2013 Electrical Systems-Close Call & Lessons Learned Thread

Amberwolf was talking to me this morning about having a central thread where good electrical system design could be discussed. Items that were outside the scope of our 2013 Li Battery Catastrophic Events - Summary Thread that is located here: http://endless-sphere.com/forums/viewtopic.php?f=14&t=47012

I fully supported Amberwolf's idea as I also thought the gent who just had a "lessons learned" while using an automotive fuse in the system that "got crispy" would be beneficial to us all. To me his experience is the kind of thing that can happen to any of us. However by pooling our field experience the shortcomings of products and design approaches can be shared when used in our application environment. His knowledge would sure benefit more than a few. Thread here: http://endless-sphere.com/forums/viewtopic.php?f=3&t=47192 Also Amberwolf's recent AC connector failure, documented here: http://www.endless-sphere.com/forums/viewtopic.php?p=693832#p693832

Right now these nuggets are scattered throughout the forum. Perhaps this thread will help, and we can fine tune the format as we go... For now lets just post the issue here and discuss it here. Pictures of the problem always help.

While reworking 5S packs to make a 10S 36V Pack I worried about something coming loose and arcing.

Initially I applied liquid electrical tape, that doesn't bond well with the solder used on the zippy PCB and require 4+ application.

I've been looking at plasti dip as a cheap way to colour match my rims and it's pretty much the same thing.

I checked with my IR temp gun and it doesn't heat when drying, so it shouldn't disturb pouch cells.

I will engross an entire pouch (as this only touched the top PCB) at some stage and see if it eats away at the foil over time


Also this stuff gets you higher than Joe Cocker, so apply it in the open or in a super duper well ventilated area.

You just recieved a PING battery, let's imagine. It has two leads, a black and a red. You want to put on a couple of anderson connectors. there are two loose wires. STOP!

You have the potential for an explosion, right in front of your face, measuring 10,000 degrees, with a force the same as a large firecracker, and the power of an arc welder, which will spray molten copper all over a 2-3 cubic inch area, including your eyes if they are that close. And it WILL happen! No matter how careful you are, you'll end up touching the loose live terminals together. This may destroy your BMS, and might destroy your battery or even your eyes. Ask me how I know! I've been soldering for 40 years, I've soldered surface mount electronics, i've got a half dozen soldering irons, I ought to know how to solder something live safely, but trust me IT ISN"T SAFE TO PUT A SOLDERING IRON ON A LIVE CIRCUIT WITH LOOSE WIRES unless everything else is insulated thoroughly and carefully. If the live battery terminals can touch, THEY WILL touch. BANG!

Before you put the Andersons on the end of a battery cable, USE THE FUSE, LUKE! Yes, install a fuse.

Step 0: put on your safety glasses. Yes do it, any time you are in the shop, any time you are soldering, and anytime you are working around a live bare battery wire that isn't made up. They cost $2. Just put them on. No goddamn excuses.

Step 1: pull apart the two battery terminals, and put electrical tape over the ends so they can't touch each other. Better yet use heat shrink tubing.

Step 2: Carefully cut the positive lead in two, where you will insert a fuse.

3. Insert an automotive 30A fuseholder, or if you are using a bigger battery, a fuse rated for your battery voltage and current. Note that automotive fuses are only rated for 32V! Solder the leads, and use heat-shrink tubing on them to make a water-resistant seal (better than electrical tape) or use liquid electrical tape. Liquid tape is good, but will not successfully cover small sharp wire points.

4. REMOVE THE FUSE

5. You can now easily work on the Anderson terminals, safely, without worrying about issues. The positive terminal is not live.

6. Now replace the fuse. If it does not go bang, check the andersons with a voltmeter. If you have proved your battery is all hooked up + to red - to black and everything is OK, then plug it in and ride off into the sunset.

LATER:

Going to fix the Anderson connectors? Adding diodes, a kill switch, christmas lights, police siren, a toaster, whatever - REMOVE THE FUSE FIRST! don't ever solder on a battery wire that is live unless all the other terminals are insulated. Don't think you can keep loose bare terminals from touching and going bang. Murphy will make them touch. Remove the fuse to disconnect the battery terminals from power, and work on DEAD terminals.

if you have a BMS on the circuit then the overcurrent protection will shut it off faster than a fuse. of course it has to be working for this to happen. no shorted mosfets accepted.

after chroot had his fire i did recommend people not ever run the B+ battery lead through the BMS. if the BMS requires the B+ to run the circuit current then the lead from the battery to the BMS should be fused. 500mA.

the red charging lead to the B+ terminal should have a fuse at the battery. no more than the charging current +50%.

I've installed a 50A fuse in-line for my 104V 15AH 32S Headway battery system.

Prior to buying the fuse I purchased a 125 V max DC 50A circuit breaker which trips immediately at 65A from a solar power / wind power supply house.

Before I installed the circuit breaker I noticed its cautionary label stating that it is "polarity sensitive - Do not reverse feed".

Fortunately I had a Eureka moment before installing the breaker and realized that regen probably = reverse feed.

I guess these are designed to be mounted with one breaker in drain-only wiring and a separate breaker in isolated charging wiring.

My Headways are rated at up to a 2C charge rate = 30A regen charge, and I've noticed peak regen current of between 20A and 25A for periods of <15 seconds.

What would have happened to the breaker during 15 seconds of peak regen out of 45 seconds almost continuous regen down a steep hill?

Anybody need a 125V 50A DC breaker that is polarity sensitive? I don't expect to build a system without regen. (edit: sold 4/2/2013)

Kinda in relation to Samd's inline fuse holder meltdown...

Very early on in my e-bike building 'sperience I found that the (cheaper) inline, cylindrical fuse holders, made of a semi-translucent white type plastic are crap under heat, and melted one, never went back to em and started using the black plastic type auto blade fuse holders... not a problem since (note only using these types on <60v systems.. above that going for more beefy stuff.)

I also used multi pin connectors for phase and hall wire connections close to the wheel for easy disconnect... MADE FROM THE SAME TYPE OF WHITE SEMI-TRANSLUCENT PLASTIC... I have since had three failures of these connectors, where once more the white semitransparent plastic is totally crap under heat, and hot phase wires managed to melt the plastic internally and cause internal shorts within the connectors...

Conclusion... avoid using connectors made from semi-flexible, white semi translucent plastic where heating may occur.

Joe

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.

But having connector shells that don't distort or melt under mild heat is important too. :lol:

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.
... :lol:

Click to expand...

Damn you and your superior Logic!

Joe

This may apply to eVehicles... I got called over to the neighbor's that her diesel Kubota tractor was arching, smoking and smoldering... Unbolted and pulled the negative battery cable hoping the battery wouldn't blow up when it arced coming off.

Failure mode was a short in the solenoid on the last startup. Starter stayed engaged and on. Got hot smoked, and shorted internally which resulted in the battery leads going up in smoke.

Here is the lesson learned and a great safety feature that I did not notice. The positive conductor was 24 inches long and a gauge heavier than the 12 inch long ground conductor. The positive battery lead to the starter wraps around the polymer diesel fuel tank. It got out of this short with no noticeable damage. The short ground lead, that is away from other wires and the fuel tank basically burnt in two and smoked all the insulation. Son of a gun if these guys didn't design in a "fusible link" in the starter circuit. Props to Kubota... and a lesson I just learned.

dnmun said:

if you have a BMS on the circuit then the overcurrent protection will shut it off faster than a fuse. of course it has to be working for this to happen. no shorted mosfets accepted.

Click to expand...

Sorry. What actually happens is a big chunk of the BMS mosfet comes flying out of the battery case, and the now-shorted mosfet can't control anything. ASK ME HOW I KNOW THIS!!!!

Experience, and the theory that current flow above the MOSFETs power handling spec starts spitting out white smoke faster than the Vatican at Pope-picking time, is how I know! I can post a photo of the aftermath if you want. Short circuit = crater in control electronics and loud noise. Then cursing. Afterwards gnashing of teeth. Then emptying bank account for new BMS.

And how fast is a fuse? Depends on the fuse, and the current. Fuses let through a specific amount of energy (I^2*T), so more current = less time, all the way down to nearly infinite current and immeasurably small time. At very high currents, fuses can be as fast or faster than electronics. Protecting electronics outside their safe operating area.

But the real point is pulling the fuse gives you a manual disconnect that can allow you to safely work on dead wires, rather than trying to keep live wires separated, which you will eventually fail at doing even if you are Thomas Edison! Pull the Fuse, eliminate the danger of loose live wires. Basic electrical safety.

i could have used a fuse on either of these packs. may have prevented my burn?? (what fuse would be acceptable for this) http://endless-sphere.com/forums/viewtopic.php?f=2&t=31627&hilit=184v and for the record i think andersons suck

Lemlux said:

I've installed a 50A fuse in-line for my 104V 15AH 32S Headway battery system.

Prior to buying the fuse I purchased a 125 V max DC 50A circuit breaker which trips immediately at 65A from a solar power / wind power supply house.

Before I installed the circuit breaker I noticed its cautionary label stating that it is "polarity sensitive - Do not reverse feed".

Fortunately I had a Eureka moment before installing the breaker and realized that regen probably = reverse feed.

I guess these are designed to be mounted with one breaker in drain-only wiring and a separate breaker in isolated charging wiring.

My Headways are rated at up to a 2C charge rate = 30A regen charge, and I've noticed peak regen current of between 20A and 25A for periods of <15 seconds.

What would have happened to the breaker during 15 seconds of peak regen out of 45 seconds almost continuous regen down a steep hill?

Anybody need a 125V 50A DC breaker that is polarity sensitive? I don't expect to build a system without regen.

Click to expand...

How these breakers work is similar to their a.c. relatives, they both use a ( thermal) overload component and a (magnetic) short circuit component in the devices, difference being the dc breaker has usually better arc seperators and/or as in yours an electromagnet which sucks the large dc arc into the seperators, connect battery + to - breaker terminal and the internal electromagnet is reversed and it will repel the arc and harder to extinguish it under higher voltage conditions, it could continue and burn the breaker disconnecting a heavy arc.
I wouldnt usually worry overly about reverse currents during charge and regen, i'd concern myself primarily with short circuit unwanted discharges, order of a few 100 to a few 1000A, these do the damage.
These 3kA+ breakers are just brilliant at this job, for the same cost or less of the correct 100V fuse you could have got a domestic unpolarised solar breaker lately and it doubles as a main switch.
You can also rate the breaker low say 40A if you run at 40A battery amps mainly, if the breaker is 'b' curve it will just carry this, but what else is excellent is, carry 100A bursts for maybe 10 secs or more within its design, a fuse would be stressed or blow.
Short the breaker output at 300A and 'pop' a couple of mS as demo.
https://www.youtube.com/watch?v=OnOsSmbXCoQ&feature=youtube_gdata_player

Fuses are totally unsuitable for main battery supply protection, for more reasons i could count on both hands.

So this is bad?

Pretty gold plated barrells and can nail it down, fixes 2 out of 10 of the issues compared to a breaker.
Can you link to its specs?

Car audio product link to that fuse holder.
http://www.sourcingmap.com/inline-agu-fuse-holder-gold-plated-for-car-amplifier-p-37518.html
http://www.bcae1.com/fuses.htm
Have a read what this bloke has to say about the AGU fuse type holder you mention.
Also read the intro about the voltage rating, max 100V dc, usually 80V, over this they tend to go ceramic with filler and becomes an expensive, fuse like this one to do the job.
http://www.altestore.com/store/Enclosures-Electrical-Safety/Overcurrent-Devices-Fuses-Breakers/Fuse-Fuseblocks/50-Amp-250-VAC-125-VDC-Class-R-Fuse/p4160/

megacycle: Thanks for your comments. I think I can summarize by concluding (1) that you don't think that my fused approach is effective, (2) that the greatest potential downside of the polarized breaker I didn't use would be a spike of several hundred or thousand amps to my electronics, and (perhaps) (3) a current code compliant non-polarized breaker is probably the way to go.

Your comments also caused a member who doesn't use regen to contact me and buy the never used breaker from me today.

Here's a picture of the label on the breaker.



Interesting that the breaker was made in Lesotho where, according to Wikipedia, 40% of the population earns <USD 1.25 per day, the "International Poverty Line".

Lemlux said:

megacycle: Thanks for your comments. I think I can summarize by concluding (1) that you don't think that my fused approach is effective, (2) that the greatest potential downside of the polarized breaker I didn't use would be a spike of several hundred or thousand amps to my electronics, and (perhaps) (3) a current code compliant non-polarized breaker is probably the way to go.

Your comments also caused a member who doesn't use regen to contact me and buy the never used breaker from me today.

Here's a picture of the label on the breaker.



Interesting that the breaker was made in Lesotho where, according to Wikipedia, 40% of the population earns <USD 1.25 per day, the "International Poverty Line".

Click to expand...

Yeh wondered where Losotho was, i'm use to European and Australian breakers, in Australia, The GE ones i got, not sure, the Moeller breakers says Austria, yours looks bigger like 'Square D' mine are to fit standard 'din rail'


Even the specifications on the label were confusing me, as they rate differently.
Our specs say 50A breaker would trip at 50A in a given time frame.
If its tripping instantly at 65A it wouldnt be b or c curve, to quick.
I cant even find its spec's anywhere.

1.If you have a fuse rated at above your battery rated voltage and will allow your rated controller current to flow, can sustain the overload bursts and can catch a partial or full short circuit in mS, to limit damage, isnt in a crappy holder causing crap connection issues, can indicate its blown, you have means of detecting a fault after the fuse has blown and there is a dc switch, rated at the continous and overload rating before it, to turn the power off to change it, so if it does blow again, its not happening when your fingers are inserting it into the holder, bang KFC+$20+ walk, qed.
Oh and got a spare $20 for 2 spares, i say go for it i'll use my sub $10 mcb,

2. The polarised breaker is fine, (its not electronic) if connected correctly, as a fault during regen, more than likely happen at your fets, not at the battery and the breaker should be rated to not trip during normal regen levels anyhow, because that could be nasty, just like your battery fuse blowing during regen, forward current is usually a lot higher than regen.

3.Either or, but preferrable if your prone to distraction, like i am sometimes, BANG!!.

I use these with solar where the voltages are a lot higher and opening a solar circuit makes the panel voltages go to maximum open circuit voltage and low current, so it may need to open a 200Vdc@10 amp, a long, low current arc tends to burn through stuff, so the breaker snuffs the arc as it switches off? as operating it as a switch it is still sprung to trigger.
Battery systems different, when the cb trips, will probably be just under battery voltage, as the battery is recovering from the short circuit on opening, but current will be 100's of amps normally, enough to explode stuff and so it trips in mS to restrict the I2t power damage 8).

I would avoid the use of silicone wire for ebike use all together due to the poor abrasion resistance of the sheathing material. You can literally cut it with your fingernail and over time any chaffing due to vibration and hard surface will wear through and create possible shorts. This has happened to one of my packs. I'm in the process of rewireing any section that might be prone to such abuse. Best to use automotive or marine grade wiring from the start.

megacycle said:

Lemlux said:

Your comments also caused a member who doesn't use regen to contact me and buy the never used breaker from me today.

Here's a picture of the label on the breaker.

Click to expand...

... yours looks bigger like 'Square D' mine are to fit standard 'din rail'

Even the specifications on the label were confusing me, as they rate differently.
Our specs say 50A breaker would trip at 50A in a given time frame.
If its tripping instantly at 65A it wouldnt be b or c curve, to quick.
I cant even find its spec's anywhere.
...
2. The polarised breaker is fine ...

3.Either or, but preferrable if your prone to distraction, like i am sometimes, BANG!!.

I use these with solar where the voltages are a lot higher and opening a solar circuit makes the panel voltages go to maximum open circuit voltage and low current, so it may need to open a 200Vdc@10 amp, a long, low current arc tends to burn through stuff, so the breaker snuffs the arc as it switches off? as operating it as a switch it is still sprung to trigger.
Battery systems different, when the cb trips, will probably be just under battery voltage, as the battery is recovering from the short circuit on opening, but current will be 100's of amps normally, enough to explode stuff and so it trips in mS to restrict the I2t power damage 8).

Click to expand...

I'm the member who bought the breaker from Lemiux. I tried a test today with a 6s 5ah turnigy pack with the breaker ground plugged in with 18ga (fusable link?) wire and bullet connectors and a bullet socket on the battery side and 14ga stranded wire on the plus breaker side, the idea being to test it by shorting the stranded wire to the bullet socket. Got a nice spark and vaporised about 5mm of stranded wire but it did not trip the breaker. Tried it three times but no trip. I think the stranded wire is hitting the socket one strand at a time vaping each in turn. Maybe I should tin the ends being shorted and then retest.

But, it would be nice to know that this breaker is not some sort of slo-blow apparatus so if anyone does find specs that would be helpful.

Questions:

- What curve would be best for this application anyway?
- Is the din rail size smaller?

-dg said:

I'm the member who bought the breaker from Lemiux. I tried a test today with a 6s 5ah turnigy pack with the breaker ground plugged in with 18ga (fusable link?) wire and bullet connectors and a bullet socket on the battery side and 14ga stranded wire on the plus breaker side, the idea being to test it by shorting the stranded wire to the bullet socket. Got a nice spark and vaporised about 5mm of stranded wire but it did not trip the breaker. Tried it three times but no trip. I think the stranded wire is hitting the socket one strand at a time vaping each in turn. Maybe I should tin the ends being shorted and then retest.

But, it would be nice to know that this breaker is not some sort of slo-blow apparatus so if anyone does find specs that would be helpful.

Questions:

- What curve would be best for this application anyway?
- Is the din rail size smaller?

Click to expand...

I'm wondering if your pushing enough current if its a 'b' curve it should trip in a few secs at around 125A+, instantly maybe 150-200A, if its 'c' it will push the trip curve over so it sustains a higher level of overload. Have you got anything to measure current?
I'll post somes common curves later' busy at moment.
Twisting your strands tight together might help your test.

http://www.diynot.com/wiki/electrics:fuses_and_miniature_circuit_breakers:mcb2
Found the link informative easy to follow. Had the following curves diagram.

Hope its of some help -dg.
Your circuit must supply about 3 x 50A, if a 'b' curve to trip instantly.
Youd need a good charged battery capable of 30C too, to reach instant tripping.
Your charged lipo sagging under fault feeding the circuit, could be like 22/0.1=220A short circuit.
Make sure your short is slapped on solid, as your pushing your battery limits,
if your fault is adding 0.1ohm it might not trip instantly either.
Have a back way of isolating circuit incase its another curve or doesnt trip.
I would advise wearing welding gloves or similar, if your testing as you have been, could get burnt using those small cables, heating up, ups their resistance too, be carefull of burns and splatter.
Oh and i wouldnt test it too much,
It could end up pitting the contacts if you trip it too much.

To keep the cables from torching I cut one shorter then the other. So if I had a battery one lead would be 4 inches long with .5 inch of conductor exposed and the other will be eight inches with .5 inches of conductor exposed so in order for them to touch you would physically bend the longer wire to touch the shorter one. I then do the oppsit on the wires I am conecting the battery to so if the (+) was the long wire coming off the battery then the (+) I am soldering too would be the short wire and vice versa for the (-).

ok, I'll bite

If you leave your batteries connected to your controller for months on end it turns out they will drain down to 0 volts.

megacycle said:

Hope its of some help -dg.
Your circuit must supply about 3 x 50A, if a 'b' curve to trip instantly.
Youd need a good charged battery capable of 30C too, to reach instant tripping.
Your charged lipo sagging under fault feeding the circuit, could be like 22/0.1=220A short circuit.
Make sure your short is slapped on solid, as your pushing your battery limits,
if your fault is adding 0.1ohm it might not trip instantly either.
Have a back way of isolating circuit incase its another curve or doesnt trip.
I would advise wearing welding gloves or similar, if your testing as you have been, could get burnt using those small cables, heating up, ups their resistance too, be carefull of burns and splatter.
Oh and i wouldnt test it too much,
It could end up pitting the contacts if you trip it too much.

Click to expand...

I missed seeing this earlier. Thanks, it is very helpful. Now I have to decide if I am brave enough to test it.

Yeh 24V with a low Ah, low C battery is fairly lame for tripping a high A rating breaker.You got nanotech's?
It should fit with your motor load anyhow, if your pulling less than 30A battery, maxed out, you could even go a 16A or 20A and get the close protection breakers are good for, i'm thinking 40A really for guys hovering @ 100A set ups.

120V slammed across that short from my good 8Ah flightmax, sounded like a pistol shot, a neighbours dog up the road went off, good job it was midday, can hear it in the video background :lol:.