Phev Insight

[Mike Dabrowski 2000]

Peter,
The sense series resistance is 180 ohms cold, and when heated by a heat gun, it shot up to 2K
http://www.99mpg.com/workshops/mikessat ... st_2.jpg,0
They seem to be thermal resetable fuses, so I suspect that a simple .6W resistor divider should work fine.
Anxious to see how you make out, as we have been talking about trying that BCM fakeout for some time.
Good Luck

[uhtrinity]

I hate to disagree with Mike, but when I tested the sense resistors (if that is what they are called), they acted like normal resistors. I tried heating a few. They actually started at 180 - 190 ohms and dropped to 160 ohms hot. Is it possible there could be a few different kinds installed in packs? Or could you have mixed up your sense resistor and PTC notes?


Will be interesting to see how the system handles a larger pack and will answer the question of how large of can the Insights BCM go. Also, are you going to let the stock system do some charging? I can't imagine trying to fully charge a battery that size. Just charging under 8Ah seems to take forever

As far as faking out the BCM, if you can use either 40 or 50 cells, then you wouldn't have to as you would have 11 points to tap from (the 11th point is the final (-) if you are counting all (+) terminals. I haven't checked the Insight pack closely, but the Civic pack adds a 12th by tapping both sides of the cut switch. I will know for sure about the Insight wiring when I start my visual load gauge next week

[retepsnikrep]

Originally Posted by uhtrinity

I hate to disagree with Mike, but when I tested the sense resistors (if that is what they are called), they acted like normal resistors. I tried heating a few. They actually started at 180 - 190 ohms and dropped to 160 ohms hot. Is it possible there could be a few different kinds installed in packs? Or could you have mixed up your sense resistor and PTC notes?


Will be interesting to see how the system handles a larger pack and will answer the question of how large of can the Insights BCM go. Also, are you going to let the stock system do some charging? I can't imagine trying to fully charge a battery that size. Just charging under 8Ah seems to take forever

As far as faking out the BCM, if you can use either 40 or 50 cells, then you wouldn't have to as you would have 11 points to tap from (the 11th point is the final (-) if you are counting all (+) terminals. I haven't checked the Insight pack closely, but the Civic pack adds a 12th by tapping both sides of the cut switch. I will know for sure about the Insight wiring when I start my visual load gauge next week

I'll measure the resistors in the BCM subpack monitoring leads myself, I'm happy with your values for the ptc strips and I'm going to move the other 3 or 4 thermistors from an old pack and spread them around the new one.

Sadly the voltages don't work out even enough to use 40 or 50 cells (I would have liked to have done that) so I'll fake it with a PD made from 1% tolerance resistors this will mean the bcm thinks pack is in perfect balance and prevent any balancing forced charging as opposed to just low pack forced charging which I can disable with MIMA anyway as I wish. Of course if I do need to use all 50 cells in the end then I will just tie the leads to the cells at 5 cell Intervals. Yes the IMA battery cut switch does have connections both side to the bcm.

I'll see what the bcm makes of the pack but may have to do the BCM current monitor sensor output divide by 10 trick, we shall see.

Yes I will let the IMA system charge as normal when braking, descending hills etc, but I will make much more use of assist obviously and try to run the pack down on journeys and charge at home. I'll use MIMA to stop it force charging when low. Seems like MIMA is a pre-requisite for any sort of larger pack mod or you will find difficulty mangaging or using it's extra capacity.

Peter

[Mike Dabrowski 2000]

When I get a clear area on my bench, I will remeasure the resistors. The pack we were looking at was from a 2000 insight.

[uhtrinity]

So, on the sense resistors, my take is they are strictly there so the BCM can get "two subpack" averages. This would allow it to sense if the pack is out of balance voltage wise or to pick up shorted cells. The math shows that each resistor would have to dissipate about 1/4 watt or more @ 40 mA or higher. The voltage sense wires are also a very small gauge, I would think if they were meant to handle any balancing charging they would be a higher gauge. The 180 ohm resistors seem like they are there for current limiting purposes only, in this case 40mA - 50mA per subpack pair which would be 360 ohms.

When I build my real time voltage load display I'm thinking of using a multi meg ohm voltage divider to route battery pack values to the front of the car. Every 14.4 volt (18 volt peak) subpack pair will produce a voltage value to the front of about 1/2 (7.2 volts / 9 volts peak). I will be tapping off of the C connector on both packs using the same values the BCM sees. The other option I have considered is using optoisolators, though they cost more. I'm planning on 10 gauges per battery pack for a total of 20. It should look pretty and give me the real time balance monitoring I need

[Roger Crier]

Well done Pete!

Make sure you write down everything you do, because people will be queuing at your door once you get it up and running. Did the Battery vehicle society members join in and get the price down or did you just go for it anyway? Oh and just to stop everyone in their tracks, how many squillion quid were the batteries in the end?

How much assist amps and regen amps do you recon those batteries will stand?

I must come and pitch a tent in your garden so I don't miss any "Chernobyl" moments!

Rog

[IamIan]

Originally Posted by Roger Crier

Oh and just to stop everyone in their tracks, how many squillion quid were the batteries in the end?

As far as I know uhtrinity is using a HEV civic battery pack to add his capacity... The two that I bought were $500 ... or $250 each... they should be able to handle up to the 100 Amps of Assist and 50 Amps of Regen that the HEV Civic hits them with... and should have about 6Ah of capacity... In his Parallel setup the current ( 100 amps or 50 amps ) should be getting split very close to evenly... I am currently doing tests on my two HEV Civic battery packs... although my current setup only goes up to 1.8Amps of charging and 10 Amps of discharging.... I have not seen much need to upgrade the testing system to go higher up to the limits of 50Amps Charge or 100 Amps Discharge.... Since after I am satisfied with my testing I plan to use them at no more than 20 Amps discharge and no more than 2Amps of charging.

[retepsnikrep]

Originally Posted by Roger Crier

Well done Pete!

Make sure you write down everything you do, because people will be queuing at your door once you get it up and running. Did the Battery vehicle society members join in and get the price down or did you just go for it anyway? Oh and just to stop everyone in their tracks, how many squillion quid were the batteries in the end?

How much assist amps and regen amps do you recon those batteries will stand?

I must come and pitch a tent in your garden so I don't miss any "Chernobyl" moments!

Rog

The BVS bulk buy went down the drain as not enough interest to get a decent price.

The 50x40ah Li-fepo4 cells cost me £2350 or about $4600, the queue at my door might not be quite so large now

I think they will take the full 100A assist and 50A regen, I would not have bought them otherwise.

Hopefully the only "Chernobyl" moment will be me hoping about when I break the 150mpg UK economy figure

[retepsnikrep]

Just to help everyone out on the various threads regarding the IMA Battery here is the schematic page from the manual as a jpg file.

http://www.solarvan.co.uk/insight/imabattery.jpg

Peter

[retepsnikrep]

Been working on this a bit today.

Disassembled my old spare battery pack.

Removed thermistors from the 4 special subpacks, just need to bury those thermistors in the new pack and straight plug into BCM.

Am now cycling the 20 removed subpacks with my robitronic charger Charge rate 6.5A, discharge rate 20A, 5 cycles per subpack. I shall record on them capacity and IR etc I'll then sacrifice a few of the weakest and remove the PTC strips. I shall bury the ptc strips in the new pack and try to sandwich each recovered strip inbetween 10 cells to give the same warning as it does in the standard IMA pack. I'll add the pot as per unhtrinity to adjust the threshold once assembled. The ptc strip appears to be the final safety cutout in the event of a single cell overheating. I suspect the general pack temp is just measured by the thermistors which then control the fans. I'll set the ptc trigger point at about +50C to suit the Lithium cells.

I also removed the 12 subpack resistors which are soldered onto the end board, each is about 170-185R They look about 500-600mw rating.

Now looking at faking the intersubpack voltages with a potential divider, the question of the current consumed by the BCM subpack monitor inputs is very relevant. Each twin subpack monitor input goes through two 180R resistors so 14.4v (two subpacks) divided by 360R gives about 40ma max current if the resistance of the input is negligble. But how is that voltage measure inside the BCM? Optoisolator? I suppose I could open up the back of my car and remove on of the connections to measure the current

I looked at an array of 10x6.8k resistors (68K) total this gives a max current of 2ma through the PD, is that enough to drive the subpack monitor inputs? This also gives about 3w of heat disipation over the entire PD. If I was to raise the available current to 20ma by droping the overall PD resistance by a factor of 10 then 30W of heat would have to be disipated

So I'm looking at using my full quota of 50 cells instead which means a tap every 5 cells, easy. Voltage is a bit high but I shall try not charging cells to max V, If I stick to 3.75v per cell (instead of 3.85v) that gives a fully charged pack voltage of 187.5V Can it stand it? The nominal voltage will be about 3-3.2v under load a fair bit less I expect. 140-160v. I expect the fully charged pack voltage will fall away as well once the charge is terminated.

Thoughts?