Need expertise on wiring Turnigy packs?

[liveforphysics]

It's all trivial my friend.

Don't even need to spend a thought on it.

 

[EVDragRacer]

Now that I removed the cobwebs from my brain, it seems the term series and parallel is only described for wiring one controller to two motors. I guess each controller will wire separately to each motor to produce 4000amps and the max voltage allowed by the motor. In that case how is the throttle arrangemnet? I would hate for one motor/controller to "fight" the other.

It's all trivial my friend.

Don't even need to spend a thought on it.

 

[EVDragRacer]

"The Zilla works excellently with single motor systems, but realizing that many of the quickest EVs use dual motors, the Hairball interface also has a option for safe automatic Series/Parallel shifting of dual motors for even more power and efficiency. "

http://cafeelectric.com/zilla.php?zenAdminID=74843299a81299efa3d61314338bea89&zenAdminID=74843299a81299efa3d61314338bea89&zenAdminID=74843299a81299efa3d61314338bea89&zenAdminID=74843299a81299efa3d61314338bea89

 

[MitchJi]

Hi Luke,

Also, if you are willing, I need diagrams drawn up to explain more in detail of what you are recommending. I am willing to pay for support, please contact me at:
ecedraev@gmail.com

My friend, the last thing I would do is try to get a fellow racers money. Racing does a plenty good job taking our money on it's own.
I will absolutely work on pack development for you, and I happen to have a CNC machine and sheets of Poly-Carbonate, and sticks of copper bus-bars.

The only thing I would ask in exchange for helping your project is that you freely and openly share the things learned in your journey, failures and successes, improved procedures and methods, problems experienced with designs etc...

Openly share your project details and work, and you've got my full support.

Very very generous 8)

 

[x88x]

"The Zilla works excellently with single motor systems, but realizing that many of the quickest EVs use dual motors, the Hairball interface also has a option for safe automatic Series/Parallel shifting of dual motors for even more power and efficiency. "

http://cafeelectric.com/zilla.php?zenAdminID=74843299a81299efa3d61314338bea89&zenAdminID=74843299a81299efa3d61314338bea89&zenAdminID=74843299a81299efa3d61314338bea89&zenAdminID=74843299a81299efa3d61314338bea89

/\ This. I know White Zombie uses a single Zilla 2K with hairball to drive two 9" motors in a similar configuration to yours, so if the controllers can just be run in parallel to bump it up to 4kA, then it should be fine. ..then later add another one to bump it to 6kA...repeat as needed.

 

[EVDragRacer]

White Zombie has been running in series only since his new Dow kokam pack. In my opinion he would do much better with two Zilla's, but his car would have no grip. Our Camaro has plenty of grip, so we are willing to try 4000amps at the wheels. I am not sure any suspension could handle 6000amps. I would love to try it though!

"The Zilla works excellently with single motor systems, but realizing that many of the quickest EVs use dual motors, the Hairball interface also has a option for safe automatic Series/Parallel shifting of dual motors for even more power and efficiency. "

http://cafeelectric.com/zilla.php?zenAdminID=74843299a81299efa3d61314338bea89&zenAdminID=74843299a81299efa3d61314338bea89&zenAdminID=74843299a81299efa3d61314338bea89&zenAdminID=74843299a81299efa3d61314338bea89

/\ This. I know White Zombie uses a single Zilla 2K with hairball to drive two 9" motors in a similar configuration to yours, so if the controllers can just be run in parallel to bump it up to 4kA, then it should be fine. ..then later add another one to bump it to 6kA...repeat as needed.

 

[x88x]

Yup. IMO their next upgrade needs to be the rear wheels/tires...considering that even at 2kA they're struggling to put the power down.

EDIT:
Hmm, 6kA at 370V comes out to almost 3,000 HP..in a Camaro..yees, this must be done.

 

[EVDragRacer]

Okay this is the latest. I have received some of my 10S packs and decided to dissect one pack. The 10S pack is one 5S built over another 5S packs. The tabs are all welded together and the soldering looks good. The problem is the large positive and negative wires are very small 12 AWG, not large enough for the pack size. We must increase wires size for safety reasons. The problem is I do not want to solder near the tab fearing I might heat the pouch cell. I was thinking of cutting the 12 AWG 1" away and then create a solder connection to a large size wire. Not sure if the 1" of 12 AWG will still create a problem? If I can replace the positive and negative wires with larger wires, then my only worry is the balance wires, will they also need to be upgraded? That would be a pain in the azz. I really want to use these Turnigy packs, so I must find a solution. Any suggestions guys?

I purchased the second Zilla 2k today.

 

[x88x]

Not sure if the 1" of 12 AWG will still create a problem?

Yes.

Not nearly as much as the full distance would, but it will still restrict current flow. That's so odd that they made it with 12AWG of all things...you would think at the very least they would have put 10AWG on it.

I've never done it, so I won't comment, but I know I've seen other people around here do soldering directly on pouch tabs. Maybe one of them will drop by.

 

[dogman dan]

What's the c rate of that pack? Figure out if 12 guage is big enough for that amperage, a burst of that c rate. More likely, if you crunch the numbers you will find that a few inches of 12 guage leading to your buss bar is fine. I bet in use, you never even reach that max c rate of the individual pack because you are going to be paralelling a bunch of em. Build the pack in 10s sections, and each individual pack in the section won't be carrying so much current. Then really big wire can be used to connect each bussbar, where the big load really is. If each section is all paralleled packs, only the bussbar sees the whole current. So the individual pack can easily carry the load through it's stock piece of wire for a few inches.

It's hard to draw a picture in words, but the idea is all paralell connections on the pack wiring. The big load is going to then go through your series connections where you connect the bussbars to each other in series to get the big volts. So the large loads are all going through gigantic wires that connect the bussbars.

I wouldn't worry about the balance wire guage. With proper pack management, those will only see milliamps. They won't be doing anything during a run, but they will slowly even out the cells while charging.

 

[x88x]

What's the c rate of that pack?

They're the 5Ah 45/90C Turnigy nano-tech's. So, at a max current level of 4kA they should be putting out 80C, or 400A each. IDK, personally, even for just the short amount of time that this thing is going to be running I wouldn't want to trust 400A to a single 12AWG wire.

 

[liveforphysics]

What's the c rate of that pack?

They're the 5Ah 45/90C Turnigy nano-tech's. So, at a max current level of 4kA they should be putting out 80C, or 400A each. IDK, personally, even for just the short amount of time that this thing is going to be running I wouldn't want to trust 400A to a single 12AWG wire.

It's strange yours are fitted with 12awg.
All my nanotech packs are factory fitted with 8awg.

I've got a battery design meeting I need to prepare for the rest of the day, but this evening I will try to spec some design elements for you.

 

[dogman dan]

Yeah, I thought that was strange too. I expected 10 or 8 guage on high c rate packs.

You are doing the math so I ask an ignorant question. Given the total size of the pack you are planning, are you really going to be pulling 400 amps per pack? If so, start planning a larger pack. Or start planning on only a few runs out of the pack. Something has gotta give at those amp rates per 5 ah pack. Cells will puff, etc. Max c rate discharges won't go 500 cycles. That pack has to be big enough so that the amps coming out of the wires they come with is tolerable at least. So in the end, I don't see you having to replace the pack wiring so much as add paralell batteries till you get a more reasonable c rate from each.

I do understand though, the desire to make the thing as light as possible, and not expensive for no reason. it's a race not a ride to work. But shaving stuff too thin in racing is just going to break before the finish line.

 

[EVDragRacer]

Because of the lexan battery boxes, and the use of a pouch cell leaves plenty of room for additional weight.

This question may seem simple to you guys, but I need your professional opinions. My original plan was to use enough packs for 25AH, light but powerful. But, now I am worried about the wire size, what can I do to relieve stress on the wires? I still want to push 3000 to 5000amps from this pack. If I added more in parallel, say jumping from 25AH to 50AH or even 75AH, will this make it possible to use the 10AWG wiring?

 

[EVDragRacer]

Sorry guys, I typo the wires are 10AWG on all packs.

Okay this is the latest. I have received some of my 10S packs and decided to dissect one pack. The 10S pack is one 5S built over another 5S packs. The tabs are all welded together and the soldering looks good. The problem is the large positive and negative wires are very small 12 AWG, not large enough for the pack size. We must increase wires size for safety reasons. The problem is I do not want to solder near the tab fearing I might heat the pouch cell. I was thinking of cutting the 12 AWG 1" away and then create a solder connection to a large size wire. Not sure if the 1" of 12 AWG will still create a problem? If I can replace the positive and negative wires with larger wires, then my only worry is the balance wires, will they also need to be upgraded? That would be a pain in the azz. I really want to use these Turnigy packs, so I must find a solution. Any suggestions guys?

I purchased the second Zilla 2k today.

 

[x88x]

Ah, ok, 10 isn't quite as bad, but I still wouldn't be comfortable if it were my stuff. If it were me, I would replace the single 10AWG with two 6AWG cables..but I like overkill, so.. *shrugs*

 

[EVDragRacer]

Guys, have been checking my new packs daily with my iCharger, all okay will keep checking. I have posted some questions on battery wire size related to relieving stress with more in parallel, anyone have any comments?

 

[Jeremy Harris]

It might be an idea to get the wire gauge thing into perspective here.

10g wire has a resistance of about 0.00075 ohms per ft, or 0.000063 ohms per inch.

If you leave 1" tabs of 10g wire on the cells, then pull 100A though the cells in each pack, these short lengths of wire will lose around 0.63 watts each - is that really a problem?

They won't even get slightly warm at that sort of current, and the volt drop will be just 6.3mV, barely measurable. For comparison, if you run 0000 gauge cables (around 0.00005 ohms per foot) then you'd lose around the same power in about 15 inches of it as you would in one inch of 10g.

Even if you ran at 200A per pack you'd only see a power loss in an inch of 10g of around 2.5 watts, still not enough to get near to heating it up appreciably, let alone fuse it.

Length is just as important as gauge when it comes to wiring high current stuff up, and very short lengths of smaller gauge wire aren't really a problem as long as you keep away from fusing level current densities. To fuse 10g wire takes around 350 amps or so if it's in a reasonable length, far more if it's a very short length attached to a fatter cable that will conduct heat away from it.

Jeremy

 

[EVDragRacer]

Okay, that helps thanks. But, I am still a little confused on how amps and volts will draw from small wires with a small AH pack?

Questions:

* If the pack/wires are used for only 15 seconds (or less) is it possible to keep the pack to a lower ah rate, like 25ah?
or
* Will it reduce the heat and "stress" on the pack wires if I add more batteries in parallel to up the ah, lets say 50ah or 75ah?

It might be an idea to get the wire gauge thing into perspective here.

10g wire has a resistance of about 0.00075 ohms per ft, or 0.000063 ohms per inch.

If you leave 1" tabs of 10g wire on the cells, then pull 100A though the cells in each pack, these short lengths of wire will lose around 0.63 watts each - is that really a problem?

They won't even get slightly warm at that sort of current, and the volt drop will be just 6.3mV, barely measurable. For comparison, if you run 0000 gauge cables (around 0.00005 ohms per foot) then you'd lose around the same power in about 15 inches of it as you would in one inch of 10g.

Even if you ran at 200A per pack you'd only see a power loss in an inch of 10g of around 2.5 watts, still not enough to get near to heating it up appreciably, let alone fuse it.

Length is just as important as gauge when it comes to wiring high current stuff up, and very short lengths of smaller gauge wire aren't really a problem as long as you keep away from fusing level current densities. To fuse 10g wire takes around 350 amps or so if it's in a reasonable length, far more if it's a very short length attached to a fatter cable that will conduct heat away from it.

Jeremy

 

[Jeremy Harris]

Okay, that helps thanks. But, I am still a little confused on how amps and volts will draw from small wires with a small AH pack?

Questions:

* If the pack/wires are used for only 15 seconds (or less) is it possible to keep the pack to a lower ah rate, like 25ah?

The battery/cell rating itself is fairly independent of time, so a cell rated at, say, 90C, will deliver 90 times it's rated capacity pretty much for as long as it can (which will only be around 40 seconds) The ultimate wire rating is time dependent to some extent - that fusing current I quoted of 350 amps for a length of 10g wire is the one minute fusing time for a few feet of the stuff in free air. That gauge of wire will take a lot more than 350A for shorter time periods, maybe two or three times as much for 15 seconds.

* Will it reduce the heat and "stress" on the pack wires if I add more batteries in parallel to up the ah, lets say 50ah or 75ah?

Yes, that's a good idea. The greater the capacity of the pack, the lower its internal resistance and the less stress that it will feel when asked to deliver a very high peak current for a short time. For example, Killacycle has a 9kWh battery pack but only uses around 0.6kWh per 8 second run down the strip. The extra capacity is there to give them the high peak current they need.

Jeremy

 

[liveforphysics]

When you parallel the cells (with a proper balanced bus design), they behave just as if you had a 50Ah cell rather than 10 x 5Ah cells.


As far as conductors go, remember, even just air (no wire at all) can handle millions of amps of current once it has a conductive path (which could be as simple as a smudge of soot or a line from a pencil etc).

You're looking at 3.2mOhm/ft for your 10awg wire resistance.

My bus design uses 2.5" of wire per lead (could be shorter, but 2.5" lets it be easy to service if needed).

This means .67mOhm per lead. If you're drawing 400amps per cell through a 0.67mOhm resistance wire, you're getting 0.268v of voltage drop across that wire, and it's heating at a rate of 107watts.

If you operate at 107watts for 10 seconds (hopefully it will be less ), then you've added 1.07KJ of thermal energy to that wire. Let's pretend that wire has no conductive cooling from the bus it's plugged into, and pretend the silicone has a specific heat value of zero, and is a perfect insulator, so we can work the worst case scenario temperature rise.

So, you're 2.5" piece of 10awg wire has a volume of 3.34cm^3 of copper. The specific heat value for copper is 3.45J/cm^3*K.

So, every 11.52J of heat into this 2.5" piece of wire raises it's temp 1degC. So, if you draw 400amp from each pack for 10 seconds, and add that 1.07KJ of energy into the wire, it goes up by 87degC. So, in the worst working temp rating for silicone wire I could find is 150deg C.
So, as long as the wire temperature is below 63degC (145degF) to start with before the run, you won't even be up to the working temperature rating of the wire after 10 seconds of 400amps per pack.

So, the motivation if you wish to change the wire gauge comes down to higher performance through a reduction of voltage sag.
Let's see how much sag and power loss from the pack leads you're going to get as 400amps/pack.
So, 0.67mOhm on each side of the pack makes 1.34mOhm for each 5S cell group, times 20 cell groups. 26.8mOhm x 400amps = 10.72v drop across the 370v pack due to pack lead wires, or roughly 2.9% of the pack's total voltage.


So, lets see what the difference becomes if we were to change to 8awg or 6awg wire.
If we changed to 8awg (from the cell connection forward, no 10awg wire at all) then it would be 8.49v drop across the 370v pack, or roughly 2.3% of the pack's total voltage.

If we changed to 6awg (from the cell connection forward, no 10awg wire at all) then it would be 4.22v drop across the 370v pack, or roughly 1.1% of the pack's total voltage.


So, no we know 10awg wire can handle 400amps easily for 10 seconds.

We know what the loss will be from the 10awg wire (2.9% pack power).

We know that if the wire was all changed to 8awg or 6awg, the potential power to gain would be +0.6% (8awg) and +1.8% (6awg).

Personally, I would write off the 10awg leads as acceptable pack conduction loss, and focus efforts on areas to yield higher potential gains.

 

[liveforphysics]

So, step 2 is to buy lots like this:

http://cgi.ebay.com/Copper-Flat-Bar-375-x-1-x-36-Class-II-/140471204841?pt=LH_DefaultDomain_0&hash=item20b4bc7be9#ht_1355wt_1139

You're going to want enough to make 11 x 12" pieces (may other sources etc, this was just a quick search on ebay).

Then you're going to drill them like this:

3mm boarder, 8mm holes 29mm centers starting 7mm down from the bottom for the top row and 7mm up for the bottom row.

Then you're going to buy 18 packs (so you have a few extras) of these:
http://www.hobbycity.com/hobbyking/store/uh_viewItem.asp?idProduct=6580

Next, on each pack, stay consistent, and call one set of leads "1" and one side "2". Label them with a colored tag or shrink on the wires or whatever. Label the associated balance tap set in the identical color. Cut the positive wire of pack lead set "1" to be 1.25" long. Cut the negative wire of pack lead set "2" to be 1.25" long. Strip 1/4" of the wire insulation back on each side, slip a 1/2" piece of oversized (so it doesn't touch the wire and shrink to it during soldering) heatshink around the wire and slide it up as far as it can go and tape it to keep it there.

Get a solder pot. Seriously. You're going to want a solder pot, it's dirt cheap these days anyways:
http://cgi.ebay.com/New-ML-21C-Mini-Lead-Free-Solder-Pot-/300511352229?pt=LH_DefaultDomain_0&hash=item45f7df55a5#ht_3635wt_1139

Use a higher melting point solder than good old 60/40. As much of a PITA as it's going to be to get good connections, you're going to want to go with a high temp solder for reliability under your extreme loading/thermal situation. I would go with 95Pb 5Sn for this. It's not eutectic, so you're going to need a little board to hold the wires against while they cool so you don't move them, or you're going to risk bad joints. But, it will keep your connection together up to ~312degC.

So, dip each little wire stub up to about 1-2mm away from where the insulation starts, then take them out, push them side by side while sitting on a little chunk of something high temp tolerant that won't stick (a scrap of wood would be fine even). Then inspect the joint to ensure good wetting and no faults, slide the shrink down over it, and now your 2x 5s pack is a 1 x 10s pack.

Now cut the remaining negative lead from set #1 and the positive lead from set #2 to be 2.5" long, slide a 1/2" piece of shrink on them, have an 8mm male connector end pinched in an alligator clip (or pin-vise or whatever you like to use), and have a heat gun running on it to keep it above the solder's wetting point. To confirm, temp (and ensure you don't have a cold solder joint, poke the connector end with a dab of solder and confirm it flows nicely in the little cup for the wire. Dip the stripped end of the wire into the pot, watch it become wetted, then pull it out and place it into the male cup, and turn off (or aim-away) the heat gun. Keep it extremely still (because it's not a eutectic solder) until it's cooled.

For balance tap paralleling, it's time to make a BUNCH of these until you never want to even see a JST connector again. Fortunately, they are cheap, so it's mainly just an investment in time (and it's very easy work).

You need them to build them that tie together 10 x 5S JST extension plugs into a common female 5S JST end (like each pack has on it now). You will need to make 20 of them, so you are REALLY going to be sick of making them when you're done. And they had better all be perfect too! So checking each one after you finish would be a really good idea.




Then, you CNC a sheet of lexan to have places for each of the 11 x 12" bars to mount.

Plug-in each bar, and leave a couple studs standing up threaded into the copper, so when it's time to lay that sheet of lexan down on top, you can just reach across the top of the edge cells and line-up the studs with the holes in the sheet on each end. Going to want a buddy or two for that part, so they can catch a thread on top of the studs with a nut as you line them up. Then draw all the nuts down to lock it in place, and put a rubber cap (or something) on each nut to insulate. Then fish each of the 20 x 5s balance tap leads up through the slot machined on the edge of the pack, so you've got a place to monitor and/or balance the pack as needed (if you ever need to balance, I kinda doubt it, but it's CRITICAL to monitor).

Then you bolt as many giant leads onto the edge copper bars as you plan to use. I would go with at least 4 x 0000 cables from each end of the pack going up to the controller.

I will leave you to figure out how to do the box that the cells use for structural support on the bottom side, as it's pretty chassis and your skills dependent. Maybe a screwed together lexan box, maybe a custom fiberglass enclosure or whatever. It doesn't really matter much, but what is important is to make sure across each end of the pack on the edges that will be the flat sides of the cells you have a big heavy 1/2" aluminum plate with some big 1/2" holes on the edges, so you can fit long pieces of all-thread on each corner with nuts on the edges and crank down the pressure on the cell stacks in the pack. Put a sheet of 1/8" rubber on the part of the sheet that will contact the cells to avoid chaffing/shorting. Then CRUNCH down on that pack, your cells will thank you, and it's almost impossible to put too much force on them (but use some judgement lol).

That's a lightweight fairly simple to DIY method to put together an electric car pack that should be good for >2000hp in a dragster.

And if you don't have a CNC hookup, I can help with that for making the lexan top sheet.

 

[EVDragRacer]

Hello Everyone,

I sincerely appreciate everyone help and advice in building my battery pack. I thank you for designing the battery box, but I need to connect each pack by a connector and not use a buss bar. This is required for reduction of weight and quick change pack design. I am still fuzzy on how each 10s 5ah block connects, here are my questions?

* Do I connect each 10s 5ah block in series until I reach my voltage needed, then connect the "last" positive and negative in parallel from each string until I reach my AH needed?
* In regards to the balance plugs, each 10S pack has two balance plugs, how do I get these all connected properly for a BMS and charging?

Thanks again!

Ron

 

[liveforphysics]

First, this isn't a commuter, its a racecar. BMS has no place anywhere near your pack. BMS = massive issues with nothing to gain.

Second, you can charge from totally empty to full in 6mins (with the right generator). You could top-off after a run in 2-3 mins at the worst case.

Who needs to swap packs when you can charge that fast?

If you want to be able to change the size of the pack to make it smaller (which seems like a very poor choice), you would just split this pack into 2 batteries each 5p, so split this down the middle. This way you only need to unbolt 2 big leads, unfasten the clamp holding the removable pack and lift it out.


The balance taps get paralleled by the jst connector plugs I showed you in the pics above. If you're doing 2 split packs, you just make a pair of 5 into 1s rather than a single set of 10 into 1s.

 

[liveforphysics]

If sizing the gen for charging has you puzzled, its very easy think about as horsepower seconds.

If you want to take 1000hp out of your pack for 10s, then a 100hp gen running for 100secs, or a 50hp gen running for 200sec, or a 25hp gen running for 400sec would all charge you back to full.