[frodus]

Thanks again for the prompt replies. Admittedly I'm in a bit over my head so I appreciate the constructive feedback. I will work on a simple drawing tonight that lays out what I'm attempting to do and how I have it planned in my head.

To comment on your last topic about the controller needing to handle 200V+ : https://www.evwest.com/catalog/product_info.php?products_id=482&osCsid=0td1go2hqvsm45hr94ue5t33c0 This is what I was planning to purchase. I may be mistaken but I was under the impression the controller could see more volts than it's rated for but not use them all. This way I could feed 200v+ to the controller but it would only output the 180v the motor can use.

I was also told by the owner of HPEVS that motors don't "see" voltage, only amperage and that's regulated by the controller...so the controller feeds amps to the motor and the motor is then being "run at a specific voltage".

Is this accurate?

That's if the controller can handle over 200VDC, which it will not. Absolute max for the 144V X1 controller they use on the Netgain is about 180VDC.

See if the place you got the batteries will take them back, unused. It sounds like you may still be far off from needing batteries and it may not be a great design voltage for 2 of those modules in series, 3 strings.

The LG modules have cell voltage taps and temp sensor taps you can use, you don't need to take them apart to attach a BMS. It seems like you may not fully understand what a BMS is, and how it works and how it is connected. Take a look at Zeva BMS manuals or Orion BMS manuals to see how they do it.

You needed to do way more research before you started purchasing items. You're not in over your head, you just haven't gotten your "design" to a point you were ready to purchase. It needs refinement.

 

[swo]

some motors you can over volt but may not last as long
also have to check your controller voltage/amperage


if no batteries in there now could just charge these to 65v and put in now use your old controller and motor just to see how it works...

 

[LDT3]

The Pacifica modules are certainly not welded shut, I have one apart on the bench

Can the LG units be easily re-wired?

Problem is that for my application 2 modules in series (120V) would be too low, and 3 in series (180V) would be too high. 37 cells in series would be ideal giving me 140.6V, additionally I'd wire 3 cells in parallel and would only have to get a single BMS which can handle 37s.

 

[x.l.r.8]

3p becomes problematic, 2P means each module is 8S
So 2P 32s fits in 4 modules.

 

[LDT3]

3p becomes problematic, 2P means each module is 8S
So 2P 32s fits in 4 modules.

I see your point - but I was thinking of re-packaging:
Could you remove 1 cell from each module and then re-wire so that each module now becomes 3p5s? Effectively 19V per module?


Later you could take the spare cells, parallel cluster them in 3's and string them along. It's makes sense to me on paper but without seeing the inside of the module I don't know if it's feasible or not. If it can be done, it would be an ideal and inexpensive solution to my design.

 

[x.l.r.8]

Not really, the tabs are at their limitations on 2P unless you make a link to connect the P groups across 3 modules.. They need to be substantial but They only keep the P groups balanced. You could join 3 modules together in 3p and series then to another 3 modules joins at cell level. You can take a number of cells easily from the modules you have plates to keep the pouches together .
As you can see from the first picture, 2P makes the module tight for joining the tabs. Unless you ultrasonically weld them your stuck with solder as spotwelders do not work. Making it a 3P will simply not work. I removed the redundant copper tabs from the faceplate and widened the openings to fit the tabs through. The redundant tabs make ideal clips to hold the pouch tabs to solder them.

Picture 2 shows linking 2 modules to make a 2P 16S module, they are side to side as I plan on mounting them to give maximum exposure to the cooling plates. But you could stack differently. As long as the pack is balanced it will pull evenly through the series connectors so the P group plink needs to be good enough to allow group balance if there is an uneven load on the S connectors. The danger being is if you have a S connection the load will increase across this link. You have the option of making it a) fusible or b) big enough to carry a fair amount of power. It would be worth measuring the internal resistance of each 37S before doing this as if one is considerably higher you need to rectify that before making the P connections. 3rd picture shows each leaf of the module after the tabs have been separated. Each hold 2 pouches either side of an aluminium heatsink. So when together 2 pouches are in contact with each other on one side and against the heatsink on the other. When you remove the pouches you are left with just the trays picture 4.
If you revert back to picture one, a 2P has 4 tab connections and uses the heat sinks as they were intended. Making them 3P would require extending the tabs.

 

[LDT3]

Thanks! - These internal pictures really help, based on what I'm seeing I may just leave the cells as they are and follow your suggestion in photo#2, setting up three modules' cells in parallel this way, hence having to get only a 32s BMS for 6 modules arranged in 3p2s.
My interest in breaking apart the cells and going to 37s cells was to use the higher voltage Hyper9 motor - it might be more trouble than it's worth. Looks like these modules can easily handle the higher amperage of the lower voltage Hyper9 anyways.

 

[LDT3]

Picture 2 shows linking 2 modules to make a 2P 16S module, they are side to side as I plan on mounting them to give maximum exposure to the cooling plates. But you could stack differently. As long as the pack is balanced it will pull evenly through the series connectors so the P group plink needs to be good enough to allow group balance if there is an uneven load on the S connectors. The danger being is if you have a S connection the load will increase across this link. You have the option of making it a) fusible or b) big enough to carry a fair amount of power.

I'm definitely going this route (#2), now that I've opened the modules I see that it will be pretty simple to do (although somewhat tedious).
I'll be soldering the balancing P wires to the small copper tips on the top of the cell tabs. What would be the right gauge wire for this? Would 16awg silicon wire suffice?
Mind you, I will still have a 1/0 cable connecting the terminals of my 3p sets of modules together.

Originally I figured that I would simply parallel the modules together, however I've been advised that parallel strings of cells can cause problems.

Thanks again.

 

[LDT3]

Picture 2 shows linking 2 modules to make a 2P 16S module, they are side to side as I plan on mounting them to give maximum exposure to the cooling plates. But you could stack differently. As long as the pack is balanced it will pull evenly through the series connectors so the P group plink needs to be good enough to allow group balance if there is an uneven load on the S connectors. The danger being is if you have a S connection the load will increase across this link. You have the option of making it a) fusible or b) big enough to carry a fair amount of power.

I'm definitely going this route (#2), now that I've opened the modules I see that it will be pretty simple to do (although somewhat tedious).
I'll be soldering the balancing P wires to the small copper tips on the top of the cell tabs. What would be the right gauge wire for this? Would 16awg silicon wire suffice?
Mind you, I will still have a 1/0 cable connecting the terminals of my 3p sets of modules together.

Originally I figured that I would simply parallel the modules together, however I've been advised that parallel strings of cells can cause problems.

Thanks again.

 

[e7c]

As you can see from the first picture, 2P makes the module tight for joining the tabs.

Do you have any photos of the 2P packs complete? How has this worked out for you?

Thank you very much for the photos you provided they provide much more insight into these packs than most out there.

I am in the similar boat as this thread's OP. I wanted to run the pacifica packs as a "parallel series", as even evwest now recommends on their site until I encountered this very strong counter recommendation in orion's BMS wiring manual. This doc talks about it in depth: https://www.orionbms.com/manuals/pdf/parallel_strings.pdf

So that leaves me trying to figure out a good way to do 96s2p to avoid those issues. I considered your "option 2". But it ends up being a lot of work to fab up bridges for these (192 of them). It seems better to do it your way and avoid bridging.

 

[e7c]

I'll be soldering the balancing P wires to the small copper tips on the top of the cell tabs. What would be the right gauge wire for this? Would 16awg silicon wire suffice?

Did you ever end up doing this? Do you have any photos? I think the parallel wires should be able to handle the max pack current but I am not sure. This is what really turned me away from this approach, as that means like 4/0 wire which is expensive and hard to work with.

 

[EVmattyP]

Big thanks to all those who have contributed to the discussion. I'm learning a lot just from listening to you guys banter back and forth.

I want to shift the initial topic slightly to see if this resolves my problem.


Each A7 module has 16cells in series, correct? We've been discussing how to balance charge the cells within each module as I understand it. What if I only wish to treat each module (I have 6) as a singular battery and want to configure my final "pack" a 1S6P (modules not cells) so I should have a 60.5V @ 270Ah or approx. 16kwh pack.

Could I somehow buy a BMS that would only monitor those 6 modules? Would that cause problems based on how I want to connect the modules? Would there be a restriction in how many amps a charger could throw at a "pack" in this configuration? Is that based on the BMS used?

On a different topic, EVWest rates each module as being able to discharge 200-400A continuous with a peak rate of 800A for 10s. If I connect all 6 of my modules in a parallel configuration how does that affect the output rating, in terms of amps? Does it stay the same or does it multiply/stack like the Ah rating does? so 6 modules in parallel would be capable of outputting 4800A for 10s?? That seems highly unlikely but I wanted to make sure in case I needed special wiring to handle an increased load.

Thanks again

 

[john61ct]

Each A7 module has 16cells in series, correct? We've been discussing how to balance charge the cells within each module as I understand it.

However you end up doing it, you definitely should be at least regularly checking at that level, and be able to restore balance as it goes out, even if manually not "live" while in use.

> What if I only wish to treat each module (I have 6) as a singular battery and want to configure my final "pack" a 1S6P (modules not cells) so I should have a 60.5V @ 270Ah or approx. 16kwh pack.

> Could I somehow buy a BMS that would only monitor those 6 modules?

Yes in theory anyway no problem.

You choose what goes through the BMS protection, and to that extent yes BMS must be able to handle the max amps +X% safety factor.


> EVWest rates each module as being able to discharge 200-400A continuous with a peak rate of 800A for 10s. If I connect all 6 of my modules in a parallel configuration how does that affect the output rating, in terms of amps?

6x 200A = 1200A

Series boosts voltage leaves amps as is.

Your motor controller settings let you set a maximum Amps cap.

 

[EVmattyP]

> EVWest rates each module as being able to discharge 200-400A continuous with a peak rate of 800A for 10s. If I connect all 6 of my modules in a parallel configuration how does that affect the output rating, in terms of amps?

6x 200A = 1200A

Series boosts voltage leaves amps as is.

Your motor controller settings let you set a maximum Amps cap.

So if each module is connected + to + and - to - that would be parallel so my voltage would stay the same and total Ah capacity goes up to 270Ah, correct?

Are you saying that I can configure the controller for (hypothetically) 800A of draw from the battery pack and that would be approx. 133A from each pack?

I'm fairly confused about motor voltage as I was under the impression that the controller just passes through the input voltage as it sees it from the battery pack. Now that I think about it more it seems like the controller is determining the output voltage based on how many amps it can pull from the pack. As the amps drop so would the voltage, am I on the right track?

 

[brian_]

I'm fairly confused about motor voltage as I was under the impression that the controller just passes through the input voltage as it sees it from the battery pack.

No. For a first approximation, assume that the power from the battery into the controller is the same as the power from the controller into the motor. Power (for DC, in watts) is just amps multiplied by volts. For an AC motor the calculation is more complex, but the principle of power in (approximately) equals power out is still true.

This is why in the low-speed range where the motor current is limited (to protect the motor or by the current capacity of the controller), as the vehicle speeds up and the motor voltage and so power rise, the battery current and so power rise (with a constant battery voltage).

The controller typically is unable to boost voltage, so the battery pack voltage is the limit for the voltage which can be provided to the motor.

 

[john61ct]

Lower speeds starting out need lots of power, torque. Also climbing hills.

Voltage is needed as speed rises, not torque any more.

Power / amps usage varies a lot minute to minute, voltage scales more gradually with speed.

Is all that correct?

 

[brian_]

Lower speeds starting out need lots of power, torque. Also climbing hills.

Voltage is needed as speed rises, not torque any more.

Power / amps usage varies a lot minute to minute, voltage scales more gradually with speed.

Is all that correct?

I would say that's close...

Torque to the wheels is just another way of expressing driving force (multiplied by the tire radius); you need force to overcome drag (more aero drag at higher speed), and lots of force to accelerate or to climb (regardless of the speed).

Voltage is mostly proportional to motor speed, due to back EMF. There is also a smaller component of voltage needed to drive current through the resistance of the windings, and that voltage is proportional to current. The need for more voltage with higher speed is not instead of current or torque.

The result is conveniently straightforward:

• motor torque is roughly proportional to motor current
• motor voltage is roughly proportional to motor speed
• electrical power is current multiplied by voltage
• mechanical power is torque multiplied by speed (motor torque by motor speed, or wheel torque by wheel speed... same thing)

In motor design, in gearing, and in battery configuration there is always a trade-off with no right answer: whether it is speed and torque, or current and voltage, you can't get around the need for power.

 

[john61ct]

That all hangs together well, very satisfying, thanks

 

[john61ct]

So if each module is connected + to + and - to - that would be parallel so my voltage would stay the same and total Ah capacity goes up

Yes, and serial multiplies voltage keeping Ah the same.

Watt hours stay the same whatever the layout.


> Are you saying that I can configure the controller for (hypothetically) 800A of draw from the battery pack and that would be approx. 133A from each pack?


Yes not sure if all controllers let you set a limit, maybe in watts power rather than amps current?

 

[brian_]

... not sure if all controllers let you set a limit, maybe in watts power rather than amps current?

I don't know about aftermarket controllers, but OEM EV controllers definitely include power limits. For instance, from about 2700 rpm to 10,000 rpm, a pre-2018 Leaf is limited to 80kW by the controller programming, not battery voltage or motor performance.