EMW Charger testing the final chapter (I hope) (Update: NOT!)

I had already completed my low voltage (12v) tests and so far had discovered a reverse orientated IC blowing a 2n2222 and possibly the IC as well, replaced the IC and transistor and carried on testing with a lot of help and guidance form Valery I found a wrong value resistor out by a factor of 10 (27K instead of 2.7K that it should have been) replaced this and completed the low voltage tests with a fair amount of confidence that it was working as expected.

So I proceeded to HV testing and things were looking good, Connecting my 75v DC 45AH battery to the AC inputs went well, All measurements checked out fine.

Apart from the spark upon connection which scared the S#!t out of me (even though I was expecting it)
Moved on to the full AC test and this also went well (initially), Input voltage was around 288V (correct for 240Vac I think, RMS?) as shown on the LCD panel, Input current was set at 10A in and out as in the notes, Output voltage was still at zero as expected.

Started the charger and the dummy load lamps (2x60w incandescent bulbs in series) lit up as it entered the CV stage with an audible click then I got the duty screen, however when the values started to change the display problem (not clearing screen as values change) reared its ugly head again, I had been working around this with clrscreen statements.

As I could not see the readings without this being fixed I pulled the plug (Bad Idea?) and disconnected the load as in the instructions.

I need to fix the display refresh issue before I can move on.

When I plugged in AC power a second time the dummy load lit immediately!

And trying to repeat the test it responded differently, it just showed the Type 1 & 2 screens and then says charging complete without showing the duty screen, I tried disconnecting the dummy load and powering on this tripped my circuit breaker!

I then tried switching it back on (load connected again and it lit up again) and seeing if I could cancel the charge somehow, whilst going through the menus I heard a snap and saw a whiff of smoke, powered off immediately.

Smell check narrowed it down to the driver board I think (defiantly not control board) no obvious burn signs, so I guess it is back to low voltage testing again?

Low voltage test failed, now it won’t even power up, the only life I get is two flashing lights on Arduino red green in quick succession and nothing else LCD is blank. (Shit!)

Removed driver board quickly and checked over the board with my nose trying to identify which component had fried and my top lip revealed that U4 was red hot, so I am guessing this is fried and possibly the A3120 as well (Valery mentioned when these go they tend to go together)

SO I have 2 problems to solve now:-

Sort the display refresh issue so I can see the readings on the screen and repair the damage to U4 the DC-DC convertor that now seems to run very hot (I am guessing this is fugged) and probably replace the A3120 that is connected to U4, I may also need to remove and test the supporting components and replace as necessary.

Placed an order at Digikey for a replacement DC-DC (U4) and a few other bits for the next charger build, enough to get free delivery.

 

AC Input showing inrush resistors in series with live and neutral

 Output Diode package in white, Input Diode Bridge in black

 Driver board showing connections (removed from HV board for clarity of photo) 

 HV power stage board showing white inductor connection and black +ve -ve inputs

 same HV board from slightly different angle showing sense wire connection

  same HV board showing other inductor connection in white hall effect sensor and output terminal

Controller board in situ

Prepping for MOT

So the final pieces of the jigsaw are to fit a couple of automotive relays.

(1 for the vacuum pump for the brakes and 1 to disconnect the PSU/DC-DC convertor when the ignition is switched off as this causes the 12v battery to drain to overnight.)

And to get the EMW charger testing completed, get the charger working and charge the 3rd (Front) battery box.

It was my dad’s 70th birthday this last weekend so Saturday evening was already booked.

I spent most of Saturday morning working out how to use 12V automotive relays and fitting both the rear PSU/DC-DC convertor cut off relay and the front vacuum assist brake pump relay, wiring it through the vacuum sensor.

I tested it by wiring the relay direct to 12V switching on the relay permanently and adjusting the vacuum sensor sensitivity down until it settled a little.

I may need to service the vacuum pump or the vacuum chamber on the car as it did not seem to hold the vacuum for very long before the sensor detected a drop in vacuum pressure and kicked on the pump again, however it did definitely respond to me pressing the brake pedal as expected and refreshing the vacuum ready for the next push. Great :-)

So all I have left to do here is find an 12V ignition wire somewhere near the front of the car and route it to the positive input on the relay so the vacuum pump only runs when the ignition is on.

However there was never a need for 12V ignition in the front of this rear wheel mid-engine beauty, so I will need to go hunting under the dashboard for an ignition feed to finish this.

The car is booked in to the mechanics to have its final fix before MOT on Tuesday of this week (24th) Brent just needs to make a battery tray from some angle iron  for the 12V  Battery and fix all the HV cabling to the body proper as it is just secured to brake pipes at the moment with zip ties.

So Sunday morning I got down to charger testing.

Honda Beat EV takes its first drive

Saturday 14th September, got stuck into the car early on Saturday morning as Steve was planning to visit and give me a hand getting the car going, removed the throttle position assembly and took it indoors to do a full breakdown and check (on Steves advice) I can believe I did not do this initially but I guess I thought there was not much that could be done wrong here.

How wrong was I, after emailing Steve last week he had advised I took the TPS assembly out and diagnosed it on the bench to make sure It was putting out a clean 0-5V signal from zero throttle position to 5V max throttle position, plus I had a sneaky feeling that I may have put the magnets in wrong (when I got the TPS assembly from Steve it had no magnets or sensor fitted).

Spent half an hour or so re-soldering the sensor to the throttle cable to allow for better insulation of each TPS sensor wire, it’s a very small sensor, and the hole it fits into is not much larger either so it leaves very little space to insulate everything.

Wired it up to my bench PSU and gave it 5V monitored output signal and waved the sensor at the magnets, nothing not a dickybird!

I was beginning to wonder if the TPS was stuffed when I discovered that the hall effect sensor only works when it is in close proximity to the magnets so I fitted the sensor in its housing and screwed this into the TPS assembly and voila I had needle movement on my DVM, spent the next 30 minutes discovering that I had the magnets 90 degrees out of phase and one set reverse orientated, corrected all of this and I had a good signal coming out of the TPS, Time to put it back in the car.

Refitted the TPS assembly and fired up the Soliton for a recalibration of the throttle, there was no apparent change until I touched the throttle after recalibrating and heard the pleasing sound of the Kostov whirring into life

WICKED, IT WORKS! “Quick love get the camera” I shouted to my partner and here is the result.

EMW Charger testing continued

Carried on testing the EMW charger after I charged the car last night with Steve’s Zivan, I checked the voltages on the A7520 and found 0V on all pins bar pin 6 & 8 which had 4.82V on them, reported this back to Valery and await his advice.

Also I removed the charger box from the boot last night after marking the final hole that needs drilling to accommodate the +ve feed to the battery pack.

I had to change this slightly as I originally thought I could connect both +ve & -ve from the battery connections on the Soliton to the charger but of course the contactor set (which is not in the boot) would prevent this so I cut and crimped an additional length of cable to give the +ve feed from the charger the battery side of the contactor set.

Steve has said he may make it over tomorrow (Saturday) so hopefully I may get the car moving this weekend.

Need to remove the Throttle Assy and diagnose on the bench to get a proper 0-5V range coming out of the sensor, I must have the magnets fitted wrongly or something will investigate tonight.

Also need to move the shunt onto the negative battery terminal and hopefully Steve can help me setup the “Large Cycle Analyst” as well on Saturday as I have not a clue what to do with it.

1st Charge (from Steves Zivan) @ 20:20 on 12/09/2013 (2 packs only @ 148V)

Borrowed Steves Zivan last night he also gave me the Cycle analyst and the Shunt (which I mistook for a fuse and put inline with the +ve before charging, DUMMY!)

Steve corrected me with a ”Fuse what fuse! You don't mean the shunt I gave you do you?” DOH!
Still as the pack was at empty I luckily did not cause any damage to Steve’s Zivan.

It responded as expected and I left it charging for 1 hour as Steve recommended that should be good for 15 miles or so.

Disconnected the Zivan and switched on the car connected laptop to Soliton and checked status, “Pack Low Voltage” messages had disappeared so I assumed I would be good to drive (Wrong)

I then spent an hour or so trying to calibrate the throttle sensor again, I think I may have the TPS mechanics badly setup or something, when I got the TPS body it had no magnets fitted and no sensor, I  bought the sensor and wired that myself (Correctly according  to datasheet AFAIK) however I guessed at the magnet orientation I am checking this with Steve now.

The TPS body I am using is from a motorcycle and has been bolted into the car by Brent the mechanic this is all fine, however when I disassemble the TPS body, the part that moves inside is a “C” shaped metal bracket with two small recesses for the magnets at each end of the “C” these recesses are just the right size to fit two magnets at each end of the “C” however I guessed at the orientation so I may have inadvertently created a confusing magnetic field for the hall effect sensor and this may be why I am getting inconsistent readings from the throttle .

I am checking this also with Steve, in fact I think I will need to get Steve over soon to give me a hand setting up the cycle analyst as well as ironing out the last few things to get me moving.

Charger Delays

Fixed the garbled screen by adding loads of CLRSCRN statements before the PRINTSTR statements, still not sure that the LM211P is not fried, but Valery thinks not (I ordered one anyway just to be sure) everything follows the test procedure normally until I get to the “Short the output terminals” section which should then do a Zero Volt calibration, instead it just skips straight to the Power config (not right)

Ok so I am progressing with the charger slowly can’t get it to do the Zero Point calibration at the moment and I have been emailing Valery for advice.

Still Steve has offered me his Zivan to charge and test the car it can’t do a full charge but if I charge it for an hour or 2 it should get the car running adequately to get it down to Brent for the final fitting and MOTing I hope.

Brent needs to weld me a 12V battery holder bracket and fit all the HV cables properly to the car (rather than Zip ties to the Brake pipes as they are now) he also needs to fix the contactor to the car.

But that should be about it, Damn it’s getting close now.

Can’t wait to see what this baby can do.

12V charger Testing begins

On Sunday after the show I spent the day testing the charger, swapped out the blown 2n2222, reoriented the LM211P and hoped for the best, fired up and “It’s alive”, well the LCD panel lit and displayed the 4D systems logo, no smoke or flames so looking good in my opinion.

A bit more reading and it looks like I need to upload firmware to get it going, wired in the firmware upload switch and tried again no joy! the Arduino flashes when I connect but does not upload, remembered a jumper under the Arduino checked the Schematic and it looks like the link supplies the LCD with power, The instructions refer to using the programming switch to turn off the LCD so you can upload software, so I tried it with this jumper removed and voila it worked! YEY

So now when I power up the charger I get a screen asking me to select the battery type showing “LIFEPO” which is exactly what I should see, YEY.

However when I select the battery type it moves to the next screen without clearing the last garbling the text on screen every step just makes it worse so I will need to fix this first before I can test safely.

Fired off a message to Valery and posted on the diyelectriccar forum as well, spent the rest of Sunday prepping the +ve and –ve output cables and realized that now I have fitted the contactor the battery terminal is no longer in the boot with the charger (Dammn!) Still I put it where it is on Steve’s advice to keep the cable as short as possible that way eliminating any long live cables however now considering the charger position I will need 2 meters more “220V Permanently LIVE” cable to connect the charger to the battery otherwise I would need to leave the car on all night whilst I charged and I would be putting my charge power through the contactor set not a good idea me thinks, I will need to check all of this with Steve before wiring up.

So close I can smell the rubber burnin ;-) Two more weeks and she should be road legal :-)