Sunday, 22 September 2013

Power Supply

To make my Armdroid as portable as possible, decided it would be sensible to make up a dedicated Power Adapter instead of relying on my Bench Power Supply....

I bought a slimline AC Adapter with a regulated rated output of  12V, 5.0A from CPC - I think costing around £17 + VAT.   I could have chosen a 4 Amp adapter, but oddly, was more expensive.

I cut off the 2-pin connector and soldered on the GX-16 style connector, making sure the pins have the correct polarity:

On my Armdroid, +Vcc is on right hand side, GND left, as the connector is viewed plugged in.  Yours might be different, so it's worth double checking by following the cabling in the base to the 7805 power voltage regulator to determine polarity.   The 7805 has the following pin-outs:

The regulator on my Armdroid is bolted to the chassis, yours might be mounted directly on the PCB, but you can identify the +12V /+15V line by tracing the cable to the Input pin.  A multimeter can help with this.

The completed power supply :

Magnet Madness

This morning...  Reinstalled the sensors and cable tied everything back into place.   The spacer supports are not very good, I'll probably replace these with new lengths of aluminum tube and make a better job sometime.

This is rather puzzling.....

First paragraph of the assembly preparation talks about installing magnets into the reduction gear slots, with the exception of the hand gear, which needs no magnet...

... and fair enough, no magnet was installed there on my Armdroid.

The magnets are required for operating reed switches - a magnetic field causes the contacts inside to close, or open if they are normally closed.  When the magnetic field ceases, the reed switch returns to original position, making or breaking the electrical circuit.

But wait.....  What's this reed switch for ?

It's positioned directly in front of the hand reduction gearing in the middle - the very one which has no magnet inside!

That supporting bracket is described as "hand switch bracket" (part no. 16) in the parts listing.

What possible sense is there for having a reed switch without a magnetic field to activate it ?

The controlling computer will never get sensible feedback because it's state will always remain the same.

Perhaps I'll temporarily add a small magnet and experiment later what happens to figure out this discrepancy....   Have I discovered another mistake in the manual/blueprints, or maybe the designer's changed their mind implementing  feedback here?

Any ideas, anybody ?

Saturday, 21 September 2013

Rearming - Part 2

Nothing is particularly easy......

Having reassembled the arm, I soon discovered a serious problem...  The forearm and shoulder joints was completely jammed solid after a couple of turns on the motor pulleys.   It was not until the next day, I could see exactly what the problem was....  The reduction gears had been catching on the motor bolts.

I actually remember swapping these around the other month back to keep consistent with the blueprints:


However, reading the manual, it's clearly printed in black & white "The bolts which are next to the reduction gears should be placed nut out to prevent the reduction gears catching on the end of the bolts."

The blueprints are wrong -  I've been finding plenty of mistakes with this drawings!

So, I had to once again, remove the arm, sensors, timing belts, reduction gearing, swap these bolts around, and start all over again....   hopefully, second time lucky....

Nut & Bolt swapped around as per the instructions :

Start of reassembly...

The reduction gearing was re-installed, and all moving parts lubricated with one of these precision oiler pens:

This time, I double checked every motor turned without any problems before lowering the arm into position.   I have slight play on the shoulder gearing which I suspect is a consequence of the hairline fracture.

Having then reassembled the rest of the Arm, I noticed from the exploded diagrams I'm missing a couple of washers - part number 120.  I really can't face taking this apart again, so I'm going to leave for now....

Gripper cable reattached to the tension spring mounted on the Shoulder Pan.

I also have been checking tension of all cables - left hand side wrist was loose, so a good pull with a pair of pliers and re-tightened grub screws did the trick!   Also cut away excess string leaving roughly 1cm in all cases for future adjustments.

Here are a few pictures of the Armdroid now :-

It's really exciting to see things coming together !

As you can see, just need to reinstall the reed switch sensors, and cable tie all cables back into position.

Turning all motors by hand, all joints appear to be behaving themselves, a little more lubrication will be applied, and hopefully, that completes the mechanics.

Wednesday, 18 September 2013

Rearming - Part 1

Decided it would be really nice to attend this months Raspberry Jam at Bletchley Park and take my Armdroid along to demonstrate....

This will be a tall order with less than two weeks to go, and considering the nature of the venue, we should be controlling all this from a Raspberry Pi !   Until this weekend, very little consideration has been given to this aspect of the project work.

So on that note, we're going to start reassembly...

Installation of motor pulleys - replacements made over the weekend installed into front, and originals at rear, carefully checking for end-float in the motor shafts as we proceed:

I've been applying Thread Lock to all grub screws

The cracked reduction gearing is still an issue... however with a little drop of superglue to reinforce things, this will allow me to reinstall them for now.   I'm still waiting for the new spur gears to arrive, so this is only a temporary measure:

New timing belts roughly placed in order - Medium, Short, Long, Short, and finally Medium.  Although they have a tendency to spring about the place, they can be easily repositioned once the axle is bolted in place.  All moving parts are lubricated with WD-40, but I also have a Precision Oiler pen for hard to reach parts such as bearings.

As you can see, they went in pretty easily after tweaking things about:

Last job for the evening - installing the reed-switch sensors:

Starting with the motor support bar stiffener - this was fitted loosely, and was not tightened until the switch support bar and sensors had been re-fitted.

Further adjustments will definitely be required - one sensor is touching a reduction gear which means friction will rear it's ugly head

Final adjustments will have to wait until the arm is installed !


Here are a few pictures of the Armdroid shoulder so far :-


Tomorrow, the arm will be bolted into place, and that means I can start fine tuning everything.   If all is well, I'll do another bench test, this time posting the video on YouTube.

I've also been busy setting up my Raspberry Pi with a development environment to start writing the controlling software.   I'll update the blog later with more details on that....

Sunday, 15 September 2013

Motor Pulleys - Upper & Fore arm

Today, we're heading back into the workshop to modify the stock 14T pulleys to replace the broken ones on my Armdroid.

These motor pulleys are located at the front of the shoulder assembly and are used to operate the Upper and Fore arm functions.

First problem was how to reduce the overall length, and hold this tightly in the Lathe to perform necessary machining without damaging the pulley:

New - slightly longer which will cause alignment problems without modification

....and this picture shows how we got around that problem

We temporarily bolted the motor pulley onto a 40mm cap-head bolt and secured with a Nyloc.  This allows the workpiece to be installed into the lathe without damaging the pulley flanges.

We also carefully checked the center alignment with a steel rod - necessary for boring out to accept the motor spindles.

After this we started machining away the excess length.  This resulted in a small section which could not be easily machined, but that could be drilled out later!

Now for drilling (after removing the temporary bolt)...

We was a little cautious, and did this in small increments, starting with a 3.8mm drill bit, followed by 3.9mm, and finally 4.0mm.  I wanted these pulleys to be tight fitting onto the motor, and tight is exactly what we got!

Drilling pilot holes for the grub screws:

The mincraft drill is not really suitable for this kind of work, but with a little perseverance and patience we got there!

Finally, tapping both holes to accept M3 grub screws:

This procedure was repeat to make two complete pulley replacements:

Still need a little tidying up, but they are exactly the same as the originals, and almost ready for installation.

Saturday, 7 September 2013

Restringing - Part 3

Final day of restringing....

Using the old string as a guide, cut two new lengths of string for the fingers - one is doubled up, the other single, and both feed into the tension equalizer:

This was the tricky bit...  Clamping the string into place with these grub-screws which had a nasty habit of spinning and pulling on the string...  at that same time, keeping an eye on the lengths to ensure they are even.

I found a steel rod helped keep things in place until tightening.    Ideally, you need two Allen Keys for locking these grub screws - you tighten one clockwise, the other is tightened anticlockwise, and should counteract things.   I only had one, and certainly was a struggle!

With the equalizer in place, I could now make a start on the remaining string which is the grip action cable.

The gripper cable starts from the rear drum (locked by grub screw),  feeding forward, before returning to the idler pulley next to the drum where it all started:

The double ended finger string should be hooked around the smaller pulley, but kept falling off without any tension - will be sorted later

The end of the string will be hooked onto the shoulder spring, so I have simply clipped this to keep in place until then.  Masking tape keeps the reduction gear from rotating, and also keeps excess string out of the way.

Arm ready for installation: