[valerun]

Originally Posted by Zak650

I made a 9 pin female header for the LCD Screen.




I still need to make a support for the top of the screen, but it's easier to remove the screen if necessary.

Great tips, Zak - thanks for sharing! We will be including your tips / photos into our build manuals if you don't mind...

We are starting the next batch of 5 units now - will be posting pictures as we go. After done, will post a full photo/video set on site.

Thanks,
Valery.

[Zak650]

Hi Valery,

No problem using the pics, glad you like the ideas. You might include a long enough strip of the female header material for this option. I do believe an upper screen support is needed. The screen also could simply be glued to the enclosure and stay there if the control board is removed. If something happens to the lcd it would certainly be a lot easier to fix without having to unsolder it's pins from the control board.

[rwaudio]

Originally Posted by Zak650

Assembly process photo: Keeping Arduino pins vertical for soldering - Note: no pins in A6,A7 holes.
Love my new Hakko FX-888 Soldering Station and 4 tips, well worth the money.

I borrowed your idea and tweaked it slightly, by raising the PCB off the desk you can use the PCB itself as the guide without the need for stand offs.

[rwaudio]

And some more build progress. I'm not using the standard EMW enclosure so my assembly is slightly different than the instructions. I'm building the enclosure around the heatsink, using the heatsink itself as part of the enclosure. I'm quite impressed by the physical layout of the high power pieces along with the driver board. It would be nice if the main control board was a bit more integrated or lined up nicer with the driver board since they have the most board to board wires. The overall layout of everything that mounts to the heatsink is very well done though and is a very efficient use of 3D space without limiting access to the mounting bolts for the IGBT's or rectifiers.
(Reminder, this is the air cooled PFC version of the charger)




The only "issue" I have with the kit is many of the parts don't fit the footprint on the PCB, I believe mine was one of the first kits so lets just hope I got leftovers from the previous design. Val, make sure you update to the proper components to fit the PCB and make assembly simpler and nicer.




I used my CNC router to drill markers holes for the main layout, the holes were then tapped as M4, M5 as required along with the shallow temp sensor hole.




I made a small change to the diode block "wiring" I used a pair of 0.1875" x 0.5" x 0.9" copper blocks drilled with the appropriate holes to both mount to the diode block putting the two diodes in parallel as well as having a hole sized to the 8 Awg wire. This is where a good soldering iron comes in handy, a low power iron would not solder 8awg to a copper block. The legs on the Hall sensor are also too big for the PCB holes and must be trimmed down slightly to fit, don't take off too much material though!!




Instead of mounting the Inductors with the leads facing up I decided that one connection from each inductor could be soldered directly to the PCB. This simplifies assembly and will reduce the wiring a bit but it will slightly complicate servicing the power PCB if it's every required to replace an IGBT or something else.




Instead of the bent steel straps as per the instructions I decided to build a "beam" to clamp both inductors to the heatsink. This beam will also become the mounting point for the "top" of the enclosure as it's shown in the photo.
I made it the same length as the heatsink so the sides of the enclosure will bolt up to both the heatsink as well as the ends of the beam.




As you can see there is nothing fancy about the "beam" it is made up from some aluminium I had laying around. 1.25" x 0.125" C channel (1" internal) gives the perfect spacing to straddle the inductors without interfering with the wires. A piece of 1" x 0.5" and 1" x 0.375" bar stock fill in most of the channel so that it both sits on the inductors where I want it to and doesn't sit on the plastic part. I'll add some screws to the middle and ends to make it one piece so there should be minimal flex when it's tightened down to hold the inductors with the single center bolt.


IMAGE REMOVED BECAUSE IT SHOWED THE DRIVER BOARD INSTALLED BACKWARDS WHICH WILL BLOW UP YOUR PFC IGBT AND MANY OTHER SMALL PARTS!

And finally a general view from the other side with the driver board etc. Right now everything is dry fit for testing and assembly. Once I'm happy with the whole layout I'll be mounting everything permanently and using heatsink compound and loctite.

The final charger should be fairly compact at 10.25" x 7" x around 9.5 -10.5"




This is the initial rendering of the top panel, all of the "divots" would be through holes for airflow, the display and buttons would be front and center(ish), this would mount to the heatsink in the same orientation as the driver board with the heatsink fins open at the bottom. You will notice the drilled/tapped holes in the sides of the heatsink where all of these panels will attach.




This is the initial idea for the bottom panel, it would cover the full size of the heatsink with mounting for the fans with some through holes for wiring since the power brick would be mounted just above the fans in the open area. Since I have limited space two of the screws that mount the panel to the heatsink would be attached through the fans themselves.


One warning, when assembling the main power board be careful of the polarity of C11, there are two + signs within the circle the larger one is correct, the smaller one could throw you off. I had actually installed it backwards initially but caught it with a quick double check of the PCB file before soldering. This would be a big boom at some point!!

The one question I do have though is about PWM fan control, is this implemented already via the Fan out on the main control board or is that simply to control a relay to turn the fans on/off, if so what decides when to turn the fans on/off???

[valerun]

Originally Posted by rwaudio

And some more build progress. I'm not using the standard EMW enclosure so my assembly is slightly different than the instructions. I'm building the enclosure around the heatsink, using the heatsink itself as part of the enclosure. I'm quite impressed by the physical layout of the high power pieces along with the driver board. It would be nice if the main control board was a bit more integrated or lined up nicer with the driver board since they have the most board to board wires. The overall layout of everything that mounts to the heatsink is very well done though and is a very efficient use of 3D space without limiting access to the mounting bolts for the IGBT's or rectifiers.
(Reminder, this is the air cooled PFC version of the charger)

The one question I do have though is about PWM fan control, is this implemented already via the Fan out on the main control board or is that simply to control a relay to turn the fans on/off, if so what decides when to turn the fans on/off???

WOW this is such a nice build and GREAT photos - thanks for sharing!!

PWM Fan control is implemented but as-is it will drive fans with <1A total current. Depending on your fan config, this might or might not work. You can also always swap out the PN2222 transistor with a power FET (something like 10A 200V unit and then drive any fan combo with that. Just remember to take power for the fan before the 12V regulator or you will risk burning the regulator.

To simplify the assembly and make the unit more failsafe, in the fully built units we wire the fans to the 15V supply directly so that they are on whenever the charger is running. You could also run them in parallel to the input / output relay control for a bit finer control.

V

[rwaudio]

Originally Posted by valerun

WOW this is such a nice build and GREAT photos - thanks for sharing!!

PWM Fan control is implemented but as-is it will drive fans with <1A total current. Depending on your fan config, this might or might not work. You can also always swap out the PN2222 transistor with a power FET (something like 10A 200V unit and then drive any fan combo with that. Just remember to take power for the fan before the 12V regulator or you will risk burning the regulator.

To simplify the assembly and make the unit more failsafe, in the fully built units we wire the fans to the 15V supply directly so that they are on whenever the charger is running. You could also run them in parallel to the input / output relay control for a bit finer control.

V

Thanks!
Once I get the charger up and running I think I'll play with the different methods of powering the fans and see what works best. Ambient temp in my garage is typically 15-25C in the summer and 0-5C in the winter.

A bit more progress..



So this is the basic design made with V-Carve Pro (CAM software with some very basic CAD functions)




Aluminium panel after drilling marker holes. I don't have any coolant on the CNC machine so drilling through holes for this would overheat the drill bit and cause issues, so I stick with 0.05" deep marker holes.




After milling and engraving but before cleanup.




Cleaned up, paint filled, buttons installed and the display sitting behind the cut out just to make sure the opening is large enough to view all of the text. The protective film is still on the display so there is some glare etc. The display is quite nice once the film is removed. As you can see my engraving depth wasn't perfect on the EMW 10kW part, this is due to slight overtightening of the mounting screw in the upper left corner. The table is MDF so this type of thing is somewhat hard to avoid on my DIY CNC Router. I also should have reduced the engraving depth of the button labels but it works.




Test fit on the enclosure to make sure my measurements were correct and there is enough clearance for the display and buttons. Looks good so far!

I have the fan mount panel also designed and the tool paths ready, I should be able to mill that this weekend as well. Then I have to decide on fan size (120mm x 25mm/38mm) to finalize the design of the side panels. This is important because the side panels are also the "legs" and will determine how far off the "floor" the fans sit for proper airflow.

[rwaudio]

It's been a productive day, a bit more progress...



The fan bracket / bottom panel of the charger.





Some 120mm x 38mm fans for test fitting. I will probably go with 120mm x 25mm fans instead.




The basic shape of the charger is coming together nicely.

[PStechPaul]

That's a really nice looking build. Maybe I should look into getting a CNC machine. But for anyone who does not have access to one, I have found an on-line source for specially milled and engraved panels such as this. I have used www.frontpanelexpress.com, and a panel like this might be about $50, and shipped within 2 weeks. They have their own design drafting software.

Conversely, if you would be interested in some small quantity prototype and production panels, the company I do design engineering for might be interested. I can tell that you are conscientious and have excellent skills and quality. Maybe you can contact me by PM to discuss details, if you are interested.

[jackbauer]

Experiencing a bit of a weird problem with the charger. I have been using on street charge points quite a bit over the past few months. Most are limited to 16A at 230v and the charger has read the pilot signal and adjusted power to match with no problems. More powerful 32A charge points are being installed and last Sunday I attempted to use one. The charger calibrated correctly and began charging at full power (32A@175v on the dc side). However about 10 seconds later the charge point shut off power and displayed "current overload" on its display. Tried a few more times with same result. I contacted the power company and they kindly provided a log file from the charge point. Apparently the charger was drawing 42A from the supply. Now this didn't make any sense as I us a 32A circuit breaker in the charger and on my home power connection for the car and they have never tripped.

Connected up a clamp on ammeter at home and started charging. 28.2A. Connected my UNI-T power analyser and got a 0.94pf and roughly the same power and current readings. Finally I tried a newer clamp on ammeter and it read 42A! Next tried a home energy meter which has a little split core CT and it claimed i was drawing 9kw. It also turns out the on street charge points derive current from an energy meter so it looks as if this is what is causing the problem. So I'm trying to figure out why this is happening. Any help much appreciated.


@rwaudio very nice build indeed!

[PStechPaul]

Many clamp-on ammeters do not respond to the DC component or distorted waveforms that may be caused by switching supplies, especially if they do not have adequate PF correction circuitry and line filters. But the DC component is a likely culprit. True RMS meters are a lot more expensive, generally $100 or more, and they can go out of calibration especially at zero reading.

It would be helpful to see a scope display of the waveform. And also a schematic or basic conceptual drawing of main components, switching frequency, etc. It's not too surprising that a 32 amp circuit breaker did not trip on 42 amps. They are generally designed to NOT trip at the 32 amp rating and are guaranteed to trip at maybe 125% of rating, or 40 amps, within a time frame up to an hour. The trip current and time may also be affected by temperature and other factors.

The 9 kW seems reasonable since 230V and 42 amps at 94% PF is 9.08 kW. I haven't closely followed the entire thread but I'll see what I can do to help if you provide a link to the appropriate documents. Open source is great!

OK, I found the schematic and converted it to a PNG image and PDF:



I'm not sure exactly what you are doing with the PWM but it appears to be a basic inductive current mode buck converter. There may be a problem with reading the output current accurately since the LEM module output is simply filtered to the average DC level, while RMS may be different. But the output appears to be well filtered with 1000 uF so the waveform may be close to pure DC. And I'm not sure how the LM211 creates the PWM signal, although obviously it changes the duty cycle based on output current. I assume also that the voltage module tells the processor when the desired float voltage has been reached and throttles back the PWM to maintain that level. Just thinking and typing and tossing ideas into the ring. Obviously it works and the purpose of analyzing the circuit is to determine what might be causing your problem. OK, enough for now. I just need to get more familiar with the circuit and the application.