Using pullups will work. Are you amenable to having the version pins as #defines in config.h as the TLE5206 branch does?
I think we can ignore pins 52 & 53 for a start, they’re always low on versions 1-3 and not used on the TLE5206 board.
I’ve opened a PR with what I’m talking about.
The existing version detection setup uses two of the pins that would be useful for SPI, and aren’t needed to identify any of the boards in present circulation. At the same time, adding a couple more version detection pins so that the routine will be ready to accommodate future versions makes sense.
I’m getting way to far behind in my forum reading…
Count me in for a group buy, too, if it’s not too late. More than willing to contribute to the effort even if the maslow is still collecting dust in the otherwise empty shop. Soon as the post-snowmelt mud firms up a bit…
So just how fast will the current motors spin at 40V, or rather how much can they be juiced up before the magic smoke departs? Will the current limiting allow the extra voltage to help with getting the motors up to speed quicker, like the equivalent use to overcome stepper motor inductance?
one more thought on this. Can we get LEDs on the input pins to the board and
some easy way to put some on the output of the MOSFETs so that it can be a
simple “are the LEDs blinking together” diagnostics to find blown MOSFETS (and
other simple diagnosis, like loose power cords )
I’ve been looking at a daughter board that could be plugged into a motor circuit at either the motor or board end. It would make measuring voltage easy, and could allow a current measurement as well. LEDs on that would resolve between a bad drive, a bad cable and motor and a bad motor. A red/green LED across the M+ and M- would give the info you’re describing. A final plus is that a quadrature encoder divider would fit there nicely.
On the TLE5206 board itself, the chips have thermal and over-current shutdown. I’ve driven the output directly into an ammeter to see what high current would do - the output drops, but is normal again when the meter is removed. The ‘blown chip’ issue should be much rarer on these boards.
I put the generously sponsored TLE5206 Board for a test run the past few days.
A 12V/10A from china made sure I will get ~ 7-8A without magic smoke.
It’s been great 4 Maslow holidays!
Starting with calibration I ended lucky, as nothing that was not already crap, was damaged.
A 1 year unnoticed flex in my motor mount mock-ups became visible and has prevented other parts form breaking. I think I will keep a ‘predetermined breaking point’ in mind with this shield, for a new topic.
The most feared Maslow-Panic-Moment is when nothing stalls and all forces are unleashed.
A human error entering a wrong number during calibration was likely the reason for me to witness how my steel motor mounts can flex, twist and bend. This I’ve I tried 2 times in a row before I discovered a -10something in the settings for the motor distance.
A sanity check of all numbers we have is required and even with different frame dimensions, sled weights and modifications, we know the upper and lower limits what makes sense. Before the ‘Pull Tight’ there needs to be a "Are we Green?’.
The second day I hooked the chains on the marked position, set chain length manually, adjusted all settings that I know to their hand measured value and did a new calibration. This time my already twice compromised motor mounts (bent back) would still show a flex but survived the 3 seconds.
I did the calibration cuts, entered the measurements and got through.
Knowing frame and motor mounts have exceeded their lifetime, the TLE5206 Board test was the goal the last 4 days and to run it, I choose something simple and that will still work with crappy machine.
I ran the board 6 hours and fingertip checks on the board felt always cold. (I admit I have my board in the flow of the AC)
I have fried 2 motor-shields in the past and thing I have 1 that has not decided yet if it is good and bad. I do not see this happening any time soon with my ~17kg sled. I’m confident I can pack on more weight (hypothetical).
With truncation turned off in the settings, I ran 2 files (~6hours) with 8 decimals behind the point in one go.
Due to a surprise gift from my co-worker (from the land of Kamasutra) of 11 sheets 1200mm x 800mm of OBS? Whatever, it’s free so I’ll cut it, I only tested around the centre and not at the top of a full ply sheet.
just checking (because it’s something I’ve been nervous about with the latest design tweaks), even when you were pulling hard enough to bend up the motor mounts, you didn’t bend the linkage at all, right?
It might, if you found one with the right specifications. I looked last time they were mentioned but didn’t see any that seemed right. The TLE5206 chip has active over-current and over-temperature protection, but the stock boards could benefit.