Mars 5 Ultra

[Chris333]

I need a BFH!

I will not print at all now. Hit print. The build plate moves down. About 1" from the resin it stops and slowly goes down in increments. (like it is trying to find the screen). Then about 1/4" before it ever touches the resin I get the warning "resin has exceeded maximum level, printing stopped".

There is a sensor in the arm the build plate clips to. It is there to sense when the plate hits the resin. And to tell when it hits the glass so it can push (spring plates) down.

https://m.media-amazon.com/images/I/71oZcDMxDEL._UF350,350_QL80_.jpg

[Chris333]

I took the build plate off and removed the vat and ran a print. Same exact thing happens. No vat, no resin. Warning that I have too much resin.

Ordered a new printer.

Put the build plate and vat back on. Now it is working    WTF

[Sokramiketes]

Have you figured out if the sensor is normally open or normally closed?  Can it be permanately wired together or clipped?

[Chris333]

I cannot find anything about it , but the little video shows it on top of the arm.

But the way this thing has spring plates. I think it needs the sensor to even work. Unlike other printers that lower down till a flag is tripped. Mine has the flag at the top.

I have been thinking about buying another because they way I've been going I'm making new designs before that last one prints. So not mad I bought a new one.

I need to find someone (smarter than me) who has written all the smart stuff out of the code.

What is the type of sensor called so I know what to search for? Stress sensor?

[Sokramiketes]

Well, reading that the answers on reddit seem to point to putting it on a stable surface (like the ground?!?) to see if that improves the error... it might not even be a standalone sensor.  Rather, something worked out in the feedback loop of the motors somehow. 

So, stable surface and a reset?

[Sokramiketes]

Seems like there's a firmware update too.  FYI

[JeffB]

Well, reading that the answers on reddit seem to point to putting it on a stable surface (like the ground?!?) to see if that improves the error... it might not even be a standalone sensor.  Rather, something worked out in the feedback loop of the motors somehow. 

So, stable surface and a reset?

 

My guess is that if it's been working for dozens of prints, it's not the surface it's mounted on.  If something depends on a sensor of some kind and then fails to operate flagging an error for that sensor, the best place to start is that sensor and anything associated with it (sensor, wiring, connection, etc...). 

The first step would be checking the wiring and connection.  Maybe a firmware update, but I hesitate to believe that would solve the problem.  Replacing the sensor itself is easy if you can get one.  Otherwise, if you can identify the signal from the sensor (is it on/off or does the current change due to a state change), you can look at that. 

In this case, I think it's a relatively "dumb" component...  I don't think there's any logic in the component itself other than changing state.  But of course, I could be wrong.  Though I have a hard time believing this is some sort of CAN/ASI buss type arrangement with a "smart" sensor.

Jeff

[Chris333]

I cannot find any firmware on their website.

When it first came out they had a video that showed the head of the printer with a cut away showing the sensor inside. When the build plate lowered it can tell went it first touches resin. Then it can tell again when it touches the glass. It then pushed down even more to level out all 4 corners of the spring plate.

https://us.elegoo.com/products/mars-5-ultra-9k-7inch-monochrome-lcd-resin-3d-printer


Right now I'm no longer getting the warning. But nothing is sticking to the build plate either. I think it is the sensor. I know these sensors are used for other high tech calibration equipment. I'm just drawing a blank as to what they are called.

One site said the Mars 5 Ultra had 2 different motherboards and show photos of both to point out the differences. My motherboard doesn't look like either photo.

[Chris333]

https://ae01.alicdn.com/kf/S55b98f22fffc47f984a020b0396482b85.jpg_640x640q90.jpg
https://buildyourcnc.com/cdn/shop/files/Screenshot_202024-08-13_20191623-800.jpg?v=1730087329&width=1445

[JeffB]

https://ae01.alicdn.com/kf/S55b98f22fffc47f984a020b0396482b85.jpg_640x640q90.jpg
https://buildyourcnc.com/cdn/shop/files/Screenshot_202024-08-13_20191623-800.jpg?v=1730087329&width=1445

I stand corrected...  That's no "dumb" sensor.  Way overcomplicated for what we need on these machines.  I guess they've run out of space for improving resolution or detail reproducibility, so they're going with bells and whistles to differentiate machines now.

Jeff

[Chris333]

Just to prove I'm not on drugs    yet.

There is a plastic cover over the head of the machine. Here it is slid back. (well forward)


This is the underside. There is "something" with wires epoxied to the arm.



BUT more importantly when I took the build plate off to take these photos I noticed a tiny 1/8" circle of cured resin on the plate. That is why nothing is sticking. Running a print now and I can hear the FEP so I know it's working!

So the factory reset fixed it I guess.

[peteski]

Well Chris, to me that photo seems to sow that the sensor or sensors are strain gauges.  These are used on those inexpensive digital scales which can have resolution of fractions of gram.

Basically it's a semiconductor material which reacts to minute stretching/contracting changes to the material (metal) it is glued onto.  That hole which looks like 4 large holes drilled through provided thin enough cross section of metal for the strain gauge to be able to measure minute defection it.

Here it will sense when it first hits the surface of the resin, and  the same when it hits the bottom.

This is not a simple on-off switch. There has to be a circuit which amplifies the strain gauge's signal before presenting it to the printer's microprocessor. You also cannot replace the strain gauge itself. The entire metal piece has to be replaced and the new one might have to be calibrated by some factory calibration routine.  Or the electronics to which the strain gauge is connected to might be misbehaving. Probably on the motherboard.

Is it possible that you accidentally bumped hard into that piece?  That could have distorted is slightly, causing a problem. That would cause it to misread the actual environment.

[Chris333]

Sweet learned something new.

The build plate clamps and unclamps right to this area. So yeah it gets bumped all the time. But after doing that and before every print begins the machine calibrates the sensor.

It is working now after resetting to factory. And that BTW is just me switching "on" slow printing because "fast" is default. Supposedly all this does in change the speed the vat tilts down.

[Chris333]

Looks like it can be used as a pressure gauge for ohms that moves has it flexes. So the calibration might just be zeroing the value it's giving?

But the photo Elegoo posted makes it look like a complicated computer chip.

[peteski]

Looks like it can be used as a pressure gauge for ohms that moves has it flexes. So the calibration might just be zeroing the value it's giving?

But the photo Elegoo posted makes it look like a complicated computer chip.

Not sure how that strain gauge is used in the printer's circuit, but I suppose that when the build plate has nothing on it, that might be used as a zero value. When the build  plate touches the surface of the resin (after previously moving through air), the strain gauge will detect that event, since resin is more viscous than air, so it will be like you doing a belly-flop in a swimming pool. 

https://en.wikipedia.org/wiki/Piezoresistive_effect#Piezoresistors

Well, the strain gauge itself is likely a silicon based device which interfaces with electronics in the printer and those interface with the printer's controller. The sensor is affected by ambient temperature, so that has to be compensated for too. So if you look at it that way, it is a fairly complex setup.