The anatomy of a Volt EVSE

[Henry_FL]

Well I have my 100 dollar ESEV, it has a solid ground fault indicator ( which I knew before purchase ) and have started the process of taking it a part.

My initial goal is to fix or disable the GFI feature and to then modify the board for 240V operation

I've spent a few hours tracing things out, looking at part numbers and have a good feel for how this is designed ( I am EE with 30+ years design background )

So here is my first impressions and comments

1) The unit is pretty easy to open up. Just use a flat blade screw driver and pry the top cover off. It is welded, but doesn't take much force ( on this unit anyways ) to break the weld. To put back together would just need some silicon to hold to cover back on

2) Removing the PCB is a major pain-in-the-ass. One has to cut some wires, remove some screws, but one screw is un-accessible as its under the bump where the J1772 cord enters. I ended up drilling a small hole to get a screw driver to remove the hidden screw

3) Once I had the PCB out and flip over, my jaw dropped with all the components. I have reviewed the open source design and it has about 1/20th the part count. These units have a relay for all 5 J1772 wire, a low-side current measurement, which is one of the reasons you can't just make the neutral the other phase for 240V, GFI, dual power supply's, a pair of atmel processors and associated parts

Summary, this is a single phase design and cannot have split phase 240 ( US ). It is though worth nothing but a firmware change should able to handle 230 vac as that's what is used in the EU ( still one neutral one hot ). the internal supply's are rated for the higher voltage.

I've heard from other forum member's that most of the units sold, all have the ground fault issues, in varying degrees, as i first guessed these are earlier prototypes

For 100 bucks, its worth the parts if you want to build your own charger and as time allows, I'll do my best to fix the GFI and then come up with a method for 240V operation, for now, its a time filler / fun time project

IMG_2056.jpgIMG_2062.jpgIMG_2063.jpg

[MichaelH]

Thanks for sharing. Very interesting. Good luck on fixing it. For the amateur this obviously comes under the heading of, "don't try this at home."

[DonC]

The guy who did the EVSE upgrade mentioned that the Leaf EVSE will handle 240 but the Volt EVSE wasn't sufficiently robust to do so.

[Noel Park]

Thanks for sharing. Very interesting. Good luck on fixing it. For the amateur this obviously comes under the heading of, "don't try this at home."

Yeah, we don't need any more Volt garage fire stories at the momnt, LOL. Careful Henry!

[DCFusor]

This is nearly identical to the one in the 240v voltec. The differences I see are that the 240 one has two relays (one for each leg of the power) - yet I see three in this...and nothing much else. Can't tell without a higher rez picture, but on the 240v one, the GFCI sensor is just a current transformer with the two mains leads going through the center of the donut, and that'd be real easy to disable - either short the secondary output wires to one another, or just reroute the mains wires so they don't go through the donut core and therefore can't trip it. I suppose if you wanted to get fancier, and just make it less sensitive, you could discover a resistor value to "burden" the secondary of the current sense transformer some more and accomplish that.

The 240v one used a small generic switching supply to get the isolated low voltage for the uP and pilot circuit stuff. It runs straight off 240 of course. Don't know what's in the 120v one there, but really the only other thing you should need to do is make that low voltage supply happy some way - either run it between one leg and neutral, or get one happy with 240v in the first place.
The 240 unit doesn't even have a place to connect a neutral to it.

And, if you do, you'll have something no one else has yet - a 240v unit you can switch "down" to 8 amps. The "high" setting is the same on either unit, in amps...

Oh, there's one other big difference in them, which I found out by how they act. The 120v unit can and does detect a floating neutral, even in the absence of any ground fault current.

[Henry_FL]

This ESEV has 5 relays, one for hot, neutral, ground and the two control signals, silly, but must be a UL thing, why would one have to have a relay on the ground to the J1772 connector?. In control signal relays are much smaller.

It slow work without schematics and conformal coating to figure out part numbers and the interconnection, the GFI is a Farichild RV4141 but its not wired as the application notes, since there is no need for scr and latching relay that a regular gfi has. The gfi is dual core, so yes it does both netural and ground checks.

I'll send DCFusor some hi-res pictures ... be interesting to know the differences in the two chargers

[Voltmeister]

Henry,

Once I had the PCB out and flip over, my jaw dropped with all the components. I have reviewed the open source design and it has about 1/20th the part count. These units have a relay for all 5 J1772 wire, a low-side current measurement, which is one of the reasons you can't just make the neutral the other phase for 240V, GFI, dual power supply's, a pair of atmel processors and associated parts

From the looks of things, that parts count is amazing!

[dtaubert]

This is nearly identical to the one in the 240v voltec.

Here is a picture of the front side of that board:

[Henry_FL]

Thanks for posting the 240V picture,

Its appears to be a completely different PCB, the input section is all different, the position of the relays on the other side its different ... I'd say it uses allot of the same parts, but there totally different boards

[Fulgerite]

Henry... THANK YOU for posting those photos! It looks like it would not be too difficult to replace the 16 gauge power cord from the top side of the circuit board without removing it. (Assuming you can scrape off all that silicone goo.) Great post!