Mike's SCARA Robot

[macardoso]

Can you not just buffer those limit switches with an appropriately chosen and arranged transistor circuit? Should be way simpler than replacing the ones that are hard to get to?

I definitely could, but it would require me to add a new circuit board somewhere. I don't have a good spot to add one, nor do I want a hand soldered breadboard hanging somewhere (not that the wiring is all that pretty right now, but at some point, I'd like to put this into an enclosure.

I have to pull the T2 sensor out anyways, and I'm considering machining a new code wheel for the T1 joint sensor, so I'll likely be pulling it all apart anyways. Bit of a bummer, but it will be cool to see further inside the joints. Doesn't look too hard to do, other than dealing with the weight.

The gearboxes are supposed to get repacked every 8000 hours of operation with Harmonic Drive brand SK-1A harmonic drive grease. Bet that is overdue... Wonder how much a tube would cost if they'd even sell it to me.

 

[macardoso]

As a fun side note, there is a brake release button on top of the robot which has been non-functional this entire time. Once I got the wiring figured out, I found out the brake button was working again. That was on my list of things to address later, so it was a nice surprise

 

[macardoso]

So I am out of luck on using the original homing sensors, rewired for sourcing. They worked exactly as expected, except for one thing... The current limiting resistor (800 ohms) in front of the emitter drops almost all of the voltage from the power supply. All that is left after it is the forward voltage of the photo interrupter emitter, a measly 1.20V.

Even the robot manual is wrong (not the first time), calling this pin on the sensor cable +24V.



Without modifying the circuit to place the resistor after the circuit, the best I can do is source 1.2V and pull down to 0V. So my "rewire the original sensor" plan is now dead. Time to replace it - or replace the main PCB (I don't want to go there yet - expensive).

The best options I have seen are Omron EE-SX67, EE-SX77, EE-SX87, EE-SX95, or Panasonic PM25/45/65

I need the sensor to have PNP outputs and accept 24VDC. I would prefer an integral cable. Prices seem to be in the $10-40 range.

None of these sensors have the same form factor, so I will need to machine an adapter. hopefully I can come up with a solution that fits in the same mounting holes without needing to machine anything on the robot.

Edit: one final option is to create a tiny interface board that hooks up to the homing sensors inside the robot and amplifies their output.

Edit: My front runner is the Panasonic PM-F25-P. It is tiny, but I will definitely need to machine an adapter plate. It is 1/3 of the cost of the Omron sensors so I'd like to be able to make it work.

 

[Lo-Fi]

You could probably get away with surface mount components for adaptor boards. It's only signal level, after all. https://www.seeedstudio.io/ or similar are pretty darn cost effective too.

 

[macardoso]

Here is a model of the new sensor (left) alongside the existing sensor (right). I will need to make a tiny adapter to adjust the height and position to get the beam located at the correct spot while using the original mounting holes



Here's a top down view of the sensors. Notice how the new sensor (right) sits further back from the original mounting holes in order to get the beam window in the exact same location.



Here is the sensor mounted with the adapter.



And the actual adapter. This part is TINY. The mounting holes are for M2 screws. I'll need 3 of these. My old concern is if using both of the mounting screws could impact the code wheel. If so, mounting it with only one of the original screws would be acceptable. I'll probably make them from Aluminum.



And a very quick and dirty drawing.

 

[Papa Charlie]

Looks like it would work except for the one mounting hole in the adapter that is partially impaired by the new sensor mounting hole. I would change from the impaired adapter hole to a pin that would protrude into the PCB to maintain alignment. That would also allow for more surface area for the head of the fastener to mount the sensor to the adapter.

 

[hman]

Thanks for the info on the Panasonic! Can't count the number of Omron EE-SX670 series (670, 671, 672, etc.) photosensors I used on machines at HP. They have a nice, narrow beam, and come in a variety of mounting options. Always on a purchase order, so "funny money," and more-or-less unaware of how costs tend to mount up. Good to know there's a more ffordable alternative.

 

[macardoso]

Looks like it would work except for the one mounting hole in the adapter that is partially impaired by the new sensor mounting hole. I would change from the impaired adapter hole to a pin that would protrude into the PCB to maintain alignment. That would also allow for more surface area for the head of the fastener to mount the sensor to the adapter.

Yeah, I'm space constrained and unsure how that will pan out. One option is to use only the original mounting holes and not use the hole thru the new sensor. The sensor would be retained by the overlap of the screw head. Another option is to use only the new mounting hole on the left, and the original on the right. I could improve this my turning a tiny grub screw to fit the original hole and have a turned locating diameter to keep the sensor straight.

I think I can make it work with how it is drawn though.

The adapter bolts directly to the robot body into two tapped holes.

 

[macardoso]

Thanks for the info on the Panasonic! Can't count the number of Omron EE-SX670 series (670, 671, 672, etc.) photosensors I used on machines at HP. They have a nice, narrow beam, and come in a variety of mounting options. Always on a purchase order, so "funny money," and more-or-less unaware of how costs tend to mount up. Good to know there's a more ffordable alternative.

That's pretty neat, do you still work there? Honestly, with how expensive industrial sensors are, these are super cheap by comparison. I really hope these end up working out for me.

 

[hman]

I accepted a "golden handshake" in 2005. There had been several rounds of both voluntary and involuntary RIFs. And though I really enjoyed working there, it appeared that things were slowing down ... and the offer I accepted included HP's continuing their part of health insurance coverage until I got on to Medicare. I still miss a lot of the wonderful folks I worked with.

PS - we had a goodly number of the blue Seiko SCARAs in the calculator assembly area. I didn't work with them directly, but did design and build a part feeder for one of the calculator lines. A Seiko picked the part up from the feeder and placed it onto a PC board. Here's a scan of an old photo I have of one of my part feeders, with the Seiko in the background.