I really do appreciate the time and trouble you take to answer these questions, and in the process help educate us all. Many thanks @argTutorial on RCDs first, conclusions at the end:
RCDs detect faults by measuring the difference between the current in Line and Neutral - they should be identical, and if different some current has 'leaked' away to earth due to some kind of fault; in the worst case, that leakage is through the body of someone who is touching the faulty wire/car and their feet standing on the ground.
Conventional RCDs (type AC) will only detect where the leakage is 50Hz AC - which is what you get for simple cases where the insulation has failed and the live cores are partly in contact with earthed metalwork, or the wires are bare and a person has grabbed hold of it. These RCDs are very simple to make - just a transformer core with a turn around the two wires and another winding to activate the mechanism. This is all you need for old-fashioned appliances with just motors, resistance heaters etc.
Most modern appliances have electronics in them and will often have a power supply that rectifies the AC supply to generate some sort of DC. If you get a wiring fault inside the appliance, the leakage current is unlikely to be pure 50Hz AC, it's more likely to be pulses of DC (which you can think of as AC at a different mix of frequencies). You need a Type-A RCD to detect this kind of fault. Fortunately, pulsed DC will still activate a transformer, so a Type-A can be built in almost the same way as a Type-AC, just adding a little bit of electronics to the same sort of transformer as a Type-AC. That sort of electronics has trivial cost nowadays, so the Type-A hardly costs any more than the Type-AC. Given the widespread use of appliances with switchmode power supplies, and the minimal extra cost, IMO we should really be using Type-A everywhere - and the industry is slowly moving in that direction, with some brands of consumer unit now being Type-A by default. The regulations have been a bit slow (IMO) in that they have only been requiring Type-A in specific situations of higher risk - but that has included EV charging since 2013. This is sensible as many plausible faults in a typical design of EV charger would generate pulsed DC, and EV on-board electronics are a high risk area (high vibration, risk of un-noticed damage in traffic accidents etc.).
There remains the possibility of a fault that is pure DC without any AC component. Such faults are rare - even if a typical charger/SMPS circuit is generating pure DC at its output, there's not usually any obvious failure mode that will couple that DC back into the supply wiring. However, if you do manage to create a pure DC fault, then it's very bad. Not only will standard Type-AC or Type-A RCDs fail to detect the fault (because transformers don't work with DC), that transformer can become 'saturated' by the DC so that it no longer performs its normal function. This means that the DC fault could sit there un-noticed for a long time (because it doesn't trip anything) and all that time the RCDs are no longer providing their protection against more normal faults (like a person touching bare wires). You need a Type-B (or Type-EV) RCD to detect these pure-DC faults, or a special circuit in the chargepoint to do the same job. Detecting pure DC faults is much more complex since you can't use the transformer trick that magically cancels the L and N currents - it's notable that most Type-B RCDs are significantly bigger than Type-A equivalents to make room of the necessary circuitry. Not £300 worth of extra - that's a premium being charged to reflect the small numbers of these devices currently being made - but could easily be double the manufacturing cost of a Type-A.
Regarding the saturation issue, this can be solved by simply making the transformer bigger - though of course manufacturers don't want to make them bigger than they need to be because that costs money. The specification of Type-A requires it to tolerate at least 6mA of pure DC while still responding to any AC/pulsed DC on top of that. This then sets the specification for the internal DC detection on new-generation chargepoints: if they detect leakage > 6mA and shut down, then that guarantees that any upstream devices won't get saturated and they can continue to be Type-A. This also gives the distinction between Type-B and Type-EV RCDs. Most standard RCDs have a sensitivity of 30mA - being chosen as a compromise to be likely to trip in the event of a current big enough to kill you in a touching-bare-wire scenario, while not tripping on the small amounts of leakage that occur in normal usage (most particularly due to interference suppression components). Type-B RCDs just add DC sensitivity to their existing AC sensitivity at the same level (so typically 30mA) hence they will detect significant DC faults but don't guarantee to keep the level low enough to prevent Type-A devices becoming disabled - going the Type-B route, all your RCDs need to be Type-B. Type-EV on the other hand has general Type-A behaviour (at 30mA), plus DC protection at (6mA) so as to permit upstream Type-A devices. In most cases, EV charging is installed on a separate circuit and there isn't an upstream RCD so the choice of Type-B/EV is unimportant, but sometimes it isn't. One particular case in point is houses or farms with a TT earthing system and a 100mA time-delayed RCD protecting the whole installation against earth faults (that are normally handled by the supplier fuse on other earthing systems). For an installation like that, the Type-EV is more suitable than the Type-B.
So, we have a rare fault possibility and an expensive way of solving it - how does this relate to EV charging?
As above, a pure DC fault inside the car's charger is relatively unlikely - not much more so than your washing machine or phone charger. However, there's a specific failure mode specific to EV charging. The control pilot wire in the charging cable has 12V DC applied to it relative to earth, through a 1K resistance. If the cable becomes damaged and shorts the pilot to neutral, approximately 12mA DC will flow in the neutral. That fault is harmless enough in itself, but the 12mA may be enough to saturate a Type-A RCD. If the cable damage has also exposed the Line wire so you can touch it, now you have an electrocution situation with the RCD disabled. You'd still have to be pretty unlucky to get in this situation, but you have to admit that it might happen.
So, what if you've got an existing Tesla (or other brand installed before 2019) with only Type-A RCD?
- There's a theoretical risk of electric shock from damaged charging cables that your installation doesn't protect you from. You could spend £££ to upgrade, or you could take care to monitor the condition of your cables. If you used to use an electric lawnmower in the 1970's, that was a much bigger risk!
- If it was installed pre-2019, then there's nothing in law pressing you to upgrade; it's a personal decision on the cost vs safety benefit.
- If it was installed 2019 onwards, then any Electrical Installation Certificate claiming that the installation complies with BS7671 is technically fraudulent. You could consider going back to your installer to get it fixed, but if the work was priced on a time-and-materials basis then you will probably end up paying for it anyhow. If it was a fixed price bid and they won it by quoting a lower price by using inadequate materials, then they deserve to be called to account.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Tesla Wall Connector - Type B / Type A-EV RCD
- Thread starter pow216
- Start date
@Baldrick I just want to thank you for your spreadsheet of chargers. I had been searching on-line for a comparison of the options and found nothing as informative. This is a great example of the ways in which Tesla owners seem to be willing to help each other.
Don't think this has been posted already: Open Energy Monitor have a type-B RCD for sale to go with their EVSE (presumably could be used with others): Type-B RCD 1P+N (Chint NL210-63-263/30)
£157.50 inc VAT. Looks like it takes up 3 slots of a consumer unit.
£157.50 inc VAT. Looks like it takes up 3 slots of a consumer unit.
Looks like the best price so far. That’s exactly the same device as I linked to in post #13 above, but a different retailer and about £40 cheaper. Out of stock at the moment though...
MrBadger
Badger out
Fullerene
Member
I am not sure my understanding is entirely correct - can I put this to the Audience (@arg - is that you on the IET Forums...?)
The model in stock according to the data sheet has:
Tripping sensitivity
30mA - additional protection against direct contact.
Rated current In(A) 25, 40, 63
Rated sensitivity I△n (A) 0.03, 0.1, 0.3
I read this to be 3 different models breaking AC at 25, 40 & 63 Amps with AC leak ratings of 30mA, 100mA & 300mA, DC leak being 30mA across the board
Am I correct in thinking this is above the required 6mA DC leak for regs or am I reading the datasheet incorrectly?
Last edited:
There’s (at least) 9 models described by the datasheet - permutations of the load capacity and sensitivity. Yes, this is the 30mA sensitvity version we are talking about , but it’s type B.
I tried to explain this point in my long post above. To meet regs, you either need Type B at 30mA sensitivity (or better), or you need Type A at 30mA sensitivity plus DC detection at 6mA sensitivity.
The reason for the 6mA is not that you need that sensitivity in itself, but because the Type A is only guaranteed to deliver its rated 30mA performance while any DC is kept below 6mA. So the DC protection in modern chargepoints is not a safety measure in itself, it is there to make sure the Type A RCD doesn’t get disabled.
MrBadger
Badger out
Is it best for this to go at main fuse end, or closer to the charger?
I'm going to need another fuse box for this, and i'm short of room on main fuse board, so it may be better closer to the charge point, but that will leave me with a length of unprotected cable between Henley block and rcd - I didn't check if it was an rcd or rcbo, if latter I would be able to run mcb from an existing fuse box, but that still probably would not provide fault trip on the cable.
I'll get a sparky to do what is needed, but i'm after some idea of a plan to preempt them.
I'm going to need another fuse box for this, and i'm short of room on main fuse board, so it may be better closer to the charge point, but that will leave me with a length of unprotected cable between Henley block and rcd - I didn't check if it was an rcd or rcbo, if latter I would be able to run mcb from an existing fuse box, but that still probably would not provide fault trip on the cable.
I'll get a sparky to do what is needed, but i'm after some idea of a plan to preempt them.
There’s too many combinations to give somple universal advice. However:
- You almost always need overcurrent protection (MCB or fuse) at the source end.
- The RCD required to protect the chargepoint can be at either end; there’s some arguments in favour of the chargepoint end, but subjective.
- In addition to the requirement to protect the chargepoint, the cable may need RCD protection (at 30mA fast-acting), depending on the cable type and how it is installed, also earthing system.
- With two 30mA fast-acting RCDs in the same circuit there’s no way to make them discriminate (in the event of a fault, either or both RCDs may trip, rather than just the one closer to the fault). Discrimination isn’t mandatory (unless lack of it can cause danger, not applicable to the simple case of a single circuit), but is desirable. This can be a reason for putting the chargepoint RCD at the source end (to avoid having two). Again, a matter of taste.
But there’s lots of other ways to do it.
Sirricardo
MSM LR 19" Tow....The Unicorn.
Hi All,
Can someone tell me why most chargers available in the UK are limited to 32Amps?
I'm pretty sure Tesla chargers in the US can do 60Amps and I have seen posts reporting early version UK tesla chargers doing 40Amps.
Cheers,
Sirricardo.
Can someone tell me why most chargers available in the UK are limited to 32Amps?
I'm pretty sure Tesla chargers in the US can do 60Amps and I have seen posts reporting early version UK tesla chargers doing 40Amps.
Cheers,
Sirricardo.
The cars are different in Europe vs N.America. N.America have single-phase input at up to 80A (depending on model, age and options), corresponding European models have three-phase inputs at up to 32A per phase.
Early version UK cars definitely can’t take more than 32A on any input. There was speculation that you could go above 32A single phase by paralleling the phase inputs (though that would risk overload of the neutral), but I am not sure anybody ever demonstrated that and Tesla never supplied any equipment to do so. The original UMC did parallel the phase inputs but was limited to 32A; the Wall Connector doesn’t parallel the inputs unless you install it contrary to the instructions. More recent cars have supported 32A single phase with internal arrangements to reconfigure the chargers and remove the need for external paralleling.
As to why it’s like this, the UK supply network offers larger single phase supplies than most places in Europe, but even here going above 32A is challenging and large loads are better supplied at three phase. USA is unusual in commonly having 400A supplies at single phase available to homes.
It would be possible, but nobody does it - neither the cars nor the chargepoints are commonly sold in Europe for >32A single phase.
You can get 63A three-phase chargepoints, but they are always configured as tethered so you can’t then use a type1-type2 cable to charge a US import Tesla. Some people have bodged such an arrangement to charge Roadsters at 63A single phase from public chargepoints - Tesla didn’t change the chargeport on EU-model Roadster (the connector is a proprietary Tesla thing anyhow, as it pre-dates the standardisation on type1 in N.America/type2 in Europe)
If you imported the car and the relevant chargepoint gear from the US it would work here (though you’d have trouble with CE marking if you wanted to do so commercially).
Note that 50A showers and 50A EV charging are rather different loads due to the expected duration of use (though the showers are already a bit of a problem in diversity calculations since you can leave them running indefinitely, pouring the energy down the drain).
Maximum available single phase supply in the UK is typically 80A or 100A depending on where in the country you are, with the only option for more being to switch to three-phase. Still, you can often manage 2x 32A chargepoints, perhaps needing load management to reduce the charge rate while showers etc are in use.
So from a purely UK perspective it would be useful if the cars supported a bit more than 32A, but most other parts of Europe struggle to get even 32A single phase from a domestic supply. Even here, you won’t be able to have unfettered 32A charging once all your neighbours have EVs, so it would only be a short term thing.
Sirricardo
MSM LR 19" Tow....The Unicorn.
Ok Cool, thanks for the comprehensive reply.
Going back to your previous statement about a parallel configuration...is this feasable at 16A per phase if the neutral is 48a rated.
Is there a regulation preventing this arrangement if the supply can handle it?
Will the 3 phase type B breakers (for dc leakage?) function properly?
Cheers S.
It is feasible, but I have not seen anybody report the results having tested it to see what the car will do with a pilot signal above 32A and paralleled single phase - and note that there's at least three different charger configurations (original Model S, facelift S/ModelX, Model 3) which could all behave differently.
No.
Yes.
MrBadger
Badger out
Stumbled across 48A/11kW limit on Model 3 LR AC charging - SR has lower maximum, 32A/7.7kW. Not a huge/any difference over the 7.7kW that 32A would give?
Onboard Charger
apologies, but this page refreshes to error page after a few 10's of seconds. But it reiterates other resources that match these rates.
Onboard Charger
apologies, but this page refreshes to error page after a few 10's of seconds. But it reiterates other resources that match these rates.
Similar threads
