What main contactor to use for AC system with regen?
 

I have searched around and i am finding it difficult to select a main contactor.

The main concern i have is that most contactors seem to have polarity preffernces. The main contacts being labelled with + and - or A1 and A2. Because the current can flow in either direction in an AC setup with regen i don't want to damage the main contactor.

Several data sheets for contactors mention things like magnetic blowouts which i have heard don't like reversed current flow.

Anybody with an AC motor setup like to share their contactor experiences?

Vacuum contactors work fine for both AC and DC. They do not use magnetic blowouts. They are a bit more expensive than a DC contactor. And of course you'd need 3 phases instead of 2.

I use a DC contactor from the battery pack to the controller. If you shut off the power to the controller, the motor (AC or DC) will coast (or the AC can regen, depending on how your controller is set up).

An AC contactor (perhaps with an overload block) would go between the controller and motor. Are you using the contactor for reversing?

The main contactor should be on the battery pack and there for should be a DC contactor, non that I have seen care about positive or negative and work fine for regen.

I use an Albright Contactor.

http://www.albrightinternational.com...ct%20Range.pdf

I had a look at the Albright site. They talk about polarity sensitivity and magnetic blowouts.

Blowouts enable the contacts to switch D.C. voltages of 48v or higher. Fitting of blowouts usually makes the contacts polarity
sensitive and where this is the case the positive markings ‘+’ on the top cover of the contactor must be observed. Magnetic Blowouts
can be used with certain ‘P’ type switches, but we do not advise them suitable for frequent current switching applications. The suffix
’B’ denotes the fitting of magnetic blowouts.
Note:
The SW200 series is fitted with magnetic blowouts as standard therefore the option is to not have blowouts, this is indicated by
the suffix ‘N’. Also relevant to the SW200 range is the suffix ‘R’ which denotes non polarity sensitive blowouts are fitted.

So if i read that correctly they put in the magnetic blowouts to help deal with voltages over 48v. This normally makes the contact polarity sensitive unless non polarity sensitive magnetic blowouts are used.

I am not using the contactor for reversing, it is just for isolating the battery pack from the controller, DC-DC converter etc.
This contactor will see current from the battery pack to the controller and when in regen will see current from controller to battery pack. This is why i was worried about contactors that had polarity markings on the case for the main contacts.
The Vacuum contactors you mentioned, would these be like the contactors from Kilovac?
I was looking at the Kilovac contactors on eBay and in the photos they have polarity marking on the main contacts.

Contactors aren't cheap and i would hate to get the wrong type.

The power between the batteries and controller is always DC, even on an AC system. The polarity sensitive Albright will work.
Regen current will never reach normal current levels. Besides, AC systems do not typically fail in the same manner as DC (short circuited) where blowouts are necessary.

Per Mechman600's post, the battery bank is isolated from the controller by a DC contactor (I would add 'or a vacuum contactor').

The polarity is an issue when you are trying to break the flow of current, like during a short circuit of the battery or when the power electronics on the controller fails.

For a main contactor, you connect the + battery cable to the + on the contactor and it's hard wired that way, polarity doesn't change so a polarity sensitive contactor will work just fine.

I think Astro is wondering about the change of polarity when in regen, when current is flowing back to the pack, and how this affects the polarity of the contactor's blowouts.

This is a non-issue because contactor blowouts are meant to break high amounts of DC current in the event of a DC controller failure, as stated above. However, unlike a DC controller, an AC controller is highly unlikely to fail in a closed/shorted state, making a contactor with blowouts unnecessary.

I think the confusion here is thinking polarity changes while regen.
The polarity (ie. direction of flow) does not reverse during regen. It continues to flow around the circuit in the same direction. Only the motor is now pushing - to the battery instead of pulling from the battery + .

Hope this clears things up.

I was indeed thinking that the direction of current flow reversed during regen. The motor and controller basically acting like a charger.

I had a quick look at this site Understanding Regeneration | Roboteq and it says that current flow reverses during regeneration.

Even though the amount of current during regen isn't as high as the currents when the motor/controller are acting as a load. I was still worried that the potential across the contactor terminals would be the reverse of what was marked on it.

But from reading the other comments is sounds like with the contacts closed the reversed current flow won't cause any issue.
It is only if the contacts try to open whilst high regen currents are flowing that the reversed flow may cause issue.
In normal operations, when high currents are flowing the main contactor should never need to open so being in regen mode or not makes no difference.

Take your time deciding on a contactor. I just lost around 16 hours of work (over the last 2 days) on the PCB, because it got corrupted by DesignSpark! They are working on it, to see if it can be recovered. Most of the time was spent on the layout for the driver section. Well, I finally decided on a good driver section layout. They can't erase it from my mind! The schematic part is done now. So, it's just laying it out and routing it again, and hoping it doesn't happen again. I reinstalled designSpark. Hopefully that will help.

One nice thing is, if there is a short circuit condition on any of the igbt half-bridges, the offending hardware overcurrent will turn off all high and low side IGBTs in a couple microseconds. A separate desaturation detection circuit that's on each igbt (6 in total) will also turn off all igbts. Also, the software will stop making sine waves in less than a microsecond if any of those faults happen. If there's a failure of igbts so you get a dead short to ground, there should be a fuse that blows quickly. I can't think of a situation where regen will be a problem with the contactor. Especially, since the controller would have control of the contactor, and would know if regen was happening anyway, because of the current feedback.

Ouch, having to do all that rework is horrible.
Best to think of it as a chance to improve upon what you lost. :)
Not sure how to put a positive spin on the lost 16 hours tho.

I've finally got a new layout that is BETTER than before! Hurray! It was worth losing the other one. Well, they are similar to each other, but this one is a tiny bit better. The driver section symmetry is so glorious I can hardly stand it. haha. I want to tell it to stop being beautiful, but it just won't listen!