Enginer Plan B

Can I ask a stupid question:

How come the battery packs can not be connected in series in such a way that the voltage is 240V?

 

Enginer Plan B and more information as requested

I will sugest you to go to the Enginer website and get interiorized with their principle of operation and also to read the ample content of CalCars for more in depth information on any PHEV conversion.
Regards

 

Re: Enginer Plan B and more information as requested

I know how the system works overall.

I'm not familiar with how many cells are in the battery pack however. Are there multiple 48v battery packs connected in parallel? Or just one battery pack containing 16 cells?

I know the converter will stabilize the output voltage, of course that is a good reason for its inclusion.

 

Re: Enginer Plan B and more information as requested

That depends on your definition of a cell. Enginer kit has 48V nominal pack, which consists of 16 "cells" in series. The size of each "cell" can vary based on specific product used, it can be a single prismatic cell of 40AH ( 2kW kit ) or 80AH ( 4kW kit ) or smaller pouch type cells grouped in parallel to acheve desired AH capacity and enclosed in 2 metal boxes ( 8 "cells" each ). This is all driven by price, Enginer finds cheapest cells in the world to keep their cost low, hence many of unhappy customers ( setting aside all BMS issues, which is whole 'nother story ).

You could take 76 LiFePO4 cells instead of 16 and build a 240V+ nominal pack, this way you don't need DC booster. That is what PIS kit does and some others as well. The problem here is that you must use smaller cells, or you won't fit them in Prius. Plus, the pack quickly becomes expensive. The upside is very high efficiency ( no DC boost heat losses ), PIS easily runs at 10kW and peaks at 20kW at times. This requires good quality cells, like CALB, Winston or GBS.

So, bottom line, you have 2 technical paths to take and you get what you pay for, no miracles here.

P.S. For simplicity sake I am not talking about spoofing Prius's CAN messages to get even higher MPG, I was just responding to battery related questions.

 

Re: Enginer Plan B and more information as requested

Thanks for the info. It covers almost all of what I was wondering about -- specifically the use of multiple smaller capacity cells to achieve the 240 voltage without having to boost voltage, against the use of smaller number of larger capacity cells, and their respective pros and cons.

I also suppose it is much easier to balance 16 cells vs 76. Do you always need to have an even number of cells?

 

Re: Enginer Plan B and more information as requested

You don't need to have even numbers of cells when they say balance the cells they mean all have the same SOC (state of charge) if you have one charged more than the rest it will reach full before the rest have charged properly and if you have a low cell it will be flat before the others have been fully utilized, so you see an unbalanced pack has significantly less capacity and dies earlier due to overcharged and undercharged cells. The enginer system is great for the price, but it is the cheapest by far. I think it is great they are giving a chance to many who couldn't afford to have a plug in of sorts as well as getting many more people involved.

 

Re: Enginer Plan B and more information as requested

Nah, it doesn't matter, its all about cost. Its cheaper to manage 16 cells than 76, since you need less BMS channels/modules. Battery quality is also all about cost, some cells are made better than others. Even best BMS can't save crappy cells from early grave.

Cell count is purely based on desired pack voltage, doesn't have to be odd or even.

 

^^ Thanks for the info. I thought even was required because BMS likes to use buddy cells, but now I think this is just a method of getting 48v out of 32 cells.

 

Is there nothing special about pairing up cells in parallel versus stringing them all in series when it comes to protecting other cells from being damaged by a bad cell?

 

That's right, there is nothing special. Once you connect 2 or 3 or N number of cells in parallel, they become one "cell" for all intends and purposes, with capacity that is sum of all cells in parallel. BMS sees them as one cell and they cannot be managed separately anymore. For example, take 32 cells of 40AH each. You can make 32S1P pack, by connecting them all in one series string and you will end up with 96V nominal 40AH pack ( 3.8kWh energy capacity, usually rounded up to 4kW for nice round figure ). Or, you can parallel pairs of cells and end up with 16S2P config, 48V nominal 80AH pack ( same 3.8kWh of energy ). You trade voltage for current and vice versa, but pack size in kWh doesn't change. Again, you get what you pay for.

When cells die from abuse or being poor quality to begin with, they usually increase internal resistance ( IR ) up to the point of open circuit, hence contributing less and less current until zero. In 1P scenario that means the pack can't be used until bad cell is replaced. In 2P scenario this means 2nd cell in parallel to the bad one will pickup the slack, but overall pack capacity will be reduced to the "weakest link". Unfortunately, this means remaining cell will become overloaded and will die soon after, so in the long term there is very little benefit from 2P config from reliability point of view. The benefit of multi-P config is to increase pack AH size, while keeping same voltage. If DC booster is designed for 48V input, then you can achieve various pack sizes by using larger cells or pairing smaller cells, but still keeping 16 series connected "cells".

There is nothing special about Enginer's 48V pack, my guess is that DC booster design was swiped from existing 48V AC inverter, modified by removing AC stage. What would be really nice is to design DC booster with higher input voltage, like 96V or 120V, hence less conversion losses and higher efficiency. Then we could build 32S1P or 40S1P packs using 40AH cells, to keep the pack size/cost reasonably small, yet end up with more efficient system.

 

What is the strategy in buddying up cells in the 16S2P configuration. What is the criteria to select two cells to be in a pair. Do you first measure the voltages of all discharged or charged cells, then select similar voltages?

 

I've only worked peripherally with boost and buck converters. I wish I had the "expert" we had with us when building fuel cells a while back.

It seems to me that the challenge of 5kW power handling capacity with a 10x voltage increase is quite substantial. Enginer's runs 48 on the input up to 480 on the output. This must come with immense downside (mainly on conversion efficiency), I'd think. So, where the Prius only needs 242 volts, is there a rewiring scenario which makes use of an available DC converter that does not have so much dynamic range and is much more efficient? What is it's efficiency in the first place?

 

I believe the converter that Enginer employs boosts 48V to 240V in order to charge the Prius' 201.6V battery.

 

The same converter is used for many applications. I believe one of the hybrids runs at 400 plus volts.

 

No particular strategy, just pair them in parallel and they will equalize themselves. No need to do any selection, assuming all cells are the same size/make/model.

No, Prius OEM pack is at 240V, so DC booster goes from 48V to 240V. I'm not aware of any hybrids with 400V battery. Perhaps you refer to final voltage at the motor, after Toyota's booster, but I believe that is around 500V.

 

A standalone VFD is a very inexpensive yet robust component these days, commonly available up to 75 hp.

There is a DC bus (before final conversion to variable frequency AC) within each one that one could tap into. Some are programmable, and in fact if you ran it at 0 hz, the current of one of the three phase legs will simply be unchopped direct current.

The only question is if it is:

AC (step up/down)-> AC (rectified)-> DC (invert&pwm)-> AC
or:
AC (rectified)-> DC (boost)-> DC (invert&pwm)-> AC

If it is the latter, then the hardware contained within has the desired DC to DC converter. But the former makes sense, since using a transformer is inexpensive and efficient.

 

Commonly available VFDs have neither a DC/DC or a transformer. A transformer at 60 ot 50 Hz would be too bulky and heavy. The Prius inverter is somewhat unusual in having a 20 kW DC/DC incorporated into it. Note that it only carries battery current (up to about 100 A and hence about 20 kW), not the full motor power (over 50 kW). Even so, it needs a pretty heavy inductor. A full power inductor would be heavier and larger again.

 

I have a VFD right here that takes 120V single-phase AC and outputs 240V 3-Phase AC. Somewhere the voltage is stepped up or boosted.

I also have a transformer the size of a deck of cards that steps up 120V AC to 240V AC. It runs 1500 watts.