1
\$\begingroup\$

I need to make a power switch (on the very cheap) for an electric vehicle.

The battery bank is a 8s (29.6 Vnom) LiPo pack, powering an ESC which drives the motors.

Max continuous current to be drawn through the switch is approx 120 A. Peaks can be expected at about 180 A or so, but very infrequent/hardly at all.

The constraints are that the circuit must be powered from the battery pack, and not draw much/any current when in the off state. There are also some dimensional constraints (ie that the entire solution shouldn't be bigger than like 50mm x 50mm x 20mm or so).

I would imagine that I would need a mechanically actuated switch at the rated voltage, that enables a contacter of some sort (mechanical/electrical/solid state).

I've been doing some research, but I need a little advice as to what I should go with for a super cheap solution.

So far I've come across:

Thyristor circuits

MOSFET switching circuits

Solid state relays

Mechanical relays

Some of these might require a DC/DC converter or something to get the voltage down to useful levels to engage the switching circuitry.

Does anyone have any other ideas and could someone suggest what the best (cheapest) option would be to go for here?

Thanks

\$\endgroup\$
  • \$\begingroup\$ Is this for safety isolation, or just turning it off when not in use? \$\endgroup\$ – pjc50 Mar 23 '16 at 12:30
  • \$\begingroup\$ Does the ESC have any "standby" facility that you might have overlooked? \$\endgroup\$ – Andy aka Mar 23 '16 at 13:01
  • \$\begingroup\$ @pjc50 It's mostly for just turning it off when not in use. I have to unplug the battery to charge it, so that is my safety isolation. I will probably leave it (not completely unattended) in the off state, but I'll probably include a fuse to make it a bit safer. \$\endgroup\$ – makepeace Mar 23 '16 at 13:04
  • \$\begingroup\$ @Andyaka My ESC doesn't have a standby facility. It drew my batteries to 0.5 V per cell when I left it plugged in for a day, which was a sad day. I managed to recover 7 of the 8 cells though. \$\endgroup\$ – makepeace Mar 23 '16 at 13:05
  • \$\begingroup\$ No under-voltage lock-out threshold? \$\endgroup\$ – Andy aka Mar 23 '16 at 13:07
1
\$\begingroup\$

You only have two practical solutions

a) A mechanical switch

b) MOSFETs

Thyristors and SSRs are out due to their minimum 'on' voltage which would create too much heat to dissipate in your required volume.

... and if it's for system isolation, only (a) will do.

You'll note that (b) is plural. For super cheap (relatively), I expect you would be using multiple TO-220 type packages in parallel. Fortunately, MOSFETs share nicely when in parallel and fully on. With sufficiently low RDSon, you could manage the heat produced by 180A within your volume.

The industry is producing many excellent new FETs with VDS in the 70-100v region for 42v car electrics, these would be a perfect fit.

To hold them on, FETs require no power, which would allow you to design the drive circuit fairly economically. If you switch the negative rail, so have the positive available for the gate supply, you may need little more than a potential divider to stay within the max gate voltage.

Your drive may need to switch them quickly, to avoid excess dissipation if switching on or off under load.

schematic

simulate this circuit – Schematic created using CircuitLab

The above might give food for thought. R1 and R2 pot the battery voltage down to 50%. The gate voltage range must be correct for the FET, usually >10v for best RDSon, <20v for gate breakdown, though check the specific FETs you end up using. LiPos should keep the terminal voltage within a 2:1 range. If you are fussy, you might reduce the value of R1 and parallel a zener with R2. Put a LED in series with your zener, or use several LEDs in series as your zener is an alternative.

Q1 and Q2 speed up charging/discharging the gate for on-load switching, omit one or both if this will never happen. I've shown a mechanical on/off switch, though electronic control is easy enough to arrange.

Switching the negative rail like this means that if you routinely ground the negative terminal of things, the battery pack and the powered electronics cannot share the same ground. There is the scope to get things wrong.

A high side switch would require some sort of voltage booster to operate the gates.

\$\endgroup\$
  • \$\begingroup\$ Thanks very much @Neil_UK. The MOSFET(s) option looks like the perfect solution. I won't be switching under extreme loads - probably a 2 A at the maximum. Could you perhaps suggest a circuit? I don't mind designing one myself, but it would be nice to have something to work from. \$\endgroup\$ – makepeace Mar 23 '16 at 13:06
  • \$\begingroup\$ I think you need to design for the possibility of switching under high load, in case of faults; you don't want exploding MOSFETs to add to your fault. PS there are various electric car design forums that might be useful: diyelectriccar.com/forums \$\endgroup\$ – pjc50 Mar 23 '16 at 13:19

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.