[Frigzy]

Hi everyone,

I see a strong interest in off-the-shelf on-board charging module among the electric vehicle enthusiasts. Me and my friend, firmware and electrical engineers, who worked over 10years for the leading EV companies, realized that maybe we can design one and bring it to you guys.

The idea is to develop a ready-to-go charger, that would take AC voltage at the input and generate DC voltage at the output. The charger should support digital communications (supposedly CAN) with full set of measurements/configuration/safety mechanisms.

The charger can be used together with a charging port, that we're also going to develop. It's going to be a number of different charger port options for different charging interfaces (J1772, COMBO, chademo).

So, it's going to be a complete off-the-shelf charging solution. You only need to add a BMS and a battery to get your vehicle running (well, maybe a motor and a drive inverter also).

Here is the proposed specification:

Parameter                  Min  Max    Unit
============================================
Input voltage              85   300    Vrms
Output voltage             100  430    Vdc
Output current                  45     A
Output power               0    7000   W
Output ripple (full load)       1      Vpp
Ambient temperature        -40  +100   C
Efficiency                 90   95     %


Communication bus   CAN
Protection          Undervoltage, overvoltage, overcurrent, output short circuit
Safety              high-voltage interlock
Measurements        Input and output voltage/current, temperature
Certificats         UL, FCC
Isolation           4000V between input and output. 4000V between input/output and a digital interface

The charger is going to be built around proper automotive-grade electronic components with emphasis on reliability and safety.

Unfortunately, we can't make this an open-source project due to a number of reasons, most major of which are:

• MCU, CPLD and other key components are going to be specialized, high reliability, automotive grade ones, i.e. less known to the general public. While I'm sure, there are plenty of high-skilled FW/HW developers, who worked with, say, TI C2000 architecture on this forum, it's unlikely, that the project can get any meaningful contribution/recognition from the open-source crowd.
• It may not be safe to work with exposed high-voltage / high-current circuits without a proper HV training.
• Some of bring-up/calibration procedures may require high-end lab equipment and tools.

The positive side is: it's going to be a non-expensive, high-quality, off-the-shelf and ready-to-go on-board charging solution for your EV project.

It would be wonderful if you guys can suggest on viability of the project (i.e. would you be interested in getting one) and what features/specs you'd like to see.

Since it's a very early stage of the development, we can try to accomodate pretty much any of your dreams

 

[Frigzy]

The topology is going to be
AC input => PFC => 450V DC => resonant LLC => DC output .
The charger will consist of 3 identical boards, 2.5kW each. We may also make a smaller/cheaper version with just one 2.5kW board.
Right now we're working on the PFC stage. It's all very flexible at this point, please write bellow your suggestions or what you'd like to see.

 

[ga2500ev2017]

I guess I'll ask the obvious question. What advantage is your charger going to bring to the table over a repurposed Tesla Gen 2 charger with open source control hardware/firmware that's being developed in one of the other threads here? It's 10 kW of power, can be stacked, is automotive grade, and has firmware that can be modified.

Two items that would personally be of interest to me is an add on DCFC port to a current non DCFC car and a medium power CCS/ChaDEmo charging interface. It would be useful to add a fast charge port to Fiat 500e for example. Getting an 80% charge in 20 minutes would redefine the range ability of the vehicle. And topping that off with a 20 kW home charging port would make it even better.

Just my suggestions

ga2500ev

 

[Hollie Maea]

What would really be useful is a variant that goes up to 800VDC. While there are a number of ~400VDC chargers available (plus Tesla salvage ones that are favored by people on this board) there is almost nothing available at 800VDC which is becoming more common now for high performance vehicles.

Obviously that's more work, but there is a real hole in the market for this.

 

[Frigzy]

I guess I'll ask the obvious question. What advantage is your charger going to bring to the table over a repurposed Tesla Gen 2 charger with open source control hardware/firmware that's being developed in one of the other threads here?

It' a very good question - thank you for bringing it up!

I believe, the advantages over the Tesla gen 2 charger are:

1) Tesla charger was never meant to be an off-the-shelf charger. That means zero technical support and a lot of uncertainty. If Tesla decides to trim the firmware and add, say, authentication between the ECUs - there is not much a DIY person can do.

2) We can implement any number of custom, DIY-oriented features. Say, communications over a user-friendly RS232 / RS485 and built-in battery charging logic with programmable parameters / charging curves. Tesla charger is basically a dummy power supply, that produces desired voltage/current at the output. Charging process is controlled entirely by the BMS. (I'm not sure if above-mentioned open-source controller does the charging - please correct me on this).

3) Gen 2 was designed in 2012-2013 using the technologies of those days. There are much better/efficient electronic components available in 2018. Say, PFC stage of gen 2 is a typical configuration with a diode bridge. We're going to implement a totem-pole GaN bridge-less PFC and we're looking to get 97-98% of overall charger efficiency. That means twice as less heat, much smaller size and air-cooled option which, I hope, DIY-ers will get excited about.

4) We're going to make a smaller, single-phase version for small vehicles / motorcycles. It's going to be absolutely miniature, light and cheap.

5) We potentially will provide CHADEMO / CCS vehicle-side charging ports, that will add support of DC fast charging into your project.

P.S. 7kW is a very conservative number. In reality, we're designing PFC/LLC stages to be capable of 3.3kW. It's not promised, but in the final design we may get up to 10kW of total charging power.

 

[Frigzy]

What would really be useful is a variant that goes up to 800VDC.

While it's definitely possible to develop a 800V charger, it makes everything more difficult (read: expensive) in order of magnitude. As of today, I don't see too many 800V vehicles/chargers around and it's hardly within a DIY scope. Maybe it would be a good future project - let's start slowly

P.S. I just realized, that there is nothing that would stop from using 2 of 400V chargers to get 800V output. Outputs of the charging modules are isolated. You can combine 2x 7.5 kW chargers to get [email protected] or 2x of smaller, 2.5kW chargers to get [email protected]

 

[OliverH]

A 800 V option would make sense with corresponding battery sizes (100+ kWh) which are not usual in DIY projects.


An off the shelf charger would be nice, but keep in mind that there're countries which request CE/ EMC approvals.



Full respect what you intend to build up.

 

[Hollie Maea]

While it's definitely possible to develop a 800V charger, it makes everything more difficult (read: expensive) in order of magnitude. As of today, I don't see too many 800V vehicles/chargers around and it's hardly within a DIY scope. Maybe it would be a good future project - let's start slowly

P.S. I just realized, that there is nothing that would stop from using 2 of 400V chargers to get 800V output. Outputs of the charging modules are isolated. You can combine 2x 7.5 kW chargers to get [email protected] or 2x of smaller, 2.5kW chargers to get [email protected]

I know it's harder, which is why there's a market hole. Just something to consider, especially if you get a lot of "what sets your charger apart".

And while there aren't too many 800V systems currently on the market, that's changing, as it is the best way to get very high power. Porsche is going that way, and there are enough other companies looking that way that there is active work going to get the CCS standard up to 800V. Not sure how much that will trickle down to the DIY space, as most people are content with Tesla drivetrains these days.

Two chargers in series would be a solution, a company named CurrentWays used to give that capability for their chargers but they no longer are in the market. One downside is that 15kW is more than most people need, so there's a lot of weight, volume and cost consumed by going with two (relatively) large chargers in series.

Right now the only 800V charger available on the market that I am aware of is made by EDN and costs over $7000USD for a 11kW unit. And it's big and heavy.

PS, to be clear I think a 400V charger would also be great if you do a good job on it and the price is right.

 

[Frigzy]

An off the shelf charger would be nice, but keep in mind that there're countries which request CE/ EMC approvals.

I only have experience with UL certification. However, I believe that there is an expedited, simpler process of CE certification for products, that already carry a UL mark. Also, there is a "UL for Europe" mark, which should technically allow to sell products in Europe. I will post updates once we get to certification stage.

 

[Frigzy]

One downside is that 15kW is more than most people need, so there's a lot of weight, volume and cost consumed by going with two (relatively) large chargers in series.

We're going to produce smaller/cheaper 2.5kW version of the charger. 2 small chargers in series will provide 800V @ 5kW. We will do our best to get chargers reasonably priced. There is definitely going to be a beta-testing program - we will give a few chargers away with a massive discount or even for free in exchange for feedback.

---
PFC schematic is finalized, working on the LLC part. The hardest part, of course, is a transformer.

 

[OliverH]

I only have experience with UL certification. However, I believe that there is an expedited, simpler process of CE certification for products, that already carry a UL mark. Also, there is a "UL for Europe" mark, which should technically allow to sell products in Europe. I will post updates once we get to certification stage.

To sell an electric device in Europe you need CE approval.

 

[MattsAwesomeStuff]

To sell an electric device in Europe you need CE approval.

Ditto US/Canada.

Also to legally plug it into a wall.

That means twice as less heat

Some might even say "half".

 

[piotrsko]

Ditto US/Canada.

Also to legally plug it into a wall.

To legally sell in most jurisdictions it must be approved, but not all. There is no requirement to cert to plug in most communities, otherwise you cant build or wire your own electronics.

 

[Frigzy]

Ditto US/Canada.

CE is a European certification mark. In US there are UL/FCC. Not sure about Canada though.

 

[brian_]

CE is a European certification mark. In US there are UL/FCC. Not sure about Canada though.

In Canada, common electrical certifications are CSA and UL Canada. Radio-frequency equipment which might require FCC (Federal Communications Commission) certification in the U.S. would typically require certification in Canada, but FCC certification is generally accepted by Canadian authorities to make it easier on manufacturers.

Note: A company which has "Canada" stuck on the end of its name is usually the Canadian subsidiary or operation of a company from outside of Canada; that is true of UL Canada.

 

[carguy911]

I am driving my fifth and sixth electric and would love an affordable, programmable charger. Put me on your list for purchase.
Phil

 

[Frigzy]

I am driving my fifth and sixth electric and would love an affordable, programmable charger. Put me on your list for purchase

Sure - I will post updates here. So far we're still designing a PCB for an engineering prototype. It should work the same as a production charger, but it's bigger and has a lot of test point, debug connectors, etc.

 

[DaveBlack]

Sure - I will post updates here. So far we're still designing a PCB for an engineering prototype. It should work the same as a production charger, but it's bigger and has a lot of test point, debug connectors, etc.

How is this project going?

Ill need a charger for around 130v @ 30amp and a bit like others the options are limited and or expensive!

Dave

 

[Frigzy]

How is this project going?

Ill need a charger for around 130v @ 30amp and a bit like others the options are limited and or expensive!

Dave

Sorry, work & family are taking their toll and I was not able to work on the project recently.

Now I'm slowly getting back to it. Schematic is ready at this point and I'm working on the PCB layout for the first engineering prototype. Everything needs to be double-checked and reviewed. I'm committed to send the board out for manufacturing by the end of the month.

The charger is going to be designed for 300-400V output at the best efficiency, but this can be customized at manufacturing stage.

 

[Frigzy]

837 components, 610 nets, 6 layers.. routing is going to take a while.