Commit 1bb8fe20 authored by David Preiss's avatar David Preiss
Browse files

Update README.md

parent 3e2cc054
......@@ -2,15 +2,25 @@
## Week 5 - 4/1/21
This week I took a first stab at one of the initial motivations for building the dynamometer, but that I've neglected so far. That is looking at stepper motor torque limits as we approach saturation and thermal limits for the motor. The plot below illustrates this by sweeping through increasing current chopping setpoints (remember that open loop stepper motors are driven as constant current devices). What we see is a surprisingly linear response in peak holding torque (keep in mind these tests are conducted at 0 rpm).
This week I took a first stab at one of the initial motivations for building the dynamometer. That is looking at stepper motor torque bounds as we approach saturation and thermal limits for the motor. The plot below illustrates this by sweeping through increasing current chopping setpoints (keeping in mind that open loop stepper motors are driven as constant current devices). What we see is a very linear response in peak holding torque (all of these sweeps were conducted at 0 rpm). The exception to this occurs at 3000mA, where we see what looks like a dramatic reduction in peak torque. In reality this is the power supply running my absorber (the brushed DC motor) hitting its current limit (ie. the test motor does not stall here). So with that said I am publishing this very wrong data early (but such is the nature of hitting class deadlines). So I have no reason to suspect that this motor doesn't continue off linearly into the 4A or 5A regime, which would pose a lot of interesting questions as discussed below.
![currentSweep.png](./images/dyno/currentSweep.png)
[Here's a link to the stepper that I am testing.](https://www.omc-stepperonline.com/nema-23-bipolar-1-8deg-1-9nm-269oz-in-2-8a-3-2v-57x57x76mm-4-wires.html) It's one of the more commonly used NEMA23 steppers used for DIY applications, rated for 1.9Nm at 2800mA. Fortunately this lines up extremelyt well with our own tests! At 2500mA we acheive a peak torque of 195N*cm. Below is an image of stepperonline's speed torque curve (you can disregard the 36V vs 48V difference as this has minimal impact at 0 RPM).
![23HS30_2804S.png](./images/dyno/23HS30_2804S.png)
So this plot gets at the question of, is power consumption for NEMA-packaged steppers thermally or saturation limited, and how close are we getting to saturation limits? (I think they're almost certainly thermally limited). The first thing to look for is if we can see the effects of saturation happening before thermal limitations kick in. There's a neat [paper from the biomimetics lab](https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7827048) that lays this question out in the context of BLDCs (they also discuss the possibility of hitting a curie temp limit, but that typically happens well past saturation).
If we are very far from saturation then there's a good argument to make for having drivers that tap into more peak current headroom that would otherwise burn up the motor in steady state (especially for steppers with high inductance when you aren't in the peak ~3A regime of most drivers anyways) that would otherwise go unused. Ultimately this gets at trying to level the driver playing field between "variable reluctance hybrid motors" (steppers) and "3-phase PM motors" (BLDCs), and then get definitive numbers for low RPM torque output, because I haven't seen anyone publish data along those lines. So the question boils down to the tradeoffs between a motor with high pole count + low reluctance flux path (again a stepper) vs. everything BLDCs have going for them: external rotors + way more PM material + better cooling, etc.
It's very likely clearpath and all of the closed loop stepper companies do this already, but depending on how this data pans out, there's a neat case to make for taking cheap off the shelf steppers and driving them with parameters very tuned to their performance (on the boundary of melting windings) via the dyno, and seeing how much more performance you get.
Next steps to look into here are:
* Recalibrate my load cell as I've moved a lot of things around since its initial calibration (despite the fact that my data seems to line up well with the vendors).
* Add a thermocouple to the logging setup such that we can also get measurements getting at thermal limits.
* Switch over to a weaker NEMA17 form factor for the test motor, such that I have ample absorber torque to make sure I can defeat the test motor at any reasonable holding current.
## Week 4 - 3/25/21
This week I focused on improving data analysis (see the nicely interpolated speed-torque curves below), speeding up test times, and slogging through the endless process of finding bugs and improving repeatability.
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment