Blink / Rav4 Blows Out a Contactor Pin (with gory pics)

[TonyWilliams]

CHAdeMO/SAE Combo will easily take the >10kW stuff, so why bother with 70A AC?

-Phil

Except the cat is already out of the bag with Tesla Roadster and Model S with "twin chargers" using the 70 - 80 amps.

 

[surfingslovak]

I think the industry folks realize that these things should have been designed to effectively NEVER burn out, even with marginal quality players like Ecotality.

I believe the message might have reached the intended recipient earlier today. Given that eTec, which was acquired by ECOtality in 2007, reportedly had extensive experience with electric charging infrastructure, it's surprising to hear all the complaints about Blink stations. This thread is just another example.

That said, I'm really impressed that the Honda Fit EV detects the overheating charge port condition, and terminates the session if needed. Perhaps all EVs should be able to do that.

Note that when you go to eTec's old website, it does not redirect to ECOtality, it displays the default GoDaddy hosting page instead. That doesn't look very professional.

Click to open

 

[surfingslovak]

Mike, good on you to contact CS. It certainly won't hurt that they have your inquiry on record now. I could be wrong, but I'm under the impression that 877 NO GAS EV and online chat were operated by a 3rd party vendor on Nissan's behalf, and most, if not all of the agents, were located in Colorado. That's not to take away from them or the quality of their work. I'm only mentioning it to provide additional context, and it might be a good idea to confirm that in a future call.

I take it back, the TeleTech support center servicing the Nissan account appears to be located in Amherst, OH. TeleTech reportedly has a working relationship with several automakers, including BMW, Ford, General Motors, Hyundai, Infiniti, Jaguar, Kia, Mitsubishi, Nissan, Volkswagen, and Volvo. That said, this thread continues to see significant activity, and at this point there are few OEMs, which have not visited here. I was particularly glad to see REMA a few days ago.

 

[MikeD]

I Googled "evse + Blink + melt", and got this thread and at least one other related to it on the first page. Other keywords would probably do the same, so if one were looking for instances of hardware failure during an EV charging session, it is hard to miss.

On searching with "evse + melt" I saw an unrelated thread about a Volt and its portable EVSE. One comment was in part "We don't need speculation, we require facts when something like this happens in order to potentialy assign fault. Yes the car and the EVSE could indeed be the cause, (unlikely as IF they were drawing excessively it SHOULD have tripped your breaker) but there's really no way to know without a detailed reconstruction and through investigation of the facts.". Seems reasonable to me.

Another post about what may be an EVSE design problem: http://livingleaf.info/2011/12/melting-evse-plugs-not-with-leaf/" onclick="window.open(this.href);return false;

I don't like to see ANY such design/manufacturing problems, because they can make potential EV buyers change their mind particularly at this opinion building stage -- but such problems need to be identified if they truly exist and corrected early on.

So far I've not aware of additional instances of charging failures involving the Blink or the Rav4 as in the OP, but problems like the alleged "resistive heating at bad wire terminations" varies as (amps * amps * ohms), so as sales of higher charging rate cars increases next year any such actual problems should become evident.

 

[TonyWilliams]

Had a call from Michael Martin at Blink yesterday. It was a quite informative, no BS call. Strictly the facts as he knows them, which I appreciate.

There are about 14,000 Blinks with the REMA handle out there, and this is the first melted pin. Toyota / Tesla uses a Yazaki J1772-2009 on the Rav4 EV. They had a conference call planned today with all the above. To date, he stated that the pins on the REMA appeared to be terminated correctly (unlike the pin that Phil showed).

Yazaki has had some insulator caps come off the pins on LEAF (mine was melted, so hard to guess what was going on there). Also, there have been bent pins, and the Honda Fit had problems on some Blinks at IKEA, and those units were sent to REMA.

Interestingly, the REMA is rated for 71.4 amps, and not 75 or 80 that the standard allows. He didn't know why there was this limitation.

We talked a bit about smart meters for EV's, and how Bluetooth is not working, and it looks like PLC (which is already working in Europe) will be the communications standard for that.

 

[Ingineer]

Thanks Tony!

How could the Rema be rated for 71.4A when it's clearly only going to only accept #10 AWG wire? Also, the sticker on the one I examined was clearly labelled 30A. Something doesn't smell right... Maybe it's burning! :shock:

-Phil

 

[surfingslovak]

Thanks for the update. FWIW, Yazaki was the most frequent visitor to the this thread from all the OEMs. Now we know why!

 

[Ingineer]

The other thing is that, as yet, almost all Blinks have been running at most around 18A, the bulk have been at a cozy 16A. On the unit I examined with the badly crimped pin, My thermal camera instantly spotted a discrepancy between the 2 pins, even on a Leaf only running at 16A.

Once we get a bunch of new Leafs pulling almost twice the current, I bet we start to see some burning.

-Phil

 

[davewill]

Thanks Tony!

How could the Rema be rated for 71.4A when it's clearly only going to only accept #10 AWG wire? Also, the sticker on the one I examined was clearly labelled 30A. Something doesn't smell right... Maybe it's burning! :shock:

-Phil

Perhaps they use different pins when using bigger wire and that's why the plug on the Blink has a sticker saying 30a instead of a rating stamped into the plastic? It kind of makes sense that you'd want to match the pins to wire size, after all. A too big pin would probably make for bad connections, too. Of course, if the pins aren't capable, then the plug can't be "rated for 71.4".

 

[Ingineer]

Perhaps they use different pins when using bigger wire and that's why the plug on the Blink has a sticker saying 30a instead of a rating stamped into the plastic? It kind of makes sense that you'd want to match the pins to wire size, after all. A too big pin would probably make for bad connections, too. Of course, if the pins aren't capable, then the plug can't be "rated for 71.4".

My point is the Ecotality rep told Tony that they are rated for 71.4A, when clearly all the ones they use are rated for 30A. The handle will never have a current rating, as it's just a piece of plastic, it's the pins and the wire which determine the current handling. Every Blink I've ever seen uses a Rema with 10AWG rated for 30A. So the 71.4A figure comes from what? Oh, maybe they have another model using the same plastic rated for that, but that's immaterial, as Ecotality doesn't use that model.

According to their website, they only offer the 30A unit.

-Phil

 

[colby]

Ok.. Had to make an account to get in on this conversation. I've been monitoring it for a couple of weeks.

I am the guy from facebook who's Fit EV had charge port overheating issues. When Honda came out and connected their notebook to my car, I saw the charge port temp getting up towards 75C and then the car stopping the charge for safety issues. When I first got the Fit EV, 1 out of 7 Blink units at my work were overheating. Today, 6 out of 7 will hit 75C in just 20 minutes. These stations are used for 8-12 hours per weekday with about 2-4 insertion cycles per weekday. The other day I had to stop a RAV4 driver from plugging in to a defective station and told him to use L1 for the time being.

My point is that while maybe only one car has melted it's charge port, these REMA connectors are getting very hot and more are likely to melt, especially in cars that don't have temperature monitoring. The J1772 spec states that the connection should not get hotter than 50 degrees above ambient. This is presumably where the Fit EV is shutting off charging. REMA connectors are not meeting J1772 spec.

When a connector doesn't have an overheating issue, the enclosure is about ambient temperature, maybe a bit more when pulling 30A. A connector that overheats will quickly get noticeable warmer at 30A. At 16A, the connectors get a little warmer, but nothing scary.

And I can confirm that I have spent a lot of time looking at REMA connectors on Blinks and all of them are labeled that they are rated for 30A (though they clearly can't handle that).

I sent an inquiry to Ecotality a couple of weeks ago, but haven't heard anything back. A Honda engineer has told me that the issue is "currently under investigation by RIMA and others." I would sure hope so!

 

[Ingineer]

Thanks for taking the time to post that Colby.

Sad to say, This is what I have feared after my own investigation. Though with what you report, I cannot dismiss that there is a possibility that there is damage in the pins on your Fit's inlet which could be causing accelerated wear on the stations you connect to. Though I'd still err on the side of the Rema pins not being engineered properly. The temperature rise is caused by the resistance in the connections between the EVSE and the car. My theory is that the poor crimps are heating slightly, but still within spec. Unfortunately the poor crimp is expanding the crimped metal and also allowing oxygen into the heated connection which begins to increase the resistance, resulting in even more heat and oxidation of the interface and this rapidly becomes a thermal runaway. Each charge is another nail in the coffin! Tony's situation supports this theory. Since the stock RAV has no thermal monitoring, this will start to occur more often as time goes on, and hopefully we'll hear about some of it.

It's unknown whether Nissan will have thermal monitoring in the Gen 2 Leaf, but if they don't it could well be a disaster for the EV project. As soon as people start plugging them into their Blinks, we'll have lots of melted connectors like Tony's.

-Phil

 

[Ingineer]

For all the OEM's that may be reading this thread: In your next generation J1772 handles, incorporate a simple thermostat in series with the handle proximity switch as I have done in the previous RAV4 upgrade. This is low-cost, and can likely be easily implemented with minimal changes. This can also be implemented in any inlet by the addition of the thermostat and the 330 ohm resistor. At the prices people are charging for handles/inlets this is an insignificant cost, and could well prevent disaster. There is much prior art, so this is not patentable.

I suspect this could be added to any handle in production without any tooling changes. A more elegant solution is to signal the EVSE that a thermal event has occurred so it's UI can also alert the user to the problem as well as throttle the charging. This could be accomplished by using a diode on the pilot line so that no additional conductors are required in the cable. In most EVSE designs, this could then be implemented with a simple bit of software added, even on "dumb" EVSE's.

I'll be happy to consult with any EVSE manufacturer to add such a capability to their design.

-Phil

 

[chris1howell]

For all the OEM's that may be reading this thread: In your next generation J1772 handles, incorporate a simple thermostat in series with the handle proximity switch as I have done in the previous RAV4 upgrade. This is low-cost, and can likely be easily implemented with minimal changes. This can also be implemented in any inlet by the addition of the thermostat and the 330 ohm resistor. At the prices people are charging for handles/inlets this is an insignificant cost, and could well prevent disaster. There is much prior art, so this is not patentable.

I suspect this could be added to any handle in production without any tooling changes. A more elegant solution is to signal the EVSE that a thermal event has occurred so it's UI can also alert the user to the problem as well as throttle the charging. This could be accomplished by using a diode on the pilot line so that no additional conductors are required in the cable. In most EVSE designs, this could then be implemented with a simple bit of software added, even on "dumb" EVSE's.

I'll be happy to consult with any EVSE manufacturer to add such a capability to their design.

-Phil

Hi Phil,

I would love to add this to the OpenEVSE Hardware/Software. I think OpenEVSE would be a great platform for advancing the safety of EVSE/EVs.

 

[colby]

Chris--

I added thermal protection to my OpenEVSE last weekend. I put three NC thermostats in series with the CP and put them up against the areas that get hottest (the line in to fuse connection, fuse to relay, and relay to J1772 cable). If any one of them overheats, the pilot signal is disconnected and charging stops. Once the device cools charging starts again. This can be good or bad...

I was considering using thermistors and reading the value with an ADC to throw a trouble code at a certain temperature, but I wanted the overtemp system to work if there was a failure on the MCU side (either software or hardware).

There may be a more elegant way to do this than I did. Another way would be to use NO thermostats on the proximity line (assuming that the cable routes this back to the EVSE), connected to a resistor (I calculated the value last weekend, but don't remember it off the top of my head) to ground, to simulate a trigger press.

I used 75C thermostats, but I think I want to reduce that a bit, since I'm measuring through insulation.

 

[Ingineer]

Hi Phil,

I would love to add this to the OpenEVSE Hardware/Software. I think OpenEVSE would be a great platform for advancing the safety of EVSE/EVs.

I'm happy to help! I'll shoot you an email. I'm sure we can work out a good method.

-Phil

 

[Ingineer]

Chris--

I added thermal protection to my OpenEVSE last weekend. I put three NC thermostats in series with the CP and put them up against the areas that get hottest (the line in to fuse connection, fuse to relay, and relay to J1772 cable). If any one of them overheats, the pilot signal is disconnected and charging stops. Once the device cools charging starts again. This can be good or bad...

I was considering using thermistors and reading the value with an ADC to throw a trouble code at a certain temperature, but I wanted the overtemp system to work if there was a failure on the MCU side (either software or hardware).

There may be a more elegant way to do this than I did. Another way would be to use NO thermostats on the proximity line (assuming that the cable routes this back to the EVSE), connected to a resistor (I calculated the value last weekend, but don't remember it off the top of my head) to ground, to simulate a trigger press.

I used 75C thermostats, but I think I want to reduce that a bit, since I'm measuring through insulation.

Colby, I do not recommend you do it this way. This could subject your relays to arc damage, and even a fire and even possibility of damage to your on-board charger. (tiny chance, but still) The best way will be to patch the software. I'll work with Chris on this ASAP.

-Phil

 

[GRA]

Chris--

I added thermal protection to my OpenEVSE last weekend. I put three NC thermostats in series with the CP and put them up against the areas that get hottest (the line in to fuse connection, fuse to relay, and relay to J1772 cable). If any one of them overheats, the pilot signal is disconnected and charging stops. Once the device cools charging starts again. This can be good or bad...

I was considering using thermistors and reading the value with an ADC to throw a trouble code at a certain temperature, but I wanted the overtemp system to work if there was a failure on the MCU side (either software or hardware).

There may be a more elegant way to do this than I did. Another way would be to use NO thermostats on the proximity line (assuming that the cable routes this back to the EVSE), connected to a resistor (I calculated the value last weekend, but don't remember it off the top of my head) to ground, to simulate a trigger press.

I used 75C thermostats, but I think I want to reduce that a bit, since I'm measuring through insulation.

Colby, I do not recommend you do it this way. This could subject your relays to arc damage, and even a fire and even possibility of damage to your on-board charger. (tiny chance, but still) The best way will be to patch the software. I'll work with Chris on this ASAP.

-Phil

Yes Sir, nobody but your 'run of the mill, how fast will it go and what color can I get it in' ignorant car buyers on this forum. :lol: And Nissan wonders why they've gotten so much flak for not giving their Leaf customers any more information about its operation than is necessary with their ICE cars. :roll:

 

[TonyWilliams]

Here's my two cents on thermal monitoring; I would absolutely want a fault light that left no doubt that there had been a thermal issue. Also, prior to a complete shut down at 70C, perhaps a 55C threshold that reduces current (by changing the ohm value to the pilot pin) from say whatever it is to 30% less.

Chris--

I added thermal protection to my OpenEVSE last weekend. I put three NC thermostats in series with the CP and put them up against the areas that get hottest (the line in to fuse connection, fuse to relay, and relay to J1772 cable). If any one of them overheats, the pilot signal is disconnected and charging stops. Once the device cools charging starts again. This can be good or bad...

I was considering using thermistors and reading the value with an ADC to throw a trouble code at a certain temperature, but I wanted the overtemp system to work if there was a failure on the MCU side (either software or hardware).

There may be a more elegant way to do this than I did. Another way would be to use NO thermostats on the proximity line (assuming that the cable routes this back to the EVSE), connected to a resistor (I calculated the value last weekend, but don't remember it off the top of my head) to ground, to simulate a trigger press.

I used 75C thermostats, but I think I want to reduce that a bit, since I'm measuring through insulation.

Colby, I do not recommend you do it this way. This could subject your relays to arc damage, and even a fire and even possibility of damage to your on-board charger. (tiny chance, but still) The best way will be to patch the software. I'll work with Chris on this ASAP.

-Phil

 

[TomT]

Unfortunately, I don't think there would be an easy way to do this as an add-on since it is the pilot signal duty cycle that controls the maximum allowed charging rate.

Also, prior to a complete shut down at 70C, perhaps a 55C threshold that reduces current (by changing the ohm value to the pilot pin) from say whatever it is to 30% less.