Speak EV - Electric Car Forums banner

1 - 20 of 75 Posts

·
Registered
Joined
·
3,849 Posts
Discussion Starter #1
Just got an Ioniq 38 kWh, arrived home from dealer in evening, 22 miles left, needing 100% charge for trip at 10 the following morning. Plugged into Wall charger (dumb Rolec with Viridian Manipine Ecu with 32A resistor wired in) and had supper. First ever charge at this current, Ampera's been fine at 14A for years on this. Came back later to find car says Charge stopped unexpectedly, check the EVSE. It had tripped at the mini-CU thing. Reset the trip, started charging again, checked again a bit later and it had tripped again. It added about 18 miles in these 2 short sessions, and I estimate that's 15 mins of charging on each session. A later charge on a Rapid had it starting around 27 kW, and slowly ramping up to 34. So I would not be surprised if the car does a slow-start charge process, whatever ampage is selected.

I swapped out the 32A resistor on Mainpine & put the 16A one in, started a charge & went to bed as was now midnight. Fortunately that worked fine, car 100% the next day. Phew!

Qn is, what's happening? There are lots of "32" numbers around. In the attached pic I'm pretty sure it's the middle switch that tripped, happy to do a retest to confirm.

Am I likely to have a larger earth-leak at higher currents? Is the car actually drawing maybe 33A & that's enough to trip? Is a change-in-current-drawn (maybe by 1A, whatever) likely to cause some surge that trips it? Anyone else got an Ioniq tripping a 7kW EVSE like this? While y'all digest this (Thx in advance, guys!) I'll swap that 32A resistor back in, stick my current-clamp round a wire & take notes.
136444
 

·
Registered
Joined
·
467 Posts
If it's the middle one in your picture that's tripped then it suggests it's not an overcurrent. The main supply switch (left hand side) is rated at 32A so that should trip on overcurrent before the 40A rated centre unit. Therefore more likely to be a residual current problem that's tripped the earth leakage protection. Have you tried the car on any other ac points? Don't think the dc charge behaviour will tell you much as this will likely be related to either the on-board charger in the car or (as you say) the charge point itself leading to higher leakage currents at higher charge rates.
 

·
Registered
Joined
·
4,796 Posts
Your "middle switch" is the RCD and, unless faulty, should not trip on current alone before the left hand switch. So that implies a current being detected in the Earth.
I presume that you are using an untethered lead and normally that would be my initial concern. However I fail to understand why that works at 16 Amps if faulty. Have you tried a public 32 Amp charge elsewhere?
 

·
Registered
Joined
·
3,849 Posts
Discussion Starter #5
Not tried a public AC, but there's one in Sainsbury 1/2 a mile away, so I'll test that. Good suggestion thx! Yes, I have untethered lead, virtually brand new as car's only done 5k miles & sat in dealership a lot. Could put a megger on the lead if that would help? Off to retest it. Sods law says it'll behave now!
 

·
Registered
Joined
·
3,849 Posts
Discussion Starter #7
Ok, good-ish news I think. Fault is replicatable. It's the LHS "B32" switch that's tripped.

Opened up EVSE, put 32A resistor in, 200/600A range current-clamp looped around brown wire at EVSE car socket. Power up, plug car in at 27 mins past 10. Charging starts at 21A, fluctuates a tiny bit for a couple of minutes, no doubt the charger is getting battery up to temp, checking stuff/cell balance, whatever. Then, over about 90 secs the current ramps up smoothly to 31.5A and sits there rock-steady. Depending exactly how I hold the current clamp round the wire, or let it rest, I can get that to read anything from 31.1A to 31.6A. After 10 mins, this is so steady I thought I'd see what the Earth current was, so set to 2/20A scale & try measuring that. Again, depending on exactly where it's held, I can get anything from 0.2A down to 0.08A being displayed steadily after a few secs. There's a load of wire around of course, and it's not rated for low current measurement obvs, and I have no idea of what sort of noise/wobble you get with these things. Looks like +- 0.2 A could be typical error margin.

After 23 mins, am thinking about going for a coffee as it's been rock-steady at 31.3A for a while now. Then, Clunk, it trips out at 10:50 precisely while my eyes were on the current display. No sign of the car changing anything at all there.

So I think I'll stick in a resistor for a tiny bit less current, and see what that does. Is this B32 thing using temperature bimetallic strip thing to measure current? What's the tolerance on them? Am curious to learn this stuff!
 

·
Registered
Joined
·
161 Posts
From memory a B type MCB will hold 32A forever. As others have said, it‘s the RCD, not the overcurrent device that’s tripping. RCDs work by measuring the current flowing in to the circuit (along the live leg) and the current flowing out of the circuit (along the neutral leg). It should be the same (residual current = 0). If there is a leakage path to earth, some of the return current will flow through that back to either the PME connection at your meter or direct to the substation, bypassing the RCD and leading to a measurable residual current.

If the difference between the current out and current back exceeds 30mA (the 0,03A marking in the picture) then the unit will trip.

Not immediately obvious to me why it trips after so long or only at higher current.
 

·
Registered
Joined
·
4,905 Posts
Is this B32 thing using temperature bimetallic strip thing to measure current? What's the tolerance on them?
They are usually thermal plus magnetic, the latter being for very high current instant trip. Yours is clearly tripping on thermal, which it shouldn't - they can normally run continuously at about 20% over rated current.

Two obvious possibilities come to mind.
1. The breaker is faulty.
2. One or both connections on the breaker are not well made and are heating up, which heat will of course get conducted into the mechanism and add to the normal heating.
So, if you feel capable, first check the connections are good - clean, fully inserted and tight (usually around 3nm). If that doesn't work get a new (decent quality) breaker - they aren't expensive.
 

·
Registered
Joined
·
3,849 Posts
Discussion Starter #11
Sounding like a thermal thing to me, which fits with both my charging attempts on first night starting ok & lasting a couple of mins until I was sure all the right lamps were flashing on car = happy. Also fits in with a 16A charge going ok for hours. I've got a USB thermal camera on my Android phone, so I'll break that out & see if it appears to warm. Am also wiring up a multi-turn pot so I can adjust the current a tad. Great fun, this stuff! Thx for all the help, absolutely invaluable.
 

·
Registered
Joined
·
4,905 Posts
Am also wiring up a multi-turn pot so I can adjust the current a tad.
Looking at a B trip curve dropping it just a couple of amps would probably be enough.
But check those terminals anyway. If they aren't good they won't get better being run hot and you could end up with a melty, black mess.
 

·
Registered
Joined
·
3,849 Posts
Discussion Starter #13
Just done another 20 min test, tripped out at the end. This time I pointed a thermal camera at the visible front-face of the CU stuff. The RCD on RHS is running near-as ambient, which seems to be around 9C at the mo.

The MCB showed 15C, prob about 5 mins after starting charge while I faffed with variable resistor & got camera setup. Then the front face temp very slowly ramped up steadily, 16C, 17C, ... until it was showing 20C at the hottest point I could focus on, then it tripped, anothe 20 min charge completed. I think this is pretty conclusive. Next I'll shut down the house, have the cover off & check terminals; if I can safely do a test with ir camera I will, before checking tightness of terminals. Don't know if this cheapo (£100) IR add-on camera has the resolution to identify which terminal might be getting warm, or not. It's only about 64 pixels square res, but proving to be a v useful bit of kit. Maybe every Ev owner should have one - it's potentially a life-saver, and in this case it's proving to be a useful diagnostic tool.

While playing with the variable resistor, it's clear the Mainpine is limiting the max current it'll signal to 31.4 amps near-as, I can't get the car/Mainpine above this. At the other end, I can get the car charging at 5.1A ! Std says 6A, maybe Mainpine's saying 6A also, but the Ev's happy to go as low as 5.1 A. The car closely tracks my changes within a second or so, as it should. No surprises there. The resolution is infinite I think, I can tweak in increments of 0.1A if I want.
 

·
Registered
Joined
·
3,849 Posts
Discussion Starter #14
Just taken a peek at this page: I'm looking at the 3rd setio nin blue box.
http://protekuk.co.uk/index.php?route=product/product/download&product_id=312&download_id=14

"Application exampleThree MCBs are to be installed side by side inside an enclosure. The ambient temperature inside the enclosure is at 40°C. The required maximum load current drawn per MCB = 29A.Determine whether 32A MCB would provide suitable solution.• From the Temperature Derating table, The nominal trip current (In) of 32A MCB operatingat 40°C is derated to 30.4A• From the Grouping Factor table, Three MCBs installed side by side have a groupingfactor K(g) = 0.95• Combined effect of temperature derating & grouping factor, 30.4 x 0.95 = 28.88AIn this application the 32A MCB would trip at 28.88A and therefore be unsuitable to supplythe max 32A load that is required. In this case a 40A or higher MCB should be selected instead."

This MCB derates acc to temperature, as ambient temp inside enclosure of 40C drops the tripping-current from 32A to 28.88A. This has to be what's happening here, maybe exacerbated by not-tight-enough terminal heating everything up.

Theirs is rated 32A at 30C in enclosure, but drops to 30.4 at 40, so maybe my enclosure's getting to about 33C inside & trips at 31.5A?

This looks very tight to me, but maybe that's as it should be, and maybe this enclosure should be running at near-as-dammit ambient temp? But if they're usinh a bimetallic strip to monitor temp, that by definition has to be consuming current & heating up, so will inevitably warm the enclosure. But how much warming's normal?
 

·
Registered
Joined
·
4,905 Posts
You don't have any MCB grouping and your box is apparently less than 20deg since the front of the MCB only reaches 20. So I don't think that ambient conditions are the problem. In fact that all sounds pretty cool - many of us have these mounted in warmer places and get through summer without then tripping.

I don't recognise that branding so I suspect it's just a cheap and nasty breaker. They will do for typical domestic use but this sort of application is testing the limits and you need decent stuff.
(Eg. See the "Rolec Charger - Dangerous???" thread.)
 

·
Registered
Joined
·
3,849 Posts
Discussion Starter #16
Just IR camera'd it with the lid off. The terminals are fine, fully tight, and the wire at top is slightly warm. Bottom has chunky slab of copper joining it to the RCD, no probs there. The Lhs side of this, however, is getting hot. Charging at 31.5A, it's getting up to 43C about half-way up, so looks like internal heating somewhere in the middle. Reducing the current to 16A brings this down to around 29C after 10-15 mins, so this is looking stable at least. So I'll leave it at 16A for the moment, and get a good quality replacement tomorrow.

I do wonder if there needs to be some sort of regular inspection required for these EV setups; mine's been running un-checked for 5 years just fine, but once I ask for 32A it's clearly pushing some items towards their limits. What with crappy RCBOs in older Rolecs, and a recent bad gge experience described in another thread, how many more of these are we going to see when Ev volumes really take off? I almost feel there should be an approved list of mfrs & part-numbers for the MCBs, RCBOs, Contactors, RCDs etc used for these. How do these cheap & nasty items get allowed in the first place?

Thx for the help guys, fantastic response from you all! :)
 

·
Registered
Joined
·
3,265 Posts
For the cost swap out the MCB.
They can age, especially if they have had heavy duty cycles.
It's not unknown to have duff ones straight from the factory.
 
  • Like
Reactions: T67M

·
Registered
Joined
·
3,265 Posts
Just IR camera'd it with the lid off. The terminals are fine, fully tight, and the wire at top is slightly warm. Bottom has chunky slab of copper joining it to the RCD, no probs there. The Lhs side of this, however, is getting hot. Charging at 31.5A, it's getting up to 43C about half-way up, so looks like internal heating somewhere in the middle. Reducing the current to 16A brings this down to around 29C after 10-15 mins, so this is looking stable at least. So I'll leave it at 16A for the moment, and get a good quality replacement tomorrow.

I do wonder if there needs to be some sort of regular inspection required for these EV setups; mine's been running un-checked for 5 years just fine, but once I ask for 32A it's clearly pushing some items towards their limits. What with crappy RCBOs in older Rolecs, and a recent bad gge experience described in another thread, how many more of these are we going to see when Ev volumes really take off? I almost feel there should be an approved list of mfrs & part-numbers for the MCBs, RCBOs, Contactors, RCDs etc used for these. How do these cheap & nasty items get allowed in the first place?

Thx for the help guys, fantastic response from you all! :)
Cross posts.
Good decision. But even those temps in and of themselves are not unexpected. Its a whole heap of current for a sustained period.
 

·
Registered
Joined
·
3 Posts
I'd agree wih reducing the current setting so you have more of a margin between the nominal 32A breaker and the nominal 32A you're setting the car to draw from the supply. Also agree with Mikegs about the possibly unknown qualities of the MCB.
Everything has some tolerance - or variance in manufacturing, and cheaper MCBs may not have had so much testing done to ensure they meet the specs.
Have a look at this (MiniatureCircuitBreakers_primer_EN_201601250852395217.pdf):
there's a graph on page 22 which shows that MCBs are not supposed to trip for 2 hours at 1.13xthe nominal current, but should trip at 1.45x the nominal curent in less than 1 hour.
So either the MCB is over-sensitive or you're not seeing the true current, perhaps if there's a high frequency component then the mcb (which uses a heater to cause the trip at low overloads) might be correctly responding to a current just over 32A this but your meter might not be seeing the peaks.
Can you reduce the current to somewhere around 28A or does it only allow steps at 16 or 32A?
 

·
Registered
Joined
·
3,265 Posts
I'd agree wih reducing the current setting so you have more of a margin between the nominal 32A breaker and the nominal 32A you're setting the car to draw from the supply. Also agree with Mikegs about the possibly unknown qualities of the MCB.
Everything has some tolerance - or variance in manufacturing, and cheaper MCBs may not have had so much testing done to ensure they meet the specs.
Have a look at this (MiniatureCircuitBreakers_primer_EN_201601250852395217.pdf):
there's a graph on page 22 which shows that MCBs are not supposed to trip for 2 hours at 1.13xthe nominal current, but should trip at 1.45x the nominal curent in less than 1 hour.
So either the MCB is over-sensitive or you're not seeing the true current, perhaps if there's a high frequency component then the mcb (which uses a heater to cause the trip at low overloads) might be correctly responding to a current just over 32A this but your meter might not be seeing the peaks.
Can you reduce the current to somewhere around 28A or does it only allow steps at 16 or 32A?
Whilst this suggestion would certainly get around the issue its not really a proper solution.
There are multiple 32 amp EVs charging merrily around the country.
It's better to spend £5 to get the full charging ability of your car.
 
  • Like
Reactions: mikegs
1 - 20 of 75 Posts
Top