How can I evaluate efficiency of a H-bridge driven DC machine under different control methods? I can measure input voltage and current of H-bridge and output speed and torque produced on the shaft of DC machine (I have torque transducer). I have samples of all variables every 50us.
I want to evaluate actual efficiency during chosen interval. When evaluating efifciency as (output power / input power), zero samples are causing a problem. Should I use moving average filter for all variables in order to avoid zero samples?
Karol,
The good part is that you have the hardware all in place !
All you need to do is to employ identical moving average filters for each of the four signals.
The filters should be so designed that at a time, they average across an entire operating cycle of the H-bridge. For example, if your H-bridge has an operating cycle of 20ms, and you want to average each quantity across 100 samples each cycle, then the MA filter should be averaging across samples of 0.2ms, and at a time it should work with 100 samples.
The filters will have some dynamics of their own, but once these settle down, you should get correct power values to compute the efficiency with. So a steady state condition is essential for the computation.
Hope this helps.
-Sanjay
Hi Karol,
The longer you can wait, the better it goes - as a matter of fact :-)
Recall the filter dynamics I mentioned in my first reply ? They will progressively decay as you wait !!
You can go for higher order MA filters, but I doubt if they will be of additional use in any way !
Cheers !
-Sanjay
- Badges
- Science topic
- Similar topics
- Engineering
- Control Theory