Admin Group
Joined: 05/06/2004
Location: AustraliaPosts: 5186
| Posted: 07:28am 31 May 2011 |
Hi Guys
I'm starting a new thread to talk about a Maximite project I'll be working on. Actually I have a few Maximite projects on the go, but this thread will be about the power plotter.
As I mentioned in the Maximite thread, I'm putting together a alternator power to RPM plotter. The idea is I use the Maximite to quickly plot a power/rpm curve on the screen, and save the figures in a excel file on the SD card. I've spent the last week playing around with the Maximite code and a interface to find a combination that give good figures on the screen. At first I was using an old servo motor and battery as the test subject, but it was a bugger of a thing to test with, the electric drill struggled to load it up correctly and give a nice range of RPM's.
So I ditched that and put together a smaller test setup consisting of a stepper motor with crank handle, a couple of bridge rectifiers and resistors as the load/shunt. The stepper motor generates 2 phase AC, a few volts, enough for testing.
One of the problems I was having was erratic plots, the watts readings were very scattered. The software averages the figures for the RPM readings, so after cranking the stepper up and down a few times the plot would smooth out, but I wanted to see why it was so erratic in the first place. Me trusty old cro showed the problem. Bottom trace is the RPM input to the Maximite, I'm using the same clipping circuit I used on the PicLog, and as you can see it gives a nice clean stream of pulses. The top trace is the DC fed into one of the Maximites analogue inputs. Centerline is 0 volts, 1volt per division. A lot of AC hum. The Maximite is very fast at reading its analogue inputs, so it sees a voltage jumping up and down, explaining the erratic plot.
I did try some software averaging to smooth out the readings, but found a couple of 10uF caps across the inputs made all the difference. I do wish I could sync my camera with the cro. Trying to get a shot without the retrace is difficult.
The program now reads the data in and plots a power curve on the screen. I'm using the Maximites period input to work out RPM, the frequency input only updates every second so thats a bit slow for this job. One little problem is the period input reads to milliseconds, meaning for a alternator with lots of magnets like a F&P, the RPM resolution spreads out noticably at higher RPM. For example, a 5ms period works out to 461rpm, a 6ms period is a big jump to 384rpm. Once we get into the lower RPM,s, the readings are much closer, ie 10ms = 230rpm, 11ms = 209rpm. For lower magnet count alternators its not a problem. I'll use some maths to fill in the RPM gaps anyway.
So thats where its at. With the little problems sorted, next its on to making it all look pretty on the screen and give some meaningfull data.
Glenn
The best time to plant a tree was twenty years ago, the second best time is right now.
JAQ