2013 Electrical Systems-Close Call & Lessons Learned Thread

Status
Not open for further replies.
I'm an electronics geek, love motorcycles, atvs and snowmobiles. In the past 15 years I have run across my gamut of used and abused vehicles and witnessing some scary opportunities for electrical shorts and fires. So, I'll toss my advice on here in hopes of helping. I used to be an IT guy who services mainframes, server clusters, and the typical batch of retail bench repair nonsense.
I'm helping a distributor for e-bikes in the 450W range for an Anglican church ministry in Vancouver, WA USA. Using simple frames and AGM batteries.

1. Wrap your harness, it's basic and easy to do. Secure that wiring with tape for water protection, and adding some abrasion resistance. Zip tie were if flops and moves from vibration to prevent chafing and flex breaking of the wire strands. Leave about 1/4" of the harness exposed for testing and identification of the wires by color code if you like. (Don't get me started about chaffing and the aircraft industry problem..)

2. Whenever possible, use tin coated copper wire. Nothing like stripping the wire and finding green oxidized copper wire that ask as a free heater unit to your circuit. The water and air gets under the insulation, and makes the wire brittle. This is a long term problem and will get you eventually. Use silver wire like the old aircraft of WWII for high amperage and high voltage sections. That leads to number 3.

3. Coat the backside of the Molex or nylon connectors with RTV silicone, Permatex black, to prevent the pins inside from getting wet, and keeps the pins and their retainer tabs from moving around. To date I have replaced hundreds of bent pins, loose pins, pins that fallen out of the connector, and we are talking about cheap Chinese pins that was sometimes a bit on the thin side for the voltage and amperage going through the circuit.
If you accidentally immerse your machine in water, or have an exposed loose conductor, it's hard for it to arc to something else, if all the other metal bits are coated. Even a couple plies of cheap electrical tape has several hundred volts of insulation resistance.

4. Heat shrink is your friend, heat shrink with RTV squirted in there before curing adds strain relief to support the wire bundle before it goes into the connectors. Sort of an impromptu molded cable. Heat shrink with adhesive lining which is a bit harder to find, is just as good. 3M is not the only supplier, there are plenty of generic suppliers. Get a simple $14 heat gun at Harbor Freight to get it right.

Cooling:

5. Get some airflow around that controller, it might make the difference in the long term to save your precious transistors. The trick is not having the biggest metal heat sink on there, but how fast you can draw the heat from it. Airflow helps the metal get rid of the heat.
Think about a small DC fan to push air from a small hole in the front to the rear of your controller/electronics. Just like the PC industry learned, the first Pentium processors were woefully melting and re-flowing the solder on the neighboring video card and sometimes thermally cooking the voltage regulators nearby. Hot and cold cycles break wires in cpus, transistors,

Extreme Mods:
Copper shims to draw the heat from the transistors to the heat sinks. Look at the Pentium 4- 4 core and better heat sinks for PCs. Duplicate it.

Oil cooling is an option for controllers on sinks that are sealed. The high end gaming PCs went to the old school of copper sinks and water cooling.
I doubt the controllers need a pumped system with a radiator yet. But, I bet there is someone out there with a 8-50 kw controller that might need that.
If the heat sink is just warm to the touch you are ok, even better when it's at room temperature.
http://www.youtube.com/watch?v=6sP45uBj4-k

Thanks megacycle for the fuse chart. It was informative. I am still not sure what size fuse or breaker to put on a 450W 48v or 60v bike that is only pulling and average of 10-30 amps since the controller is pulsing the motor. I think I will get an o-scope on the B+ wire and see what's the average current and size the fuse and breaker from there.
 
Thanks Mechanerd for your info too very informative, i've been in the game, mainly electrical for 30 years and some of what you mentioned i did'nt know the reason for tinned conductors, i never really new why, even seeing them often thanks for that, that's one of the main things, never ever stop learning in electrical/electronics.

With MCB overcurrent really have to allow for uphill powered senario's, generally i would say, if your controller is limited you could go by that as your peak rating, because you might have to allow a fair while to get up a hill, unless you never leave the flat.
I suppose your max amps is logged if you have the CA too and could follow that, but that's after the event :? my last controller was to 45A, now if thats going up a hill flat out, i really would'nt want it to trip with a car up my butt :cry: looking at a 32A it would be running at about 1.4In, looking at the graph, logarithmic :? so it's tricky to work out, between 1 & 30mins you could abruptly stopped
Now we can go into a trick with breakers, Edit, some of them are adjustable, you could tweak this to a 35A, but i'd have to have a thread on it. Might be better utilised for ev sized like motorbikes and cars if there's interest, as i know of it and the operation and can see most of the issues, but never actually had to do a small one.
I'm guessing, but not 100% that most mcb's are standard guts, in the standard range of values, 10-63A, from the good manafacturers and it interests me to know, because this could mean we could have 1 circuit breaker constantly variable settings up to possibly 63A+ continuous, burst to close to that 1.4In rating close to 90A this could be at voltages up to about 220V for most standard breakers 20kW sustained peaks.I just would'nt want to see them tripping to often with that though :twisted: :cry:

Before i get to excited by the above concept, it is still easy with a bike in my case to fit a 40A in the standard range, as it comes down to what i'm trying to protect, it has cabling to handle the continuous and short circuit currents, the controller was a $200 bucks and really i want an improvement on a close rated fuse which would have to be minimum 45 or 50A and be stressed by any peaks values, possibly to a point after a lot of hours it could just fail in more than likely one of those peak moments.
 
Actually i've been a bit slack in regards to using the given formula's for working out values and i may go into that if i start an mcb calibration thread to mod a few breakers.
 
Here is a thought guys if your working on something live wear washing up gloves. I do something like that working on 415v ac live. Get one of these BP connectors over the cable after you chop the exposed copper off.
55490505.jpg





What is a good ide ais not taking the negative near the positive and you do your anderson plug first leaving two tails. One (+)will go to your switch/fuse/relay and the other (-) to a disconnect terminal. Doing that keeps the positive and negative away from each other. Its better to do that and use a clamp to hold the cables to the surrounding frame so there cant be any pulling on the wires to make them come out of terminals

076.jpg
 
How do you go working live these days Chris?
Use to be go, now if you need like a variance of the WHS regs.
Reg 157, which you can only use for test and commissing with a spotter and only were there's no other way, or your a liney for a network, even then the minimum PPE washing gloves would be LV certified and leather riggers over that.
$6k individual or $30k corp.
So be safe out there bud, like what you'd have to fork out if you arc blast a board and take the main out and/or get burnt.
With the new regs, the latte sippers maybe looking for examples to justify their big spend on the WHS harmonisation role out.

PS love the blueys +1 on always using the BP's.
 
megacycle said:
How do you go working live these days Chris?
Use to be go, now if you need like a variance of the WHS regs.
Reg 157, which you can only use for test and commissing with a spotter and only were there's no other way, or your a liney for a network, even then the minimum PPE washing gloves would be LV certified and leather riggers over that.
$6k individual or $30k corp.
So be safe out there bud, like what you'd have to fork out if you arc blast a board and take the main out and/or get burnt.
With the new regs, the latte sippers maybe looking for examples to justify their big spend on the WHS harmonisation role out.

PS love the blueys +1 on always using the BP's.

240 & 415v commercial and residential switchboards I work on. I dont do high voltage because I am not that sort of electrician ( and dont want to glow in the dark) but 120v dc like in this thread is more a short circuit high current concern. Energy companies and OH&s are always crapping on about something and they change the rules every time they change their undies then later some of the changes go back.

If these guys have nothing else then one or two layers of rubber gloves is a good start if they are concerned or have nothing else handy. Its better than working on it bare handed.

Lol i wouldn't work on live ac with washing gloves I would use the proper ones from electrical wholesalers and a rubber mat on the ground if I was really concerned particularly on live busbars or the line side of a main switch because that is merciless if something stuffs up. I hate doing it anytime and prefer to Isolate supply wherever possible, live is last choice believe me.

These guys are worried about short circuit current. I would be worried about it too and take precautions to insulate cables against short circuit contact wherever necessary.

They are just a couple of safety tips that have kept me alive this long. With a little care and prior planning you can avoid the short circuit phenomenon fairly easily.
 
Which side of this rock you livin Chris, you in NSW?
I know what you mean with the WHS, but more butt covering is needed these days though, there's more office walla's than us guys doing the work, justifying their existance and i could'nt afford the 30k.
The leather over gloves are used to stop the rubber copping the arc blast or snagging on a sharp on a live one.
I had to wear them working for the network, hard to work in when your testing and jotting down results :?

Yeh the battery stuff is mostly arc flash and protecting against a short, sunglasses and even those tight fit cotton gloves might be better than nothing if your so inclinded.
 
megacycle said:
Which side of this rock you livin Chris, you in NSW?
I know what you mean with the WHS, but more butt covering is needed these days though, there's more office walla's than us guys doing the work, justifying their existance and i could'nt afford the 30k.
The leather over gloves are used to stop the rubber copping the arc blast or snagging on a sharp on a live one.
I had to wear them working for the network, hard to work in when your testing and jotting down results :?

Yeh the battery stuff is mostly arc flash and protecting against a short, sunglasses and even those tight fit cotton gloves might be better than nothing if your so inclinded.
look at this wisdom..

I%2Bknow%2Bnothing.jpg


I see nothing,
I was not there..

Confession is good for the soul but is often a bad career move.
 
"On some occasions I look the other way because in war I do not like to take sides".

"Makes wonderful strudel". :D
 
mechanerd said:
1. Wrap your harness, it's basic and easy to do. Secure that wiring with tape for water protection, and adding some abrasion resistance.

As a dedicated e-bike mechanic, I am not so convinced that telling people to tape wrap their harnesses is such a great idea. Some of the cocoons of tape I have had to cut through to get to the fried connectors inside give me recurring night mares :D

In my personal opinion, tape, (electrical, duct, cloth) sucks the proverbial, creates maintenance issues and is extremely hard for novice to get "right"... maybe looking neat but ultimately creating more hassles than not taping. I'd be willing to go so far as to say that using tape on or around electrical connections is a temporary solution at best and if its use can be avoided it should be.

Apart from that, Mechanerd, Love your advice.

Joe
 
Not quite eBike related, but electrical as in panels and safety related. ... I got asked to "help" at a commercial facility today to trouble shoot some electrical issues. Here are three of my top pet peaves in opening up a panel, as in don't do this...

  • Don't let the wiring from your breakers press against the front panel edge right where the panel cover screws go in. I hate when the cover screw punctures the insulation as you are removing the cover.
  • Even worse, when you cut your main #3 cable 8 inches short for the panel, make a splice right at the entrance bung nut with split bolts. Tape the heck out of the mess, and then have the panel cover pushing directly on who knows how many layers of tape to seat correctly. How I hate shorting out the incoming mains with no line fusing...
  • Finally when the service entrance is a story above in another commercial space, and you expand your operations, Don't do stupid sh_t! Like have a 100 amp main upstairs, with #3 feeders, then decide you need more power and just change out the 3 phase main from 100 amps to 225 amps on the same #3 feeders! The feeders should have changed to 4/0!!

What a mess...
 
Looking for defects like that all the time Dave, as electrical workers here in Aussie we are now obligated by law to note down the defects on certificates.
The certificate we fill out, puts the onus back on to the installation owner to have the issues fixed, obviously can often be contentious and the large fines for non compliance are written on the document.

Example, contractor comes to that board. First thing i'm supposed to do myself should be suit up, fire resistant overalls, gloves, visor and at least let someone know i'm there, before i even open the switchboard.
If the breaker wiring was a mess, to me, it would be notable, as a non dangerous defect that needs attention, could be fixed by internal ducting and/or cable ties but needs an outage, unless i've got good reason.
The main cable would be classed as immediately dangerous and we're given the power to disconnect the feeder for safety reasons, in my time i've done or arranged to have this done on a few times, eg once when a rusted top of an outside switchboard had dropped in and taken a phase out, man my spincter goes when i see potential mains shorts within inches of my face. Usually i'll make the area as safe as i can and report it immediately.
Feeder with incorrect protection might not be immediately noticable unless the place was having a proper regular inspection, very few do, unless they're large concerns, they all should do and wish they all did, keep us in work for ever.
I think most sparkies would find defects on any installation, often subtle like incorrect discrimination or co-ordination of protection, distribution of loading, man where do you stop.
Thermal imaging, breaker testing, earth loop impedance, rcd testing,

Anyway Dave, sounds like some good out of hours work for someone :wink:
 
amberwolf said:
A better conclusion to draw from that is not to use connectors with insufficient contacts/too much resistance/poor connection (to wire or to the other contact), so the heat doesnt' get generated in the first place. ;)

Not sure that would have saved me Amberwolf. Some advice from Luke/LFP regarding the arcs in fuses and how they don't quench in fuses, versus circuit breakers.

I still have that fuse and the internals still show a small bit of metal that was flung out when the fuse tripped then ended back up across the internal fuse bridge. In the end it heated the fuse casing to the point of ignition, but there's still 2volts apparent across the terminals. The fuse didn't cut the circuit in the end.
http://www.endless-sphere.com/forums/viewtopic.php?f=3&t=47192
 
Samd said:
Not sure that would have saved me Amberwolf.
Probably not, but that is not what my statement was in regards to, or in reply to. ;)
 
Not so ebike related but.....

A long time ago I asked one of the techs to tidy up the cabling that festooned the lab. 5 mins later he was done but had also managed to cut the lines to the phones, faxes and modems! So, I'm up a ladder trying to sort out the mess, I've found where the phone line is cut but my hands are full of wire strippers, junction boxes, screwdrivers etc. Anyway, not wanting to lose the cable among all the dead stuff I held it in my teeth....

It wasn't so much the electric shock that made me fall off - more the surprise.

And yes, I did know it was live.... Just not thinking.
 
mechanerd said:
2. Whenever possible, use tin coated copper wire.
One caveat to this: do not tin multi-strand copper wire with solder if you will be clamping the wire in a screw or crimp terminal. The solder creeps over time under the pressure of the terminal and loosens the connection. Crimping then soldering is acceptable.
 
Two things
Caliban said:
It wasn't so much the electric shock that made me fall off - more the surprise.
:lol: , sorry could'nt help it, hope you were non the worse for the experience, done just as silly stuff too.
Tom L said:
mechanerd said:
Crimping then soldering is acceptable.
When soldering a big connection, heat the connection and draw the solder through the connection from the opposite end, same for silver soldering, it's drawn through.
 
Tom L said:
mechanerd said:
2. Whenever possible, use tin coated copper wire.
One caveat to this: do not tin multi-strand copper wire with solder if you will be clamping the wire in a screw or crimp terminal. The solder creeps over time under the pressure of the terminal and loosens the connection. Crimping then soldering is acceptable.

Yes, I agree. You put pressure on the soldered wire with a screw terminal or something you are just asking for it to crush and fracture over time.

What I meant by tinned copper wire is , is stranded wire that is electroplated with tine during the manufacturing process itself. It tends to be more weather resistant, and is nominally no more expensive over the long haul.
 
in february,

I was wiring up a multiconnector for charging my 14S pack while in the bike.
used a 26pins aviation connector for this. 15 pins for balancing, 4 pins each for the charging current (1420i -> 9A)

Wired it all up, mounted my batts, everything was looking good.

Than I decided to double check, by measuring the voltages of each balance wire, to see if they were in the correct order.

Got to about the 8th tab, when my hand slipped, and my multimeter tip caused a short. Was a small spark.
As a reaction I quickly pulled the tips of the multimeter away from the connector, but the spark jumped the gap, and went to another tab when my test leads were to far.

The whole connector began arc welding.

This was very very bad, all my lipo was in the framebag, al ec5 connectors were secured with a tiewrap to prevent coming loose when riding.

Also, this was inside the house I'm building, lots of dry wood a few meters away.

So, I did the only think I could do, grab the bike with my left hand at the handlebar, right hand under the saddle, and threw it outside.
Now, there's no easy route out from where I was, there's pallets of plasterboard, tools, lumber, ...
Took me about 12 seconds I guess, before I got it out.

By that time, the connector had melted in a puddle of plastic, extinguishing the fire.

Now (and here's the fun part), the connector was in a waterproof box, under the saddle.
28u31gw.jpg


Damage to connector
waizyw.jpg


left hand at the handlebar, right hand under the saddle, and threw it outside

I never felt the heat, doctors say I must've burnt through the nerves almost instantly.
Had 3 surgeries, hand is working and looking fine again now, thankfully.

this was after the first 2 hour clean-up, removing all the black, fried bits. Leathery fingers after that.
2nsxrg9.jpg


After the first surgery
30dj8r9.jpg



What really bothers me is the fact that I'm very tedious when it comes to wiring. Everything triple-checked, heat shrink, very good solders, ...
If I had just trusted my skills, and not had the urge to measure again, it would've been fine. I'm sure the connector was wired up correctly.
Bummer.

Luckily the pack survived.
 
wannesd said:
Luckily the pack survived.

Dude, you are a legend!

Mate, I understand that sucks and glad the hand is coming back, however, Truly, to have the sheer determination to remove the bike from your house, basically manhandling the beast whilst your fingers fried, Despite the fact that the arc extinguished itself... Is the stuff that heros are made of.... And aint it a bizzare thing, that when it came down to "do or be done" your choice of "do" probably took milliseconds to kick off and then all automatic and getting it done.

Humans are amazing

e-bikers are amazing humans :D

Joe
 
Glad you are on the mend! Nightmares happen "so fast..."

So, do you think the secondary spark vaporized enough metal to create a sustaining plasma? Was the connector shell on the bike female contacts? 14S only about 60 volts... unexpected failure mode for sure.

This is great information, thanks for sharing it.
 
Yes, the male part was plugged in with the contacts at the backside showing, to get better accessibility for taking measurements. The two are now permanently fused together :wink:
 
http://endless-sphere.com/forums/viewtopic.php?f=6&t=48388&hilit=House+fire&start=25

This person recently lost his house, car, ect. be sure to treat Lipos with respect and take the necessary precautions. I cover my setup with a fire blanket when charging, always balance charge, use a fire alarm and keep a fire extinguisher nearby. Store packs in a fire safe area. I use an old office quality metal file cabinet.
 
Jeebers kfong... for every LiPo house fire, there are thousands of LiPo users Who Do Not Have House Fires.

I appreciate scare-mongering is fun though

Joe
 
Status
Not open for further replies.
Back
Top