31/08/2009 Progress

Battery protection circuit

• Re-drilled circuit board and added "jumpers" ... so that ground is switched for all components.
• Tested, and there is NIL leakage when circuit is "off". In other words, total and complete isolation (which was the idea !).
• This was a LEARNING EXPERIENCE.
• I hadn't thought about "reverse" leakage through components such as a capacitor, 555, and transistor.
• I could have tested for this on the breadboard, before committing to the circuit diagram.
• Learning lesson is:

1. Write down clear objectives (ie no leakage ... no possible wrong connection ... voltage and current levels ... different input voltage ranges).
2. Include "heat dissipation" in the testing criteria.
3. Make a "testing checklist" ... to test circuit before committing to final build.
4. Then do final build, and once more test against the checklist.

• More examination of the circuit, and I realised that the "Seat Switch" part of the circuit didn't work.
• This was because there was a logical error in the zener regulator circuit for the 555 and relay part.
• This error of logic actually leaves the relay and 555 operated simply through voltage dividers. As luck would have it, the voltage dividers work just fine.
  Worked out how I could salvage the PCB (cutting links, and adding dividers) so the Zener would work.
• However ... trying to calculate values so the Zener would provide a stable 12 Volts ... gives very high wattage through resistors, and high wattage through the Zener.
• Ultimate decsion ... is to leave the voltage divider circuit in place. If it doesn't provide enough current for the relay at low voltage ... then simply remove one of the four 150 ohm resistors. See excel spreadsheet of values here:

• The other problem, is that the Zener after the Contactor gets very, very hot.
• This is unexpected. The Contactor draws 207mA at 27 volts. This means the Zener drops 12V at 207mA, which is 2.4W. This is probably why it is hot ! It is a 5W Zener, so it won't break down, but dissipating the 2.4W is a problem.
• I've move the circuit items, so the Zener rests against the Cement 5W resistor ... the idea being that the heat will dissipate here.
• I've used heat transfer silicon paste all over the 5W resistor and the side of the cement resistor. This could be a problem, as the stuff doesn't seem to have set. If it drips onto the board, it might short something. Yuk !

Battery Balancing

• Assume: Battery internal resistance = 0.006 ohms, and battery difference (voltage potential) = 0.070 V.
• If hook batteries together with "perfect" wire ... then power is 0.07/0.006 = 11A !!
• With 1 ohm resistor between them, power is 0.07/1 = 70mA
• Discoved standard "hook up" wire has resistance of 0.117 ohms / metre. See here.
• Therefore, if balance taps are standard hookup wire, and are 10cm each ... then between each cell is resistance of 0.2 * 0.117 = 0.023 ohms.
• Directly hooking balance taps could produce 0.07/0.023 = 3A !!
• Adding another 30cm of hookup wire between each balance tap, means 0.07/0.0581 = 1.2A !!
• Hookup wire will take this much power (but not happily). Max is 1.5A.
• Calculated that 6 core alarm cable has resisance of 0.078 ohms / metre (ie better).

Driving

• Hooked it up for the boys to drive in the afternoon.
• When Tommy was driving, the accelerator jammed fully open !
• Brakes weren't strong enough to hold the Kart.
• He went crashing through the front gate, down the driveway, and across the road (out of control !).
• This was a LEARNING EXPERIENCE:

1. I should have refused to allow them to drive it, until I had installed all safety features.
2. I should have installed brake microswitch to controller "stop".
3. I should have installed Kart control, including start and emergency stop switch.
4. I should have insisted that all safety features were installed before the boys drove the Kart.
5. I should have listed all safety features, and made a checklist ... and thoroughly worked through the checklist before letting anyone drive the Kart.

See Safety Feature Checklist

Kart Progress

• Connected brake lever (switch) to controller. Works well !
• Moved accelerator to other side handle (to prevent locking).
• Took seat off, and mounted controller onto frame (beautiful !)
• Did wire soldering controller to motor (beautiful !)
• Taped up wires, to make them look better (excellent !)

30/08/2009 Progress

Kart itself

• On Saturday morning (yesterday) finished the steering, and Kart done (except brakes).
• On Saturday afternoon, connected it up (Jury Rigged for power) and ran it. Superb !
• Sunday morning, added the brakes.

Battery protection circuit

• Sunday ... worked on the controller box.
• Realised that I'd made a mistake with the controller box. There is a slight leakage of power back through some of the components (transistor ? 555 ? more likely capacitor !)
• Worked out how to fix it.

29/08/2009 Progress

Electric Kart

• Poppy finished the steering etc and delivered the Kart in the afternoon.
• I "jury-rigged" the power, so the boys could drive the Kart.
• Excellent fun ! Truly amazing thing !
• The boys love it. It is (of course) completely mad !

29/08/2009 Progress

EV Batteries

• Checked balancing. And started balancing the next "bank"
• Created board to connect all balance taps together, and soldered on cable to remotely connect: (a) 'fuel' guage, (b) BMS, (c) charging "shunt" connectors.
• Tested my circuit idea for switches to be 12v potential only, not 42V (Note to self ... didn't thoroughly test or understand this last time around).
• Tested circuit idea for 75 ohm resistor on the "stop" switch. Doesn't work !
• Tested 4 x 110 ohm resistors (instead of 2 x 220 ohm). Still far too hot !
• Tested 4 x 150 ohm resistors. Burning hot to touch. No good !
• Good news ... is that the Zener is cool.
• Tested 3 x 150 ohm + 1 x 220 ohm. Not enough power for circuit.
• Rough calculations: LM555 = 15mA. Relay = 12V, 282 ohms = 42mA. Capacitor etc (say) 5mA. Total 62mA. Measured = 64mA
• Drop over the 4 resistors is 30volts, at 64mA = 1.92W ... or 0.48W / resistor
• Better relay is 331 ohms, 36mA. Save 6mA. Power = 1.74W being 0.43W / resistor
• Decision = buy 4 x 150 ohm resistors,

Things I want to do today

• Cut the 'balancing' board and the circuit board.
• Cover balancing board in insulation. Bind safely.
• Etch the circuit board.
• Drill the circuit board.
• Mount parts ?

28/08/2009 Progress

EV Batteries

• Created the final main power wire (10AWG into 4, to connect to battery).
• Created the small "mini-board" balancer with 1 ohm resistors. Plugged it in. It works (not too hot).
• Worked out a solution to the "Accurate DMM" problem ... namely that Poppy already has an accurate DMM. Can just borrow that.
• 1200 watt power supply turned up. Hooray !
• Thoroughly measured and tested balancing board effects. Excellent !

Things I want to do today

• Re-create the circuit board, with better traces (straighter, etc) ... and better positioning (narrower).
• *** TEST *** my circuit idea for switches to be 12v potential only, not 42V (Note to self ... didn't thoroughly test or understand this last time around).
• *** TEST *** that my theory on why the delay circuit works is sound.
• *** TEST *** what happens when diode is not used.

27/08/2009 Progress

EV Batteries

• Good news ! In the process of trying to work out how a 555 can better work with a PNP transistor (and after much experimenting on a breadboard) I found an excellent circuit here.
• Breadboarded the circuit (see below) and it works perfectly !

• Bad news ! I am going to have to create a new circuit board ... but this is also good, because next time around I will include headers for better wire connections. Not happy with the wires going into the box last time around.
• I have studied and think I understand the circuit now ... all except the diode.
• Have sent email to the author, and also created a thread on a circuits forum (here) to which I received a reply, that the diode seems only to be there to discharge the capacity on turning off power (which will not apply ... if ground is switched instead of positive).
• ...
• Finished first draft of the new circuit board for the controller.
• Put all components on the breadboard, and hooked them together. They work perfectly ... except that there wasn't quite enough current with 4 x 220ohm resistors before the first 12V zener. Taking down to 2 x 220 ohm resistors works perfectly. Will need to replace all with 100 ohm resistors (or similar).
• Also - didn't get to check the 75 ohm resistor in series with "stop" button (didn't have such a resistor).

26/08/2009 Progress

EV Batteries

• Circuit components from Farnell turned up. Excellent business ! Will definitely buy all future components from them, unless I need them in under 24 hours. Brilliant supplier. Much better than Jaycar.
• Next order, will get LM338 but in TO-220 package, not TO-3 (and try to get it to work better). Don't want to be beaten by the LM338.
• Checked the Forum for responses on the shunt charger circuit. Yes, all good.

.

• Got a BD682 (PNP transistor) to replace the BD681 (NPN)
• Installed in circuit ... doesn't work.
• It seems the PNP transistor needs to be raised to the +ive rail it is switching, in order to turn it off.
• So if trying to switch using a 555 ... which raises and lowers to its own rails ... then it isn't possible to completely switch off the BD681.
• This is a particular limitation applying because the 555 output toggles between its own rails.
• This does not apply to a NPN transistor, which can be turned off by simply driving it to ground, and then turned on with its simple threshold voltage (which, from memory, I think is 2.5V or something).
• Interesting lessons. I am going to have to re-create the Kart Control circuit.

25/09/2009 Progress

EV Batteries

• Purchased the RSP-1000-48. Very cheap. In Australia (Hooray !).
• Purchased a new LM338K, and some resistors (to balance the pack safely).
• More research, and I'm more certain that previous LM338K was killed either by temperature (soldering) or a short (lack of suggested diodes when large external capacitors in use). Will test this afternoon.
• Also discovered a new source for IC's (Farnell - much better !).
• Tried the new LM338K ... and killed it. That's two dead ! I think the reason is, that the "Economiser" circuit that is built into the Contactor, does not play well with the voltage regulator. It seems likely that there is some kind of feedback, or something, which is stuffing up the regulator.
• More thinking.
• I came up with a great (temporary ?) solution. The Economiser circuit changes resistance. It wants a high inrush current, and then low current after that. I can use the Economiser itself as the resistor in a zener regulator circuit. The Zener will always provide the balance resistance required so that the Economizer sees the other voltage (so ... for 12V Zener, full battery = 30volts ... empty battery = 18volts. Perfect !).
• Tried it, and it works.
• Modified the circuit board to work, and now all I need is a BD682 (PNP) instead of the BD681 (NPN).

24/08/2009 Progress

EV Batteries

......Charger

• Got reply from Computronics ... that RSP-1000-48 is in stock !
• Emailed back asking for total price, so that I can pay immediately.
• This is the best ever !
• Also ... worked on 'Shunt' circuit, and other circuits required.
• Posted suggestion on rc-groups (here) and got reply.
• Further communication, trying to work out how to get it right.

......Control circuit

• Put wires on the circuit, and circuit in box.
• Thoroughly testing ... discovered that the LM338K is putting out full voltage ! Darn !
• More testing, and the LM338K is stuffed.
• But it had worked perfectly on the breadboard. It seemed to work on the circuit board (but I hadn't tested the voltage).
• Testing and thinking and studying, gives 2 possibilities: [1] the soldering heat was too high (I did have difficulty trying to solder to the can) ... or [2] kick-back from capacitors in the "hidden" circuit in the Contactor, frying it (idea from application notes to device).
• Decision: Buy another LM338K, and test it on the breadboard with the diodes suggested, and not through the contactor, then test with the contactor, then mount it without soldering it so badly !

22/08/2009 Progress

EV Charger (temporary, until better solution created)

• Completed analysis of the most simple, safe (low component count, fast to make) current limited charger for the batteries.
• The absolute easiest thing to do (no balancer, no charger) ... is to create a current limited supply at exactly 4.1volts, and apply individually to each paralleled group of cells.
• This will take time, but is easy to do, fast, and gives an immediate result.
• Meanwhile, I can order the parts for a more sophisticated high power balancer for very fast charging of the whole pack ... and work out what I can do for a power supply with current limiting.

......List of parts required (and able to be bought on weekends, locally)

• ZV1630 LM338K ($16.95)
• ZT2298 2N2222A NPN Transistor ($0.35)
• RT4642 500Ohm 25 Turn Trimpot (vertical) ($2.75)
• RR0548 100Ohm Resistors ($0.46)
• RR0557 240Ohm Resistors ($0.46)
• RR3206 0R1 (0.1) Ohm Resistor (5W) ($0.40)
• RR3208 0R22 (0.22) Ohm Resistor (5W) ($0.40)
• RR3214 0R47 (0.47) Ohm Resistor (5W) ($0.40)

......Note - the resistors are 5W, because low ohms only come this large, locally !

21/08/2009 Progress

EV Charger

• Thorough research ... found that some switchmode power supplies have current limiting if overloaded ... so can just choose a power supply with max rating (according to specifications sheet) being the maximum charging rate required.
• Meanwell RSP series is perfection (constant current limiting, resistor controlled voltage, remote on/off, and seperate 5.0V power supply).
• Meanwell PSP series is still good (constant current limiting, but with less features)
• Meanwell SP series (cheaper, and without features) has "Fold back current limiting"
• !! NOTE !!! Other, cheaper supplies, have "hiccup" protection, meaning completely switches off, waits, and re-starts ... which would be complete and utter crap for charging.
• Lamda supplies also look good.
• Have sent email to Computronics, who import the Meanwell power supplies.

......NOTE

• Discovered on "Procon" site, that the "Manson" adjustable power supplies (same as my existing adjustable power supply) can be used in parallel for higher voltages. Also, the "top of the line" Manson power supply (9602) can be controlled externally by (eg) Pic microprocessors. Remote on/off. Remo
• Advantage #1: You get two additional very good power supplies, completely adjustable, and usable for other projects.
• Disadvantage #1: There are two power supplies, meaning double the things that can go wrong.
• Disadvantage #2: The supplies are not "current limited" in the way required, which means that a microprocessor is required to make them work ... which is another thing that can go wrong (with catastrophic consequences !!)
• Disadvantage #3: It can never be just "plug in" and so, cannot be used by the boys.

Fast Lipo Charger

Design of a fast Lipo charger

Introduction - the Objectives

• Any child can plug in the Go-Kart, to be safely charged.
• Charger has numerous fail-safes, so that it cannot "burn" like the iCharger 1010B+ did (see here)

Part #1 - the Balancer

......Unsatisfactory situation

• 'Standard' solution (eg iCharger) seems to run cycles of charging through the main leads, and then discharging high cells, until the pack comes into balance.
• This takes !! FOREVER !! Because the discharge is slow (1A max).
• Also, this will not work very effectively when a large pack is paralleled. The slow discharge rate, means things will take practically forever, to come good.

......A proven alternative

• After searching, I found an excellent solution by "Dan Baldwin" for a very simple High Current balancer. Here is the thread for the circuit.
• At Post #332 is a picture of 12 of these circuits, protecting a 12 cell pack.
• At Post #276 is (unbelievably!) a Target 3001 file (hooray !!)
• Searched and Futurlec (Australia) has all the parts. Cheap too !
• The most important things, it seems, are to heat sink the adjustable zener, and to carefully check the circuit before adding it to batteries.
• !! NOTE !! If the Zener breaks (and this is the kind of risk that might apply to all circuitry) then the voltage to the battery can exceed the safe limit.
• !! THEREFORE !! There needs to be a back-up system, such as the Chargery sensors (on the taps) to sound the alarm (or kill the system) if something goes awry.
• Another possibility (slightly different circuit, maybe not as good) is here and explanatory (user) thread here and here ... but it is really just the same as Dan Baldwin's design ... connected to a large FET that cuts power ! Still inherently unsafe if one of the zeners croaks it !

......Proven Alternative #2

• From Endless-Spehere, Fetcher's "Even Newer 4 to 24-cell Battery Management System (BMS)" found here.
• See adaptions by "methods" for Lipo, here.
• Basically ... just replace the resistors before the LM431 (changing set point) ... and replace the "TC54" with a 3.0volt version.
• Advantage - is that the boards are pre-made (already with holes etc).
• Disadvantage - is that they will take longer to arrive, and they don't "Shunt" as hard as Dan Baldwin's, and they don't have the capacitors to stop "ringing", and there's no "over-charge" cut-off switch.

Part #2 - the 'Fail Safe' Charge Guard

This must protect against over-charging, if balancers break

• Possibility #1: Chargery BM6 (alarm at 4.22volts) ... "last resort"
• Possibility #2: A second adjustable zener, connected in parallel to discharge circuit ... and connected directly to "alarm" circuit for disconnection.
• Possibility #3: A digital circuit reading voltage (eg a PIC AXE).

Part #3 - the Power Supply

• Must be a CC/CV supply ... meaning Constant Current (reduced voltage) ... at Max C rating for batteries, until certain voltage is reached, then switching to the final voltage ... which means that current will reduce

......Possibility #1 - really good switchmode power supply

• See "methods" power supply here. Just a plain "Switched Mode" power supply, that has max amperage matched to the battery pack.
• Best supply is RSP-1000 ... see here (Australian supplier) is $362.00. Specifications state that "Overload" is 105-125% (ie 22A to 26.5A) ... and protection type is: "Constant Current Limiting" and "recovers automatically after fault condition is removed". !! PERFECT !!
• Specifications sheet also describes how can use a potentiometer on output pins, to adjust maximum output voltage ! !! BRILLIANT !!
• Have emailed them for response.
• !! NOTE !! cheaper supplies use "Hiccup" protection on overload. Not good.

......Possibility #2 -

• A fully programmable and controllable power supply, with circuit controlling the current, and voltage.
• Involves "better" quality power supply, and very reliable circuit / programming ... and failsafe.
• This permits charging of smaller packs, and charging of larger packs. Also (of course) permits errors (if used wrongly).

Part #4 - the Controller

• Input from Balancers (turns off, when all are shunting)
• Input from Balancers (reduce current, when shunting)

d

20/08/2009 Progress

Sad

• iCharger 1010B+ Burned out this afternoon, while discharging a single 5A Turnigy (5S).
• It shouldn't have done this.
• It would seem that the large capacitor exploded inside the case, and then started smoking (a direct short, it would seem).
• Clearly, the design does not have enough "safe-guards" in it, and is dangerous to use without complete and total supervision (ie during all 3 hours of a charge !!!).
• Clearly, not a satisfactory solution for large-scale Lipo use.
• Purchased from HobbyCity, so we'll see what their warranty arrangements are like. Hmmm. Doubtful (but we'll see)

Glad

• Installed all the components on my circuit board this morning.
• This afternoon I added wires etc, and tested it.
• Worked perfectly, first time. Hooray !
• This is the device to isolate the battery, and do pre-charger before connecting (through Contactor) ... and to allow immediate and complete battery isolation if get off the seat, or press the "stop" button. See design here

• While at work today (slow day) I thoroughly researched and designed a "high power" fast charging system for the Lipo batteries (see Topic Here)

Press-n-Peel PCB making

Last time I did this (about 4 years ago) it all went swimmingly smooth, and I had no problems at all. This time, was a disaster.

The main change (I believe) is that I have different Laser printers now. A different brand (and, obviously, different toner).

It seems the temperature required for the PCB transfer is much higher. So the instructions for a "Brother MFC" laser printer are:

1. Print the PCB in "Mirror" (it goes upside down, on bottom of the board).
2. Thoroughly scratch the surface of the PCB with 600 wet & dry paper (NOT steel wool).
3. Then wet ... but don't touch with fingers (NO OIL !)
4. Then put a chopping board (wooden) on ironing table, with board on top of that.
5. Then iron on nearly full heat (2/3rds the way up the "steam" setting).
6. Press down hard. Edges need special attention.
7. Do this for 3 or 4 minutes.
8. Leave the board alone while it cools down by itself.

15,16 August 2009 Progress

Electrical Progress

• Discovered "RC" circuit is not quick enough to power the huge Contactor relay. Probably because the Relay has an "Economizer" circuit in it, I believe the relay is expecting a "Switch" type jolt of power. Actually, probably banking on this in order to draw down the relay contacts.
• Tried other kinds of RC circuit modifications. None successful.
  So then decided to use 555 Timer (because it gives a sharp signal).
• Lots, and lots of stuffing around.
• Eventually figured it out, and got it to work perfectly !
  Also ... solved the low voltage power problem (ie 12 volts). Simple Zener diode as power supply, with a series of small (low watt) resistors to "share the load".
• This is effective at the higher voltage, and gives a nice stable 12V power supply for the 555 and for the relay. Effectively, it is a very rough regulator, or "pre-regulator" ... except there is no need to use an actual regulator to get smoother power (555 is robust).
  Tested, and re-tested.
• Then re-drew the PCB (yet again) to include the extra components.

Final Schematic
Final PCB (printed to "Print-n-Peel" PCB paper)

15,16 August 2009 Progress

Kart progress (physical)

• Spent all Sunday afternoon at Poppy's house, cutting & welding.
• Rear brake fitted :)
• Frame created :)
• Rear wheel created :)
• Motor mount created :)

......Work still required:

• Front wheels
• Steering
• Seat

14/08/2009 Progress

Electrical Progress

• Calculated that LM338 will fit.
• Re-drew the PCB to include the LM338 (which is huge !)
• Re-drew the PCB with outputs on the 'left' hand side (for easy plugging)
• Purchased LM338 and 2 filter capacitors recommended by Spec sheet.
• Re-drew the PCB to include the filter capacitors.
• Investigated Pre-regulator (instaed of using LM338). These crude circuits waste a lot of power, especially when the difference in voltage is high. There is a "pure" burning of power in transistors, which will need to be heat-sinked.
• Thorough testing (by drawing over) the circuit ... and discovered +42V being applied to the ground pin of the fuel gauge !!! Would have got there anyway, through capacitor. So added diode. Probably not needed (as there is no ground ... just +ives attached to Guage) ... but there is unknown potential inside the device, to stuff it up !
• Discovered fault in PCB software ... pins on the TO-3 LM338 are wrong !!! Very serious fault, corrected in PCB sofware. Whew !!!

Kart Progress

• Pillow block bearings turned up. Excellent ... but heavy !
• Kilovac relay turned up. Excellent ... and light (especially by comparison with the bearings).

Other notes

• Dropped back the hire car, and collected by car after its repair.
• Enormous difference between the two cars. Interestingly ... the smooth (and gutless, and whiny motored) hire car meant I was very passive in traffic. The STI makes you an aggressive driver. The difference is quite dramatic !

13/08/2009 Progress

Electrical System Progress

• Continued reading on LM2576.
• Discovered it won't work (maximum input voltage is 40V).
• DC-DC converters are all enormous !!
• Searched thoroughly, and could not find a DC-DC regulator that would handle input between 30V to 42.4V (ie the range of my batteries) and output at 12V ... and would handle 3.4A "inrush" current ... and run happily on 130mA (hold current)
• The answer is the LM338K (3V-37V 5A Adjustable) ... it is in a TO-3 package, so just have to work out how to use this package.
• Printed the datasheet

Kart progress

• Printed the plans to go to Poppy's (for welding).
• Went to Poppy's house.
• Worked all day, and completed the Axle ... as well as buying steel for the remainder of the Kart.
• Axle is good. Mounting for brake and gear is the best ... wheel has a slight wobble in it (darn ! but may be able to fix this).
• Got a major headache (probably not enough food / water / caffeine).

iCharger connections

iCharger Charging and Balancing connections

image/picture

12/08/2009 Progress

Electrical System Progress

• Found and downloaded iCharger wiring.
• Added picture of connections (below)
• Ran multi-meter across iCharger break-out board (unexpected results !)
• Selected connectors for iCharge to Cart
• Worked out how to make a battery pack
• Worked out why the regulators keep burning up (too much voltage ! 7815 has maximum of 35volts input).
• Anything with fixed output (12V or 15V) has too-low input voltage.
• An adjustable regulator with enough power (5A) is $16.95 ! --- AND --- is in the TO-3 package (bugger !)
• The LM350K (3A, TO-3 package) can only output 1A at a voltage difference of 30volts ... for $13.95 !
• The LM338K (5A, TO-3 package) has current limit circuitry, and can supply up to 12A for 0.5s ... but only 1A at a voltage difference of 30volts ... for $16.85 !
• This leaves only the LM2576 ... which will give from 4.2Amax to 6.9A max (depending on voltage difference). Part ZV1460, $16.95.
• Purchased the correct switches, and the correct relay.
• Purchased 8AWG and 12AWG wire (the closest I could get to the batteries) to practice soldering on.
• Read the full spec sheet for LM2576 ("Simple Switcher 3A Step-Down Voltage Regulator").

Battery box progress

• Bought Aluminium, and determined the final shape of "battery box".

Cart Progress

• Print Cart steel plans ... to show Poppy, and take to Poppy's house.

11/08/2009 Progress

Circuit Progress

• Research and thinking about the circuit.
• Purchased more parts, and fried another 7515 (these things are really fragile ... they won't handle anything over 1A very well, or at all !)
• Learned further about high power darlington transistors (very high hFE switching ... so seemingly very twitchy on small or "start up" loads).
• Bought more parts (including spares ... in case 555 circuit is required for timing)
• Got the timer circuit to run (hooray !) working on full 42V. Might be too slow to start up on lower voltage ?
• Figured out I can use a "button press" to short the circuit, and turn it off.
• Confirmed by measurement, that once turned off ... absolutely 0A draw (of course !).
• Recreated schematic, very readable !
• Discovered a better relay (shorter ... fit in box better ... and same power).
• Created the schematic and pinout (to scale) in Target 3001.
• Installed it in the schematic, and discovered a potential mistake.
• Completely finished the new version of the schematic, and PCB.

Need to Buy

• Part SY4062 - being DPDT relay, 5A, $5.95 ... model FRF6 relay, spec sheet here: http://www.ficrelay.com.hk/details/FRS6.pdf

Corrected Schematic, and PCB

Circuit Progress

• Research and thinking about the circuit.
• Purchased more parts, and fried another 7515 (these things are really fragile ... they won't handle anything over 1A very well, or at all !)
• Learned further about high power darlington transistors (very high hFE switching ... so seemingly very twitchy on small or "start up" loads).
• Bought more parts (including spares ... in case 555 circuit is required for timing)
• Got the timer circuit to run (hooray !) working on full 42V. Might be too slow to start up on lower voltage ?
• Figured out I can use a "button press" to short the circuit, and turn it off.
• Confirmed by measurement, that once turned off ... absolutely 0A draw (of course !).
• Recreated schematic, very readable !
• Discovered a better relay (shorter ... fit in box better ... and same power).
• Created the schematic and pinout (to scale) in Target 3001.
• Installed it in the schematic, and discovered a potential mistake.
• Completely finished the new version of the schematic, and PCB.

Need to Buy

• Part SY4062 - being DPDT relay, 5A, $5.95 ... model FRF6 relay, spec sheet here: http://www.ficrelay.com.hk/details/FRS6.pdf

10/8/2009 - Progress

Circuit Progress

• Calculated pre-charge resistor
• Refined PCB design (including diodes to protect relays)
• Created list of parts.
• ... THEN ... worked out need to relay switch GROUND (not positive) ... because then can easily switch on the monitors (fuel gauge etc).
• Re-configured the circuit to work switching GROUND (not positive).
• Purchased all parts required to make the 'LIPO Protection' circuit.
• Tested that the circuit will work with switching ground instead of +ive.
• Tested selected relay with circuit ... works perfectly !
• Tested Gauge and alarms with multimeter (even that drain, ruins the circuits).
• Tested Gauge direct through relay to ground ... works perfectly !
• Tested Gauge and alarm together ... works perfectly !
• Purchased switches (decided switch types).
• Tested voltage divider to run relay (realised a mistake in circuit, and fixed).
• Tested relay circuit (works !).
• Testing transistors and delay circuit ... managed to fry a regulator (7515).
• Realised that relay won't fit in the little "Jiffy box" I've got ... so completely re-drew the PCB so that it will fit horizontally (very tight fit).
• Too tired, went to bed at 1.00am

Things to do:

• Test circuit with actual parts (and at actual voltage).

Circuit Progress, 8-10 Aug 2009

Circuit Progress

• Tested the "voltage divider" part of the circuit. Works !
• Tested the "latching relay" part of the circuit. Works !
• Tested the "RC Delay-on" circuit. Works !
• Downloaded PCB software (Target PCB 3001)
• Completed PCB layout

Circuit Design

Objectives:

• Batteries protected ... from slow discharge, or charging when connected ... by being totally and completely isolated automatically when not in use.
• High voltage isolated ... otherwise things will melt, or zap people ... this is important because I am likely to increase the voltage in this or later designs.
• Sparks eliminated ... using pre-charge ... important because controller will change and voltage and amperage will increase, in this or later models.

Component List

• Part ZT2193 ... being BD681 NPN Transistor
• Part SY4052 ... being relay DPDT 12V DC - Coil 285 ohms 2x5A @ 30VDC
• Part RR0556 ... being 220 Ohm 1/2 watt 1% resistors
• Part ZV1515 ... being 15V 1A voltage regulator
• Part RR3274 ... being 100 Ohm 5 watt 5% carbon film resistor
• Part HG9980 ... being 5 sheets "Press N Peel PCB film"
• Part HP9514 ... being Blank Fibreglass PCB 150mm x 75mm
• Part NC4254 ... being Ammonium Persulphate
• Part SP0753 ... being Momentary "on" switch, with 12volt illumnation
• Part HB6004 ... blue Jiffy Box
• Push button (normally open)
• Push button (normally closed)

Component Info: Relay

• Jaycar SY4052 ... being relay DPDT 12V DC - Coil 285 ohms 2x5A @ 30VDC
• This is a "FRM1-2C2" relay. Spec sheet at: http://www.khalus.com.ua/data/chips/chameleon-relay/pdf/frm/frm1.pdf


• ... 285 ohms means 42mA to run it.
• "Must operate" at 75% of 12V = 9V. "Absolute Maximum" coil voltage = 110% of 12V = 13.2V
• Means "voltage divider" circuit of a series of resistors ... 220 ohms, 220 ohms, 220 ohms ... then switch ... then relay coil. This means minimum 9.04Volts through the coil, and maximum 12.66volts (when battery fully charged).
• At R3 = 42v max, At R2 = 32V, At R3 = 22V, at switch at relay = 12.9volts max, and 9.2volts (when batteries are at 40V) ... so each resistor sees 0.455 watts.

Components: Circuit Breaker

• Thermal circuit breakers are very slow (4 seconds to turn off ... how long for short circuit ?)
• A Magnetic circuit breaker will trip very fast (10ms for 1.5 times rate current, 2ms for 4 times rated current, or higher) - see Farnell part 9540768 ... 25A, 80VDC. $66.39.
• Circuit breakers are "hydro" (to slow down the trip speed ... slightly delayed).
• (Perhaps) shouldn't use under-rated ... because they won't be able to stop the spark (ie cut the power) at higher voltages, and higher amperages.
• Magnetic circuit breaker: http://www.marinenutz.com/products/Paneltronics-A-Frame-Magnetic-Circuit-Breaker-%252d-CE-%252d-25-Amp-%252d-Single-Pole-%252d-Black.html

......Alternative #1: $26.95 ... Jaycar CAT. NO. SF2265 - being 100A Amp Circuit Breaker ...

• Maximum 50VDC
• Will trip at 110A.
• !!! Will trip within milliseconds at short circuit !!!

......Alternative #2: $8.85 ... Jaycar CAT. NO. SF2261 - being 25 Amp Circuit Breaker ...

• Maximum 50VDC
• Will last indefinitely at 25A.
• Will trip within 1 hour for 31.25A (125% maximum)
• Will trip within 4 seconds for of 37.5A (150% maximum)

Battery wriring

Safety wiring

......Result of learning ... want

• Fuse - to protect against short circuit, and controller destruction.
• Contactor - to fully isolate batteries (prevent pack destruction if left on accidentally)
• Seat switch - which uses contactor to isolate batteries, if no-one is sitting on the Kart.
• Big Red Kill switch - uses contactor to isolate batteries, in emergency
• Brake microswitch - preventing acceleration while braking.

......Learned about Contactors

• A "Contactor" gives perfect mechanical full isolation.
• Kilovac EV200 (US$65) will break over 4000A at 36volts.
• Gigavac GX200 (US$??) will do the same, maybe cheaper.
• Takes up to 12ms to release (so too slow to prevent life-threatening injury).

......Learned about circuit breakers

• Can be used as a "disconnect" switch.
• There are a range of excellent circuit breakers, designed for 12volts, and 24 volts (car applications) ... but these are "thermal" ... and may allow "over current" for some considerable time before they will switch off.
• Really only considered as "slow short-circuit" protection.
• They won't protect batteries.
• Seperate fuse is required.
• Need seperate "Over Current" protection (seperate device).
  A battery system of 12 to 48 volts works fine with a circuit breakerrated up to 160 volt d.c.
• Search "DC Circuit Breaker" on ebay.com
• Best here = http://www.solarseller.com/dc_circuit_breakers__dc_circuit_breaker_ul_listed_to_125_volts_dc.htm
• Alternative (more expensive) supplier of the Airpax CD-125 type = http://www.cloudelectric.com/product_p/cb-cd125-24.htm
• Poorer example = a 60A breaker http://au.farnell.com/eta/4130-g211-k4m1-70a/circuit-breaker-70a/dp/8567190 takes 1/2 second at 600A to blow !!! $45.70
• Airpax CF-125 breaker (UL listed up to 48 volts, amperage range select 10A through to 110A) ... eg $39 for 60A.

......Fuses

• ANL fuses are OK for low voltage (under 80 volts).
• On www.ebay.com.au - search for "ANL Fuse"

......Learned about Semiconductor switches

• Using Mosfets ... to try and shut down the voltage faster than a physical contactor ...
• .... see UQ design for "Sunshark" ...
• ... but in reality, this is just duplicating the mosfet control in the controller (ie no more reliable than the mosfets in the controller) ...
• ... and doesn't give full electrical isolation.
• Therefore is not useful !

Power Connections

Issues:

• "switch on" spark.
• Switching off (to make sure batteries don't completely drain).
• Disconnecting batteries ??? to charge them in place.
• Connecting for charging

03/08/2009 Progress

Progress

• Arranged to meet Poppy (my Dad) on Thursday 13 August 2009, to:

1. Buy Steel
2. Make rear axle
3. (if time permits, and bearings arrive) make frame

• Cannot do anything before then (unfortunately) because I am too busy during the week (one staff member down, and car in for being fixed) ...
• ... and too busy on the weekend (Rugby matches, Basketball matches, Bucks party for Emmy's Fiance, Hen's night, and kitchen tea).

Final Drawing. All done.

• Takes into account bearings actually purchased.
• Takes into account all other things required for the cart.

Bearings - Final decision (purchased)

Rear Axle

Final axle - precise measurements (from measuring parts).

01/08/09 Progress

Components weighed

Total current design weight = 37.48kg

• 5.50 Front wheels + steering assembly
• 7.15 Wheel
• 0.67 Axle
• 0.38 Sprocket
• 0.52 Chain
• 0.71 Brake disk
• 1.15 Brake caliper & handle
• 5.25 Motor
• 5.34 Batteries
• 0.23 Controller
• 8.33 Frame**
• 1.50 Seat**
• 0.25 Steering wheel**
• 0.50 Other components**

** = estimated

Steel measured

• 3.5m of 75x25 RHS, 1.66mm wall at 2.38kg/metre
• 200x2700 of plate steel, 4.00mm or 5.00mm
• Frame itself weighs 8.33kg without remainder components

Bearings

• Needs to have some kind of mounting arrangement, otherwise cannot get it into the axle. That is ... needs to slide in then mount
• 6004 bearings are ID 20mm, 12mm width ... but need to be "pressed" ... and need to stop the axle moving. eBay source seller is zaccsbearings.
• Source of the other bearings (http://www.smallparts.com.au) is too expensive.
• Located another source - http://www.vxb.com.
• Thoroughly researched Pillow Block bearings (easier to mount, by far).
• Found evidence of successful use for powered Go Karts - on the forum http://www.diygokarts.com
• Checked prices in Australia (double, or more than double).
• Ordered from http://www.vxb.com, using PayPal.
• Needs slight revision of frame design for Kart.

Frame - steel or fibreglass

Decision - is that at present (according to current abilities to fibreglass) must be steel chassis.

Steel:

......+ Very strong

......+ Very fast to make

......- A bit heavy.

......- May have difficulty buying the 1.6mm wall 75x25 RHS.

......- Cannot start now. Need to wait for car to be repaired, then go with Pops to buy.

Fibreglass

......+ Completely funky

......+ Can start today

......+ Extremely fun to make

......+ Very satisfying, if the outcome is good.

......+ May be very lightweight.

......- Takes a long time to build

......- Probably more expensive

......- Only maybe saves 2kg (down from 37.48kg to 35.48kg)

......- Potential strength problems with high-torque motor etc.

Almost final drawing

This drawing has:

• bearings properly drawn (to scale) and in place
• Disk brake properly drawn to scale, and in place - with caliper
• Gears drawn to scale, and placed
• Motor reproduced to scale, and in place
• Battery pack sized and in place (under the seat)
• Chassis properly and finally drawn.