Forums | Radical Creation

As I am progressing forward with my longboard build, I am finding more and more areas to improve. One of these areas is the power switching system. A typical electric vehicle of any kind will have an anti spark switch. This is a solid state circuit that will slowly provide current to the motor controller to gracefully and safely charge the capacitors. After a short time, the circuit will provide full current to the controllers up to the rating of the circuit. These also come with a mechanical latching switch used to activate the circuit. They also typically have built in fuses or other safety limiter in case of an over current draw.

The anti spark switch I got was from Alien Power System and is rated for 120 Amps and up to 14s. I will be pulling up to 100 amps continuously at 10s voltage. The area I am looking to improve is the activation bit currently handled by the mechanical switch. This switch design is not ideal for the longboard application due to vibration and shock. The latching mechanism is known to break down and even unlatch while riding. While this can be replaced with a heavy duty rocker switch, it can still be knocked off while riding.

The best solution I have found is a soft latching switch circuit. This utilizes a momentary switch to latch a logic signal high or low. A slightly modified version of this circuit will yield a push on and hold off configuration. This is ideal because it is the most reliable and most robust of all the options.

I put together a quick circuit on a breadboard and interfaced it with a mosfet. I will post a full schematic once I finish it. Anyway here is a video of the working circuit.

youtu.be/UR5QIhaAFDc

It seems to work perfectly and I will next look to reduce the size of the circuit by making it on perforated board and possibly a custom pcb.

Updates to come!

I drew up a rough schematic of the circuit as shown below:

The next step is finding components with proper ratings. The main two to note are the mosfet and the voltage regulator. The mosfet has to be able to switch at least 60v at low amperage while the voltage regulator has to regulate to the maximum value of the mosfet gate. Alternatively, a regulator could be used with a voltage divider if it regulates to higher than the desired gate voltage.

That's it for now,
Updates to come once I find proper components.

While testing the existing anti spark switch on its own, I experienced a failure. The mosfet circuit seems to have gotten stuck shut and will no longer respond to switch presses. Due to this newly discovered unreliability, I will be making my own anti-spark portion of this circuit and have it pcb'd as one board. Here is the updated circuit:


Updates to come!

I have been testing out the Vedder-Fetcher style anti-spark switch and although it has worked longer than the other cheap switch, I am starting to have problems with it. The switch is not broken (Stuck on) but after rides where lots of current was pulled, turning the mechanical switch to the off position does not turn the anti-spark off. I have found that switching it on and off several times in succession is the only way to get it to turn off. Also, when it turns on after being off for a while, the voltage on the output slowly climbs to the battery voltage over about 20 or 30 seconds. I am now trying to isolate the cause of this abnormal behavior to understand it and fix it before implementing into my own design.
Browsing on endless-sphere, I saw Fetcher posted this:

fechter wrote:That circuit has a track record of blowing FETs. If your controller caps are small and the voltage isn't too high, it might survive. I think it would actually do better without the capacitor. Seeing how the riba spot welder handles turn on surges of thousands of amps, a fast turn-on can work. I'd also add a 12v-15v zener across the gate to source as a transient protection.

I may try building a test circuit with the added zener diode for added surge protection. I have also started a topic on Vedder's site about the issue and I look forward to reading his response. I will have to work out this issue before PCBing my own board.

Update to come.