I think my new Magnetic Filament Monitor is dead.

dc42

A screen shot in one of your previous posts showed that you are running firmware 2.05. Please upgrade to firmware 2.05.1. It might fix the issue, although more likely it won't.
Please keep a close eye on the AGC value reported by M591. If it goes much above 100 then we'll replace your filament monitor.

shinook

Ok. I’ll update the firmware.

shinook

Update the firmware.
I keep getting random pauses from “too much” or “too little” movement. Though they aren’t as common as the good old “ sensor not working”. My AGC is still in the low 90’s. And every time I check the status of the monitor(while its working) the results are between 99% and 103%.

Honestly, it’s causing more problems than its worth. If I can’t get it to stop pausing my prints, I’ll need to find a more reliable monitor.

dc42

@shinook said in I think my new Magnetic Filament Monitor is dead.:

I keep getting random pauses from “too much” or “too little” movement.

What do you have the min and max allowed percentages set to?

shinook

70% to 130%

shinook

It’ll work great for multiple prints, then pause a 3 hour print 5 times.

dc42

Have the "sensor not working" errors gone completely, or not?

shinook

No, I get them several times a day.

shinook

I wish that it at least kept the calibration data from before a pause. Or gave some data about the pause. But it deletes it all and makes trouble shooting it from a logical data driven method near impossible.

shinook

But, at this point, with the AGC still showing good, I think it’s just got something wrong with the brains of the thing.

dc42

My guess is that there is occasional corruption in the received data. This isn't necessarily due to interference, it could just be caused by the processor load causing interrupts to be delayed. There is a parity check on the data, but of course that will only detect a single bit error.

Does your system include a PanelDue, and if so, does it use the default baud rate of 57600 or a different one?

shinook

No, I’m using a. Maestro with the LCD 12864F-3.

shinook

If you think it will help, I could unplug the display and see if that helps. I only use it as a display, and not to control the printer.

shinook

I tried disconnecting the LCD, and was able to get 88% of a print done before it paused.

shinook

So, if nobody has any other suggestions for me to try, can I return this and get another one? Or will I need to find a different filament sensor to try that doesn’t cost so much that wont pause my prints for unknown reasons.....

I don’t want to give up on it, but I’ve been fighting this for too long with zero progress. I can’t keep having my prints pause for no reason, its lowering my productivity and costing me money.

dc42

Please provide a M122 report taken after you have done quite a lot of printing with the filament monitor active. I'm interested in the number of parity and framing errors reported by the filament monitor. If it's zero then we need to send you a new filament monitor. If it's a high number then that suggests my theory about undetected data corruption is correct.

shinook

I’m having trouble doing “a lot of printing with the monitor” due to the pauses. I’ll try to get something today, but this is super frustrating.

shinook

OK. I'm done. I can't print with this thing. Its constant pauses for random "too much" "not enough" "not working". Here are the M122's that I'm able to get today. either let me return it and try another one, or give me something else to try. But as of now, I'm disabling the filament monitor and unplugging it. I have prints that need printed so I can get paid. Constant pausing due to my filament monitor is unacceptable, especially for a monitor that I paid premium money for.

console 3.txt console 4.txt console 5.txt console 6.txt console 7.txt

dc42

Thanks, looks like you are getting quite a lot framing errors (but interestingly, no parity errors). So I think it's likely that the errors are caused by reception of bad data that isn't detected as such. I was already planning to change the firmware to ignore isolated data points that are out of line with other data. I'll bring this work forward, I may well start on it tomorrow.

Please post your config.g file, it may help me to reproduce the problem.

Does the filament monitor cable run close to the extruder motor cable (or any other stepper motor cable) for an appreciable distance?

Walternate

At what point are you going to say this is a bad unit and give the man a new one? Is there a rash of mfm's going bad due to firmware? I have one. Works fine. From a support perspective, this looks to be bad hardware. My two cents. It works, then it doesn't. Then it does for a bit, then stops. No change to anything on the system. Now he's at a halt. In addition, shinook is an absolute duet fanboi. He got ME into the mfm even though I didn't want to. So glad he did because I love it. Let's move this guy forward.

@dc42 said in I think my new Magnetic Filament Monitor is dead.:

Thans, looks like you are getting quite a lot framing errors (but interestingly, no parity errors). So I think it's likely that the errors are caused by reception of bad data that isn't detected as such. I was already planning to change the firmware to ignore isolated data points that are out of line with other data. I'll bring this work forward, I may vebe start on it tomorrow.

Please post your config.g file, it may help me to reproduce the problem.

Does the filament monitor cable run close to the extruder motor cable (or any other stepper motor cable) for an appreciable distance?