Thursday 16 February 2017

Weighing for Stall Torque

So a while back, Alastair and I, made a watering system for the greenhouse. It was fantastic. For the three days we used it.

It was in fact, the end of Mine and Louise's greenhouse cycle as we'd got the allotment late and had nothing left to grow, however in the last two weeks we've started planting out cauliflower, red onions and parsnips in the spare bedroom so will shortly need to use the system again.

So I called back the technical support of Alastair, and alas during testing (We're well disciplined technicians) the closing of the tap by the servo was failing. It appeared that the servo was struggling on its return either because the tap was stiff or there wasn't enough power.

Alastair - being the Einstein he is - suggested we tested the force the servo would need in order to turn the tap. So our first test was to thread string around the handle roughly 1cm from the pivot point. We then hung a plastic cup from this and slowly added water.

Test 1: Cup on a string and add water
********************
Lesson 1: If you use a knife to put holes in your cup and then add water. The water will eventually reach the holes and pour all over your counter.
********************
The water wasn't heavy enough to move the tap at all, and in fact didn't make turning it any easier.
Let's move on. 

The next thing to try was a random ball of lead that weighed roughly 600g. This still was unable to turn the tap on its own. However once we added a hammer to the top it quickly turned. Great? Not quite. So we now weren't seeing the effective weight power we needed to turn the tap because it was imprecise between the weight of the lead, and the weight of the lead and the hammer. 


********************
Lesson 2: Do not put lead on the scales you use to weigh your food. Lead poisoning is a thing.
********************

Additional to our problems, attaching string to a tap in the way shown below, has added a cm to the distance from pivot making the maths more difficult and the whole thing a bit wild.

Knots are tricky

Eventually we managed to find some scales to make it all easier, some old style cooking scales.

We removed the top, held the tap steady and pushed down on the metal point of the scale until the tap slowly began to turn. We then measured the constant weight shown as the tap was turned. To add accuracy we repeated the test until we had some averages.

We found we needed around 820g of weight to turn the tap. We had been using the Tower Pro SG90 Mini Servo, 37-1330 which hasStall torque - 1.8kg/cm (4.8V). Specs.

What's that I hear you ask? What is stall torque? It's a good question. At this point I had to ask Alastair what stall torque was and more importantly. Why we cared.

 ********************
Lesson 3: Find out why you are doing things before you do them.
********************
So, "Stall torque is the torque produced by a mechanical device whose output rotational speed is zero. It may also mean the torque load that causes the output rotational speed of a device to become zero, i.e., to cause stalling." Thanks Google!

So we were finding out how much stall torque power we needed for a 5v servo (4.8v) in order to make the tap turn. The answer was 820g to make it go down/open. However in reverse, we had to push it up with other factors such as going against gravity, water flowing etc. we would need additional power and force. Evidently, since the current servo was struggling to do this, it needed more than 1.8kg of stall torque to close the tap.

So we decided to just up the ante and use a more powerful servo. Which we had planned to do when we first saw it wasn't working.

To conclude, it was an interesting and exciting waste of time which I thoroughly enjoyed.