I am having a problem and I need some help...

[dak664]

There is no need for a huge load or accurate current measurement to test for bad connections. Cheap multimeters have a 200 millivolt scale, for 0.1 milliohm resolution with around an ampere of load and you will have plenty of time to make voltage measurements without draining the battery pack. A bad connection between a battery terminal and bus bar will show up as a comparatively large voltage difference, like 10x or more.

Aluminum is a good conductor but making reliable connections is hard. It is softer than copper so deforms under a screw head, and the connection gets looser with each temperature cycle. When it is loose enough to allow air to penetrate, a high resistance surface layer of oxide will form. Copper is much more forgiving.

 

[SlyCayer]

Hi, but how do I create a 1A load? That's my problem.

 

[Jeremy Harris]

Take a look at Ohms Law. This defines the fixed relationship between resistance, current and voltage. As an equation it can be expressed as:

I = V/R (where I is the current in amps, V is the voltage in volts and R is the resistance in Ohms).

So, to get 1 amp with a 40V battery you need a 40 Ohm resistor. To get 10 amps you need a 4 Ohm resistor.

The load resistor needs to be able to handle the power (as heat) that running that much current through it will generate. To work this out you just need to know that power = volts x amps. In this case, 40V at 1A is 40 watts, 40V at 10A is 400 watts.

You can measure the resistance of a load with your multimeter, but it will read low, as most heating elements have a resistance that increases as they get hot.

For this test you don't need a high current, so if you can find a household heating element that has a resistance around the right value you should be able to use it OK.

Jermey

 

[ebkr]

The best adjustable dummy load I use on my 12V SLA cells is the 300W inverter and some construction lights as a plain active load which I can change gradually using the dimmer or just other wattage bulb. The inverter costs only like 30 bucks and the 250W light is 15 or so. Can't go cheaper then that. But it takes only 11 -14V so you'd have to test your whole battery string by string only or something like that. So just take it as an idea. Let me know if you need any further explanations... If you don't want to mess up with strings, but rather test the whole battery you can use some high resistance wire from the cheap heat gun or hairdryer and simply cut the coil proportionally to your voltage. 110V is 100%. If your battery is 80V use 80% of coil. heat gun is 10-15$ dirt cheap and the coil is already nicely wound on the heat resistant thing, so you don't have to use bricks or anything else. You can even continue to use the heatgun so you don't have to kill it, just use a portion of the coil as the load. My coil gun is 1.75KW which makes quite decent load I'd say.

 

[SlyCayer]

I have a rice cooker that is 400 Watts, that would make a good ~10A load at 43V? Again sorry for being slow in this, I just don't want to screw up my wife's appliances or screw up my $5,000.00 battery pack.

If the rice cooker will do, I put the + of the battery on the + of the rice cooker and the - on the -? Correct me if this will not work.

 

[dak664]

Forget the rice cooker and any other appliance. They may have a solid state control that could be damaged. Just use a trouble light with a 60 or 100 watt incandescent bulb. There is no plus or minus on the ac plug, polarity doesn't matter.

 

[SlyCayer]

Ohh,

I will do that tonight.

 

[SlyCayer]

I just got my spot welder done(will post some videos and comments about my spot welder build soon).

I need to make sure I am gonna order the correct copper for my bus bars. is 1/8" X 3/4" Flat bar enough for my bus bar that will carry almost 500 amps? Or should I got with the way more expensive 1/4" X 3/4" flat bar?

Please let know if the 1/8" is good enough. Also, anyone know any good source for copper?

 

[Jeremy Harris]

1/8" x 3/4" copper bar has a resistance of around 0.28mohms per metre, or 0.085mohms per foot. You want to keep the voltage drop as low as you reasonably can, so as you don't waste too much power, but at 500A a 1ft long buss bar of this size will only drop 500 x 0.000085 = 0.0425V volts, or waste 500^2 x 0.000085 = 21.25 watts, which is probably just about acceptable. If you have long buss bars, then you can multiply up these figures. For example, a 4ft long 1/8" x 3/4" buss bar carrying 500A will drop 0.17V and lose 85 watts as heat, so will run warm at this load.

If you went up to 1/4" x 3/4", say for long buss bars, then the above values are halved, so a 4ft long buss bar carrying 500A will only drop about 0.085V and lose about 42.5 watts as heat.

It also depends as to how long you think you'll be running at 500A. The buss bars really only need to be sized for best efficiency at your average current, so if this is lower, then the average losses will be lower.

My guess is that 1/8" x 3/4" should be OK if the lengths are short, maybe going to the 1/4" x 3/4" if you have over a few feet of the stuff.

Sorry I can't help with a supplier, I'm across the other side of the planet from you and copper's heavy and costly stuff to ship long distances. The copper bar I've purchased for buss bars came from a supplier of lightning conductor fittings, if that's any help.

Jeremy

 

[SlyCayer]

I can't thank you enough for all your great help Jeremy.

My buss bar will be 1/4" X 3/4" X 31" which if my calculations are wrong have a sag of 0.55V and have a waste of 27.45 watts, which seems good to me.

Has I am thinking about it, do I have to take into consideration that I actually am going through a total of 1488" of flat bar from my + to my - as my strings are paralleled and then connected in series?

If this is correct, my voltage drop will be of 2.635 V which seems high, don't you think? And waste would be 1318 watts which is also very high.

Let me know if my calculations are correct.

 

[dnmun]

you can buy copper strap from some roofing supply distributor. try googling copper roofing or gutter supplies and see what you find. or go to a scrap metal store. ask on CL and someone will find it somewhere and steal it for you.

 

[ELI_ICE_MAN]

All your series interconnections are additive, if you have parallel segments within the series you will have to calculate them as the sum of a parallel resitive circuit, then use that value in the overal series string. there are many small metal distributors on the web that can sell new copper that was left over from a larger job. E-Bay has lots of these dealers. You should be able to find exactly what you need in an auction of these left over shop scraps from some big project. I was surprised to find such a thriving market on the internet and it makes me think twice before throwing anything out

 

[SlyCayer]

If my heat loss is 1318 watts and my capacity is 3312 watts hours, it means if my load is 500 amps, I am losing half my capacity in heat? That doesn't make sense to me.

I need your help on this one Jeremy.

 

[Jeremy Harris]

If my heat loss is 1318 watts and my capacity is 3312 watts hours, it means if my load is 500 amps, I am losing half my capacity in heat? That doesn't make sense to me.

I need your help on this one Jeremy.

Sorry for the delay, we're in wildly different time zones and I missed your post yesterday.

Provided you make short connections between each paralleled bus, preferable several, then the buss bars resistance should remain pretty low. Here's a rough sketch showing what I mean, which may help:

The top buss bars are black, the bottom ones are grey and I've only shown four 20 cell sub packs. The current is shared fairly evenly between the bars, with the highest current being through the short interconnecting links. I've shown three per sub-pack, but you could increase the number if you find you still get too much voltage drop.

The key is to keep all the really high current interconnections as short as possible. Laying the connections out this way will help keep the effective length right down for the battery boxes that you have.

Hope this helps a bit.

Jeremy

 

[SlyCayer]

Ohhh, I didn't think about that. I was going to have only 1 bar connecting my series connections,

The connection between the strings should be about from 2" to 2-1/2" which will separate my amps in my buss bar in 3 if I go with your set up.

How would you suggest I do my connection between the buss bar and the bars going from + to the other - ? Should I go with a bolt or self tapping screw? Soldering? Welding?

 

[Jeremy Harris]

Bolts should be fine, and easier to assemble than trying to solder the bars up. I'd avoid self-tappers, as copper is a bit of a bugger to screw them into cleanly.

Nice short connections between the sub-packs will make for a nice, low resistance, system.

Jeremy

 

[dnmun]

i still don't see why you would not make your current carrying connections in series and then just connect them in parallel with smaller gauge sense wire.

i'm not a professional electrical engineer, and i skipped circuits because it was too hard and i didn't have time, but it would give you the lowest resistance if they were connected in series to carry current and in parallel to balance them. maybe someone can use kirchoff's rules and create the circuit equation, but it doesn't make sense to collect the current in parallel first and then conduct it in series.

 

[Jeremy Harris]

i still don't see why you would not make your current carrying connections in series and then just connect them in parallel with smaller gauge sense wire.

i'm not a professional electrical engineer, and i skipped circuits because it was too hard and i didn't have time, but it would give you the lowest resistance if they were connected in series to carry current and in parallel to balance them. maybe someone can use kirchoff's rules and create the circuit equation, but it doesn't make sense to collect the current in parallel first and then conduct it in series.

It makes perfect sense to share current as evenly as possible between the cells in the parallel groups, via low impedance interconnects, which is why everyone who builds very high current packs does it this way. Remember there are potentially high circulating currents in the cell interconnects within each paralleled group, due both to the inherently unequal current distribution in the buss bars and the variations Ri of the cells.

Jeremy