peter, have you been able to fool the insight to cope with the 10x larger pack?
To a point yes.
The bcm operates as normal but is fed fake voltages from a Potential divider so it always sees a perfectly balanced pack. The bcm counts soc/current as normal, so every 4 ah or so I activate a switch which adds another 12-15v to the PD input raising the BCM/apparent pack voltage over the positive recal threshold. It then refills soc back upto 19 bars. I repeat this ad infinitum until my lithium pack is exhausted.
Now obviously this is not ideal and the main reason I am trying to decode the bcm/mcm signals so we can fake/control the soc/etc and finally discard the now redundant BCM.
If all goes well, this Sunday I pick up a quite healthy 2000 silver 5spd Insight, paying $5400. (Test-driving the car was intriguing, interesting in a way I have never experienced before.)
The seller already gave me a spare battery pack he had. After finding a wide range of no-load subpack voltage levels, I ordered a Prodigy II multicharger. After refreshing all the subpacks in both the spare and currently active packs and finding them all hopefully healthy enough, I'd like to well + safely mount and connect the 2nd pack in parallel in the Insight. Any pics/tips anyone can point me towards regarding secure mounting points/methods for the 2nd pack?
I also plan to use a connector harness to get each subpack paired up with its twin across the two packs to maximize balancing, minimize cell overheating, restore the effective total capacity to about what it was to when the BCM allowed a 20 - 80% SOC range from a single pack, and perhaps most importantly ;> bypass the mythical Ovionics/GM/Texaco-Chevron/Cobasys 10 amp-hour limit for NiMH EV battery packs.
Per my much earlier PHEV-Light thread, I'd then like to go for at least occasional home-based charging, which the 2nd pack would make more worthwhile. It looks like 172V is about the maximum the pack should be kept at I gather, with 1.43V/cell leaving a 0.3V or so inter-cell variation allowance margin at 8.6V max per subpack. Has anyone tried a zener diode -based + current-limited arrangement where, only under home recharging, the subpacks are limited to 8.6V via a pair of 4.3V zener diodes in series (from allelectronics.com) across each subpack (or paired subpacks in this case)? I understand NiMH internal resistance doesn't vary across most of the SOC range, until dipping at full charge and then climbing steeply (which is where the zener-based voltage limiting would kick in). That should (in theory at least) allow balanced and safe but full subpack recharging.
Has anyone tried that approach and found it to work well? I realize with the pack seeing up to around 180V peak under full 50A regen all those 1W zener diodes would have be connected only under home recharging, either via a bulky >21 wire connector where the zener diodes are all on the recharger side of the connector, or a simple power connector with the zener diodes all getting dedicated normally off relay connections (4 5PSTs would be needed), powered on by the recharging power source.
Thanks for your very helpful answers you've provided to my many earlier questions.
I am really looking forward to having such efficient and interactive engineering, where energy given back by gravity and momentum can actually be stored for later use, as my everyday transport of choice.
I also plan to use a connector harness to get each subpack paired up with its twin across the two packs to maximize balancing, minimize cell overheating, restore the effective total capacity to about what it was to when the BCM allowed a 20 - 80% SOC range from a single pack,
That would not be easy.
The second pack could end up being asked for ~1/2 the assist or regen during IMA usage.
A cable for ~50 Amps for each of the 20 subpacks... would have significant design issues.
Quote:
Originally Posted by crx_rogus
Per my much earlier PHEV-Light thread, I'd then like to go for at least occasional home-based charging, which the 2nd pack would make more worthwhile. It looks like 172V is about the maximum the pack should be kept at I gather, with 1.43V/cell leaving a 0.3V or so inter-cell variation allowance margin at 8.6V max per subpack.
Voltage alone is the most basic / least accurate option when it comes to NiMH SoC determination... it increases the risk of over charging and over discharging.
Remember there is an artificial voltage elevation that happens during charging and a artificial voltage depression that happens during discharging.
It can take several hours after the charging or discharging event ends before the NiMH cell terminal Voltage settles.
Quote:
Originally Posted by crx_rogus
I understand NiMH internal resistance doesn't vary across most of the SOC range, until dipping at full charge and then climbing steeply (which is where the zener-based voltage limiting would kick in). That should (in theory at least) allow balanced and safe but full subpack recharging.
There is a difference between battery cell internal resistance and battery cell impedance.
It does vary significantly between Discharging and Charging... and it Varies depending on SoC... and on Temperature... and on Pressure.
That would not be easy.
The second pack could end up being asked for ~1/2 the assist or regen during IMA usage.
A cable for ~50 Amps for each of the 20 subpacks... would have significant design issues.
The end cables would of course have to be capable of dealing with at least 50A, but I was figuring the balancing cables would see current only from voltage differences from differing responses between the two paired cells to charging and discharging, allowing much thinner gauge cable. 5A capability in theory, 15A in practice to give a good safety margin is what I was thinking.
Quote:
Originally Posted by IamIan
Voltage alone is the most basic / least accurate option when it comes to NiMH SoC determination... it increases the risk of over charging and over discharging.
Remember there is an artificial voltage elevation that happens during charging and a artificial voltage depression that happens during discharging. ...
The graph you provids showed very clearly that voltage has squat to do with SOC (plus a bunch of zener diodes wouldn't provide temperature/pressure/this/that compensation). Perhaps it's in the overcharge region that internal resistance (or impedance... not sure how that applies to DC other than generator-sourced peaks and valleys not fully smoothened out) starts to climb per some other graphs I've seen.
So timer control with ramped-down current (pref. with battery pack temperature feedback) would be the best mimimum for home charging, with megabuck$ EV chargers perhaps being far better. I was really hoping to find a way to avoid the latter's cost.
The end cables would of course have to be capable of dealing with at least 50A, but I was figuring the balancing cables would see current only from voltage differences from differing responses between the two paired cells to charging and discharging, allowing much thinner gauge cable. 5A capability in theory, 15A in practice to give a good safety margin is what I was thinking.
As long as the second battery pack is not connected during driving.
Unless I missed something in your design...
By connecting the two in parallel like that... any time the car asks for ~50 Amps of Assist ... it will hit both for that total of ~50 Amps... which one provides what % of that ~50 Amps would be determined by the Voltage of each pack compared to each other... and the internal resistance/impedance of each pack compared to each other.... if the IMA system ever asks for its peak of ~100 Amps of Assist... you could end up with ~50+ Amps being asked to come from the second battery pack and travel over the wire you have connecting between each of the 20 subpacks.
Quote:
Originally Posted by crx_rogus
The graph you provids showed very clearly that voltage has squat to do with SOC (plus a bunch of zener diodes wouldn't provide temperature/pressure/this/that compensation). Perhaps it's in the overcharge region that internal resistance (or impedance... not sure how that applies to DC other than generator-sourced peaks and valleys not fully smoothened out) starts to climb per some other graphs I've seen.
The graphs I posted don't show Voltage at all.
They show the Ohms of the battery pack changing with SoC ... and with temperature.
And regulated current into x ohms = voltage, hence my thinking of it in terms of voltage.
Thanks for the responses.
Regards,
Roger
Seems like you are applying Ohms law... and makes sense... just don't expect Ohm's law to give you accurate predictions of battery Voltage behavior.... not all the effects in a battery are Ohmic in nature.
The voltage of a battery cell is an effect of the electrochemical potential difference from the chemicals inside.
A NiMH battery at 0% SoC... does not show 0V ... the NiMH cell will still show over 1V per cell even with 0mAh of energy stored in it... and if you put even a small load on it... the 1V will collapse to 0V very quickly... but you take the small load off... and the Voltage will rise back up toward 1V per cell.... the 1V to 0V collapse is not caused by resistance... it is caused by other non-ohmic forces.
Just as a NiMH battery pulled into voltage reversal .... the + terminal becomes negative - and negative terminal becomes positive + ... it does have a negative ( - ) voltage ... but does not have negative ( - ) resistance.
Also the Voltage seen on a battery terminals fluctuates away from the electrochemical potential of the battery the faster / higher the current flow... if you charge with 10 amps the voltage goes up... more than it should just from ohmic forces ... if you take the 10 amps of charging off... the voltage begins to fall back down as the chemical energy inside balances out and catches up.... some of the voltage change is from ohmic forces... but not all of it.
.... if the IMA system ever asks for its peak of ~100 Amps of Assist... you could end up with ~50+ Amps being asked to come from the second battery pack and travel over the wire you have connecting between each of the 20 subpacks. ...
If you think of each battery pack in R/C battery pack terms as 120S (120 cells in series, none in parallel), the heavy duty end cables and the balancing harness connector from one 120S pack would be connected to those of the other 120S pack in a way that would balance each pack's cells via connected pairs rather than cause each to instantly burst into flames (correct polarity in other words). As the two packs would sit with vehicle parked and off, those paired cells (subpacks in our case) would be balancing each other at an equalized voltage. Each pair would have an unavoidable voltage difference from the other pairs (which I was hoping the mass of zener diodes home charging arrangement would help address, but I understand now why it wouldn't be reliable enough), but the goal would be to have cumulative subpack imbalance issues cut in half to insignificance via the linked pairs.
If there's a serious voltage difference between two paired subpacks for whatever reason, under full and sustained assist, a significant current >15A could flow across, but wouldn't that require a radically degraded cell?
But it is all on the assumption that voltage equalization between pairs will significantly help reduce charge capacity degradation, based on great apparent emphasis with lipo and li-ion R/C packs on maintaining cell balancing to keep them from thermal runaway. In the R/C world I assumed NiMH packs don't tend to have balancing connector harnesses because of their relatively low-end nature vs. lithium ion or polymer packs... perhaps not worth it from a cost perspective. But 120 NiMH cells in a singe string being expected to stay balanced with such high charge and discharge currents?!?! I was figuring they could use all the help they could get, besides sharing the load with another string of 120, and assumed that would help!
Quote:
Originally Posted by IamIan
Also the Voltage seen on a battery terminals fluctuates away from the electrochemical potential of the battery the faster / higher the current flow... if you charge with 10 amps the voltage goes up... more than it should just from ohmic forces ... if you take the 10 amps of charging off... the voltage begins to fall back down as the chemical energy inside balances out and catches up.... some of the voltage change is from ohmic forces... but not all of it.
Does that voltage change (+ under charging, - under discharging) increase or decrease with cell degradation? If it increases, sort of like a lead-acid without enough water, a balancing harness arrangement would help prevent/delay an over-volt condition with the bad cell under full sustained charging and prevent/delay excessive discharging (and perhaps the voltage reversal condition you mention) under sustained full assist.
If you think of each battery pack in R/C battery pack terms as 120S (120 cells in series, none in parallel), the heavy duty end cables and the balancing harness connector from one 120S pack would be connected to those of the other 120S pack in a way that would balance each pack's cells via connected pairs rather than cause each to instantly burst into flames (correct polarity in other words). As the two packs would sit with vehicle parked and off, those paired cells (subpacks in our case) would be balancing each other at an equalized voltage. Each pair would have an unavoidable voltage difference from the other pairs (which I was hoping the mass of zener diodes home charging arrangement would help address, but I understand now why it wouldn't be reliable enough), but the goal would be to have cumulative subpack imbalance issues cut in half to insignificance via the linked pairs.
The 120S pack reads like it is connected in parallel to the second 120S pack ... with + to + ... and - to - ...
Think of it like this ... the 120 cell series pack has the IMA system ask for ~100 Amps of Assist... all cells in series may not have equal voltage but they all share equal current... that means that as long as the 120 cell series pack is being discharged at ~100 Amps ... then every single cell and subpack is also being discharged at ~100 Amps... if your secondary battery is connecting its + terminal to the + terminal of the OEM battery and the - of your secondary battery is connecting to the - terminal of the OEM battery than you are connecting in parallel... If both of those parrallel subpacks have equal Voltage , equal SoC, equal Internal Impedance / resistance ... than the Total Current of ~100 Amps would be divided equally between both of them and they would each be discharging at a ~50 Amp rate... and your connecting wire will see ~50 Amps... if your secondary battery ever has a higher SoC... or a higher Voltage ... or lower Ohms from resistance or impedance... your secondary pack could end up seeing more than ~50 Amps out of the ~100 Amps the IMA system was asking for.... or if it has a lower voltage , SoC , higher Ohms ... it might see less % of the ~100 Amps of Current.
Quote:
Originally Posted by crx_rogus
If there's a serious voltage difference between two paired subpacks for whatever reason, under full and sustained assist, a significant current >15A could flow across, but wouldn't that require a radically degraded cell?
no.
Once you connect the + to + and - to - of each battery together you have connected them in parallel and you could see ~50 Amps or more trying to go through your wiring from the secondary battery... any time the system asks for ~100 Amps of assist.
Unless you have some type of current control to limit the current flow.
The lower amount of current ... would only be the case if the secondary battery pack was not connected while the IMA system was running in the car.... every time you start the car or had it running the secondary battery would have to be disconnected.
As long as the secondary battery pack is never connected while the IMA system is in use... And as long as each 6 cell subpack is only connected to another 6 cell subpack then ... the worst case would be dV of 8.4 - 6.0 = ~2.4V dV subpack to subpack... with as little as NREL tested minimum 120 cell pack of 0.36 Ohms for all 120 cells... 0.36 / 120 = 0.003 * 6 cells in series = 0.018 per stick one OEM +1 Secondary = 2 sticks whose Ohms add = ~0.036 Ohms minimum ... for up to a worst case of ~66 Amps... without the IMA system operating... with the IMA system operating you can also be subject to a % of the ~100 Amps as discussed above.
Quote:
Originally Posted by crx_rogus
But it is all on the assumption that voltage equalization between pairs will significantly help reduce charge capacity degradation, based on great apparent emphasis with lipo and li-ion R/C packs on maintaining cell balancing to keep them from thermal runaway. In the R/C world I assumed NiMH packs don't tend to have balancing connector harnesses because of their relatively low-end nature vs. lithium ion or polymer packs... perhaps not worth it from a cost perspective. But 120 NiMH cells in a singe string being expected to stay balanced with such high charge and discharge currents?!?! I was figuring they could use all the help they could get, besides sharing the load with another string of 120, and assumed that would help!
You are correct a balancing system would be a benefit .. I have tested several packs that are out of balance.
The issue is how does that balancing system ... and connections react when the IMA system tries to operate for Assist and regen.... and what is the worst case scenario to build / prepare for.
Quote:
Originally Posted by crx_rogus
Does that voltage change (+ under charging, - under discharging) increase or decrease with cell degradation? If it increases, sort of like a lead-acid without enough water, a balancing harness arrangement would help prevent/delay an over-volt condition with the bad cell under full sustained charging and prevent/delay excessive discharging (and perhaps the voltage reversal condition you mention) under sustained full assist.
The artificial Voltage rise from non-ohmic forces has to do with the chemical reactions inside ... and that the rate of chemical reactions is slower than the rate of current the electronics can apply... as a cell degrades it's capacity drops so it voltage rise with SoC increases... while its voltage rise due to Ohmic forces reduces ... because NiMh cell fail as a short circuit with nearly 0 Ohms ... unlike PbA or Li that fail open circuit with lots of Ohms... if a NiMH cell vents it electrolyte it can also experience a similar effect to the PbA that lost its electrolyte ( water ).
An easy analogy is that of cooking a turkey ... the chemical reactions only happen so fast... if you increase the temperature / rate of energy flow ... you are more likely to burn the outside of the turkey while the inside is still frozen... chemical reactions are just slower than the rate at which energy itself moves... in the form of heat or in the form of electricity.... the electrical current will use the path of least resistance inside the cell... and those molecules along that path of least resistance will get pushed with allot more energy than they can use for their chemical reactions ... the energy will defuse out into the rest of the cell... but it just takes time for those other chemical reactions to catch up.
.. as the rest of the cell catches up the artificial non-ohmic Voltage change levels out.
... the Total Current of ~100 Amps would be divided equally between both of them and they would each be discharging at a ~50 Amp rate... and your connecting wire will see ~50 Amps... if your secondary battery ever has a higher SoC... or a higher Voltage ... or lower Ohms from resistance or impedance... your secondary pack could end up seeing more than ~50 Amps out of the ~100 Amps the IMA system was asking for.... or if it has a lower voltage , SoC , higher Ohms ... it might see less % of the ~100 Amps of Current.
I'm still not getting why every single balance cable would see such assist current vs. the 100A -rated outer main cables that see the full ~144V potential, but you cover other needs for hefty balance cables well later.
Quote:
Originally Posted by IamIan
...As long as the secondary battery pack is never connected while the IMA system is in use... And as long as each 6 cell subpack is only connected to another 6 cell subpack then ... the worst case would be dV of 8.4 - 6.0 = ~2.4V dV subpack to subpack... with as little as NREL tested minimum 120 cell pack of 0.36 Ohms for all 120 cells... 0.36 / 120 = 0.003 * 6 cells in series = 0.018 per stick one OEM +1 Secondary = 2 sticks whose Ohms add = ~0.036 Ohms minimum ... for up to a worst case of ~66 Amps... without the IMA system operating... with the IMA system operating you can also be subject to a % of the ~100 Amps as discussed above.
That drives home the loads balancing cables could easily see. I was hoping keeping cells paired together would prevent such dV rises, but equal no-load cell voltages =/= equal cell health or SOC levels.
Quote:
Originally Posted by IamIan
The artificial Voltage rise from non-ohmic forces has to do with the chemical reactions inside ... and that the rate of chemical reactions is slower than the rate of current the electronics can apply... as a cell degrades it's capacity drops so it voltage rise with SoC increases... while its voltage rise due to Ohmic forces reduces ... because NiMh cell fail as a short circuit with nearly 0 Ohms ... unlike PbA or Li that fail open circuit with lots of Ohms...
That could be enough to cancel my balancing harness idea altogether, since that happening in a cell would force a permanently excessive charge in the rest of its subpack and a permanently reduced charge in the paired subpack. It also explains how stringing 120 cells won't fail catastrophically with failing cells getting via increased ohms most of the ~144V charge voltage all by themselves; they just short themselves out of the way of the others.
Quote:
Originally Posted by IamIan
...An easy analogy is that of cooking a turkey ... the chemical reactions only happen so fast... if you increase the temperature / rate of energy flow ... you are more likely to burn the outside of the turkey while the inside is still frozen... chemical reactions are just slower than the rate at which energy itself moves... in the form of heat or in the form of electricity.... the electrical current will use the path of least resistance inside the cell... and those molecules along that path of least resistance will get pushed with allot more energy than they can use for their chemical reactions ... the energy will defuse out into the rest of the cell... but it just takes time for those other chemical reactions to catch up.
.. as the rest of the cell catches up the artificial non-ohmic Voltage change levels out.
Thanks for the very clear explanation/analogy. To avoid possibly driving you bonkers I've ordered Batteries in a Portable World to help me have a hint of a clue regarding best use and maintenance of common battery types (I use NiMH and deep-cycle PbA every day already and LiPo occasionally).
The AutoGuide.com network consists of the largest network of enthusiast-owned enthusiast-operated automotive communities.
AutoGuide.com provides the latest car reviews, auto show coverage, new car prices, and automotive news. The AutoGuide network operates more than 100 automotive forums where our users consult peers for shopping information and advice, and share opinions as a community.
