Thursday, 22 May 2014

Brighton Mini Maker Fair comes around again!

Heck that was another year already! Well not quite but the BMMF have contacted me again to submit an application to exhibit at the BMMF 2014 in September.
I really enjoyed it last year (despite nearly losing my voice) so it was a no brainer to apply again, Plus this year I have the car finished and running so I will actually be able to show the car itself as well as my 3D printer, last year despite my efforts the car was just shy of being ready.
And in fact this year it is in a much better place with all the safety’s in place and the charger fitted and working with the J1772 plug adaptor fitted in the “ex” petrol filler cap and all 4 new alloy wheels with better sized (smaller) tyres fitted, it looks the nuts :-)
Photos coming very soon.

Fitting some smaller tyres

This weekend (17th May 2014) I finally got around to fitting some smaller tyres on the two front alloys and got them back on the car (they have been sitting in the barn since I bought them and realised that although the rears were fine I would need smaller tyres to avoid the rubbing.
I have replaced the stock 13 & 14 inch alloys with larger 15 inch BBS style deep dish alloys, they were needed urgently after I first got the car on the road as I had a major lack of grip due to the massive torque now available!
I had put the original 13 inch wheels back in place after I realised my first tyre choice was just too large for the front wheel arches and have been driving the car like this for several months waiting for my local part worn tyre fitter to come up with the goods.
The car looks fantastic now with its set of 4 matching BBS style deep dish 15 inch wheels, photos coming very soon (always best to get the photos done in summer)
Even with these smaller tyres there is still a very small amount of rubbing on full lock when turning right, full lock left seems not to have the issue so I will investigate where exactly the rubbing is and see if I can do anything more about it later, but for day to day use it is now fine with the matching BBS alloys.

LVC-HVC Wiring

Ok so the LVC/HVC cut off boards have two white plugs at the top of each board (one board for each parallel bank in the traction pack)

They are labelled as shown in the image below

Online manual for these boards below

Only found this manual recently (thanks again Steve)

I “just” need to connect all the HVC/LVC board connections in a chain connect the HVC part to the charger and connect the 6 solder connections at the top (can’t quite see them in the photo above) in line with the throttle cable, between the Throttle Position Sensor (TPS) and the Soliton Jr’s Throttle input.

So this bank holiday weekend 3rd/4th/5th May 2014 I spent several hours joining all the 3 packs HVC/LVC loom cables together, I have soldered them all up as I have not got any inline 6 way connectors handy although if I need to remove the pack again (I sincerely hope not any time soon) I would have to cut these cables, so I do need to buy some inline 6 way connectors for this purpose.
As the front 3rd of the Traction Pack is a good 2 meters further forward than the other two boxes I had to run a long 6 way cable under the car to join the front battery box HVC/LVC loom, I threaded this 6 way through another second hand inner tube and fastened this under the car using many zip-ties.

Once the HVC/LVC loom was complete I wired the Throttle and Controller as shown in the picture above (to the solder terminals hidden behind the white plugs at the top) this is so the LVC cut off can cut the throttle.

It is very difficult to test the LVC in the real world without running the pack to near flat, I did test the board on the bench with a variable PSU attached to fake a cell voltage and sure enough when the cell voltage drops to 3.00v the Throttle signal line drops to 0v (cutting the throttle)

I then set about connecting the HVC to the chargers BMS pin, I ran 4 wires out of the charger box, 5v, Gnd, BMS & EOC. After much fiddling about testing this I discovered that it only seems to work if I leave the BMS & EOC pins connected together (otherwise the charger never starts at all) and then connect the HVC “sig” line to these via a low value resistor (I was advised to add a resistor inline to avoid the Methtek boards drawing to much current) I am not sure the resistor is even necessary but I put one in just in case I ended up having to use two 51 Ohm resistors wired in parallel as that was all I had to hand that worked it seemed if the resistor was even 100 Ohms the signal would just not get through or was not detected on the far side of the current limiting resistor.

So to summarise:-

1) Connected 5v from the control board in the EMW charger and fed this directly into the HVC 5v pin.
2) Connected Gnd from the control board in the EMW charger and fed this directly into the HVC Gnd pin.
3) Connected the BMS/EOC pins from the control board in the EMW charger together.
4) Connected the HVC “Sig” line on the Methtek board (HVC/LVC chain) to the BMS/EOC via two 51 Ohm resistors in parallel.
5) Connected the “Throttle” 3 Pin soldered connections to the throttle.
6) Connected the “Controller” 3 Pin soldered connections to the controller.

I have just realised after writing this up (This is why it is a very good idea to write things up) that I have not fed the LVC part of the HVC.LVC chain with 5V or Gnd!  As this would be active when the car is running (unlike the HVC part which is only active when the car is charging) I would need to source this 5V & Gnd from the Soliton controller! (Update After speaking with Steve the + 5v & Gnd

Shit I had better check this when I get home tonight as I may very well not have a functional LVC!
Anyway I did test the HVC side of the HVC/LVC cut out extensively in the car by using a variable resistor to work out what value of current limiting resistor I would need, and testing it by starting a charge and then shorting the HVC “Sig” line to ground to see if the charger stopped and sure enough if I connected it directly it worked.

My initial guess of a 1K resistor did not work so I put in the variable and dropped the resistance until it started working, checked the resistance and it was way low I eventually settled on the two 51 Ohms in parallel which should give around 25 Ohms resistance and burn around ½ watt power (2 x ¼ watt resistors) I could be completely wrong here as my electronics is still decidedly rusty!

Back on the road at last.

It has been a long couple of months but at last my EV is back up and running, this time with and HVC/LVC harness in place!
LVC is already wired in line with the throttle albeit to only the 2 rear battery boxes I will be wiring the 3rd and final (for now) battery box in tonight which will give me a proper low voltage cut out at 3.0v on the nose on any of the cells in any of the 3 battery boxes, basically if any of the cells hits 3.0v the throttle is cut, I have not tested this in place as yet but I have tested the board on the bench with a variable PSU as a test cell and sure enough the LVC “Sig” line drops to 0v when the cell pin goes below 3.0v.
Once the HVC/LVC loom is complete (Tonight) I will be wiring the HVC side of it into the BMS pin on the charger in an attempt to get the HVC working as well although this would only cut out as a last ditch safety measure as it cuts at 4.4v not 4.2v.
So I still need to upgrade the firmware on my charger to allow a CV cut out of 4.2v (currently limited to 3.99v) and also to put in some physical isolation of the chargers case from the chassis of the car, this is because when the charger is plugged in and running the chassis has some voltage on it, I would not call it live but I think the stray voltage might be either leakage from the bridge rectifier, or possibly inducted from the large toroid.
I also need to investigate a low voltage safety in the charger software as according to Steve the charger should error if it detects an abnormally low voltage and cause you to check to see if any dead cells are present, charging dead cells could result in a thermal runaway event (Yeah don’t I know it!)
And to top this lot I still need to implement the fuel gauge and the heater/demister.

BC168 balance charger wigs out!

In order for me to test and balance all of the 108+ cell packs that I have discharged, graphed, charged and balanced over the last 2 months I needed to buy a good LIPO charger and Steve recommended the BC168 LIPO Fast charger/balancer for around £90, I looked it up online and sure enough it appeared to be the bees knees of LIPO chargers (and I still think it is) after running around 80 cell packs through it (not many according to Steve he has put several thousand through his) the ground connection wire to the breakout board was showing a significant browning of the plastic suggesting it was getting hotter than it should have, I did not consider this a major issue and carried on testing the remaining cell packs whilst angling the output fan over the breakout board to keep it from getting worse, however after only a few more cells the device began playing up and would not show a reading on and cell pack on S1 line, I initially suspected the breakout board but after testing this it turned out to be the device itself and now it won’t charge properly at all the voltage sensing seems to have gone out the window as it jumps wildly about trying to sense the correct voltage, basically it’s fried. Shit!

Not convinced I did anything wrong to damage the unit so we are going to try and get an RMA on it however as it is a fairly cheap device and they come from Hong Kong it may take more time than it is worth, for now Steve has lent me his so I can at least finish the remaining packs to get my car back on the road.

Collected 42 new (used) batteries

I collected my new batteries, They are used identical Turnigy LIPO’s 5S1P’s from Steve’s Electric MX5 which he is sadly breaking (house purchase forces cashing in of assets me thinks) still I intend to purchase the Motor and Controller from it to convert my RX8 so it is not all bad news.
Carried on testing the new cells same as before
1) Charge to 4.2v
2) Discharge and graph
3) Check graphs and discard any bad one’s (Steve assured me there should be none)
4) Balance good cell packs to 3.8v and fit them in 3rd and last battery box

Unfortunately when we pulled the batteries from Steve’s car we noticed an imbalance on some of the cells in a parallel block so all of them were showing cell 2 as being down at 1.5v and after testing and graphing these cells 8 of them showed an imbalance after a single charge and discharge and graph, and the failed one was done at 0.8v on cell 2.
I pulled these 8 to one side and collected another 6 from Steve to cover any possible bad ones (only needed 6 as Steve had given me 2 extra to cover the potential bad one in the first place, And as I completely forgot to take the bad ones back to him when I collected I now have all of these extra ones to find 6 good cells from to fill the last 6 slots in my last battery box (getting close now)
Last night (24th April) I fitted all the cells in the last 3rd battery box and packed them ready for wiring, just need to charge/graph and balance 6 more of the 14 cell packs I have and fit them before I can wire up the last box and close the lid.
As soon as this is done I will drop the remaining 8 iffy cells back to Steve, There is really only 1 definite bad cell in this 8 but I want to make sure the 6 I fit are all good and well balanced cell packs, after all I have just spent 2 months testing every single cell and balancing them perfectly.

Battery rebuild begins (12th April 2014)

Started by taking all the good batteries and stuffing them back in the boxes completed one box I am 6 cells short to complete the second (although these 6 failed cell packs are no good for an EV battery they are still useable for other projects) advised Steve of the additional battery requirements (he already knew about the 36 incinerated + 1 swollen cell pack, so 5 more required.)

As I could do no more battery stuffing I moved on to fitting the buss bars back in and realised I was going to need more (new) neoprene so I put in a web order for another roll.

At this point I decided to call it a day (weekend) and spend the rest of my spare time with the partner & kids (mowing and strimming the lawn more like!)

Battery Testing Finally Completed (12th April 2014)

And by completed I mean:-

1) Ran every one of 72 x 5S1P cell packs through a discharge (to 3.5v) and graph test using the Powerlab8 Battery Workstation
2) Checked all the graphs to see if any cell packs were out of balance, made a note of the ones that were for additional test
3) Bottom balanced all Cell packs to 3.8v (Storage charge)
4) Checked all cell packs physically for any signs of swelling, (Weirdly every single one looked fine until the very last discharge test on the very last cell pack which swelled up like a balloon)
5) Ran all cell packs that failed first test through a balance charge (to 4.2v) using a BC 108 balance charger
6) Ran all cell packs that failed first test through a second discharge (to 3.5v) and graph test using Powerlab 8
7) Checked all graphs to see if any cell packs were still out of balance, Discarded any cell packs that were (5 were!)
All that lot basically took me 6 weeks of spare evenings to complete!

Battery Testing Continued

I spent several evenings this week (17th-23rd March 2014) testing batteries and have completed the first battery box 36 cell packs and the first few cell packs in the second (last remaining) battery box, so I probably have about 30 left to test, once this is done I can get onto the balancing, so far I have only had one cell pack show any signs of swelling and strangely this one seemed perfectly balanced, however Steve has recommended that I discard any that show signs of swelling, So I guess I will repurpose this one for another small battery powered project such as a 3D printed quad copter or the inverted pendulum balancing bot I am planning to build.

Let’s hope I don’t have many more that I will need to discard, I should be able to complete all this testing this weekend coming.

Battery testing begins

I began by removing the 2 remaining battery boxes that survived and getting them into my workshop, removed the lid from the first and removed all the buss-bars, LVC/HVC boards and all additional cabling until I had just the cells, attempted to pull one out, not budging, panicking a little as I was assuming they had all swollen due to the over charge I contacted Steve and asked how much force I could use on the banana leads after all I don’t want to start another fire in this box whilst I am trying to disassemble!

Steve was also assuming the batteries were probably swollen beyond use and this is the main reason I needed to get the whole lot apart to check for swelling and discard any swollen cells, after Steve’s reassurance that I could safely give them a good yank without puling the banana leads out of the cell packs, I went back for another try, this time I turned the whole box upside down and gave the end/edge of it a good whack on the bench in my workshop, flipped it back over and almost in despair grabbed a couple of banana leads and yanked!

To my surprise it popped out sweet as a nut, and of course once one was out the rest came easily, and to my great surprise and relief there was not a jot of swelling to be seen on any of them in fact they all looked good as new!

Borrowed a Powerlab8 from Steve and began testing batteries this weekend just gone 15th March. The powerlab8 is a cool little unit and does a lot of stuff but I was basically just using it to do a discharge from the overcharged state to an “Empty” state 3.5v whilst graphing the discharge on a PC that way you can see if any of the cells are out of balance or damaged in any way.

To achieve this is a little tricky though, the powerlab8 is best run from a 12v or 24v lead acid battery this allows you to do a fast charge/discharge at 20A, simulating actual usage discharge speeds, as I was not versed in the use of the powerlab8 Steve ran me through the basics via email, and after doing 4 discharges my 12V Lead acid battery hit its fully charged state (denoted by the voltage popping up to 14v, peukert effect!

As Steve explained) despite me attaching (what I thought were headlight) bulbs to the 12v to drain excess current, so after a little more fiddling around finding 12v devices that I could attach to drain the 12v and provide a load, I finally settled on the 4 bulbs I had initially, plus a 12v cooler box, plus a 12v car tyre pump, plus a single true headlight bulb wired so that it was on full beam i.e. both filaments lit, And at last this seemed to be enough, I was now able to continue testing and to my surprise at 4 x the speed as before for the first 4 tests I had only been dumping 5A rather than the 20A I had set because the 12v battery could not take any more.

So this greatly increased the speed at which I could test and by Sunday evening I had tested almost all of a single box worth of 5S1P cell packs, I think around 6 will need rebalancing for sure but I may just balance the whole lot and retest the ones that were out of balance to check there is no intrinsic issues.

This Still leaves me short the 36 cell pack’s I incinerated and 4 new LVC.HVC cut-off boards with extender boards (I had better make sure I use them this time eh! You f##!ng TWAT!)
And I am going to need to do some reworking of the buss-bars as well!

Lessons Learnt!

OK so why did I incinerate 1/3rd of my traction pack, well several reasons but basically the main (FIRST) one it is I AM A TIT! and did not wire the Low Voltage cut-off boards into the Soliton, in fact I was putting it off (here come the excuses!) due to the fact that we have been having such shit weather (rain, I AM TALKING NOAHS FLOOD!) and our house is in a flood plain and so I had spent the last few months (literally months) dealing with the flooding every weekend without fail, I was out with a spade walking fields and digging trenches to alleviate the flooding and prevent our house from flooding this went on for several months before we finally got the EA to visit with a big digger and dig us a new drainage ditch to take the massive excess of water away.

OK excuses over, the other (SECOND) reasons were that I did not wire the High Voltage cutt-off to the charger I had just recently fitted, this will be complicated by the fact that in order to test this I will probably have to remove the charger again along with do a lot more reading and testing to get this working.

Other (THIRD) reasons could include the fact that I had not fitted any sort of fuel/charge gauge in fact I had only just picked it up and was trying to wing it to some good weather (which is now finally here TO LATE! To save me from my own stupidity!)

However after speaking to Steve he feels that the EMW charger was partly to blame because it should have picked up the fact that the total voltage was below a safe limit to begin charging (which was almost certainly true, even though I did not check this and cannot prove it!) he has advised that I check that the charger does include a below minimum voltage check and won’t run (Errors!) if the total pack voltage is below the expected “empty” value. I am not even sure if the charger checks for this.

Also as I already know I am lying to the charger as it won’t allow me to set a CV of above 3.99v I need to set 4.2v

So all in all I have a heck of a lot of work to get the EV back on the road but I am determined to do it and ASAP.

“Dooms Day” EV Style (or how to incinerate 1/3 of your traction pack)

Tuesday 25th February, Left for work as normal in the EV after a normal charge the night before, got to my daughter school and dropped her off as I was leaving I got a call from my partner “you have my car keys!) Shit I thought, now I will be pushing my range to get home and back to work and home again, my EV only has 40 mile range and my round trip to work is 22 miles, my daughter school is around 2/3rds the way to my work and I and I had not pushed the charge to the max the night before as I was not anticipating any extra journeys.

Oh well I thought, better get the keys back to the missus and worry about this later (BIG MISTAKE NUMBER ONE), drove home and back to work no probs, left for home that evening and got nearly home before the car began to judder and slow, shit I thought, oh well it’s another tow (third time I have run a cell flat or the low voltage cut-off in the Soliton has kicked in I am not really sure) anyway towed the car home (only a few hundred yards) and put it on charge as normal (BIG MISTAKE NUMBER TWO, Steve has already advised me that following a run flat you should always check your cells for problems!)

Came out a couple of hours later expecting the car to be nearly charged (after all I had done extra mileage and was expecting a longer charge time this evening, possibly double my normal 2 hour charge) it was still far from charged, strange I thought Oh well I will give it more time (BIG MISTAKE NUMBER THREE) and went back in for another hour or two, I was not unduly worried as the charger throttles back the duty cycle when it nears the max voltage for the entire pack.

When I came out to switch off I realised that I was in BIG TROUBLE, there was an all pervading smell that told me something was not right, I checked the voltage in the charger and it was still far from max and still at around 70% duty, SHIT! SHIT! SHIT! Cut the power and checked the front battery box as this is the only one I can get easy access to by flipping the bonnet up.

Extremely strong chemical smell coming from under the bonnet obviously from the LIPO batteries, got my widget and checked voltages on this third of the pack and to my horror 2 banks were not reporting anything DEAD and the other two were severely overcharged at 4.4v per cell! FUCK! FUCK! FUCK! MAJOR PROBLEM, called Steve for rapid advice, he told me to watch it closely and whilst I was on the phone to him doing exactly that the smoke began! Just a whiff at first “It’s smoking Steve what the fuck do I do” Steve advised tipping a shitload of water on it.

As I was fetching the first watering can (luckily we have a load of water butts around our garden with thousands of litres) the venting began!

I am talking Lithium steam train at full tilt! I spent the next 2 hours tipping water on the burning battery until it finally gave up venting around 1:00 AM in the morning.

Went to bed with doom and gloom in my heart (oddly satisfied that I had at least prevented the whole car going up in flames!)

Spent the next weekend decontaminating the car and the area of my garden it had vented in.
Spent the weekend after that stripping the burned battery box out of the car and recovering whatever I could (not a lot) all 36 5S cells in this box were torched, the fire had ripped through the box end to end taking everything with it bar the Ali box which survived remarkably intact and the buss-bars however these will need servicing with new banana plugs (sockets seem ok) the plugs are supposed to spin freely some do not anymore and therefore will need replacing.

Also all the batteries in the rest of the car (the other 2/3rds) were also all severely overcharged to around 4.4v per cell, and after speaking with Steve it appears I will have to breakdown the entire pack test everything, rebalance all cells that are good, discard any bad cells and reassemble.

Basically a shit load of work!