Forum Replies Created
-
AuthorPosts
-
cmcgrath5035
ModeratorMock turtle = Garbled text where in CP?
Is the Gcode trashed, or ??Some other info would be helpful:
SPJS version, and OS SPJS runs on ?
what firmware were you running on tinyG? (440.20 most recent)How did you attempt to upgrade reinstall FW?
SPJS firmware updater?ModeratorMock turtle = Garbled text where in CP?
Is the Gcode trashed, or ??Some other info would be helpful:
SPJS version, and OS SPJS runs on ?
what firmware were you running on tinyG? (440.20 most recent)How did you attempt to upgrade reinstall FW?
SPJS firmware updater?ModeratorFor starters, read this wiki item a couple of times
The operation of Homing and Limits is same in G2, but G2 now supports NC/NO selection on a per switch basis.
Special note – soft limits are only enforced when a machine has been previously homed.
I think what you are saying is that your switches ar NO proximity switches, nominally 12V in ‘OPEN’ state, with voltage divider networks to reduce the 12V to 3V. I’ll assume you took into account the pull-up resistors(to 3.3v) on the tinyG board. what do you measure at tinyG2 in the NO state; 3V?
With classic microswitches, NO was not ‘preferred’ due to noise susceptibility.
But chances are that the impedance of your voltage divider significantly reduces noise susceptibility.I have played with capacitive NC proximity switches, force them to operate by holding a piece of aluminum close to them(rather than mess with forcing voltage with wire)
Is the settings.h file you are compiling with changed from the default Gshield settings?
What have you changed? It is very difficult to spot differences.
What version of tinyG2 have you compiled to (what version number)?
Or, what is the value of $fb?What are you trying to test, Homing, hard limits, soft limits?
ModeratorSee if this helps
ModeratorClick on the ‘Available as setup executable’ link to download this zip file.
You can then unzip and run the exe file before plugging in the TinyG. My file was called “CDM v2.12.10 WHQL Certified.exe”Run this file before plugging in the TinyG and it will install the driver.
This sounds like the procedure – pre-load the driver components, so that Win will find, install and start when the hardware is plugged in.
So what part of it is not working?
ModeratorYes, that is what I had in mind – the source of the AC ‘leakage’ is high source impedance and should not be an issue when you connect the tinyG USB to PC.
If you are unsure of your power supply, always a good idea to tap the shield of the usb against the outer metal of the usb connector on the PC. A slight spark (ESD like) is fairly normal, a ‘snap’ sound or heavy arc means something is amiss.
The burned USB cable item you reference was, IIRC, a clear power supply incompatibility with this application. I don’t have time to dig it up at the moment.
One lesson learned from that event, if I recall – don’t attempt to put two transformerless 12V supplies in series to get 24 V, they don’t work that way (and usually indicate that on spec sheets).
ModeratorWeird – yes, I don’t know how/why bogus stuff in CP could result in tinyG changing fuse or other internal settings. If upgrade worked once already, then fuses should have been set OK.
Unfortunately, you are on the hairy edge of debug capability.
The sooner you can get the unit headed to Synthetos for brain fix, the faster it will return.ModeratorI have seen transformer-less AC to DC Power supplies that are reference to Line (black AC) as well as White (Neut AC), Green Ac of course is connected to white back at the AC breaker panel.
So you might try measuring AC potential between your PC ground and the Power Supply output, both + and -. Typically this is a high impedance reference, so connecting the ‘- ‘of your power supply to pc ground should not create an ARC event, but make create an ‘ESD’ like spark event.
You see this sometimes when connecting a Coax feed to a modern TV powered by a transformer-less power supply.So, I don’t think the 9V ac you are measuring is a big deal, but cannot be certain from the available information. You might try connecting PC ground to Power supply ‘-‘ with a 1K or 100 ohm resistor, or whatever you might have.
If it is a ‘high impedance’ reference, the measured AC voltage across that resistor will be ~0 vac.
ModeratorThe Tinyg ground is not connected to the power supply AC ground.
Are you sure? Many transformerles power supply designs have a high impedance coupling to the Output, and some are lower impedance.
Try powering your tinyG from a 12V-24V wall wart. Not enough currrent to drive motors, but tinyG will be fine. Still see?
ModeratorIs this the first time you have tried to upgrade FW?
I think contacting Synthetos is best path.
Moderator2. Can you describe why you believe X current setting affects Y? I.E., how are you measuring current?
Here is a typical schematic page showing two of the 4 drivers.The wiper of each pot connects to one, and only one, motor driver and driver internals control the actual current limiting for that device(only)
The only things common among the current pot circuits is 3.3V and ground.
Have you tried monitoring 3.3V? Based on your description, I might suspect that your Vmot Power supply is going into current limiting when motors turn on, affecting 3.3v which is derived from Vmot input. Perhaps look at 3.3V with an Oscilloscope when doing a move or running a simple Gcode?
Current to the motors (and draw on the Power supply) is narrow pulses – detecting with just a VOM could be difficult.1. I suggest chasing #2 first. Base on your description, 2 & 1 could be explained by a dynamicaly current limiting Vmot power supply.
Very difficult to predict how significant variation in 3.3V supply might affect normal tinyGoperation.
Folks over time have reported Vmot supplies that looked OK supplying rated current to a static (constant, e.g. a resistor) load but the Supply current limiter was failing under a heavily pulsed load.Moderatorspjs 1.86 finds the usb port, but when I try the updated, specifing tinyg v8, and the 440.20 version via the online download, I get an error message saying
“could not find board arduino:avr:tinyg”.OK, that is likely a sanity check of the board by the SPJS programmer code before it initiates the FW upgrade cycle. It would be looking for a sane tinyG on the client end, not the bootloader running.
avrdude 6.0.1 is what SPJS includes and calls as well, so that is not the issue.
Try this if you want: Download
Unzip the archive into a clean directory and read the readme.txt. This is simply all the files you need in one place, but should be same versions as you have already tried.
I am doubtful my files will help. Seems something is borked in your tinyG.
If you are in the US, quickest recovery might be to send it back to Synthetos for a brain refresh.November 22, 2015 at 10:17 pm in reply to: spdir flashing, verification error content mismatch #8965ModeratorHmmm, I have never experienced or heard of an actual verification error. Sooner or later everything happens.
I’m also at a loss as to how a corrupted Gcode file could affect flash memory or fuse bits, etc.
What OS are you flashing from?
What version of avrdude do you have (avrdude –help)?You using CLI to run ardude, I assume?
If you are already in bootloader mode, and know your connection port (COMx or /dev/ttyUSB0, etc), and are running Windows, you could try the FW updater in SPJS 1.86.
I have been doing a lot of testing , seems to work well on Windows, Linux and Mac are questionable. To try it, get CP connected to SPJS 1.86 and go directly to the FW updater. I would recommend upgrading to 440.20, using the online SYnthetos download offered up by the SPJS updater.If no go,come back and I’ll offer up another alternative once I know your OS environment.
ModeratorHave a read thru this wiki item:
Recognize that these procedures are driven by the results of a $$ command.
Always check the parameter list as new FW builds are installed – sometimes new parameters are introduced(more so with tinyG2 than tinyG)
The restore from archive will replace parameters on the list with same $name, will do nothing for $newnames.By the way, since you run often in inch mode: put your machine in mm mode before launching the backup option, inch mode will fail. Also make sure parameter $tv=1.
-
This reply was modified 8 years, 10 months ago by
cmcgrath5035.
ModeratorThanks for the hints, Alden.
DJF: You may want to review this Wiki item for more details -
This reply was modified 8 years, 10 months ago by
-
AuthorPosts
