Joined
·
8,683 Posts
This is it, exactly. Thanks...
Super Capacitors
81 - 100 of 212 Posts
Joined
·
14,379 Posts
Nearly ready to install.
OK the packs are now fully assembled/cased on the bench.
They are reasonably heavy and I would say at least 1/3rd of the weight is the hardware and casing.
They weight about 14kg each according to my bathroom scales.
I plugged them into my HV charger to balance them and set charge current at 1amp. It took about 5 seconds (which is a miniscule amount of charge added) before all the balance lights were active showing the cells were all at the 3.60v. Amazing balance and consistency.
Switchboard is upgraded to 175A fuse..
Just BCM interceptor to resurrect and tweak the software..
https://youtu.be/oKevMAG6QFs
OK the packs are now fully assembled/cased on the bench.
They are reasonably heavy and I would say at least 1/3rd of the weight is the hardware and casing.
They weight about 14kg each according to my bathroom scales.
I plugged them into my HV charger to balance them and set charge current at 1amp. It took about 5 seconds (which is a miniscule amount of charge added) before all the balance lights were active showing the cells were all at the 3.60v. Amazing balance and consistency.
Switchboard is upgraded to 175A fuse..
Just BCM interceptor to resurrect and tweak the software..
https://youtu.be/oKevMAG6QFs
Attachments
Joined
·
14,379 Posts
Finished tidying ready for install.
Done the basic BCM interceptor software for tomorrow's test
Finished tidying up a few ends and added a relay to the BCM fooler + HV feed to stop the parasitic 2ma current drain through the 100k resistor matrix when the car is off.
It might not sound much, but it could drain the supercaps 800mah capacity in a little over two weeks if sat idle..
It's game over for £5000 worth of kit if they go below 2.2V
The BCM interceptor limits are set in this version at.
Max Pack V 180V. Above that regen is disabled.
Min Pack V 120V. Below this assist is disabled.
Max Pack Temp 45C. Above this both assist and regen are disabled.
The car might try and limit things even sooner than the BCM interceptor, so I suspect tweaks will be required.
Interceptor source code is attached for those interested.
It's best viewed using MicroCode Studio..
At the moment the battery temp is being passed through the BCM interceptor unmolested.
I might change that later to fake 25C plus which ensures the MCM allows maximum battery current.
I want to watch the actual temp in these initial tests on the OBDIIC&C gauge..
Done the basic BCM interceptor software for tomorrow's test
Finished tidying up a few ends and added a relay to the BCM fooler + HV feed to stop the parasitic 2ma current drain through the 100k resistor matrix when the car is off.
It might not sound much, but it could drain the supercaps 800mah capacity in a little over two weeks if sat idle..
It's game over for £5000 worth of kit if they go below 2.2V
The BCM interceptor limits are set in this version at.
Max Pack V 180V. Above that regen is disabled.
Min Pack V 120V. Below this assist is disabled.
Max Pack Temp 45C. Above this both assist and regen are disabled.
The car might try and limit things even sooner than the BCM interceptor, so I suspect tweaks will be required.
Interceptor source code is attached for those interested.
It's best viewed using MicroCode Studio..
At the moment the battery temp is being passed through the BCM interceptor unmolested.
I might change that later to fake 25C plus which ensures the MCM allows maximum battery current.
I want to watch the actual temp in these initial tests on the OBDIIC&C gauge..
Joined
·
1,545 Posts
Fingers crossed for a good outcome!
Joined
·
8,683 Posts
You're the expert, but... Maybe you should set the low voltage cutoff a bit higher for your initial testing in the car? 50 X 2.2V=110V, and 120V doesn't seem very far from 110V - when a single capacitor would only have to be -0.1V compared to the rest for it to drop below 2.2V... There's no protection on the bottom end, right?, like there is on the top?...It's game over for £5000 worth of kit if they go below 2.2V
Min Pack V 120V. Below this assist is disabled.
Joined
·
14,379 Posts
It's a fair point eq1 I'll raise it to 125v for a bit more headroom.
As I mentioned as well the BCM detected pack [Bvo] parameter voltage can be a bit (+/- 2V) off as well.
The car will likely be trying to charge them already down at that voltage.
The BCM fooler is connected at the switchboard, and there are some thick cables from the switchboard to the actual cell stacks.
Under load there will be some voltage drop in these cables,
(unlike the oem setup where the BCM voltage taps are connected at the cell/stick ends).
When assisting the voltage at the BCM fooler will be lower than at the supercaps.
When regening the voltage at the BCM fooler will be higher than at the supercaps.
That's a useful additional small voltage safety factor.
The higher the assist/regen current the bigger the voltage deviation as well
It may only be a few millivolts but it all helps.
The resting voltage of these cells will be a lot closer to the loaded voltage in this pack.
In an OEM pack (depending on it's IR) the voltage can swing 30+ volts under loads, so down to 120v when assisting at say 70A and then back up to 150+v when load removed.. Ditto when regening, so 150v resting and 185+ volts when regening at 50A.
The supercaps with their much lower IR might only swing +/- 5v between resting voltage and loaded voltage.
As I mentioned as well the BCM detected pack [Bvo] parameter voltage can be a bit (+/- 2V) off as well.
The car will likely be trying to charge them already down at that voltage.
The BCM fooler is connected at the switchboard, and there are some thick cables from the switchboard to the actual cell stacks.
Under load there will be some voltage drop in these cables,
(unlike the oem setup where the BCM voltage taps are connected at the cell/stick ends).
When assisting the voltage at the BCM fooler will be lower than at the supercaps.
When regening the voltage at the BCM fooler will be higher than at the supercaps.
That's a useful additional small voltage safety factor.
The higher the assist/regen current the bigger the voltage deviation as well
It may only be a few millivolts but it all helps.
The resting voltage of these cells will be a lot closer to the loaded voltage in this pack.
In an OEM pack (depending on it's IR) the voltage can swing 30+ volts under loads, so down to 120v when assisting at say 70A and then back up to 150+v when load removed.. Ditto when regening, so 150v resting and 185+ volts when regening at 50A.
The supercaps with their much lower IR might only swing +/- 5v between resting voltage and loaded voltage.
Joined
·
14,379 Posts
It's in and works.
Now for the details..
Install went smoothly and no IMA light etc when fired up.
Started with IMA no problem.
Done about 70 miles testing today and exercised the pack up and down as much as I could.
Hills driven up and down cycling it up/down.
Temperature
The packs haven't risen in temp above ambient as far as I can tell.
Checking pack temp OBD readings versus infrared thermometer all match.
Basically they might have risen 1C and that's it.. Very impressive.
Internal Resistance Seems very low, it's difficult to be super accurate because as soon as you start to charge or discharge the caps the voltage rises or falls linearly.
As far as I can tell voltage drop due to IR at 80A assist is about 5V.
Most of that drop could be in the wiring/fuse/contactor/main switch etc rather than in the caps themselves.
Performance
The car is happy to give max assist from full to empty. Current rises as voltage drops as the IMA tries to maintain 10kw output. Current hack means it's probably about 13kw through the range.
Capacity
Basically you have enough juice to accelerate with Max WOT assist from 0mph to 70mph.
I need to widen the voltage window a bit to push the voltage higher and extend the usable capacity. The top and bottom ends need tweaking.
Issues
1) Initially as the voltage fell away the dc-dc converter would be inhibited (turned off) by the MCM below about 140v, so the 12v battery stops charging etc. I have now given the BCM interceptor control over this DC-DC inhibit line. Now the BCM interceptor doesn't turn off the dc-dc unless the pack is below 120V. It re-enables it when the pack V climbs back over 125V. This should prevent the dc-dc killing the pack if you are sat there for a while.
2) Assist, IMA start, and autostop is disabled by the MCM and/or BCM below about 144V.
I need to force it to work at a lower voltage
3) Regen is disabled above about 172-175V so i need to push that higher or the cells won't reach the balancing voltage (180v)
4) The OEM Soc gauge seems to have a mind of it's own at the moment mostly staying on 19 bars (75%) as fixed by the BCM interceptor, but occasionally falling to one bar for no obvious reason even though Soc is still showing as 75%? I could take control of this completely with another BCM interceptor on the METTSCI line. I might do this later as i want the OEM soc gauge to reflect supercap capacity.
5) Need to make the BCM interceptor software more intelligent and allow it to manage more stuff. Need to look at some BCM BATTSCI data to identify some more flags.
I think that's it at the minute. I'm knackered now been busy all day.
Now for the details..
Install went smoothly and no IMA light etc when fired up.
Started with IMA no problem.
Done about 70 miles testing today and exercised the pack up and down as much as I could.
Hills driven up and down cycling it up/down.
Temperature
The packs haven't risen in temp above ambient as far as I can tell.
Checking pack temp OBD readings versus infrared thermometer all match.
Basically they might have risen 1C and that's it.. Very impressive.
Internal Resistance Seems very low, it's difficult to be super accurate because as soon as you start to charge or discharge the caps the voltage rises or falls linearly.
As far as I can tell voltage drop due to IR at 80A assist is about 5V.
Most of that drop could be in the wiring/fuse/contactor/main switch etc rather than in the caps themselves.
Performance
The car is happy to give max assist from full to empty. Current rises as voltage drops as the IMA tries to maintain 10kw output. Current hack means it's probably about 13kw through the range.
Capacity
Basically you have enough juice to accelerate with Max WOT assist from 0mph to 70mph.
I need to widen the voltage window a bit to push the voltage higher and extend the usable capacity. The top and bottom ends need tweaking.
Issues
1) Initially as the voltage fell away the dc-dc converter would be inhibited (turned off) by the MCM below about 140v, so the 12v battery stops charging etc. I have now given the BCM interceptor control over this DC-DC inhibit line. Now the BCM interceptor doesn't turn off the dc-dc unless the pack is below 120V. It re-enables it when the pack V climbs back over 125V. This should prevent the dc-dc killing the pack if you are sat there for a while.
2) Assist, IMA start, and autostop is disabled by the MCM and/or BCM below about 144V.
I need to force it to work at a lower voltage
3) Regen is disabled above about 172-175V so i need to push that higher or the cells won't reach the balancing voltage (180v)
4) The OEM Soc gauge seems to have a mind of it's own at the moment mostly staying on 19 bars (75%) as fixed by the BCM interceptor, but occasionally falling to one bar for no obvious reason even though Soc is still showing as 75%? I could take control of this completely with another BCM interceptor on the METTSCI line. I might do this later as i want the OEM soc gauge to reflect supercap capacity.
5) Need to make the BCM interceptor software more intelligent and allow it to manage more stuff. Need to look at some BCM BATTSCI data to identify some more flags.
I think that's it at the minute. I'm knackered now been busy all day.
Attachments
Joined
·
1,545 Posts
Well done - and more assist capacity than I expected.
Joined
·
14,379 Posts
System voltage faking for the supercaps
Thanks. I think the best way to stop the system throwing a wobbly is just to fake the system voltage at a fixed 150v.
150v gives full autostop/ima start/assist regen etc.
To do that we will have to tell the car the voltage is 150v in various places at various times.
1) The BCM fooler, this feeds the BCM voltage taps and the MCM HV input.
I will have to split this into the two sections.
A) MCM HV Input this will have to be fixed at 150V using a little dc-dc converter. (Ideas welcome) Perhaps I should look inside the MCM and try and find the voltage on the LV side of the HV detection and fake it there with a potential divider. Hmm time to get my microscope out!!
B) BCM voltage taps this will be fed by the supercaps so will vary and allow the interceptor to manage things. The interceptor can tell the MCM the voltage is 150V over the BATTSCI data line after it gets to know the real supercap voltage for it's internal use.
2) The Vpin 5v ADC line into the MCM. This tells the MCM what the MDM voltage is, and is used to compare battery voltage with MDM voltage etc. It is also used at switch on to detect when the bypass contactor has engaged and the filter caps in the MDM have charged enough to allow the main contactor to turn out without welding itself! So at switch on this is a rapidly rising 0-X voltage (less than a second). So it can't be faked with a fixed resistor but will have to be some sort of digital/analog circuit with a quick ramp up from zero to our fixed simulated 150V (5v) equivalent. I'm open to ideas for this?
The MCM has a 100k pullup resistor on this line so we could have some sort of digital 100k pot to 5V with 255 steps to make it controllable.
3) The voltage data in the BATTSCI data stream. (The interceptor can take care of this.)
The BCM interceptor can then manage the supercap pack voltage by enabling/disabling assist regen at the voltages required.
Remember the old BCM gauge?
I need to build myself one now to stream some data from the BATTSCI line directly into my spreadsheet...
Thanks. I think the best way to stop the system throwing a wobbly is just to fake the system voltage at a fixed 150v.
150v gives full autostop/ima start/assist regen etc.
To do that we will have to tell the car the voltage is 150v in various places at various times.
1) The BCM fooler, this feeds the BCM voltage taps and the MCM HV input.
I will have to split this into the two sections.
A) MCM HV Input this will have to be fixed at 150V using a little dc-dc converter. (Ideas welcome) Perhaps I should look inside the MCM and try and find the voltage on the LV side of the HV detection and fake it there with a potential divider. Hmm time to get my microscope out!!
B) BCM voltage taps this will be fed by the supercaps so will vary and allow the interceptor to manage things. The interceptor can tell the MCM the voltage is 150V over the BATTSCI data line after it gets to know the real supercap voltage for it's internal use.
2) The Vpin 5v ADC line into the MCM. This tells the MCM what the MDM voltage is, and is used to compare battery voltage with MDM voltage etc. It is also used at switch on to detect when the bypass contactor has engaged and the filter caps in the MDM have charged enough to allow the main contactor to turn out without welding itself! So at switch on this is a rapidly rising 0-X voltage (less than a second). So it can't be faked with a fixed resistor but will have to be some sort of digital/analog circuit with a quick ramp up from zero to our fixed simulated 150V (5v) equivalent. I'm open to ideas for this?
The MCM has a 100k pullup resistor on this line so we could have some sort of digital 100k pot to 5V with 255 steps to make it controllable.
3) The voltage data in the BATTSCI data stream. (The interceptor can take care of this.)
The BCM interceptor can then manage the supercap pack voltage by enabling/disabling assist regen at the voltages required.
Remember the old BCM gauge?
I need to build myself one now to stream some data from the BATTSCI line directly into my spreadsheet...
Joined
·
14,379 Posts
Inside the MCM.
Here we are looking at the MCM HV input socket and associated pcb circuit.
I shone a light through the board to reveal it's secrets.
The five chips at the bottom must be bespoke as they are labelled OKI 0CM243CA.
Google and the OKI site (renamed) below gives nothing for them at all.
It's a six pin package, but the middle pin on the input side is cut off (not present)
Product Name | Integrated Circuits | LAPIS Semiconductor
They are probably some sort of opto isolator with transistor output, but why five?
If there was just one isolator then faking the output side was probably do-able.
But with five it looks like it will have to be a low current 150v dc supply input.
Comments/ideas on the circuit and what it is doing welcome..
It looks like input current limit resistors and some filtering.. Then what?
I'll power one up and take some measurements in due course..
Looks like it will have to be something like this driven with an isolated dc-dc 12v converter.
http://www.ebay.co.uk/itm/NCH6100HV...095326?hash=item41c098855e:g:NOkAAOSwTM5Yx5JK
Here we are looking at the MCM HV input socket and associated pcb circuit.
I shone a light through the board to reveal it's secrets.
The five chips at the bottom must be bespoke as they are labelled OKI 0CM243CA.
Google and the OKI site (renamed) below gives nothing for them at all.
It's a six pin package, but the middle pin on the input side is cut off (not present)
Product Name | Integrated Circuits | LAPIS Semiconductor
They are probably some sort of opto isolator with transistor output, but why five?
If there was just one isolator then faking the output side was probably do-able.
But with five it looks like it will have to be a low current 150v dc supply input.
Comments/ideas on the circuit and what it is doing welcome..
It looks like input current limit resistors and some filtering.. Then what?
I'll power one up and take some measurements in due course..
Looks like it will have to be something like this driven with an isolated dc-dc 12v converter.
http://www.ebay.co.uk/itm/NCH6100HV...095326?hash=item41c098855e:g:NOkAAOSwTM5Yx5JK
Attachments
Joined
·
14,379 Posts
I had a short drive out this morning.
Caps had stayed at 175V overnight so no sign of any parasitic drain.
The overnight temp here meant the caps cooled down to 8C, and the BCM would not allow full current (as expected) so I will have to fake the temp to 25C+.
I'll program the BCM interceptor to fake 25C at any temp below 25C, if temp is >= 25C then pass thru the actual temp. That covers it.
I have a simple grid charger fitted to the car using two of the old cheapo 90v led power supplies. It had one diode in the + feed to prevent backwards current flow, but i added a second 400v 1A diode in the negative feed as I'm paranoid about draining things.
The charger unloaded voltage is 181v and this matched very well with the super cap balance boards. I charged them up to the present 175 max the car will allow before i parked it, I then topped them off in about 5 minutes with the led chargers to 180v and all balance lights lit. I need to wire the old battery pack 12v fan power supply into the new supercap fan system, so the fans run when the charger is plugged in for balancing purposes.
Looks like I won't be plugging it in very often as all the cell balance lights came on again within a few seconds. No obvious cell imbalance.
Manual supercap balancing using the charger might just be a week/monthly/six monthly task? We shall see.
Caps had stayed at 175V overnight so no sign of any parasitic drain.
The overnight temp here meant the caps cooled down to 8C, and the BCM would not allow full current (as expected) so I will have to fake the temp to 25C+.
I'll program the BCM interceptor to fake 25C at any temp below 25C, if temp is >= 25C then pass thru the actual temp. That covers it.
I have a simple grid charger fitted to the car using two of the old cheapo 90v led power supplies. It had one diode in the + feed to prevent backwards current flow, but i added a second 400v 1A diode in the negative feed as I'm paranoid about draining things.
The charger unloaded voltage is 181v and this matched very well with the super cap balance boards. I charged them up to the present 175 max the car will allow before i parked it, I then topped them off in about 5 minutes with the led chargers to 180v and all balance lights lit. I need to wire the old battery pack 12v fan power supply into the new supercap fan system, so the fans run when the charger is plugged in for balancing purposes.
Looks like I won't be plugging it in very often
as all the cell balance lights came on again within a few seconds. No obvious cell imbalance.Manual supercap balancing using the charger might just be a week/monthly/six monthly task? We shall see.
Joined
·
14,379 Posts
Hmm just thinking out loud there are several resistors in the MCM HV input circuit.
If I reduce them all by say a factor of 30 then instead of 150v being reqd to trigger the circuit it may well work with 5v..
R237 15k / 30 = 500R
R243 10k / 30 = 330R
R238 120k / 30 = 4k
R241 510k / 30 = 17k
Now I suspect these are also in use by the leak to chassis detection circuit so i will have to use a little dc-dc converter to power it after it's converted to 5v operation.
The only thing now are two suspicious diodes in an opposite configuration, perhaps they are used to detect a battery reverse polarity issue.
Quite how that would occur?
I'll try and transfer this to my schematic program and post it for comments.
If I reduce them all by say a factor of 30 then instead of 150v being reqd to trigger the circuit it may well work with 5v..
R237 15k / 30 = 500R
R243 10k / 30 = 330R
R238 120k / 30 = 4k
R241 510k / 30 = 17k
Now I suspect these are also in use by the leak to chassis detection circuit so i will have to use a little dc-dc converter to power it after it's converted to 5v operation.
The only thing now are two suspicious diodes in an opposite configuration, perhaps they are used to detect a battery reverse polarity issue.
Quite how that would occur?
I'll try and transfer this to my schematic program and post it for comments.
Joined
·
1,545 Posts
At point is it going to be easer/better to use a commercial, programmable three phase motor driver and BMS?
What you're doing has strong parallels to fooling and intercepting engine management signals, but in that case it usually gets to a point where a complete replacement system can do things (and do them better) that the intercepted/tweaked factory system can never do.
What you're doing has strong parallels to fooling and intercepting engine management signals, but in that case it usually gets to a point where a complete replacement system can do things (and do them better) that the intercepted/tweaked factory system can never do.
Joined
·
14,379 Posts
Fair question.
I agree at some point it would be better, not likely cheaper though IMO.
I don't think we are there yet for me for several reasons.
1) I like tinkering with the oem system and learning more about it (Useful for me and others to have a growing understanding of how it all works) It also means others can follow in your footsteps without making major changes if they want, and it retains the Insight heart...
2) A bespoke BMS and motor driver is unlikely to be cheap, the capacitor cost aside this hasn't cost much so far. I'm using cheapo balancer boards, cheap led driver charger etc. You might have mentioned it but how much is the Motec dash/ecu stuff? Plus I have loads of spares, so if I blow the motor/driver/mcm etc I can replace it for very little.
3) If you replace the BCM/MCM stuff at the back you then have to fake signals to fool the ECM at the front. Doable but you are just moving the faking from one part to another unless you also replace the ECM and go down your route..
I'm learning a lot doing this project and supercaps in a road car is pretty radical.
I agree at some point it would be better, not likely cheaper though IMO.
I don't think we are there yet for me for several reasons.
1) I like tinkering with the oem system and learning more about it (Useful for me and others to have a growing understanding of how it all works) It also means others can follow in your footsteps without making major changes if they want, and it retains the Insight heart...
2) A bespoke BMS and motor driver is unlikely to be cheap, the capacitor cost aside this hasn't cost much so far. I'm using cheapo balancer boards, cheap led driver charger etc. You might have mentioned it but how much is the Motec dash/ecu stuff? Plus I have loads of spares, so if I blow the motor/driver/mcm etc I can replace it for very little.
3) If you replace the BCM/MCM stuff at the back you then have to fake signals to fool the ECM at the front
. Doable but you are just moving the faking from one part to another unless you also replace the ECM and go down your route..I'm learning a lot doing this project and supercaps in a road car is pretty radical.
Joined
·
14,379 Posts
MCM HV Input Schematic
I've roughed out a schematic of the MCM High voltage input circuit attached.
The component types on the schematic are just ones from my parts catalogue.
The actual values are the important info. The resistors on the MCM pcb are 1% tolerance.
The 5 x OKI OCM 243CA (171) optocouplers are not listed anywhere but look like standard 5 pin types with transistor output.
The 2 x 2FV 60:12 diodes are also not listed, are they GP, Schottky or Zeners?
So my proposal to fake this input at a fixed simulated 150V input and to avoid using a HV power supply is to reduce the resistor values as below.
If I reduce them all by a factor of 30 then instead of 150v being reqd to trigger the circuit it may work with a 5v isolated supply...
R237 15k / 30 = 500R
R243 10k / 30 = 330R
R238 120k / 30 = 4k (I'll use a 3.9k)
R241 510k / 30 = 17k (I'll use an 18k)
Two of the optocouplers inputs (U2 2) & (U4 2) disappear into the LV side of the PCB, which is a bit strange, but perhaps logical when you think that if this circuit also detects HV leakage to the LV side, it must have a path to the LV side for a resistance comparison.
I would really welcome thoughts from electronic experts on what it is all doing?
Anyway I shall try changing the resistor values (paralleling others with them, to avoid damaging the pcb by removing old ones, and it makes it easier to revert to stock)
I've roughed out a schematic of the MCM High voltage input circuit attached.
The component types on the schematic are just ones from my parts catalogue.
The actual values are the important info. The resistors on the MCM pcb are 1% tolerance.
The 5 x OKI OCM 243CA (171) optocouplers are not listed anywhere but look like standard 5 pin types with transistor output.
The 2 x 2FV 60:12 diodes are also not listed, are they GP, Schottky or Zeners?
So my proposal to fake this input at a fixed simulated 150V input and to avoid using a HV power supply is to reduce the resistor values as below.
If I reduce them all by a factor of 30 then instead of 150v being reqd to trigger the circuit it may work with a 5v isolated supply...
R237 15k / 30 = 500R
R243 10k / 30 = 330R
R238 120k / 30 = 4k (I'll use a 3.9k)
R241 510k / 30 = 17k (I'll use an 18k)
Two of the optocouplers inputs (U2 2) & (U4 2) disappear into the LV side of the PCB, which is a bit strange, but perhaps logical when you think that if this circuit also detects HV leakage to the LV side, it must have a path to the LV side for a resistance comparison.
I would really welcome thoughts from electronic experts on what it is all doing?
Anyway I shall try changing the resistor values (paralleling others with them, to avoid damaging the pcb by removing old ones, and it makes it easier to revert to stock)
Joined
·
1,545 Posts
I don't know much about the Insight but I do know electronic car modification.I agree at some point it would be better, not likely cheaper though IMO.
I don't think we are there yet for me for several reasons.
1) I like tinkering with the oem system and learning more about it (Useful for me and others to have a growing understanding of how it all works) It also means others can follow in your footsteps without making major changes if they want, and it retains the Insight heart...
2) A bespoke BMS and motor driver is unlikely to be cheap, the capacitor cost aside this hasn't cost much so far. I'm using cheapo balancer boards, cheap led driver charger etc. You might have mentioned it but how much is the Motec dash/ecu stuff? Plus I have loads of spares, so if I blow the motor/driver/mcm etc I can replace it for very little.
3) If you replace the BCM/MCM stuff at the back you then have to fake signals to fool the ECM at the front
I'm learning a lot doing this project and supercaps in a road car is pretty radical.
The approach that you're taking (ie intercepting, fooling signals, etc) was the most common approach to modification of car electronic systems about 15 years ago. I always enjoyed this approach, and did a lot of it. In fact, for some car systems where (a)the software is not hacked and so aftermarket software is not available, or (b) the systems are those that very few people modify (like stability control, electric power steering, etc) this is still the best approach.
However, where there are commercially produced products that can take over the function of the factory system that is being fairly crudely hacked, then the results are invariably better going with a new system. Basically, when you're in charge of the logic, then you can do things the best way for the new requirements, rather than modifying the factory way.
Since these days there are aftermarket electric car motor controllers, and aftermarket battery management systems, all with their own software, there comes a point where intercepting the factory system will no longer be the best way of doing things.
I admire what you are doing, but hacks on top of hacks almost always end up with unwanted outcomes, or less than optimal outcomes. You're in the best place to make that judgement, but don't lose sight of the bigger picture of what is available to do what you're doing, starting from the idea of what you actually want to achieve - rather than how can I make the factory system achieve what I want it to achieve?
The MoTeC ECU and dash were expensive (although I bought both second-hand, so saved about 50 per cent) but engine ECUs and dashes to achieve much the same outcome are available for - say - 1000 UK pounds total.
Unfortunately, none of you get to drive my car but I'd be confident in saying that you'd all be quite astonished at what has been achieved - specially in areas the factory ECU system is pretty limited...lean cruise, EGR, idle smoothness, outright turbo power. Until you have full control over everything you don't realise how emancipating it is - and you wonder why you didn't do it sooner.
I repeat: I admire what you are doing (and have done), but as technology moves on, tweaking the Honda system is going to become increasingly a less-than-optimal approach.
Joined
·
14,379 Posts
I've tried the changed resistors. That doesnt work yet.... Need to do some probing with my scope next while varying the input voltage with my HV PSU.
I changed the temp setting in my BCM interceptor to faked 27C that works well.
Going to raise widen the usable capacity window by ten volts by adjusting the limits by five volts at each end..
I changed the temp setting in my BCM interceptor to faked 27C that works well.
Going to raise widen the usable capacity window by ten volts by adjusting the limits by five volts at each end..
Joined
·
14,379 Posts
Don't get lead astray by the OBDIIC&C available voltage parameters. I did LOL.
Bvo is the battery voltage detected by the MCM HV Input...
Mdv is the MDM voltage detected by the MCM VPIN input....
You don't get to see what the BCM thinks the voltage is unless you have a BCM gauge installed. (Remember them? )
The resistor changes I made to the MCM HV input give a reading of 80 volts (with my 5v isolated supply) not 150, so some more experimenting with values required.
I added a pot to the MCM VPIN input on my test MCM and it needs 2.8V input to show 150V for the Mdv parameter...
To fake this voltage it has to ramp up quickly at ignition on as the precharge contactor charges the filter caps. it then has to hold at 150v until you turn the car off.
It has an internal 100k pull down resistor on this VPIN line so we need a simple potential divide and suitable sized capacitor to allow the ramp up to 150v in the reqd time.
I'm getting an 80k resistor to +5v as the top of the PD and the internal 100k pull down resistor at the bottom with a capacitor of about 10uf. This assumes a high input impedance into the VPIN, I'll check it later. It gives a time of 656ms to charge to 2.8V. 16uf gives a one second charge time.
The advanced PDF's state that as soon as the bypass contactor is turned on if the difference between the MDM and MCM voltage is greater than 37V for more than two seconds then an error is triggered. So from that we can deduce we have to ramp up the faked MDM voltage within two seconds to match the MCM faked battery voltage....
We don't want it to ramp up too quickly to allow the precharge resistor time to charge the filtercaps.
If we rise too quickly the main contactor will be activated early and the voltage difference may burn the contacts.
I think I will try 22uf to start with. That gives a 1.4 second delay.
Bvo is the battery voltage detected by the MCM HV Input...
Mdv is the MDM voltage detected by the MCM VPIN input....
You don't get to see what the BCM thinks the voltage is unless you have a BCM gauge installed. (Remember them? )
The resistor changes I made to the MCM HV input give a reading of 80 volts (with my 5v isolated supply) not 150, so some more experimenting with values required.
I added a pot to the MCM VPIN input on my test MCM and it needs 2.8V input to show 150V for the Mdv parameter...
To fake this voltage it has to ramp up quickly at ignition on as the precharge contactor charges the filter caps. it then has to hold at 150v until you turn the car off.
It has an internal 100k pull down resistor on this VPIN line so we need a simple potential divide and suitable sized capacitor to allow the ramp up to 150v in the reqd time.
I'm getting an 80k resistor to +5v as the top of the PD and the internal 100k pull down resistor at the bottom with a capacitor of about 10uf. This assumes a high input impedance into the VPIN, I'll check it later. It gives a time of 656ms to charge to 2.8V. 16uf gives a one second charge time.
The advanced PDF's state that as soon as the bypass contactor is turned on if the difference between the MDM and MCM voltage is greater than 37V for more than two seconds then an error is triggered. So from that we can deduce we have to ramp up the faked MDM voltage within two seconds to match the MCM faked battery voltage....
We don't want it to ramp up too quickly to allow the precharge resistor time to charge the filtercaps.
If we rise too quickly the main contactor will be activated early and the voltage difference may burn the contacts.
I think I will try 22uf to start with. That gives a 1.4 second delay.
Joined
·
14,379 Posts
Bench testing my spare MCM found that 77K and a 10uf cap worked best ramping up the VPIN input in about 1.5 seconds from zero to to 2.8V (150V simulated)
Bearing in mind my efforts to also mod the HV input side of the MCM, and the resistor values I tried with my 5V isolated supply gave a voltage reading of about 80V, I'm open to ideas for what values to try next.. Higher or lower .....
R237 15k / 30 = 500R (I used a 510R)
R243 10k / 30 = 330R (I used a 330R)
R238 120k / 30 = 4k (I used a 3k and 1k in series)
R241 510k / 30 = 17k (I used an 18k)
On another note. i intend to use the BCM interceptor to also control the DC-DC high/low output line.
It's already controlling the DC-DC inhibit line to prevent it killing the caps at the low end.
If high regen takes them significantly over the 180v pack balancing voltage I will get
the BCM interceptor to turn the dc-dc on high to pull them back down quicker.
Bearing in mind my efforts to also mod the HV input side of the MCM, and the resistor values I tried with my 5V isolated supply gave a voltage reading of about 80V, I'm open to ideas for what values to try next.. Higher or lower .....
R237 15k / 30 = 500R (I used a 510R)
R243 10k / 30 = 330R (I used a 330R)
R238 120k / 30 = 4k (I used a 3k and 1k in series)
R241 510k / 30 = 17k (I used an 18k)
On another note. i intend to use the BCM interceptor to also control the DC-DC high/low output line.
It's already controlling the DC-DC inhibit line to prevent it killing the caps at the low end.
If high regen takes them significantly over the 180v pack balancing voltage I will get
the BCM interceptor to turn the dc-dc on high to pull them back down quicker.
81 - 100 of 212 Posts
