DIY Robotic Tester - BoraTronic

[Borat]

It appears that MT uses 2 optical IR sensors and cutouts in the carousel plastic to encode the position of which vial is under the home position.

Super, thanks for this info!

I have still not decided whether to use optilcal sensors (which would encode start/end position) or to use RGB sensor which can encode the same informtion but possibly in a more granular way...

For the rig moving moving parts - it will be optical switches (definitely), the one thing I am a bit hesitant is whether to have 2 switches (start/end) or just 1 end switch.. I don't think I need 2 switches for each axis, but I look forward to your comments and will decide in due time..

 

[Borat]

interesting design! Do you need a special "dark box" for the color detection or any non-trasparent plastic will do?

I have never used the RGB sensor before... As I understood - it has a built-in wide spectrum LED (warm white) and it detects/measures reflected light. I will be in a position to know once I tried it..

 

[Borat]

Putting together syringe actuator...

about 50% done..

 

[Badfish405]

Very nice. You make me feel really dumb.

 

[Borat]

Very nice. You make me feel really dumb.

Nah - if I look back into my DIY projects 2 years ago, I look really dumb! There is no shame in trying, there is shame in not doing that!

 

[geddavis]

interesting design! Do you need a special "dark box" for the color detection or any non-trasparent plastic will do?

I am going to make a case out of black acrylic for more accurate results but so far testing it just in normal light it is reasonably accurate. The sensor is fairly deep in a 3D printed housing so it does not interact too much with light anyways.

 

[Borat]

Actuator in progress, 75% complete:

 

[Borat]

 

[Borat]

Actuator is 99% ready (need to drill holes for brass nut screw):

 

[Borat]

I did a full assembly/disassembly of the actuator parts (everything can be dis-assembled into components), polished some of them, tightened screws/balanced their alighnment.. Also added screws for brass nut screw to keep it firmly in place..

 

[Ben Pedersen]

Really amazing.... I am a software developer and have a good mind for engineering, however, I do not have the tools to machine parts.. and 3D printing quality is low...

Seems that your go to material is acrylic.. What do you use to cut, refine and polish?

Another challenge is finding parts... especially because I don't know the actual names for many parts.

 

[Borat]

Really amazing.... I am a software developer and have a good mind for engineering, however, I do not have the tools to machine parts.. and 3D printing quality is low...

Seems that your go to material is acrylic.. What do you use to cut, refine and polish?

Another challenge is finding parts... especially because I don't know the actual names for many parts.

I face the same challenge of not knowing the names of parts - but after repeated googling I eventualy find an image of a the part I was looking for, from which I learn the proper name for it.. Then I google the proper name - and find a selection of parts that I was looking for..
As for acrylic - I do everything by hand: fine toothed hand saw (9 teeth per inch), drill bits, rotary sanding belt using on simple electirc drill (and rotary grinder I also have) and metal files - is what I use to cut/polish parts..

 

[vahegan]

I have never used the RGB sensor before... As I understood - it has a built-in wide spectrum LED (warm white) and it detects/measures reflected light. I will be in a position to know once I tried it..

I think you'll be much better off if you measure the light that passed through the sample rather than reflected light. I.e. not use the onboard LED, but place another white LED on the opposite side of the cuvette.

These are the optometric cuvettes I was talking about. Two opposite walls are made of thin optical glass. This is the 10x10mm size, but there are many other sizes available. I think this is much better than the round vials as used by Hanna.

 

[geddavis]

I think you'll be much better off if you measure the light that passed through the sample rather than reflected light. I.e. not use the onboard LED, but place another white LED on the opposite side of the cuvette.

These are the optometric cuvettes I was talking about. Two opposite walls are made of thin optical glass. This is the 10x10mm size, but there are many other sizes available. I think this is much better than the round vials as used by Hanna.

I am currently shining a light through the vial and then using the adafruit rgb sensor on the other side. So far this has given me very good accuracy and reliability when the sensor and led is kept at the same position in a dark housing.

 

[geddavis]

Really amazing.... I am a software developer and have a good mind for engineering, however, I do not have the tools to machine parts.. and 3D printing quality is low...

Seems that your go to material is acrylic.. What do you use to cut, refine and polish?

Another challenge is finding parts... especially because I don't know the actual names for many parts.

Personally I use McMaster car part for all my parts. Its a bit expensive but they have everything you need and it’s pretty easy to find things even without a name just by sorting through by category.

 

[Borat]

I am currently shining a light through the vial and then using the adafruit rgb sensor on the other side. So far this has given me very good accuracy and reliability when the sensor and led is kept at the same position in a dark housing.

This is very helpful! I will incorporate PWM LED signal for LED when designing connector for RGB sensor (I still have not sent the boards for printing).

 

[Borat]

There is a bit of a delay to the project - just didn't have enough time to progress.

My next step is to build flexible syringe mounts on the actuator and complete that part of assembly..

 

[Borat]

Personally I use McMaster car part for all my parts. Its a bit expensive but they have everything you need and it’s pretty easy to find things even without a name just by sorting through by category.

If you visit any ebay CNC shop - it would most likely have any parts you desire. Unfortunately McMaster is a US&A shop - and not availabe here in the United Kazakistan (aka UK). But I apprecite your feedback - and hopefuly that's helpful for others (should they decide to build their own robotic tester)!

Cheers,

V.

 

[Borat]

I think you'll be much better off if you measure the light that passed through the sample rather than reflected light. I.e. not use the onboard LED, but place another white LED on the opposite side of the cuvette.

These are the optometric cuvettes I was talking about. Two opposite walls are made of thin optical glass. This is the 10x10mm size, but there are many other sizes available. I think this is much better than the round vials as used by Hanna.

Thank you!

 

[Minhaj Qazi]

If you visit any ebay CNC shop - it would most likely have any parts you desire. Unfortunately McMaster is a US&A shop - and not availabe here in the United Kazakistan (aka UK). But I apprecite your feedback - and hopefuly that's helpful for others (should they decide to build their own robotic tester)!

Cheers,

V.

Waiting for end Result. May be u get an Order from Pak.