Implementing HV interlock (Zombieverter)

[J0hannes]

Surprised to NOT find much discussion on the topic here.
At least in Finland, the rules/regulations are such that after vehicle modifications, HV interlock needs to be functional, meaning that HV will not be connected if loop is open.

Not sure if this is just too trivial that this hasn't sprouted a discussion, but I'll be the first then

Anyone implemented HV interlock in their conversion?
Should it be a feature in Zombie or would a passive system be"good enough"

Thinking out loud, why would this not work?
Supply the +12V for contactors through the interlock loop directly or as a relay control signal. If interlock is open, there would be a led lit to indicate interlock loop open. As for zombie, this way you'd have just a precharge fail as contactors won't close...

[Bigpie]

Using it to supply the contactors 12v using a relay on the negative contactor output on zombie is how I'd do it, unless you want to have a specific error in zombie for it. For me Precharge failed would be a sufficient error.

[LRBen]

I've been thinking about this in the last few days.

The downside to running it through one of the contactor control circuit is in the unlikely event of something coming loose when driving(Or even just a loose connection in the interlock circuit) you then cut contactors with current flowing, for some inverters that's not ideal.
Obviously if something is coming loose when we are driving then we don't want to be driving for very long.

Maybe we could run it through one of the inputs pins of zombie, either 12v or to ground depending on the input pin. Then have zombie check for that on the precharge cycle. If it happens during driving, either flag an error, or if you want to be aggressive with it force 0 torque. If we are running out of input pins it could be a seperate system sending canbus to zombie on the status of the interlock circuit.

[johu]

I run interlock only in A2. Super minimal. If you open the lid of the Prius inverter you cut power to the controller and thereby the contactor block as well. So no battery voltage apart from the energy stored in the capacitor.
Touran doesn't have a single HV connector, it's all bolted. So no point to interlock.

Seems to be the point of these systems, cut/disallow HV if any connector is intentionally unplugged.

Wouldn't want to run contactor control through it physically, would make me nervous to loose drive because of corrosion or something. Cutting torque mid drive is not a safety measure but a safety hazard in my opinion. But not closing contactors in the first place while stationary seems a safe thing to do. So a pin that software can react to.

[tom91]

I agree with having the option of having a pin assigned to HVIL that does the following:

1. no closing of contactors if signal is faulty (aka not present, so for grounding pins not grounded and for 12V pins not 12V)
2. warning spot value for HVIL OKAY being set low
3. No changing of op modes (so charge to drive would be the only one that Zombie currently does)

Same sort of things can be implemented for loss of isolation resistance, we are getting more and more CAN sources for this info. Leaf BMS, BMW Sbox, VAG contactor box ect.

Forcing torque to zero is for sure a not okay way of dealing with it. Issue becomes trying to identify actual hazards that would warrant a full shutdown, which is nearly impossible to do at a global level and not a project by project method.

[J0hannes]

Ok, now we're getting some discussion Sorry this post got out of hands and is bouncing all over the topic...

Also my thoughts have matured somewhat. A brain fart would be to supply the +12V for contactors through the interlock loop directly all sorts of sketchiness in that I don't want to do that

Just to get this thought through too:
If one would want to have a passive system for HV IL, a latching supply circuit with an open circuit indication could be sufficient, right?
Logic would be such that at vehicle start the HV IL circuit would need to be closed for contactor supply to turn on at IG On.
Once we have contactor power supplied, it won't be cut if for any reason there is a connection issue with HV IL circuit.
Indication led just for the sake that you will have some feedback why pre charge fails in this situation.

Above all possible options, I'm rooting for having this as a VCU (Zombieverter) input pin AND status.

My understanding of the reason of existence for HV interlock is to prevent maintenance/assembly errors resulting in a situation that a HV connector would be exposed. So a pre contactors engagement check should be enough.

If a (HV IL) connection would be cut while driving the VCU shouldn't do anything without additional information i.e. voltage loss message from a component etc. So this gets quickly complex and therefore a minimum effort type approach is justified.

Nissan has only warning lamp lit up, as for example LBC would not give a system start OK if any battery side HV connector is not properly fit
EVB page 23: Detects the connector fit state with the function to detect the fit of the high voltage harness connector and
transfers the detected state to VCM so that the vehicle does not start with an unsteady state.

This snip is from Leaf SM EVC to get an idea what Nissan does for PDM and Inverter bus connection HV IL detection. There are just two microswitches inside the drive stack, one for each module (PDM top lid and inverter bus bar connection cover). Doesn't seem like there's nothing else there and the voltage generator and sensing circuit is in the VCU.

[Zieg]

The SCCA racing rules require the HVIL to immediately open the contactors. They also require the BMS and insulation monitor to immediately open contactors if a fault is detected. I agree with Johu that this can do more harm than good, but for now those are the rules. I did bring that up when they were looking for feedback but never got a response.

On my car, I have a set of 'permissive' NC relays that pass power through to the contactor box, and a signal from BMS or IMD will open the relay and thus open the contactors. Within the box, there is one HV plug (everything else is bolted). The power to the contactors also runs through the HVIL pins on this plug (but on the negative side, which was possible to implement with the dual economizer board I am using).

The other nice thing about this setup is that the permissive relays can be depowered during normal street driving and everything still works. If a fault is detected in the BMS or IMD, the warning light will still come on but it won't blow up my inverter.

[LRBen]

1. no closing of contactors if signal is faulty (aka not present, so for grounding pins not grounded and for 12V pins not 12V)
2. warning spot value for HVIL OKAY being set low
3. No changing of op modes (so charge to drive would be the only one that Zombie currently does)

That makes sense to me.