Relay to disconnect large inverter (Will's latest video)

[markb]

I was watching @Will Prowse's latest video (

), and I wanted to start a discussion on it here. (I can't stand YouTube's comment system.)

My concern is that relying exclusively on the relay to protect the battery from over-discharge might not be a great idea. When a relay fails, it usually fails open, but it's possible (albeit less common) that it can spot weld itself closed. If it fails closed, the battery will over-discharge unless there is some sort of backup disconnect.

Thoughts?

 

[gnubie]

If people take the advice always offered on this forum, don't buy cheapie components for safety and switching, ensure the devices are rated for the application etc, contacts welding shut won't be an issue. If a quality, properly rated relay is put through a very large number of cycles relative to its ratings the contacts will degrade but even then you have to take it well beyond its cycle lifetime before the contacts will weld shut.

 

[markb]

I haven't yet found a spec sheet for the NVX relay in the video. Even a high quality relay will can fail, so the question that needs to be considered is, what are the chances? Is the risk of damage to an expensive battery small enough to take? I genuinely don't know the answers.

 

[Walef]

Hi,
the biggest difference in Lifepo4 electrical system application performed by DIY people in Germany compared to what is published on this channel is, that BMS, OVP/UVP is managed with single components for each inviduell cell and power supply connection for large loads and charging line is either conducted by bi-directional relays (latching relays) or with SSR's solid state relays. Both options have the advantage that either no permanent signal current is drawn or permanent current draw is negligable with SSR usage.
Those latching relays don't generate the high coil inductive current as the solenoid relay shown in Will's video clip.
Regards,
Waldemar

 

[John Frum]

If you use a relay you lose all the discharge BMS protection features other than low voltage disconnect, right?
Over current/short circuit for example.

 

[Walef]

If people take the advice always offered on this forum, don't buy cheapie components for safety and switching, ensure the devices are rated for the application etc, contacts welding shut won't be an issue. If a quality, properly rated relay is put through a very large number of cycles relative to its ratings the contacts will degrade but even then you have to take it well beyond its cycle lifetime before the contacts will weld shut.

If people take the advice always offered on this forum, don't buy cheapie components for safety and switching, ensure the devices are rated for the application etc, contacts welding shut won't be an issue. If a quality, properly rated relay is put through a very large number of cycles relative to its ratings the contacts will degrade but even then you have to take it well beyond its cycle lifetime before the contacts will weld shut.

There are two main reasons a relay can fail in service:
1.) The contacts do overheat due to high current out of the relay spec limits,
2.) The relay switches high inductive current which may result to contact arcs while switching with results to contact surface damage.
If the usage application meets the relay spec. there will be a high system reliability and lifetime.

 

[Will Prowse]

If you use a relay you lose all the discharge BMS protection features other than low voltage disconnect, right?
Over current/short circuit for example.

Absolutely. That's why you should always use OCPD.

 

[Will Prowse]

And yes, you need a relay rated for the voltage and current you plan to run. If you have a large inverter and need to charge the caps, do that first. The relay in the video is continuous duty 500A. And can handle surges of 700A.

If you are concerned with longevity, you should use an inverter with a relay input and control with BMS. That is the proper way to do it.

Large relay is dead simple method that is much safer than using a FET based switching mechanism, such as a battery protect. If rated for the load, you should get a lot of duty cycles from the relay no problem. If it fails, slap another one in there.

 

[Will Prowse]

In a properly planned system, the relay should open and close a couple hundred times per year. Which is nothing considering their typical duty cycle.

 

[John Frum]

Absolutely. That's why you should always use OCPD.

Obsessive Compulsive Personality Disorder :|

 

[fat_old_sun]

not replying to this post, but have a comment on the video. At 2:53 mark, Will states that the relay coil input is resistive. In reality, it's mostly inductive and slightly resistive. If it's purely resistive (no inductive aspect), he wouldn't have gotten shocked from the back EMF.

 

[Supervstech]

Obsessive Compulsive Personality Disorder :|

Over current protection device

 

[John Frum]

Over current protection device

I was kidding

 

[Supervstech]

I was kidding

I understand... but others might not know, so I added it.

 

[markb]

Here's an interesting article on hot switching relays:

Hot Switching Reduces Relay Life

This article describes some hot switching issues that may happen when relays are placed in switching systems.

www.pickeringtest.com

Of particular note is this:

In general, vendors life test relays with resistive sources and loads.

An inverter has a large capacitance (which is why you aren't supposed to attach it to a Victron Battery Protect). I can't find a datasheet on the relay that Will recommends, so I have no idea if it's rating applies to a capacitive load. Without knowing, you might be taking a big risk with your battery! (Or maybe not.. who's to say?)

The article linked above does have a nice photo of a welded contact. Remember, that's one of the failure modes of a relay, and if it does fail in that mode, your battery will be destroyed! It would be good to see a datasheet before trusting a relay to protect my battery.

 

[gnubie]

It all comes down to using the right relay for the job. One of the very fundamental purposes of a relay is to switch loads, afterall.

 

[Will Prowse]

Here's an interesting article on hot switching relays:

Hot Switching Reduces Relay Life

This article describes some hot switching issues that may happen when relays are placed in switching systems.

www.pickeringtest.com

Of particular note is this:



An inverter has a large capacitance (which is why you aren't supposed to attach it to a Victron Battery Protect). I can't find a datasheet on the relay that Will recommends, so I have no idea if it's rating applies to a capacitive load. Without knowing, you might be taking a big risk with your battery! (Or maybe not.. who's to say?)

The article linked above does have a nice photo of a welded contact. Remember, that's one of the failure modes of a relay, and if it does fail in that mode, your battery will be destroyed! It would be good to see a datasheet before trusting a relay to protect my battery.

Oh gosh of course that's going to happen in that regard. The relay is seen as a disconnect and that's it. We are not switching it on and off multiple times a second so it should be fine. Just ensure that your relays surge rating is more than the inrush. Even with capacitance, there is a "maximum" current from the inrush surge. Just make sure you are well above that. If you are running a 300A inverter with a 700A surge relay, and the caps need to charge up after a disconnect, it should be fine. Really depends on the inverters internals.

Honestly I would rather use an inverter with relay input.

The relay will work great if it is sized well, but I would never expect it to last the rated duty cycle on the data sheet. But if you need to switch it only a few times a year, and it's within the voltage/current rating, it should be fine. What do you guys think?

 

[ghostwriter66]

I can see the "well What If" gang is out tonight ... so the 500A 12V Continuous relays that we use at work -- and in the year i have been here none of them has never failed -- and they switch allot during the day -- is RATED at:

Max Switching Power: 2000W
Life Expectancy Electrical: 100,000 Operations (at 30 Operations/min)
Life Expectancy Mechanical: 10,000,000 Operations

I just want you to let those numbers sink in ...

I will take a 200A BMS and a 500A relay over a 500A BMS any day ....

 

[Will Prowse]

Holy cow! The 30 operations/min is incredible. 10 MILLION. Those numbers are crazy. I would love to hear the clicking sound in a relay test lab haha!

And for solar usage as a disconnect, it would only switch a few times a year. Or not at all, if the system was designed properly to run the loads.

 

[ghostwriter66]

Holy cow! The 30 operations/min is incredible. 10 MILLION. Those numbers are crazy. I would love to hear the clicking sound in a relay test lab haha!

And for solar usage as a disconnect, it would only switch a few times a year. Or not at all, if the system was designed properly to run the loads.

So while I agree that perhaps 20 years ago or so that coil failure or fatigue was a thing to be concerned about (maybe), today with current technology and manufacturing tolerances the only time these coils have issues is when something or someone causes a massive voltage spike - for a sustained period... to run through it ...