Controller Isolation?

[jkrienert]

Hello!

I'm working on isolating the HV of a Numi (tm ?) 840 scooter, rear-driven by a 1500w constant 72v 3-phase hub. It originally had a to 12v Dc/dc with common ground (frame) to the battery 72v-. I opened its generic manuf. BLDC controller rated at 72v 40A batt. max to check isolation there. It also utilizes a 72v- common ground for the 12v accessories (turn signals, lights, etc), and 5v controls (instrument panel, throttle).

I've already redesigned the battery (+fused 73.6vNom 32s2p LTO with O.G. Orion BMS) with contactors, and swapped the common Dc/dc for one that is galvanically isolated. I use a 12v accessory supply (power tool battery) to soft (resistor) open the contactors to power the Dcdc and controller. The 12v system incorporates a HVIL with several safeties (inertia switch, service door switch, kickstand switch).

Can the controller also be (galvanically) isolated? Is an isolated controller superfluous since the HV traction pack is already isolated? I mainly worry about shock hazards and negative impacts to sensitive components if any ground loops would emerge... Maybe some other wiring aspects I could consider or evaluate?

Can provide photos or more description if needed. Thanks.
SEE photos of controller here:

48-72 Regen capable 48-72v controller?

Does the controller shown in the attached images support Regen? This unit was originally installed on a Numi Elite Quadhopper 840 that was manufactured in 2015. I bought this EV used in a debilitated state (dead SLA). This vehicles weight and aero, along with the hilly terrain it will be used...

endless-sphere.com

EDITED: For simplicity.
EDITED AGAIN: To link photos of controller.

 

[fechter]

The controller case is most likely already isolated. You can test this with an ohmmeter by measuring between the case and the battery negative wire. You might see a little stray leakage resistance but nothing close to zero. It wouldn't hurt anything if the case was electrically insulated from the frame as long as it gets adequate cooling.

 

[jkrienert]

You might see a little stray leakage resistance but nothing close to zero.

Any ballpark for what a little leakage resistance range might be? I'll measure as soon as I finish some other parts of the rebuild, but it might be a while till I can directly answer this.

Noting the board photo on the included link, if the 72v- uses the same board trace as the instrument and light wirings common ground (5v and 12v), would that establish a half connection to the pack and raise the risk of short, as well as generate ground loops with the other 12v starter battery?

 

[fechter]

Typically the leakage resistance will be over 1Mohm, but anything over 100k is nothing to worry about.

Pretty much all controllers are going to have the B- connected to the ground for the low voltage stuff. It's not a problem. To avoid any ground loop issues, avoid any redundant ground connections, especially on the display and throttle wires. Every ground wire should have a single trace back to the main B- connection.

 

[jkrienert]

Great tips for working this out - thanks @fechter