Recent content by wshi

This might help: http://www.diyelectriccar.com/forums/showthread.php/precharge-why-do-need-do-do-25419.html

This method works for most controllers. All you need a resistor (variable resistors are preferred) and optionally a single pole double throw switch (SPDT). You should get the variable resistor with value of 500 ohm (max). Make sure your motor controller power supply for motor hall is 5V or...

I think you should replace both your BLDC and controller. Use a BLDC with hall sensor and hall sensor supported controller this time instead. PS: Being working for controller & e-bike manufacture In the past 3 years I saw a somewhat 40% fault rate with sensor-less controller where as only 5%...

Sensor-less BLDC operation have a nature disadvantage which is the rotor position cannot be obtain easily and correctly, especially in the low RPM range (0~100RPM). This disadvantage usually causes knocking, over current and decreased controller life. The drawback may be minimized with fast...

Few month ago I did some boundary test on our 72V 24FET IRFB4110 controller with EC dynamometer.... Here is one successful record Oct. 3 2009 Room Temp. (begin/end) 31C/88F / 39C/102F Controller Case Temp. 85C/185F (torque is a function of temperature) Continuous Current Draw (begin/end)...

Hello steveo, I am just wondering did you ever do a extreme condition test for the controller? As far as I know most major e-bike manufactures require their controller to pass stall start test, which is a test they deliberately welding a BLDC on the testing base so it can’t spin, then...

No. Although sensorless is a good idea at beginning but I believe it has fundamental flaw when dealing with heavy loads and quality variance. None of our controller is sensorless. We have produced some controller using TO-247 / TO-264 and they are custom build for e-motorcycle clients.

The most recent controller are based on Renesas R8C series. Controller is configurable via serial connection and able to accept control command and output real time status/diagnostic data. Come to FET we have used STP75NF75, STP140NF75, IRFB(P)4110, IRFB4568, IRLB4030 etc. For standard copper...

Currently we only sell our product via OEM channels, but I am also open to new market and potentials. Most e-bike controllers today are feature loaded and poorly built, while my focus is bring the highest stability and quality with relatively reasonable cost. The regeneration break is built-in(2...

Knuckles, Good guess and thanks :) I come from Beijing, China and I speak Mandarin. It would be my pleasure to talk to you and your friend. -wshi

Have you looked at Google code project hosting? It's pretty cool and support SVN. http://code.google.com/opensource/

There are plenty of fake Rubycon in Chinese market (my gust feeling tell me >90% market share). Even the real Rubycon have their low temperature issue, so after all my choice is Chemi-con.

Knuckles, My pleasure! You are right I am a EE at manufacture. I started e-bike controller project from scratch 3 years ago, design and build from bottom to top, burning a lot of FETs :) Now I am living in Canada, got my dream job as computer software engineer. I am still partnered with motor...

Actually there was a labeling mistake in my previous circuit diagram. The U1 "KSA1275" actually should be 2SA1275Y which is the one I use most in production. KSA1220Y can also be used for U1 if you can't find 2SA1275Y

I agree with your analysis. Here I am posting my buck step-down power supply, which should be very similar to XieChang's design. This buck step-down design proved itself to be very reliable in past 2 years' mass production.