DIY Electric Car Forums banner

10kW / 60A DIY charger open source design

984K views 3K replies 171 participants last post by  valerun 
#1 · (Edited)
Hi All,

[Update 07/01/13]: We have improved the design over the last few months, with beefier PCBs, better drive circuits etc. As a result, this is now a 12kW / 70A charger for under $1,000! Also, we now have an option for 100A output! Finally, a custom enclosure is now an option - professionally machined and powder-coated in awesome black matte finish to show off your charger. As always, you can get more info on the latest specs / build videos / etc at http://www.emotorwerks.com/tech/electronics. Now there are over 100 of these out there!


ORIGINAL POST BELOW:

Over the last few months, my team has developed a 10kW / 60A charger design that exceeds all known chargers available on the market today (that you can put into a vehicle, anyway) - both in power and features. We just hated to pay $4000 to be able to charge our conversions in less than 5 hours...

An incomplete run-down of the capabilities:
* 10kW rated power - tested to 15kW continuous with liquid cooling
* 110 / 220VAC, up to 400VDC input
* up to 350VDC battery voltage (up to 425V with very minor modifications)
* fully PF corrected option available
* 144x144 color screen for controls & readout
* Output power control
* All battery types supported with micro-processor controlled charging regimes
* Integration features - BMS cut-off, End-of-charge signal
* Thermal protection
* Timer shutoff
* Compact design - latest units are packed into a 12x11x8-inch box (air cooled) or 10x8x7" for liquid-cooled
* User-programmable through a USB cable

We have made design fully open-source and available to all. So if you are so inclined, you can make it yourself for ~$900 all-in (including all the heatsinks, enclosure, etc, etc.).

So far (May 2011), we have made and installed 4 complete units and helped ~10 people to make their own builds. Our first unit was installed into the Fiat Conversion described at http://www.diyelectriccar.com/forums/showthread.php/noob-starting-1977-fiat-124-spider-51822.html. It is charging our 110AH 200V LiFePo4 pack at 50A in ~2.5 hours. Max current tested is 65A at lower voltages (tested up to ~100V, I would imagine this would hold up to 144V or so).

Update (August 2012): about 20 complete units installed, ~40 made from kits.

If you like, we can help you with PCBs, parts kits, assembled / tested boards, and fully assembled units. Again, details are at http://emotorwerks.com/tech/electronics.

I hope that this effort will make EV conversions a bit more affordable to people and we will have a modest contribution to getting more EVs on the roads! Please consider supporting it by letting us help you with building out your units.

As always, we welcome feedback / suggestions / questions.

Thanks,
Valery.
 
See less See more
#2,803 ·
#2,806 ·
I would think that there would be at least some of each version number. If they were never released they must be just internal prototypes. It is very difficult if not impossible to help customers with various versions of the charger when the documentation apparently does not exist or is hidden on the website with no links except what has been revealed in scattered posts throughout this thread. Moreover, some of the PCBs are not labeled with any identification as to version or date, and many of the builds seem to be different with details hidden in documentation that may no longer even exist in original form.

It seems like there have been many changes and "bug fixes" done "on the fly" and new versions made without proper testing and documentation. As I have shown, there are many unlabeled and missing parts on schematics and layouts, and this seems to persist up to V12/13, and it is unclear what the changes were for each subsequent version. There is also the debacle of having various firmware revisions. Those are among the reasons I refer to this as "a mess".

The three boards I show were in a post from one or two years ago (which I referenced), showing differences in builds which were supposedly done according to the documentation as it was at that time. It's nice that you have a link to the latest and greatest version V14, but that does nothing to help those with earlier builds, and makes it very difficult for me to provide any effective help for your customers which you have been unable or unwilling to provide. You have a complete charger from one customer and you have not responded to his requests for you to repair it or return it. Others have been very frustrated with your service and response to repeated emails. That is no way to run a business, IMHO.

Yes, you sent me some obsolete blank and/or hacked boards, which had a few parts on them, but none of the "special" parts I might need to put together a working limited power charger. I purchased some of them at my own expense, but there was still not enough to build anything useful. The boards were from various different versions and some were not included at all. Here is what you sent:



There is no identification on any of the boards as to what version they might be, but the drivers seem to be possibly V9, because they have the two ISO124s, although the one on the left was apparently an experiment to replace it with the PC917 and A7520 as shown for V12-13. The display boards might be V12, except that I have a complete board from Crackerjackz that is labeled V12, but is completely different, with two relays, only one DB9 connector, and the counterbored holes to correct the "bug" with the ground plane shorts. See what I mean about a "mess"?

The V14 boards seem to have better correlation between schematics and layouts, but some things have moved from one board to another, so the only way to use them is to have both. It seems that the best way to address problems in the field is to make available these new boards so all chargers can be upgraded to the latest version. But there are still some basic design issues IMHO, and perhaps V15 might fix them, but so far I have not found the schematics or layouts.

Valery, I am trying to help, and it seems the least you could do is supply a complete set of V12 or V14 boards and perhaps one or two of the older, smaller inductors. Crackerjackz sent me a complete control board with display and a programmed Arduino, and someone else sent me a complete charger. So I might be able to help your customers who have had problems. But my opinion, and that of several others, is that the charger will need pretty much a complete make-over with new boards and firmware, and hopefully re-use the large expensive components and enclosure.
 
#2,811 ·
I am sorry Brock - I admit we are really slow with repairs. That's of course not an excuse for keeping you waiting for 3.5 months (since Oct 22 as I see in my email box tracking the unit receipt).

Info for rest: this was a partially assembled kit purchase. After we shipped it, at the final assembly stage, there was some damage due to potentially incorrect connection between the boards and the kit was sent back to us for diagnostics & repair. That's where we dropped the ball a bit...

I will look at your unit this weekend.

Val.
 
#2,809 ·
There has been some recent private discussion about collaborating on a retrofit package for existing EMW chargers, and a new (probably modular) design. The modular design would most likely consist of a single control board and up to eight 2 kW modules that can be paralleled to get up to 16 kW. This will make it easier and cheaper to develop a 2 kW device (120V 20A), and offer field-installable upgrades. It also provides redundancy, so if a module fails, you just temporarily lose 2 kW until the board can be repaired or replaced.

Looking at the V14 design, I see only marginal improvement. But my greatest concern would be the inconsistent and inadequate support and communication offered by Valery. The retrofits are probably at least 4 months from development, testing, and shipping, and the new design might be ready by this summer. The more interest and commitment I get from those who would purchase such items, the more enthusiastic I may be to accelerate the engineering effort. Please send me a PM so we can discuss future plans privately. I have started a separate thread for this development, and this thread should remain for support of the existing EMW chargers.
 
#2,812 ·
This is great, Paul! I am hoping that you and your collaborators will have the energy and dedication to follow through on this. The amount of investment that it takes to document all the design and build procedures is not to be underestimated. The amount of documentation grows 5x when you try to enable other people to build per your design. Good luck.

In the meantime, we have discontinued kit products. Kits will only be sold to existing customers who successfully assembled their chargers already. We have a number of customers who have assembled multiple units for themselves and their customers / friends and we will be limiting future kit shipments to that segment.

This lesson is very similar to the lesson we have received in our JuiceBox product. Selling kits for the first 9 months of product availability has created an enormous load on our tech support as too many people without sufficient skills or knowledge attempted assembly and blamed us for their non-working systems. After limiting new kit sales in a way similar to the above, we have no technical issues from over 1,500 EVSE units sold since then.

Val.
 
#2,813 ·
A big part of the kit failure was lack of documentation and constant revisioning. Web pages can handle hyper level modifications since they update seamlessly, hardware needs to be pretty well sorted out and documented before being handed off.

Anyone who can solder should be able to succeed with a proven design and good instructions. Not saying everyone qualifies, just that the level of difficulty in this kit was historically bad. Otherwise it is a glorified connect the dots/paint by numbers.

The feature chasing is a big issue, and using arduino in the marketing is indicitive of that mentality, when %99.999 of folks just want a charger that works reliably, with a light to tell them when it is having a problem.

BTW, my juicebox kit started failing, and I'm not short on skills, and it was largely pre-assembled. So there are quality issues you are going to have to face. You can't put it all on the people who paid you for a kit.
 
#2,815 · (Edited)
With such poor correlation between the available documentation of schematics, circuit boards, build notes, "bug fixes", and Arduino software, with so many "revisions", it is amazing that anyone was able to build the kits successfully. And the most evident change from the 9.x versions to 11-12-13 seems to have been the replacement of the ISO124 for the highly problematic PC817 opto-isolator. I understand that the ISO124 is about $15 while the PC817 is only 50 cents, but the same thing could be done with a differential amplifier, and if you really think you need isolation, you can get an AMC1100 for as little as $4 in quantity.
http://www.mouser.com/ProductDetail...GAEpiMZZMtyhvbnOfCRF61vo0acNs%2bL1cmpG4yE6c0=

I only got involved with this charger because I saw so many people having problems with it and the support from EMW was erratic and often ineffective. Then as I dug deeper and learned more about the design, I saw many aspects of the design that I simply disagreed with, but more disturbing have been what appear to be safety issues and errors that still seem to persist even into V14 and probably the so far unrevealed V15.

The PFC circuit uses "a few inches" of copper wire instead of an actual current sensing shunt resistor, so that part of the circuit does not work as intended.

The JuiceBox does not adhere to the SAE J1772 standards for handshaking, and the charger is also non-compliant.

I had hoped to be able to put together a couple of PC boards that would replace those presently in these chargers, but there are issues on other boards as well, and I don't want to put a bandaid on a sucking chest wound. I have to agree with others who regard this as a "train wreck", and since I will need some sort of charger for my own purposes for small off-road (tractor) EVs with 24 and 48 volt SLA and FLA battery packs, so I might as well proceed with what I consider a better design, from the ground up.
 
#2,816 ·
Reality check. - Re: 10kW / 60A DIY charger open source design

Hello guys,

I haven't been to this forum for quite a while until recently.

I am appalled that Valery is getting so much bad rap from many - utterly undeservingly in my opinion.

This kit is NOT for someone who wants to SAVE money! Never was.

It is for someone who wishes to experiment, to learn something new, who is not afraid of number of hiccups and serious issues to resolve, for someone who wishes to be part of unofficial R&D.

As a reward, you get something you have built yourself and which works reasonably well.

Yes, Valery is guilty of very poor revision control and documentation, also of being too ambitious trying to add too many features at the same time or too rapidly.

Having said that, he always goes more than extra mile to help resolve the issues anyone has, to his own detriment. In reality, this keeps him from the business side of the equation. I fully understand that. He being tinkerer, as many of us are, we have the constant urge making things better and better. We hate to admit that there are necessary things such as design freeze, rigorous documentation, etc.
____________________________________________

In my case, I am not even EE. I am ME with very basic electronics skills and I was able to build the superb charger in ten days, which has been serving me for last 16 months multiple times a day!

Yes, during the build, I managed to fry few components where Valery helped me promptly to find out what I screwed up. And yes, last week I discovered I managed to do cold solder joint between main board and rectifier bridge. All these were my faults.

I made my own enclosure, I also figured out alternative (passive) way to deal with J1772 pilot circuit issue as well as analog choke fans temp control (snap thermostats).

Few months ago, I also bought the basic Juice Box kit (PCB+PSU). Yes, I had issues, but guess what: it was not Juice Box problem. It was my other car's mass production charger problem and charging handle issue, Juice Box works perfectly as well.
____________________________________________

You guys need to be little more realists and much more considerate. Valery never offered this kit as a mainstream certified charger - there are plenty of them on the market. This was supposed to be a learning experience with all the associated risks and pitfalls.

For those who are afraid or not up to the task, you should have bought fully assembled and tested charger for more $$ from EMW. Most of these issues people are reporting are self-inflicted, due to ignorance or inflated egos.

Be grateful that Valery and his team provided you with the opportunity, you made a choice. You were warned BEFORE you bought the kit that this is not for the faint of heart....

Do not blame Valery for your ignorance! :)

George's $0.02.
 
#2,817 ·
Look, Vale, I've got a lot of respect for your learning ability, watching you sorting out inductors was very instructive and you've learned a lot from the community in the process. You are eager to make a business out of it, and have taken an amazing chunk out of the pricing for chademo products as well (to the elation of many a leaf owner certainly). But I'm a diy kit kinda guy, and really appreciate people producing things in kit form, as well for something that isn't UL listed (and couldn't be at this revision rate anyway), a kit makes a lot of sense, eliminates liability and reduces costs further. So don't poo poo kits when you have a hand in their demise, as a business. And nobody buys a kit thinking their time is worthless, to be a gunea pig. This all started as a diy $200 charger. Cost is a big factor in doing things yourself.

So to whom it may concern, don't buy a kit if you can't solder and follow instructions.

Don't sell a kit if you can't write good instructions and support each version with proper documentation (and limit the rate of change if you are encouraging user contributions (i.e. arduino), else most of the users hardware is made obsolete shortly after shipping). If you are gonna mark it up substantially, you have to do your part upfront. We all know about quantity discounts.

As a business owner, I get it, but as a diy enthusiast, I get the other side of it also. The current cost of components means there are no shortage of folks wanting to capitalize on it, but on a diy forum, it seems like the modus operandi is to share knowledge, and enable each other to do more.
 
#2,818 ·
Don't sell a kit if you can't write good instructions and support each version with proper documentation (and limit the rate of change if you are encouraging user contributions (i.e. arduino), else most of the users hardware is made obsolete shortly after shipping).
The only problems I had/have with the kit are documentation and versioning related. There were no docs at all with the kit I received, and no links on the website to find any. When with the help of the community some was found, it didn't match the physical boards.

I suppose as a 35+ year experienced EE design engineer, this kind of thing makes me a little crazy. If I did something else for a living it might not bother me so much. But while managing engineering teams of 50+ people, I always had to be a real hard-ass on proper documentation and enforcing controlled versioning practice.

Managing an 'open source' project for fun and profit is not as easy as it may look, this I understand. "Herding cats" is the analogy I've always used. But there are lots of good online tools that help manage the process -- most of them free. My current favorite is Atlassian which I use extensively for my current development projects.

Bottom line is with the state of the documentation, its not possible to expect anyone no matter how experienced to build this kit without needing some questions answered -- thus creating a support burden for EMW.

BTW- Valery has stepped up and is helping with docs and has offered to provide current boards -- which is above and beyond what I expected. While the kit started out rocky, I'm quite pleased with his current support.

I was one of the Kickstarter supporters for the Juice Box and am looking to build that kit here in a couple of weeks....
 
#2,821 ·
I have a new thread on my charger design concept:
http://www.diyelectriccar.com/forums/showthread.php/new-ev-charger-design-modular-130626.html

It may be a while before I can come up with a design and PCBs that are ready to distribute, but it is helpful to read posts such as yours that identify some of the issues. Although I have found some problems with the EMW design(s), most of them are not serious, and the fact that many people have built the chargers and have used them successfully attest to the fact that they do work.

Your experience with getting a shock from the vehicle chassis is a serious issue, but I think it can be solved by adding an earth ground, and/or a GFCI circuit. The voltage on the chassis may be just from capacitance from parts of the charger to its enclosure, but there is always the possibility of leakage or a short which could present a lethal amount of current. It may be more likely to occur with 220 VAC, especially if both legs are floating, or if neutral and the high leg are reversed.

The design I propose will have isolation, but will still have an earth ground. I would suggest that anyone with this charger should ground the vehicle chassis.
 
#2,823 ·
I tried to use earth GND to chassis and it screwed with my BMS comms. Since our earth GND and N line are the same potential i guess chassis GND fell to some -18V in respect to vehicle +12V and so BMS PCB elcaps got drained, comms fell and charge was terminated... I decoupled earth and put charger on sealant blocks (rubber blocks with bolts). That solved the whole thing.
Also i noticed i got no shock from touching charger when i was charging from 3phase - 400VAC - go figure...
In the end i put a lid over my charger so none could touch it while charging and that solved the whole affair.

A
 
#2,822 ·
sexstrap; said:
1) Pre-charge i have no form of pre-charge and rely on the inrush resistors taking the whack out of power on, I think that precharge circuits have already been discussed at length on this thread but as yet I have not manged to investigate never mind implement one.
Well that is easy, you have D4 and D5 transistor outputs and if connected to 12V arduino relay module you can drive input contactor as well as output contactor. Also on newer V14 PCBs there is space for two relays that have exactly this purpose. Then you can setup precharge easy: I precharge from the AC side since it is easier to get contactor. Put one 10W 470R resistor across the input contactor poles like here: http://cr4.globalspec.com/PostImages/201101/VFD_diagram_8149454B-C374-8AEC-29C95AC3EEE0AA04.GIF
The difference is i use three 470R resistor across 3pole contactor on AC side. That way resistors and contactor are rated for 230VAC not 600VDC!

sexstrap; said:
2) I am using this charger to charge a 10KWH LIPO traction pack (not LIFEPO) and as this chargers firmware was written for LIFEPO, The CV cut off (in the firmware I am using, V12 AFAIK) does not allow for greater than 3.99v CV cut off, I want to set 4.2v cut off. TBH I did begin to follow this up with Valery but I never followed through with Valerys last request for more info, I think he wanted a copy of my firmware (does EMW not keep all the different firmware revisions?) so this has not been addressed as yet. To get around this issue I simply lie to the charger about the total number of cells in my TP to compensate for the additional 0.2v per cell which at least allows the charger to work.
Well let me show you the way...
I figured out the firmware setting to go over the 3.99V in setup. It goes like this:

// wrap at 3 (for 100s) or at 9 (for 10s and 1s)
if(x == 0 && digit[x] > 4) digit[x] = 0; // from original 3 , i change to 4 - so you can select 415 for Lipo cell
if(digit[x] > 9) digit[x] = 0;

Also be warned! LiPo cells dont like to be charged with full CC untill the end. My friend learned the lesson the hard way when complete LiPo pack went to 4.5V per cell on normal CCCV charging. BMS couldnt help even though it did its job. Of course he had setup CC 20A the whole time... Cells are now on 1/2 capacity and quite a bit bloated.
Also i found that LiPos Ri is very small when in the 3,7V zone. While at 3,95V or above Ri rises and i guess this is reason for difference from LiFe.
Solution? You have to set charger to reduce current to less than 1/10C when cells reach 4V per cell. That way BMS can do its job and charger can be stopped in time. Also i will use one EV200 Kilovac contactor to switch charger off even in case IGBT would fail closed.
You can find all data and code in post #1718. Also on my blog here: https://mazdamx3ev.wordpress.com/2013/11/21/reduciranje-polnilca/

sexstrap; said:
3) My charger never finishes charging it simply drops the duty to zero and after around 2 1/2 - hours I revisit and switch off manually, I also corresponded with Valery on this issue which he was going to look at but as with point 2 I have not followed it up as yet.
I had the same issue. I used V11 code and when i changed to V12, when at CV it started to pause and resume to CC. I was using driver board with one ISO124 amp. Then i changed to V13 driver with opto and charger stopped after CC, but it never entered CV. So i just used my reduction loop to finish charge.

sexstrap; said:
4) I would like to implement J1772 and bought the relevant (and expensive) cable and socket which are installed in the car but I don't (yet) have any form of EVSE at home I simply have the J1772 lead and plug wired directly to AC mains on a 40 AMP RCD, I use this RCD to switch the charger on (and off until I started catching up on this thread last night) I have just learned as of last night that it is kinder on my RCD to stop the charge first before cutting power, I have not been doing this due to point:-
That is easy, just use the schematic for J1772 handle circuit. Of course for that to work you have to have AC input contactor and precharge present.
https://mazdamx3ev.files.wordpress.com/2014/04/unnamed.jpg

sexstrap; said:
5) My charger seems to put voltage on my car when plugged in and charging I tried to avoid this by attempting to isolate the charger components from the case (namely the metal heat sink) and although this appeared to do the trick when the charger was sitting outside of the car (i.e. I could touch the metal case without getting a jolt) once the case was bolted to the car the problem returned and the case (and therefore car) now give me a significant belt if I inadvertently touch the car whilst it is charging (not good) I had some correspondence with Arber on this matter as he also had this issue and avoided it by properly isolating the chargers case from the car itself but this would not stop me getting a belt if I opened the boot and touched the metal case of the charger, Again this has not been addressed as yet but is high on my next job list.
Well that cant be helped. Charger is not isolated. I succesfully separated charger case from the chassis so my BMS comms could work propperly but i still got jolted if i touched the charger case and chassis at the same time. There is AC component that propagates from the charger. It is cca 60VAC and there is no real amps there so no immediate danger. It is not pleasant though.

I think both LCDs can be used as they are. You only have to select propper switch in code. I am using V12 code and really i have V11 control and driver PCBs. I have succesfully built several nonPFC units V12 PCBs using new uLCDs. The difference is they are much more sluggish than first builds.

Also i succesfully made arrangements in code to further decouple buttons as they were picking up EMI.

OH did i say i run my charger from full 3phase 400VAC and it charges at 147VDC 40A. It runs at 12kHz but i changed inductor to 6kW Ecore from one large UPS. It works.

A
 
#2,827 ·
...Well that cant be helped. Charger is not isolated. I succesfully separated charger case from the chassis so my BMS comms could work propperly but i still got jolted if i touched the charger case and chassis at the same time. There is AC component that propagates from the charger. It is cca 60VAC and there is no real amps there so no immediate danger. It is not pleasant though...
I don't get a jolt when I touch my car chassis or the charger enclosure. When I read this my car was charging so I went out and touched both at the same time, no jolt, which I would expect since I have the charger enclosure bolted to chassis with metal bolts through rubber vibration pads. Now if I touch a battery terminal and chassis ground I expect I will get a jolt from the non-isolated AC, same as I did with my Manzanita unisolated charger. Maybe you have a leakage path between battery pack and ground?

Also, since I changed (more than a year ago) the delay between CC and CV to 9 seconds, the charger always does the transition and finishes CV to 0.05C. The charger has been working fine for about 1 1/2 years now, including J1772. I've charged at around 10-12 different public EVSE's now without a problem, and I get about full power from them since I added the lines of firmware to adjust Vin from the PC817 as a function of temperature. I'm using V11 firmware.
 
#2,824 ·
I am just starting to examine the complete charger I got from John a good while ago, and I found that there is 500 pF to GND from the inputs, 900 pF to GND from the outputs, and 4.2 nF from input to output. That is about 600k at 60 Hz so the current is limited to 380 uA. But that is with the charger OFF, and the input rectifier bridge is not forward biased at the signal level of the LCR meter.

When I applied power to the 12 VDC by means of the switching supply, the display lit up and showed the initial screen for the LCD144, but the buttons do nothing and the black button just turns the display off and on. Obviously I have much more to do. I will say that the display is rather small and difficult to read, and it is not protected as it is mounted.

I will need to take the charger apart to some extent to measure voltages and learn more about the details of its construction. It appears to be a bit difficult to work on, and it would be a lot easier if the front panel were removable so it could fold down and expose the control and driver PCBs.

BTW, the reason there was no shock from the three phase source may be that the vector sum of all voltages of a balanced source is zero.
 
#2,826 ·
Thanks Paul & Arber, all this information is really helpful even if I don't fully understand it all at first glance a little research is all I need to do to put those measures in place.

Paul the new thread looks great I will follow it with interest and it (as well as building the EMW charger's) has sparked a new found interest in software simulation of circuit design, I am a complete novice at the moment but I can read & follow schematics and as I work at a university supporting science and engineering I also have access to Labview and specifically Multisim a circuit simulation package, I never realized just how complicated a circuit you could simulate using these sorts of packages, Wow!

Paul on your previous post 2825 you said

but the buttons do nothing and the black button just turns the display off and on.
You may find that the two function buttons (not the black one) have been orientated 90 degrees out they only work in one direction, Try removing them and re-orientating, I managed to second guess this before my first assembly again not in the docs AFAIK (well the ones I had anyhow).

Thanks again Arber for your detailed reply 2823 to my post 2821 when I get some time :rolleyes: I will get busy with these mods, I wonder if putting a Faraday cage around the large toroid inductor and earthing it to the AC line would remove the EMF?

I know from EMF issues with the HV cables in EV's (motor and battery) that sleeving the HV cables with a grounded copper braid (grounded to car chassis) can help eliminate EMF could we do the same for the HV wires in the charger to help eliminate this induced voltage?

I also wondered when I was building my first charger if a large mica sheet to separate the IBGT & output diode from the heat sink might help might with this, any thoughts?

(I did not do any of these mods due to my guessing might likely blow something up in the charger or worse)
 
#2,825 ·
Shouldnt matter, since i IGBT should be isolated from heatsink shouldnt it? But we still get some AC component there... Do you think it is induced or capacitive? I was wondering what, when i got zapped because it was just like static discharge it just didnt go away. And if i grounded it my BMSs comms went wild since BMS master uses chassis GND.

I measured 45VAC sometimes and 60VAC again...
I attribute this to 2kW water heater now which was leaking HV towards chassis (trough water) and i am replacing wiring now. Still have work to do, but i hope i can measure and report results soon.


A
 
#2,831 · (Edited)
Hi Everyone,

I purchased Kit about 1 year ago now, its been on the back burner for for a while.. and I'm now looking to finally see this thing work.

I'm having trouble to program the arduino chip and i was wondering if anyone here can assist me through getting this working. I'm no expert when it comes to arduino coding programming but i can understand some of the basics.

Some basic info i know for when programming my charger is it has:

-A7520 Chip on driver board
-Uses Allegro_100U Current Sensor
-My Unit also has PFC

So far i've been able to installed the usb dongle driver and the Arduino alphe 0022 programming software on my pc. Working fine connectivity wise.. no issues here.

I then download the zip file charger_2013_07_31_v12.. unzipped it and opened the charger_2013_07_31_v12.pde file and compiled it and tried to upload it.. It does upload however when i power on the lcd under the usb programmer power i don't have any display.

Can anyone help please?


I also what note .. i have some extra resistors left over in my kit (i have no idea where they go) as well as pictures of my charger build, i still need to refine the wiring a bit more.. however i want to make sure it works first!

thanks
-steve
 

Attachments

#2,832 ·
i just wanted to give a update

i was succesfully able to get the arduino programmed

ive also go through all the testing and have the charger powering on 120vac...

im not able to test a load yet.. for some reason the output measures -42v ?... this is after go through configuring settings including shorting the output..

does anyone know why this is happening?.. any suggestions?

thanks
-steve
 
#2,834 · (Edited)
Hey Val

thanks, i connected the one wire and it works!.. i cant believe one wire missing and nothing would work lol

its great to see the charger finally working!

i have some questions for you if you dont mind
...

what is the best way to setup a simple precharge? i know 2 current inrush limiters are included in the kit... do they always stay on the ac lines though? do i need another set for the dc side? if im charging a 400v battery how i properly charge without breaking anything?

is it difficult to have the output voltage increased to 425v output?

also is there anyway to program lipo type battery setting in charger? i may plan to buy kia soul ev and it use lipo type battery..

thanks for your help val

oh... one other thing i have couple resistors left over after completing charger build? did i miss something? or are they for other mods?.. ill attach a picture below
 

Attachments

#2,836 · (Edited)
I dont use inrush PTCs.
I used two relays powered by D4 and D5 as they come on V12 controlPCB. One i use as input contactor trigger and other is output trigger. That way every aspect of I/O is secured...

Precharge i made like this; one 470R 10W resistor is connected across both AC relay contacts. So charger is always under AC voltage, but initially voltage rise is cca 30V/s. This extends life of input caps and prevents fuse dropout on charger connection.
Charger has two timers one 10s for calibration and second i set for 20s (i have to precharge to 580VDC). When second timer is up, D4 and D5 relays go on but input bank is already at comfortable 500VDC so jump of 80V doesnt cause much amps...

As for output relay i dont precharge it. I have a diode on + so IGBT is the one that charges the output caps and this is setup in software.

basically my setup is like this:
http://1.bp.blogspot.com/_3A-1Nfs_4...yJYldaI5Lw/s1600/Precharging2-noContactor.jpg

http://zeva.com.au/Tech/ContactorJewelry/



A
 
#2,835 ·
Hi Valery

I returned to the charger now and i would like to ask something about V14 charger...
Why have you put almost all caps to the output side of the charger after inductor? Is there any reason why you have 7000uF on the output?

I only use 2000uF output caps on V12 charger with 2000uF 900VDC caps wired on input. It is enough for 3phase but single phase works up to 15A output. After that voltage falls off and IGBT suffers...

tnx

A
 
#2,838 ·
Ok, i got this.... Couple of months back there was a shorted cap on my driver board. It was the tantalum cap across 5V/gnd pins of the isolated amp A7510. I noticed it because i powered my driver board with desktop PSU and it read 0,5A draw! I checked isolated DCDCs and i got one LM7805 drawing current. I took op amp out of the socket and measured continuity - bingo! there was a tantalum cap shorted to gnd. Dont remember which number, but check with multimeter...

A
 
#2,841 ·
Electronics Computer hardware Electronic engineering Computer cooling Technology


Electronics Electronic component Technology Electrical wiring Electronic device


Electronic engineering Electronic component Electronics Electrical network Green


For anyone interested, we have succeeded in getting Valery's charger working, ending up with a hybrid V12 (power board and driver board) and V14 control board and arduino with some V14 modifications which are designed to help protect the IGBTs, which are the most sensitive component of the entire kit. The V14 control board seems perfectly compatible with V12 and seems to give a better result. I have attached some photos of the unit including important modifications such as the back to back zeners on the back side of the V12 driver board. We haven't tested for high current resilience yet but we would be more than happy with 30 amps rather than pushing the instrument to the limit at 60 amps. Quite alot of work and effort to get the charger going but worth it for the incredible versatility the instrument allows, no need for factory charging etc, and the instrument is relatively light compared to other models, and of course by assembling it yourself, you learn a bit about how it all works, so that if it breaks down, you can replace a faulty component or at worst the entire module without having to purchase a whole new instrument. Anyway the V14 control board gave us a good result, and is probably the cheapest module of the kit.
 
#2,844 ·
In the case of a charger, which implements j1772, and is the only thing that is going to touch the wall outlet, I think integrating that functionality in the charger makes a lot of sense. A useful abstractions from the charger is a throttle interlock as long as the cable is plugged in, and any bms inputs, but otherwise offloading the pre-charging of the charger capacitors to the car will lead to spaghetti.
 
#2,845 ·
When the vehicle is first connected to a J1772 there is no power available, only the pilot signal.
Handshaking is ideally performed by the vehicle equipment rather than the charger as it has it's own source of power (ie: can be waked up by an interrupt and return to sleep mode thereafter).

It makes sense to run a few checks and immobilize the vehicle (Throttle override) before asking the EVSE to enable AC power to the charger.
 
#2,849 ·
Hi,

Sorry if it had already been answered but it is quite hard to find something in so many pages. How can you reset the calibration in the EEPROM ?

I have got a V14 PFC SmartCharger. I did powered up the arduino part of the charger before I got a battery ready for testing. I did the first part of the calibration (short output for zero calibration) but as I got nothing to connect for the next step, I just powered off the charger when it asked me to connect a battery. Now when I try to do the calibration, the charger tells me there is 507V on the output when there is actually 0V, and no longer ask me to connect the battery.

I get something like 0.5V at the "V" and 5V at the "mV" pin of the voltage sensing harness when the output is 0V, and 0.9V at the "V" pin when the output is connected to a 46V battery. Are these normal values ? I am at the first step of the test so input is not connected yet, only the 12V supply.
 
#2,850 ·
A quick update : I found my way through the code and forced the calibration variables to reasonable values. This did not solved my problem. The adc seems to always output 1024, and that's why the charger asks me to drain 507V at the output.

The voltage at the A1 analog pin is fine (0.56-0.92V for 0-46V output) and the arduino works fine when unplugged of the charger. Something must be wrong in the code. I used V14.8. Anyone faced this problem or a similar one ?
 
#2,851 ·
I had a couple questions; hope someone here can help

i know 2 current inrush limiters in the kit... are these to be used on ac side or dc battery side? do they have to stay connected?

Where do i find how to modify the output voltage of my charger to 425v+ ?

i have bought a kia soul ev, can i use this charger to charge is lipo type battery? how do i do it?

from the assembly of my charger.. i have couple resistors left over did i miss something?

see this post for pics on my charger and resistors i had left over..
http://www.diyelectriccar.com/forums/showpost.php?p=579778&postcount=2832

thank you
 
#2,853 ·
Hi Vtecsauce

The Current Inrush resistors are meant to be used on the AC side only, inline (in series with) with the live and neutral wires, However if you can implement a pre-charge circuit AFAIK you don't actually need them, They are there to slow down the "in-rush" of current to the charger to prevent any damage being done at switch on, The reason you might want to dump them and use a pre-charge circuit instead is because they essentially waste 40Watts of power for no reason other than to limit the inrush current for the first split second after power on, After that they are essentially 2 x 20 watt bulbs burning for no reason. A cheap and dirty fix for preventing damage to the charger. That said I still have not got around to adding pre-charge to mine, and yes it is on the list of things to do :eek:

As for the changes to implement LIPO charging, I cheated and just lied about the number of cells to allow me to get the correct pack voltage, However Arber333 showed me the way in post #2823 on this thread (just a few pages back from here) Again I have as yet to implement this as my EV is in use daily.

I would not worry to much about extra components left over, I think almost everyone had a few bits left over (I was sent complete PCB's that were not needed! plus a handful of unneeded extras)

Good Luck.
 
#2,852 ·
I watched your "death bike" video, and I see that you have been doing quite a lot with various EVs, controllers, and motors. I can't help you much with the charger -yet - but I do have two of them that others have sent me to evaluate and possibly fix. I'm still recuperating from spine surgery so I'm on limited activity, but hopefully I'll get a chance soon to go through one of the chargers and try to put together a point-by-point process to check out the boards and components.

As I have pointed out in previous posts in this thread, there are quite a few inconsistencies and possible design weaknesses in this product, and even more recent releases have not fully resolved them (at least to my satisfaction). The software is another issue that some people are working on, but I have not heard anything recently. I might make an Arduino sketch that is designed to "exercise" various parts of the charger to assist with troubleshooting.

I'll follow up with any progress I may make.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top