The current issue of Circuit Cellar has this article on a DIY inverter that is about $6000 US less then commercial units. The link may go away when the next issue comes out but Microchip should have it online somewhere.

Circuit Cellar
Issue #217 August 2008

Subcategory Winner - Microchip 2007 Design Contest
INTELLIGENT ENERGY SOLUTIONS
Electric Vehicle Inverter Design
Build A System For Powering AC Induction Motors

__________________
search engine for electronic parts
Junebug USB PIC programmer kit., USB Bit Wacker, 3v0's Homepage
The 15 Minute Printed Circuit Board! (+drill time)

 

intresting project although not for a beginner

as they adress in the article there are still some safety questions that makes this only suitible for a closed circuit if it is used in a vehicel

verry intresting

Robert-jan

 

In the field of alternate energy there is little useful at the beginner level.

__________________
search engine for electronic parts
Junebug USB PIC programmer kit., USB Bit Wacker, 3v0's Homepage
The 15 Minute Printed Circuit Board! (+drill time)

 

Let's take a few simple calculation about drive AC motor from battery:
1.If we use a battery source (12 or 24 V), to obtain 310V@1000W:
Secondary current = 1000/310 ~3.5A
Primary current = 1000/24~41,6A -> this is very serious converter -> have a price more then 2000$!
2.AC motor is driven from vector or frequency source and second step is to design a power stage, to control the motor.
If we choose a voltage/frequency control of induction motor, the price of converter will be smaller then same vector control unit,but control performance of drive is poor (bad respond of dynamic torque).In other way, if we start to design a vector control, the performance of drive system will be excellent but system stay more complicated.
3.When start to design vector control system we have some pure mathematic operations:
-Clark/Park forward and inverse calculation;
-α-β -> d-q forward and inverse calculation;
- PI (or PID) loop calc.
- dynamic motor model calc to obtain torque and flux, compare with ref. values
-SVM calc
All operations take a time-> we needs of high speed F.P controller with fast ADC and capture modules.Time for calculation can be decreased if we use a space-state method and all equations are decided in matrix form.
4.Have much underwater stones when start to design a system in practice

In my opinion,price of system,which supply and control motor from battery is much expencive.dsPIC is not suitable decision for that system because is low speed and have a small memory.

 

@Bojo: You really should read the magazine article. You have made several errors in your assumptions. The authors actually constructed the controller.

 

I am still trying to figure out the vector source thing. Is not a frequency source a vector source, as vector implies an entity defined by both magnitude and direction?

Maybe somone can enlighten me...

__________________
Pay it forward.

 

Hello everyone! I am new on this forum. I have looked at the documentation on the Circuit Cellar contest. What I recollect reading in the final abstract was, that while they were able to prove the power module/driver unit and the drive electronics, they were not able to run it as a vs inverter unit.

The abstract blamed noise as the suspected problem. From what I saw in the pictures, they didn't use snubber circuits on the IGBT's and it appeared that the gate drive signals may have been run in a multiconductor shielded cable. It is really hard to tell from the pictures, but twisted pair cables (unshielded) is the industry standard. The reason being that there will be switching noise, and as long as it is of the same amplitude and phase on each wire, the noise is effetively nullified.