Welcome to the SeeMeCNC Community

mhackney wrote:Basically. It will need the HE280 connector on one long edge and pads for the 8 wires on the other. The processor chip and 5v regulator (maybe depending on processor) and an LED to show the touch and relay messages on startup. I'm also thinking a row of four 2 pin headers to use for setting the sensitivity with jumpers and another to select NO or NC behavior. I can write this up and also make a drawing of the basic format and how I'd like to attach it between the whip and hot end.

The dimensions of the board would be useful too, since I don't know how it mounts. Does it insert at a right angle to the HE280 board, or stacked?

I'm working on the specs.

Right angles is what I'm thinking.

If you choose a processor with an ADC, which is easy, you could use a ten-turn pot to set sensitivity and get 8 or 10 bits of resolution.

I understand your suggestion but in my experience a pot is a poor user interface for most end users. They have to use trial and error to tune/configure the system. Whereas 3 jumpers give 7 levels of sensitivity input, more than enough for this application. 2 jumpers might be enough.

Here is the overview on the board and how I think it should attach and operate. I have a couple of alternative ideas for the processing etc depending on costs and whether an ATTiny85 can handle both 3.3 and 5V. If not, perhaps two different boards could be produced simply populating the proper ATTiny. Or a regulator could be added. These are the types of questions that need to be answered in the next phase of the prototype.

More progress, I have the interface talking to Duet! I have the interface hooked up to the E0 endstop connector in Duet. I can trigger the probe and I see the Duet LED flick off then on. The only issue is that the Duet end stops are 3.3V and the pin output from the Arduino Micro is 5V. I don't want to risk frying my Duet so I'm not going to do any more testing until I sort that out.

I do have a 3.3V trinket that supports I2C. I used it for my accelerometer experimenting last year. It should be easy to port my now working code over to it and test. If I can get it to work, it is actually small enough that I could mount it on one of my printers and test it actually working. There are some very small ATTiny85 based boards on the market and they have everything I need to do this, they just don't have the form factor to plug into the HE280 connector. But, maybe an off the shelf, inexpensive, option where you simply splice one of these ATTiny85 boards into the 3 wires coming from the whip would be a good short term and functional option. And it could be reprogrammed by the customer if I make tweaks to the firmware. Any thoughts on this?

I've been doing a little more investigation. The ATTiny85 processor supports 2.7 to 5.5V operating voltage. It comes in an 8 pin DIP and a surface mount option. The 8pin DIP version would be very easy to prototype so I've ordered a couple and a programmer (arriving Sunday).

With this chip, I can make up a few "splice in the whip" prototypes to send out to a few folks to test and it is very small.

The attiny25 is the chip I currently use on my IR sensors, the same as the attiny85 but with less memory. As you say, it's fine on both 3.3V and 5V. But for future production, I am thinking of replacing it with a 32-bit ARM core Freescale chip, which costs about half as much as the attiny since Microchip hiked the prices of Atmel AVR processors.

Thanks David, I just need to get this thing working so I can use a Duet with accelerometer probing! It's easy for me to program and work with these devices (even on a Mac ). For this simple interface to the HE280, I don't know that there is a big market to warrant spending more on development but that could change.

But the good news is, the concept seems reasonable and I should actually be probing with Duet soon.

So I have the proof of concept working on my bench. The HE280 is interfaced to Duet (an 0.8.5) as a three wire endstop on E0. I tap the HE280 and the probe captures it, relays it to my Arduino Micro where I process the signal and output the probe event to the Duet. Duet sees the probe tap and would act accordingly if it were controlling a printer!

The next step is to get everything installed on my V1 to actually test probing. I do have a little wrinkle in that I'm using an Arduino Micro on my testbed and it is not suitable to use on the machine. I have some ATTiny85s on the way and can program and connect one of these easily. I'm not using the HE280 on this machine to power the hotend and fans since this is a test. So I need to run 12V power to the HE280 as well as the 3 lines I need to mimic an end stop. I'll have all the wiring installed and ready to go when the chips arrive.
Thanks to Matt at 713maker.com for sending me one of his HE280 robe mounts and to SeeMeCNC for replacing my "blowed up" HE280 so I could get this far.

Most of the arduino chips will operate on 3.3V. Of course, the I2C signals from the HE280 are 5V based, ironically having been level-shifted from 3.3V.

BTW, if I read the specs on that plug connector correctly, the wires aren't crimped in, they're stuck into a spring connector. Which can be enticed into releasing said wires again. Might be easier than cutting them. You've got a connectors to play with, so you tell me if that's a useful idea.

And now let me drop a shoe and see what you think:
I think you're going to a great deal of effort to adapt a specific piece of hardware, when it is probably easier and maybe even cheaper to simply replace that piece of hardware and get a more useful product that will interest more people. What if you just put an accelerometer on your little board and interfaced it for an endstop connector (much like you probably planned with your original project). Then you get something that can be used wtih ANY hotend, so long as you can mount it firmly enough. It can even be used with the HE280 by reusing the 3 harness wires as you planned. Or by running new wires and choosing not to use the stock probing feature.

Thanks Eric, yes the wires can be removed safely from the connector so the connector can be reused. I've done it a few dozen times already.

RE Shoe... I agree but the SOLE reason for doing this is for folks who have a V3 and already have the HE280. A dedicated probe board would be the next step. The other thing is, mounting any probe is a PIA for folks. The HE280 PCB has a simple mount and after market adapters for other hotends are available already. I wanted to start with this proof of concept to see if the idea actually works. I am very confident now that it does. I think a simple "spliced in" ATtiny85 can be used with the HE280 and it's worth spending another few days on that. Then I do think a universal probe makes sense but figuring out a mount that can be readily adapted is critical to success I think.

mhackney wrote:Thanks Eric, yes the wires can be removed safely from the connector so the connector can be reused. I've done it a few dozen times already.

RE Shoe... I agree but the SOLE reason for doing this is for folks who have a V3 and already have the HE280. A dedicated probe board would be the next step. The other thing is, mounting any probe is a PIA for folks. The HE280 PCB has a simple mount and after market adapters for other hotends are available already. I wanted to start with this proof of concept to see if the idea actually works. I am very confident now that it does. I think a simple "spliced in" ATtiny85 can be used with the HE280 and it's worth spending another few days on that. Then I do think a universal probe makes sense but figuring out a mount that can be readily adapted is critical to success I think.

Excellent work on this Michael.

As far as "universal" board mounting, I think a single 4 MM hole in board at one corner would work in most situations. You simply sandwich board under one of the stand-off screws.
Or possibly a single ear tab may provide cleaner mounting options?

Brian, I might be concerned about the spring board effect when probing with that. In fact, that's basically how mechanical accelerometers work - a strip of stiff material is fastened at one end and the deflection is measured at the other end.

Position the accelerometer chip near the mounting hole.

I'll have to play around with the mount you sent me and see what I can come up with. Thanks!

I had mounted my test last year vertically so I was actually using the Y axis oriented in Z.

I can't wait for you to make this happen! Put me down for at least one probably 2!

Shut up and take my money... just tell me how much. I am looking upgrade to a Duet, but didn't want to scrap the HE280 board.

Michael, Just out of curiosity:
Are you putting all this great work into the accelerometer probe out of convincion that it is a better solution than an fsr setup? After reading a lot of comments about this probe I understand that it works, but hammers on the bed while with an fsr basically you get the same without the hammering. fsr are cheap, accurate (so I thought), easy to mount (under most circumstances), and easy to understand. Are there any pro's that place the accellerometer above the fsr that I am overlooking?

DetlaCon, I have no pre-disposition if it's better than FSRs at this point. I have my V1 setup so I can do head to head comparisons on probes and controllers. Currently has FSRs and the IR Probe.

In practice, the accelerometer really doesn't hit that hard. My initial impressions were based on probing on naked glass. It makes a TAP for sure. But I think FSRs would too. Now that I've got some experience with the V3 and have FabLam on the glass, the tap isn't nearly as loud.

So the motivations are:

1) the V3 (and HE280 hotends) have a built in probe but it is not interfaced in a standard way to the controller or firmware. The idea is to build an interface that makes the probe look like a standard endstop so it can be used on any controller and firmware
2) because of the way the bed is recessed in the V3, it is not possible to mount FSRs there. The bed would have to be raised above the base plate. This adds complication and detracts from the V3's elegant design.
3) Although I love FSRs, many people have difficulty building a mounting system that works flawlessly. The accelerometer has the advantage that there is no mounting system required and, in fact, you want the bed to be held as rigidly as possible. That's easier to do than constructing mounts that allow the bed to move in Z but not X and Y.
4) a dedicated accelerometer probe (not the HE280) would have the same benefits as 3 above, the installation would be almost trivial. As Brian mentioned above, attach with 1 screw to one of the posts, run 3 wires to the controller and that's it.

But, until I can thoroughly test accelerometer probing with Duet I don't know if it is actually as good, better or worse than FSRs. I suspect it will be at least as good so combined with the mounting simplicity might be a better overall option. That said, FSRs do a great job. They are repeatable and reliable once you install them properly.

"the tap isn't nearly as loud"

That makes me wonder: could one use a microphone as a sensor? Strap it to the glass and it would register a pulse a fraction of a millisecond (sound travels 4.5m/ms in glass) after the probe hits. During normal operation nothing hits the glass. There's the normal bumping and grinding from the motors of course, but I have a feeling the tap would be significantly louder.

Others have thought of that but I'm not sure anyone has tried an acoustic approach. But, putting materials like PEI on the glass changes the properties and would need to be taken into account. And "loud" is relative. The noise from movement would need to be taken into account. Much more work can and should be done on probes!

Thanks Michael, for your insight. Sounds very plausible. And indeed fsr mounts are more difficult on a V3 than probably any other Rostock design.

i like the idea of getting rid of long distance i2c.... i'd be in.


i don't think the accelerometer taps the glass too hard, but i don't have fsr's to compare it too. it doesn't leave a mark on my pei...

I support replacing the I2C with say, RSS485, run over a quarter mile spool of rusty barbed wire before it goes to the Rambo (I kid, but there are far more resilient standards, and something like RS485 is high enough data rate even at over a kilometer of length). I2C is a fine protocol for small runs, and in it's intended use case. On a board at the end of a multiple foot wire where it's next to a very electromagnetically noisy heater is not a use case that makes ANY sense for it. Admittedly, it's easy to implement, and there is a specific 4 pin connector on the full size rambo for it.