:lol: Has anyone developed a way to charge the battery bank so when you take off there fully charged? Solar panels would need the voltage jumped to be able to charge the battery. Anyhow, we have a lot of talent in this forum so I just wanted to ask. Arturo :?:
Battery charging
1 - 20 of 20 Posts
Joined
·
4,942 Posts
No, not yet. When are you planning to start. 
See:
How insane/unlikely would it be to ad pv cells?
http://www.insightcentral.net/forum/vie ... php?t=5198
HTH!
See:
How insane/unlikely would it be to ad pv cells?
http://www.insightcentral.net/forum/vie ... php?t=5198
HTH!
The topic has under various headings been discussed at length. Although it would undoubtedly have some effect on your lifetime mileage, it would in my opinion be one of the less effective mods to achieve this. It would also decrease the life of the batteries even if done correctly (see additional post below). As such it would likely void your IMA warranty. Doing it correctly would be very costly in terms of time, effort, and materials. Doing it incorrectly could seriously damage you or the Insight.
An energy related mod that would likely save more fuel would be installing a block heater to preheat the engine in winter when mileage is at its worst.
Charging the 12 volt battery with solar or AC is also a productive project as it can cost less than 20 dollars. Having a warmer / charged 12 volt battery on a cold day will decrease the load on the IMA system, indirectly supporting the charge maintenance.
An energy related mod that would likely save more fuel would be installing a block heater to preheat the engine in winter when mileage is at its worst.
Charging the 12 volt battery with solar or AC is also a productive project as it can cost less than 20 dollars. Having a warmer / charged 12 volt battery on a cold day will decrease the load on the IMA system, indirectly supporting the charge maintenance.
If IMA charging over night was done properly the charge should never exceed 80 percent as per Honda's design. This would not decrease battery life significantly. However, assuming that this mod is to achieve some useable amount of mileage increase, the battery would have to be depleted in the course of a day’s driving before charging in order to achieve even a small difference. This daily cycling of the battery would have an effect on longevity. If this reduced battery life by 20 percent, it would, for example, reduce the life of the battery from 10 years to 8 years. Not only would this represent an inconvenience but it would also effectively cost one fifth the installed replacement cost. Assuming an installed cost of 2500 dollars this cost is five hundred dollars or 62 dollars a year. At $2.70 per gallon that is 23 gallons of gas. An Insight driven 20,000 miles per year at 65 MPG uses 308 gallons. Mileage would therefore have to rise 7.5 percent to 70 MPG. It is doubtful that the small amount of energy in the battery could achieve this. Therefore even if electrical power were free and the mod was free, it still would not likely make economic sense at today’s gasoline and battery costs.
Solar charging
It doesn't make sense if it shortens the battery life. Why does honda only go 80% on the charge? Most batterys last long by just using the upper 50% of charge. Just thought someone would have tried adding a solar panel to assist while driving so you don't end up in situations where you are charging while going uphill. Ive seen Mike 's setup on his insight.I'm going to do his setup in the future . Arturo
It doesn't make sense if it shortens the battery life. Why does honda only go 80% on the charge? Most batterys last long by just using the upper 50% of charge. Just thought someone would have tried adding a solar panel to assist while driving so you don't end up in situations where you are charging while going uphill. Ive seen Mike 's setup on his insight.I'm going to do his setup in the future . ArturoAgreed. Charging the IMA battery does not make economic sense if it shortens battery life.
The IMA system charges between 20 and 80 percent. In this range the battery internal resistance is low increasing throughput efficiency and minimizing out gassing or the chance of cell reversal in series connections.
In a lead acid system resistance goes up as the electrolyte PH goes to neutral. Deep discharge is particularly hard on lead acid batteries. NmH batteries are best used in the above 20-80 % range.
The IMA battery system nominally puts out 14,400 watts max. Mikes solar panel puts out 150 watts max. Unless the energy is stored, it won’t give much assist on hills.
Presumably you are referring to Mike's impressive plug in hybrid system using lead acid batteries and switching converter modules. However, If you are considering mounting a large solar panel, first watch the Mythbusters myths revisited issue where he tries to hold onto a sheet of plywood in the back of a moving truck. :idea:
Joined
·
4,389 Posts
Nice explanation Kip,
If I may add a few more points.
If you look at the specs on the Panasonic 6.5AH "D" cell NIMH that the insight pack is based on, you will see that they are only specified for 500 charge discharge cycles. That would not get most of us through a year of use. The battery makers found that most of the life reduction with the cells comes when they are under the 20% SOC or over the 80% SOC. When restrained within that range, tens of thousands of cycles are possible. Presently only NIMH has the durability to last the 10-20 year life of the car.
The problem with the 20-80% SOC range is that the voltage stays pretty flat within that range, so SOC determination is difficult and must involve counting the amps used and amps returned, as well as adding temperature and rate of charge /discharge compensations. According to Honda the recalibrations that many people have are due to inaccuracies creeping into this formula, where the car looses track of where the batteries SOC is, so to be safe, when the software is unsure, they recharge to some point to recover. I still don’t totally buy that one, but until we have a better theory we are stuck with it. How do they know when to stop if they don’t know where the SOC is?
When fully charged, there is an easy to see rise then dip in voltage that happens at the 100% SOC point. If each recal is followed by a 100% charge, many of the people with frequent recals would have exceeded the 500 cycles.
Charge the batteries up without the car being turned on,(solar) and you take a real risk of overcharging the batteries, and shortening their life, as well as further confusing the SOC determining system.
As far as recharging on the fly with a solar panel, while it would work, the 150 -possibily 250W that one could fit on the upper surface of the tiny car would be a drop in the bucket. When My booster pack is turned on, I am feeding nearly 2800W into the battery, and I still have to be careful with assist or I will deplete the batteries. Of course I am applying aggressive assist at 100-125 MPG, to keep the MPG over 100, even climbing hills, so the electric is doing much of the work. After using my 2800W for some time, I feel that 4000-5000 watts from the booster system would be perfect. At that rate, I could nearly run off the boost, with the NIMH HV pack only acting as a buffer.
On the AGM batteries I was advised to not let the batteries get below 50% SOC and to always recharge immediately after use to 100%
All part of the care and feeding of batteries.
If I may add a few more points.
If you look at the specs on the Panasonic 6.5AH "D" cell NIMH that the insight pack is based on, you will see that they are only specified for 500 charge discharge cycles. That would not get most of us through a year of use. The battery makers found that most of the life reduction with the cells comes when they are under the 20% SOC or over the 80% SOC. When restrained within that range, tens of thousands of cycles are possible. Presently only NIMH has the durability to last the 10-20 year life of the car.
The problem with the 20-80% SOC range is that the voltage stays pretty flat within that range, so SOC determination is difficult and must involve counting the amps used and amps returned, as well as adding temperature and rate of charge /discharge compensations. According to Honda the recalibrations that many people have are due to inaccuracies creeping into this formula, where the car looses track of where the batteries SOC is, so to be safe, when the software is unsure, they recharge to some point to recover. I still don’t totally buy that one, but until we have a better theory we are stuck with it. How do they know when to stop if they don’t know where the SOC is?
When fully charged, there is an easy to see rise then dip in voltage that happens at the 100% SOC point. If each recal is followed by a 100% charge, many of the people with frequent recals would have exceeded the 500 cycles.
Charge the batteries up without the car being turned on,(solar) and you take a real risk of overcharging the batteries, and shortening their life, as well as further confusing the SOC determining system.
As far as recharging on the fly with a solar panel, while it would work, the 150 -possibily 250W that one could fit on the upper surface of the tiny car would be a drop in the bucket. When My booster pack is turned on, I am feeding nearly 2800W into the battery, and I still have to be careful with assist or I will deplete the batteries. Of course I am applying aggressive assist at 100-125 MPG, to keep the MPG over 100, even climbing hills, so the electric is doing much of the work. After using my 2800W for some time, I feel that 4000-5000 watts from the booster system would be perfect. At that rate, I could nearly run off the boost, with the NIMH HV pack only acting as a buffer.
On the AGM batteries I was advised to not let the batteries get below 50% SOC and to always recharge immediately after use to 100%
All part of the care and feeding of batteries.
Mike, a few comment of the workings of the SOC control map or algorithm.
The SOC is determined by an integration function. As such any error in the calculation will throw off the subsequent calculation. A few small errors can add up over time to a large error. As I understand it, the problem comes in if the battery dose not get to the low or high trip indication. As the SOC wander around it can not really tell where it is until it hits the change in voltage that happens at the above 80% or below 20% full levels. The normal usage range of a NiMH battery has little voltage change (you have said this before) but the ends of the range have measurable changes in voltage. So the problem occurs when the SOC has been wandering around in the middle of the charge level for some time and then it runs in to a high or low voltage indication. Then a forced charge is initiated to find the high voltage limit and the SOC algorithm is reset.
Further complicating this is the battery controllers effort to reduce the memory effect of the battery. If the battery is used only in a narrow band (which it must be to extend the life of the battery) the battery will develop a memory and the usable band of the battery will be reduced. This makes it easier for the SOC algorithm to trip on a high or low voltage indicating the limits of the usable range. When this happens the algorithm tries to reduce the usable charge range to reflect the memory of the battery. The SOC algorithm tries to vary the limits of the usable range of charge to reduce the memory effect. Also when it has determined that the memory effect has reduced usable capacity too much it will fully charge the battery to erase the memory. If this happens too many times there is a risk of a cell in the battery dieing prematurely.
As an interesting side note, I can not always say what I think so I try to find public information to back it up. Because of the intense rivalry with Toyota every idea that went in to the development of the IMA system is patented. All patents are public information and searchable on the web, for free. Go to http://www.uspto.gov. look for the patent search page, click on quick search and patent number. 6204636 is the most relevant to this discussion . This will show you the text of the document, if you load there image viewer you can see the pretty pictures. Be forewarned, they do not write these things in English, it’s all in leaglease.
The SOC is determined by an integration function. As such any error in the calculation will throw off the subsequent calculation. A few small errors can add up over time to a large error. As I understand it, the problem comes in if the battery dose not get to the low or high trip indication. As the SOC wander around it can not really tell where it is until it hits the change in voltage that happens at the above 80% or below 20% full levels. The normal usage range of a NiMH battery has little voltage change (you have said this before) but the ends of the range have measurable changes in voltage. So the problem occurs when the SOC has been wandering around in the middle of the charge level for some time and then it runs in to a high or low voltage indication. Then a forced charge is initiated to find the high voltage limit and the SOC algorithm is reset.
Further complicating this is the battery controllers effort to reduce the memory effect of the battery. If the battery is used only in a narrow band (which it must be to extend the life of the battery) the battery will develop a memory and the usable band of the battery will be reduced. This makes it easier for the SOC algorithm to trip on a high or low voltage indicating the limits of the usable range. When this happens the algorithm tries to reduce the usable charge range to reflect the memory of the battery. The SOC algorithm tries to vary the limits of the usable range of charge to reduce the memory effect. Also when it has determined that the memory effect has reduced usable capacity too much it will fully charge the battery to erase the memory. If this happens too many times there is a risk of a cell in the battery dieing prematurely.
As an interesting side note, I can not always say what I think so I try to find public information to back it up. Because of the intense rivalry with Toyota every idea that went in to the development of the IMA system is patented. All patents are public information and searchable on the web, for free. Go to http://www.uspto.gov. look for the patent search page, click on quick search and patent number. 6204636 is the most relevant to this discussion . This will show you the text of the document, if you load there image viewer you can see the pretty pictures. Be forewarned, they do not write these things in English, it’s all in leaglease.
Joined
·
4,389 Posts
Lee
Nice find.
I have been using this SOC determination explanation:
http://www.99mpg.com/Data/downloads/rel ... nation.jpg
which as I understand it came from a technicial training class put on by Honda.
It pretty much agrees with your discription,but has no mention of memory effect, which many people incorectly felt was not an issue with NIMH.
With all of the problems with recalabrations that the insights have had, I sort of doubt that the patented memory effect compensation was included, and since the recals seem to repeat for many people whenever the SOC drops to some repeatable point, the compensation for aging is not working.
It is possible that Honda did not impliment the full algorythm on the insight, but may have on the Civic and later hybrids?
My own experience with 105K on the batteries, the last 25K, with daily 100A assist,and 50A regen, several cycles a day, and never a recal. others that use the batteries sparingly get recals frequently?
My daily usage will cycle the pack fully, light use would tend to keep the batteries near full SOC, and only use a small part of the usable range of SOC, which could induce a memory.
An interesting aside reguarding this. I was given a fluke 196C scopemeter to evaluate. It runs on NIMH batteries. I was studing the manual, and found that it had a battery restore feature that they recomend using 4 times a year where it automatically completely discharges the battery, then completely recharges it to avoid the memory effect.
Nice find.
I have been using this SOC determination explanation:
http://www.99mpg.com/Data/downloads/rel ... nation.jpg
which as I understand it came from a technicial training class put on by Honda.
It pretty much agrees with your discription,but has no mention of memory effect, which many people incorectly felt was not an issue with NIMH.
With all of the problems with recalabrations that the insights have had, I sort of doubt that the patented memory effect compensation was included, and since the recals seem to repeat for many people whenever the SOC drops to some repeatable point, the compensation for aging is not working.
It is possible that Honda did not impliment the full algorythm on the insight, but may have on the Civic and later hybrids?
My own experience with 105K on the batteries, the last 25K, with daily 100A assist,and 50A regen, several cycles a day, and never a recal. others that use the batteries sparingly get recals frequently?
My daily usage will cycle the pack fully, light use would tend to keep the batteries near full SOC, and only use a small part of the usable range of SOC, which could induce a memory.
An interesting aside reguarding this. I was given a fluke 196C scopemeter to evaluate. It runs on NIMH batteries. I was studing the manual, and found that it had a battery restore feature that they recomend using 4 times a year where it automatically completely discharges the battery, then completely recharges it to avoid the memory effect.
solar
:lol: I saw your solar set up Mike. What happens in my commute is that the batterys are 3 bars when I get to work. On the way home it's bumper to bumper traffic making it hard to charge the battery. I don't get it charged until I break out of the congestion zone which is sometimes 40 miles! If I took of with full batteries I would be able to increase the milage as the engine wouldn't have to recharge it. I see Kips been talking on this subject also. It just seems it should work if you could jack up the voltage from the solar panels to the insights battery voltage.What about lithium batteries?
:lol: I saw your solar set up Mike. What happens in my commute is that the batterys are 3 bars when I get to work. On the way home it's bumper to bumper traffic making it hard to charge the battery. I don't get it charged until I break out of the congestion zone which is sometimes 40 miles! If I took of with full batteries I would be able to increase the milage as the engine wouldn't have to recharge it. I see Kips been talking on this subject also. It just seems it should work if you could jack up the voltage from the solar panels to the insights battery voltage.What about lithium batteries?
Joined
·
4,389 Posts
Arturo
It is a case of watts.
If the whole top of the car was covered with solar cells, it would still be difficult to get the pack filled back up while driving your 40 miles.
The Ewheel would allow you to drive with the gas engine off for much of that 40 miles. Eboost would allow recharging from other batteries that you could solar charge at home. Lots of possibiliteis but not without some major modification to the car. Keep watching 99mpg.com for future developments with Ewheel and Eboost, there may be some new developments soon it those areas.
It is a case of watts.
If the whole top of the car was covered with solar cells, it would still be difficult to get the pack filled back up while driving your 40 miles.
The Ewheel would allow you to drive with the gas engine off for much of that 40 miles. Eboost would allow recharging from other batteries that you could solar charge at home. Lots of possibiliteis but not without some major modification to the car. Keep watching 99mpg.com for future developments with Ewheel and Eboost, there may be some new developments soon it those areas.
Mike,
I believe that the battery management has changed over time. The oldest patent I found talked about shifting the charge range back and forth to reduce the memory effect. From the date and the terms used it is likely this was from the Honda EV.
I can not say if the patent I cited was from the insight for sure. Usually a patent must be submitted before the technology is disclosed to the public, however this patent was already submitted in Japan. I am not too knowledgeable about Japanese patents applied for in the US, my experience is the other way around. However the patent I cited is for a hybrid car.
There are later patents that focused more on the battery memory issues. I can not tell if the newer patents apply to the HCH or Accord. Also it is possible the battery controller for the Insight was updated to include some of the newer technology. I believe the battery controller in the ’06 Insight is the 3rd iteration (can’t prove that).
It is interesting to note that a math algorithm was used for the SOC calculation and not a set of look up tables or maps. In the mid to late ‘90s I remember discussions about having to select vehicle options because the ECU could not load everything. The computing power requirements at that time where too close to the limits of the machines. On new vehicle designs today they can load the computers up with every conceivable option and use the same brain in all vehicle variations, they simply do not use some of the available capability. This is cheaper than making several variations of ECU.
I believe that the battery management has changed over time. The oldest patent I found talked about shifting the charge range back and forth to reduce the memory effect. From the date and the terms used it is likely this was from the Honda EV.
I can not say if the patent I cited was from the insight for sure. Usually a patent must be submitted before the technology is disclosed to the public, however this patent was already submitted in Japan. I am not too knowledgeable about Japanese patents applied for in the US, my experience is the other way around. However the patent I cited is for a hybrid car.
There are later patents that focused more on the battery memory issues. I can not tell if the newer patents apply to the HCH or Accord. Also it is possible the battery controller for the Insight was updated to include some of the newer technology. I believe the battery controller in the ’06 Insight is the 3rd iteration (can’t prove that).
It is interesting to note that a math algorithm was used for the SOC calculation and not a set of look up tables or maps. In the mid to late ‘90s I remember discussions about having to select vehicle options because the ECU could not load everything. The computing power requirements at that time where too close to the limits of the machines. On new vehicle designs today they can load the computers up with every conceivable option and use the same brain in all vehicle variations, they simply do not use some of the available capability. This is cheaper than making several variations of ECU.
Joined
·
4,389 Posts
I guess the only way we will ever know is to do some datalogging and get some info to base our theories on. I have the DAQ card, and the connector, all I need to do is build the analog opto isolators.
Another project on my wish list.
Another project on my wish list.
In the meantime I'm testing battery longevety by driving the old, discarded pack from dfiore. He replaced it after 147k miles of normal use and I'm now using it with grid charger at one full cycle per day (full charge, none of this 80% business). I have done 3k miles (about 50 cycles) so far. Let's see how long it lasts.
Now, to be honest, it's already worn, no surprise after 147k miles! Average cell capacity was around 4.5 Ah before I put it into my car and started the heavy cycle use. I might do another measurement after a few more cycles...
Armin
see this thread
Now, to be honest, it's already worn, no surprise after 147k miles! Average cell capacity was around 4.5 Ah before I put it into my car and started the heavy cycle use. I might do another measurement after a few more cycles...
Armin
see this thread
Joined
·
4,389 Posts
Armin
Are you getting recals like Dan did, or is it behaving differently with your heavier charge?
Are you getting recals like Dan did, or is it behaving differently with your heavier charge?
A "recal" is the indication that the pack is empty. I get one almost every day, since I purposely use MIMA to empty the pack over my commute. What might be more relevant than recal frequency is SoC level when the pack goes empty. For the dfiore battery this is somewhere around 5-6 bars, after my full charge overnight. Considering that Honda would only charge to 80%, recals should occur at a higher SoC level if the car was driven with normal IMA.Mike Dabrowski 2000 said:
But there is still some power left in the pack, before it refuses any assist. I try to be gentle with MIMA assist after the SoC drops to 1 bar.
Joined
·
4,389 Posts
HI Armin
a few questions?
Do you still have your battery monitor system in place?
Have you determined if their is a voltage differential that can be measured on the subpacks that indicates whan a recal or end of charge will happen?
Can you see a big difference between subpacks when the batteries are under load?
Does one subpack voltage start to fade quickly when a cell runs out of charge, a rate of change detection?
a few questions?
Do you still have your battery monitor system in place?
Have you determined if their is a voltage differential that can be measured on the subpacks that indicates whan a recal or end of charge will happen?
Can you see a big difference between subpacks when the batteries are under load?
Does one subpack voltage start to fade quickly when a cell runs out of charge, a rate of change detection?
Hi Mike,
unfortunately, I only have one voltmeter hooked up to total pack voltage. I wish I knew of an easy and affordable way to monitor all subpacks!!!
I guess I could start by hooking up a voltmeter to one sub-pack at a time, recording it's voltage at recal, and that way finding the weakest one. But it wouldn't be simultaneous and probably not very good.
unfortunately, I only have one voltmeter hooked up to total pack voltage. I wish I knew of an easy and affordable way to monitor all subpacks!!!
I guess I could start by hooking up a voltmeter to one sub-pack at a time, recording it's voltage at recal, and that way finding the weakest one. But it wouldn't be simultaneous and probably not very good.
Joined
·
4,389 Posts
Armin
I have a voltmeter set up to monitor a bunch of voltages on my boost system.
12V battery1,2,3,4, total of all 4(48V), boost current, total 144Vpack volts, Ewheel motor temp, boost dc/dc temp, curtics motor controller temp.
All voltages go through a 2 pole break before make rotary switch. You could use the battery back probe adapter and do something similar on the subpack taps.
I had a chance to test a pack right after a recal, and could see only a few millivolts between all the packs, right down the the 7.2V subpack level. This was without load. I would hope that we could see a difference under load.
The thread about the test:
http://www.cleanmpg.com/forums/showthread.php?t=2019
I have a voltmeter set up to monitor a bunch of voltages on my boost system.
12V battery1,2,3,4, total of all 4(48V), boost current, total 144Vpack volts, Ewheel motor temp, boost dc/dc temp, curtics motor controller temp.
All voltages go through a 2 pole break before make rotary switch. You could use the battery back probe adapter and do something similar on the subpack taps.
I had a chance to test a pack right after a recal, and could see only a few millivolts between all the packs, right down the the 7.2V subpack level. This was without load. I would hope that we could see a difference under load.
The thread about the test:
http://www.cleanmpg.com/forums/showthread.php?t=2019
1 - 20 of 20 Posts
Join the discussion
Join now to ask and comment!