Enginer PHEV Technical Information

Re: Manuals

Yikes! I'll hit 115 in the shade. 120-130 in the trunk well in the parking lot. Even on a "cool" day.

11011011

 

Re: Manuals

The off the shelf equipment used in the Enginer it is NOT designed for automotive use, this means that the design and components used in them will become unstable or not operational above 115F .

 

Does the output remain the same with different cords? The unit is most likey has a autolimited output. The use of a longer cord should limit the power to the charger (higher cord resitance). The cord may also be acting like a transformer driving the voltage and current out of phase causing the Charger to draw more current to supply the same output. Same affect a long power cord has on a table saw. Use the shortest and smallest gage cord you can find.

 

Re: Manuals

It would be nice to know the rating. I will check the web site and see if I can post a question to the company.

 

Was the above a typo? Did you mean 9A, 6A, 3A, 1.5A instead?

 

MJFrog,

Have you installed your kit yet?

JOANNA

 

The kit is installed, but not working correctly yet. Jack is putting in a new DC/DC converter this afternoon that should get it running.

 

Sorry to hear. Did you recheck the fuse? Sometimes it's the simple things that are at fault.

 

great info

 

Below is the temperature rating for the balancer. That 113° F
add a couple of small box fans to keep air flow across the units.



Hi Jon,

The operating temperature is -10 ~ 45[FONT=&#23435]℃[/FONT]


Best Regards,

Jason Wang



Chargery Power Co., Ltd.
www.chargery.com
Tel: +86-755-2643 6165
Fax: +86-755-2641 2865
Skype: Jasonwang3a
MSN: [email protected]
Email:[email protected]


[FONT=&#23435]发件人:[/FONT][FONT=&#23435] Jon [/FONT]
[FONT=&#23435]发送时间:[/FONT][FONT=&#23435] 2009[/FONT][FONT=&#23435]年8月27日 11:20[/FONT]
[FONT=&#23435]收件人: jasonwang[/FONT]
[FONT=&#23435]主题: Re: D88 questions[/FONT]



Jason



What is the ambient operating tempature of the balancer?



Jon

 

New converter installed and system is running. Hope it runs longer than last time. :thumb:

I'll post results later in the MPG Reports thread.

 

Is the Charger input and output fused?

Is each balancer fused?

Is there a full schematic of the circuit?

From the pictures above I see no fuse between the charger and the power connector.

The simply diagrams posted, I see several points that should be fused. General any device that supplies power should have fused outputs. Charger, Balancer, batteries , convertor

 

In the top left of the pic in post #1 is a 100A circuit breaker next to the charger.

I haven't investigated, but a glass fuse is included in the bag of parts.

Only what has been posted here (to my knowledge).

It's probably inside the charger.

Good idea.

 

Comments relocated from the big general thread, meant to post here was just in a hurry.

Ok, here are some more thoughts/questions/concerns. Please don't take this as trashing, I would love for their to be a reasonably priced PHEV option at last. On the other hand, I also don't want to see anyone ripped off, damage their cars, or get hurt. Based on what I've seen so far, I do at least believe that these guys are sincere. Sincerity only gets you so far though. Mostly, I love this stuff and just want to understand more about it and help move things forward

1. I love the idea of just supplying a constant current to keep the SOC up to a reasonable level. What interests/concerns me is that this is what almost every PHEV converter tried first, and eventually had to abandon due to DTC codes being generated under various circumstances. Was this simply because they were trying to put in too much current and 12A is, as Hobbit says a "magic number?"

Seems highly unlikely to me. 12A is a huge amount of current for the controllers to just ignore. As I understood current imbalance on the order of mAs between the various current sensors would throw a DTC and disable the hybrid system due to the possible indication of a short somewhere. That would also violate the claim that you can parallel up 2 or more converters to up the current. Is it true then that there is no current sensor in the regen path? That strikes me as odd, but I guess is possible. If thats the case though, why did all the other developers eventually experience DTCs? One hypothesis mentioned was because it was fairly static, rather than switching on and off like cal-cars. That doesn't make sense to me either. Wouldn't regen current be all over the place in value, and switching in and out? Also, I don't think the supplemental packs were switched in the early days. The were just going to put a <240V pack in parallel and leave it there whenever the system was on. They didn't go to a higher voltage until they got DTCs, forcing them to go around the current sensor, and had to adopt the SOC drift method.

It seems to me that either these guys have indeed put a ton of work into this and figured out the magic formula that eluded everyone else for the past 5 years, or they just don't have enough test data yet to have started running into the trouble cases. Unfortunately, based on the kind of growing pain issues we have seen so far and the apparent simplicity of the kit, I would have to lean toward the latter rather than former conclusion.

2. Another thing that worries me a bit is that this kit relies on Voltage as an indicator for SOC for both the internal NimH and the LiFe cells. In general everything I've ever read says this is a really bad idea. Several times in this thread its been written off as trivial to just read the cell voltage, and look up the corresponding SOC on the manufacturers data sheets. First of all, battery manufacturers datasheets are considered something of a joke by any EVer I've ever talked to. And thats for the reputable companies, a list that Thundersky would be pretty far down in most people's books. Second, the Thundersky datasheet glosses over a very important point. Voltage of Li-ion (and NimH) cells is highly dependent on temperature. Notice that the curves shown are all at room temperature, 25C.

Since Thundersky does not show temperature data on their data sheet, here is an example for a high quality Japanese 18650 cell:
http://www.rathboneenergy.com/batte...c/Pan_LiON_pdfs/Panasonic_LiIon_CGR18650A.pdf
I'd estimate that from 0C to 45C, these cells vary by about 0.3V at any given SOC. Assuming the Thundersky are similar, we can apply that back to the discharge curves. As an estimate, lets say that 0-45 they will vary +/- 0.15V from the 25C data. Based on the expected rate of discharge, I would say the 1C curve is a reasonable one to look at. So 100% DOD would represent about 38Ah of capacity. From there we can find the voltage at ~50% DOD/SOC, somewhere in the neighborhood of 3.15V. Over 0-45C then our actual voltage readings for 50% SOC would be 3V-3.3V. Assuming we didn't know about temperature variation and just assumed as others have claimed that you can just map Voltage to SOC, this range of 3V-3.3V would equate to an SOC range of ~ 11% to 92%. Thats a pretty wide range of reported SOC for batteries that are all really at 50%.

I would think that an uncertainty of 50% +/- 40% gives a pretty clear indication of why people don't use voltage as an indication of SOC for Li-ion. NimH is very similar. If it was simply a matter of temperature, you might just get away with a lookup table of SOC vs. V vs. T, but unfortunately this is only one of the uncertainties involved.

Now as I understand this kit starts beeping when any one cell gets to 2.8V. That does give you some margin before you hit the absolute min of 2.5V. The SOC uncertainty will also be less on the steeper part of the curves. The steepness unfortunately also works against you, as once you are on that part of the curve and still pulling current at a fixed rate the amount of time that elapses between 2.8-2.5V is very short.

This brings up another point regarding battery cycle life vs. DOD. Notice that the cycle life is quoted as 3000 cycles at 70% DOD. Sounds great, thats at least 8 years charging every day! Notice though how quickly this drops vs. DOD, as they quote only 2000 cycles at 80% DOD. That extra 10% of SOC/range costs you 33% of the packs lifespan. Still 2000 cycles isn't bad. Notice also that the voltage corresponding to 80% DOD at room temperature is somewhere around 3.075V. Well up onto the flat part of the curve. By the time you get to 2.8V, you are well beyond 80%. Since there is no cycle life data published for > 80% DOD we don't know exactly what that means but generally the non-linearity of cycle life vs. DOD continues to increase the penalty for deep discharge the deeper you go. Many batteries that are good for 1000s of cycles at 50-70% DOD are only good for a few hundred at 100% DOD.

The combination of all these effects I think illustrates why Toyota in their oem NimH pack only use 40% of the packs energy range, and avoid the steep part of the curves like the plague. Even with that huge restriction they still don't use voltage as a primary indicator of SOC, they use current integration to do coulomb counting, and monitor a bunch of internal voltages and temperatures to watch for any signs of trouble. The result is a very reliable pack over a wide range of operating conditions.

Got to get home for dinner, I'll ponder some more and see what other thoughts pop up based on todays readings.

Rob

 

A related concern is that this kit essentially uses voltage to estimate the SOC of the OEM battery as well. This is implied by assuming that it is always safe to apply 12A up to 240V under all circumstances. I'm not sure that this is a reasonable assumption, and reduction in the life of the OEM battery could result.

First we know that the OEM BMS generates CAN parameters CDL (current discharge limit) and CCL (charge current limit) based on a number of inputs including various temperatures, SOC etc. This instructs the HV controller how much current is safe to put into or pull out of the oem battery at any given time. If you've ever experienced difficulty getting into or staying in EV mode in cold temperatures, or noticed the ICE revving a lot higher to accelerate after sitting in a hot parking lot, you've run into cases where CDL and/or CCL have been set at or near zero Amps. The battery controller does this because pulling or pushing current under these circumstances would be detrimental to the battery's health and reduce its lifespan. My first concern here is that this kit has no knowledge of what the CCL or CDL state is, so it will push current into the battery whether the OEM BMS thinks its safe to do so or not.

Second is the issue of using a fixed "safe voltage." The nominal voltage of the NimH pack is frequently quoted as 201.6V (1.2V/cell * 6 cells per module * 28 modules). It looks like this is the nominal resting voltage at 60% DOD at room temperature based on this curve:
Imagerius-Pack Voltage vs temperature.gif - EAA-PHEV
This would be right at the bottom end of the 40-80% SOC range the BMS maintains the OEM battery within. Notice how this range (40-80% SOC = 20-60% DOD) stays very strictly on the flat part of the curve to protect the batteries. I think because SOC drift is not triggered until 242V there may be an assumption that any voltage below that is ok indefinitely, but I'm not sure that is true. Looking at the curve above, 240V seems above 100% SOC, particularly at cold temperatures. Also, it seems like the converter is not real precise, it seems like I've heard reports of outputs as high as 250 something. Does anyone know if the car is capable of bleeding off charge at rest? Normally the controller can prevent overcharging during regen while the car is moving by spinning the ICE with the MG. It can prevent overcharging from MG1 by turning off the ICE. I'm not sure that it has any means by which to prevent overcharging while sitting in park or stopped. Anyone know?

I also worry quite a bit about the amount of human intervention that seems to be required in terms of turning the system or charger off when the battery balancers beep etc. I realize that a lot of simplifications have been made in the name of cost, but it seems like there are a lot of chances to make mistakes with potentially serious consequences.

As several have noted I think this kit is a great opportunity for the hobbyist looking to experiment, but I definitely worry about anyone non-technical who thinks they're getting a turnkey, robust kit at a bargain price. Hopefully prospective owners are being properly vetted, and made aware of the potential issues up front.

If I can ever get my house sold I'd be very tempted to pick one of these up to play with.

Rob

 

Rob,
I like your points and it's certianly got me thinking. I'll play devils advocate and try to bring up some documentation to some of your concerns.

Not that I'm taking sides, but the balancers are single cell alerts. So you've raised an interesting concern for the individual cells. The Converter on the other hand will shut down when the Cell average is at 2.875. As I understand it this is the Voltage at a 12A load, so the theory goes that once the converter shuts off, the cells will bounce back up to something north of 2.88 perhaps.

As for the DOD, I don't know about others, but I would probably plan on timing your converter on-time. It may be a big set of assumtions, but if you knew the converter output (spec'd at 2880W) and you knew the charger charge limit (spec'd at 90% SOC) and you knew your pack capacity (spec'd at 2048 for 2KW kit) then you could calculate how long the kit can run to keep you within say 70% DoD. So my calcs say, assuming the above spec'd values hold in the field) running the 2KW kit for 25min (from a starting 90% SOC) should give you an ending SOC of 30% (70% DoD). Kludgy and full of assumptions I know, but give you something with just a $2.00 kitchen timer.

I've looked up the parts, and it looks like you can get an AMP meter spece'd for 300v and 50A for about $7. Seems like Jack could build a coulomb counter for under $50, but that's just me.

PS Dan, could you update the DC/DC Converter info in the OP to specify a "factory programmed cut off generally set at 46v" or something like that. Also be nice to have the thermal tolerances of the parts. I recall the Converter is programmed to cut out at 150F, and the balancers are spec'd up to 45C (113F). Don't know about the rest.

11011011