jv3 adding steps to the left? sensor?

[artbot]

my jv3 160sp lost it's mind today. we have two nice sized orders (with already extended delays at my request due monday).

i've had on and off stair stepping issues in the past months. cleaned strip and sensor. all good.

today the same. cleaned strip and sensor. no change.

HERE'S THE INSANE PART

just now it was zipping back and forth but printing nothing. did a test print, printed nothing. restarted the printer. now it's printing again.

printed a test file, IT STARTED STAIR STEPPING TO THE LEFT. that means it's ADDING steps. dirt can only subtract a step.

i was about to overnight an encoder sensor (the strip seems pretty pristine). but do i have something much more exotic going on here? anyone that has the slightest input. please chime in. my buyer is getting sick of "my printer is acting up" stories. all because of this
missing steps issue.

thanks in advance

 

[Malkin]

Can you install the strip upside down? Some can do that.

On a friend's roland there was a small piece of metal near the sensor that was bent and was pushing the strip out of position.

 

[SightLine]

I'm starting to worry you might be having mainboard and/or sliderboard issues with this coming back so much. It's a shot in the dark but maybe take a close look at any of the electrolytic capacitors on the mainboard, power supply, and slider board. Capacitors are reknown for failing early in electronics - failing ones can cause all sorts of intermittent issues depending on where and what they are used for. Literally 2 days ago I fixed our seemingly dead HP Laserjet 4101mfp copier/scanner. No display, only thing you would get is the main motor turning when powered on. In it's upper power supply there were 2 caps with the telltale tops bulging slightly. Soldered 2 replacements in (actually went a bit higher from 560uf to 1000uf) and it is working 100% perfect again. My thinking is caps are generally used in power filtering and stability. As many know in the computer repair feild a failing powersupply can leave a computer running but can cause a myriad of other seemingly unrelated issues.

 

[jfiscus]

After you clean the strip & sensor, I'm assuming you run a calibration (linear calibration?) in maintenance mode to have it verify the lines on the strip? When I had a Roland with a similar strip that test would tell you if you had actually cleaned it correctly. (it counts the lines going out & back and verifies that it counts the same amount of lines both ways)

Since it counts in both directions, I would assume it is having problems reading the strip count going back towards the capping station. Are you 100% sure you have the eye screwed back in the correct location and it's not angled just a bit?

You could try uni-directional printing a test file and see if it prints better, I'm not sure if it counts coming back in uni-directional printing...

 

[artbot]

thanks for all the input!... as far as linear encoder tests on the mimaki, the only one i've found is some test that the y motor just flies back and forth at full speed. it seems there would be some diagnostic test.

i have an "encoder test file" 60" of one pixel vertical lines. this test can be lengthened across the platen and can track just where the printer starts to lose steps. with this test, the printer loses steps anywhere on the strip.

it loses steps printing a single inch of this 60 inch test. and i've inspected the encoder with a loop. it's in very good condition.

the oddest thing is the adding steps. that would mean the encoder is counting extra lpi on the strip. so i'm willing to count out any thing wrong with the strip at all.

it could be a board. unless someone here is familiar with a y-motor failure acting similarly. but with it "printing nothing" then adding steps. a y-motor isn't going to cause odd aberrations in the printer. plus, the y-motor sounds very robust and quiet.

@sightline, i'll look at the boards. that is definitely not my thing. but i'll go with looking for anything odd.

thanks again. hope something obvious pops up. this appears to be an issue that could render this printer down for a long time if things don't start making sense diagnostically.

 

[artbot]

kent with macmedia has suggested that the encoder sensor is suspect. he's seen similar behavior clear up with a new sensor. he is a printer guru. kent has spoken.

 

[jfiscus]

I still think you have to calibrate your encoder strip if you haven't done that.
(have you done that?)

 

[artbot]

no. please explain. i built a test file in photo shop. it's a 60" long set of single pixel wide vertical lines. if you print this, it will specifically show by printing differing lengths of this file just where the encoder hiccup is. but that would be a physical issue on the encoder itself.

such as if the file prints perfectly (bi-directionally) but on the way back suddenly loses a pixel or to to the right at the 42" mark on the test file. you can usually reprint to reprove that result. go to that portion of the strip, and you'll find a speck of ink or dust (cat hair), etc.

if the file unilaterally loses (or adds according to kent) steps then the reader is just freaking out. evidently it can behave in many bad ways when failing.

is there a diagnostic encoder test in the mimaki service manual?

 

[ruckusman]

I think Kent is on the right track with his advice.

On the subject of suspect capacitors, don't wait until you see bulging tops, check the plastic used to encapsulate the cannister, if it's shrinking on the top - you can usually see marks where it used to be when new, they will need replacing sooner rather than later.
The plastic skrinkage is a good indication of heating from drying out which means power supply and signal noise fluctuations

 

[artbot]

will do. one other unmentioned thing is my heads have been failing (overspray or voltage issues) early like 18 months. maybe due to all the odd inks and varnishes i print? but also possibly a board is not right. but it is at the head. data swaps do not move the issue. so that has to be figured out. will do some looking.

thanks everyone. this board is like having five techs come by to look at the printer before i've even had breakfast.

 

[ruckusman]

I'm surprised you haven't been into re-capping electronics, you've been into hacking and modding everything else and replacing bad caps with a soldeing iron is eezy-peezy

 

[artbot]

i'm really an ink, coating, materials guy. that is where i spend the bulk of my r&d. i've got this flatbed build started. but event that is a cheat. stick existing printer on existing laser cnc table. i'm about to get slapped into reality when in need to sinc' the encad's x-axis signal to a servo controller and such.