct the critical design fault.

Edited 1 time(s). Last edit at 01/07/2018 04:25AM by lkcl.

... be carefull with cheap chinese components! - a friend soldered some capacitors, stated with 22µF @35V ... one of them burnt away with 24 Volts ... checked the others -- and found two of them were "reboxed" 22µF@16V ones!!!

They simply put smaller 16V caps into a bigger 35V housing

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

... found a related (German) Thread about - [www.hifi-forum.de]

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

Quote
VDX
... found a related (German) Thread about - [www.hifi-forum.de]

... it's real - it really does happen. my favourite one i heard about was a 2.5in IDE drive inside a 3.5in SATA case. i mean... this is a country where they take cement powder, put it into "milk powder" containers and *sell it to their own citizens*, killing some of them in the process. for the most part cement powder *is* actually "harmless" but when it started killing people that's when the chinese govt stepped in.

so... yeah... don't imagine it's just foreigners they're happy to swindle. this is why i've spent considerable time tracking down a factory in china where the owner is extremely conscientious. i won't give you his name or contact details because he's as rare as rocking-horse s*** and i don't want *westerners* taking advantage of his generosity! i am however happy to act as a go-between, so as not to take up his time.

Quote
lkcl

Quote
VDX
... found a related (German) Thread about - [www.hifi-forum.de]

... it's real - it really does happen. my favourite one i heard about was a 2.5in IDE drive inside a 3.5in SATA case. i mean... this is a country where they take cement powder, put it into "milk powder" containers and *sell it to their own citizens*, killing some of them in the process. for the most part cement powder *is* actually "harmless" but when it started killing people that's when the chinese govt stepped in.

so... yeah... don't imagine it's just foreigners they're happy to swindle. this is why i've spent considerable time tracking down a factory in china where the owner is extremely conscientious. i won't give you his name or contact details because he's as rare as rocking-horse s*** and i don't want *westerners* taking advantage of his generosity! i am however happy to act as a go-between, so as not to take up his time.

So are you planning to sell these shields? If so, any idea on price?

Quote
Pheneeny

So are you planning to sell these shields? If so, any idea on price?

yes - i'll need to do it as a crowd-funding exercise, as i'm not too keen on the idea of buying components etc. without knowing that people are actually gonna buy it!

which reminds me, i need to ping mike to ask him for the quote on components and assembly for say QTY 100.

in large enough volume, the similarity between RAMPS 1.4 and RD3D's BOM is close enough that really, there should only be a couple of USD extra. but that's going to be 1k volumes to get down to the same sort of pricing as you see RAMPS in the online stores.

we'll see ok? i'll let you know.

[edit] ok so i've asked mike to quote for 100 and 200 sets of components. really they should be ridiculously low, i would be really surprised if the BOM was more than $6. btw pheneeny it's entirely libre, you can get the PCB CAD yourself and get them made up if you want.

Edited 1 time(s). Last edit at 01/08/2018 05:14AM by lkcl.

Is it possible to get some naked boards ?
or a "kit" ... i really love to solder sometimes..... your project seems to be worth going with it....

thanks
Markus from germany

@lkcl:
missing fiducials for pick and place calibrations on the PCB....

Quote
Giftyakapatch
Is it possible to get some naked boards ?
or a "kit" ... i really love to solder sometimes..... your project seems to be worth going with it....

thanks
Markus from germany

yeah i have 8 left from a batch of 10 that i did, i'm back in taiwan again and mostly recovered (14 hour flights, not fun), i'll have to find them as most stuff is packed up.

yes there are no fiducials: you can see from the PCB itself how densely packed it is: there's no room. the normal way to deal with this is to use centering on at least 2 components (in the furthest corners) and pick their centre positions by eye using the magnifying capability of the pick-and-place camera.

if it was a 4-layer PCB there would be absolutely no problem because the signal tracks could be moved inside. that however would increase the PCB production cost a *lot*.

ok so i got the quote from my friend mike: PCB and component costs for QTY 100 will be around the $15 mark. the TSSOP-8 I2C EEPROM is *expensive* - over $1. if it was possible to fit a SOP-8 in that corner it would be about $0.15.

i will also need to do a small redesign: some of the capacitors are 0805 and 50V (70V would be safer, the board's designed to run up to 35V, and i heard a rule that you're supposed to double the rating of capacitors for the max voltage). you simply can't get 0805 50V capacitors in shenzhen.

Really curious how this has been going. Really impressive work so far.

Quote
CHBRIGHT83
Really curious how this has been going. Really impressive work so far.

hi chbright83, apologies, i had the PCBs manufactured, i have 8 remaining: i went to Shenzhen Maker Faire last year, it was absolutely fantastic but also slightly traumatic at the same time: carrying 50kg of equipment, changing vehicles over 20 times in under 36 hours... oh and didn't sell any or gain any orders (one stockist in canada expressed interest but did not respond to email). so it was a lot of effort and a lot of money spent... for zero financial return. this wasn't exactly encouraging me to continue.

with my other project having most of its budget exhausted to the bare minimum i can only really justify further work and time on this if there is actual genuine interest and/or financial backing. i would be more than happy to do a crowd-funding campaign for example. that campaign however *will* need to be one that is more than the usual "$2000" or "$3000" or "$5000" level, in order to financially justify the time spent.

basically i'll put the time in... *if* there's going to be a guaranteed financial benefit for me to do so. if you can think of ways in which that could happen, count me in 100% ok?

You have a working controller with uptodate firmware and all?

i got marlin up and running, yes. i didn't test the micro-sd card yet. i was in the middle of doing the reprapfirmware port.