[1036Project]

I officially own a lithium insight! Probably the cheapest and roughest build around, but lithium nonetheless.

I'm using the Mystery Ford cells I found on battery hookup.com.
Being that I don't have the full specs of these batteries, I didn't want to take the time to do a clean install. They may not be sufficient for the insight or my driving habits.

The batteries came already compressed in housings. I wish I would have left them that way. Curiosity got the best of me and I took one apart to get a closer look. They were compressed much tighter than I had assumed. I tried to get them back together using ratchet straps. Not only did this not work, it damaged 2 cells in the process.

I'm obviously not going to be putting damage cells in my car. I used allthreads and poplar board (eventually to be replaced with aluminum plates) to compress the undamaged cells back together.

I am usually hesitant to make permanent changes. Instead of cutting the harness to make a BCM Fooler, I bolted resistors to the orange end board from the stock pack. I then wrapped the end board in cardboard to prevent shorting.

For HV wiring, Walmart had 4-gauge 24-in ground wires for 5.99. They leaves a lot of excess wiring, but it saved a lot of time cutting, stripping, and crimping.

Cost breakdown:
Batteries $290.43 shipped
Resistors $5.99
HV Wires $23.25
Hardware $18.69
Misc wires and hardware ?$10

Total 328.36$

The total install time was around 7 hours (plus countless hours of research). About 2 hours of which was spent recompressing cells. Probably around another hour to figure out why I was getting a P1445 IMA code. It turned out to be the relay starter mod I had done previously.

after removing that little piece of wire everything is working great.

Currently I am unable to afford a BMS for this system. To minimize risk (in my completely nonexpert opinion) I'll be keeping the batteries as close as possible to 50% SOC, monitoring pack voltage, and frequently checking cell voltages to see if any are drifting.

DISCLAIMER: using lithium batteries without a BMS can cause injury or even death!!! You could even end up burning your car to the ground!!!

I would love to hear y'all's thoughts, comments, and concerns.

 

[retepsnikrep]

All projects are interesting. Even V.001 Mockups/test setups etc!

If those are LiPo cells and the same size as the Honda ones, then you can likely assume the same general capacity. ~5ah or so.

You know the serious risk you are running without a BMS and a 'prototype' setup.

How many cells do you have in total?
What are you Max/Min pack voltages?
Did you balance/check the pack cells manually before starting?
Do you have a BCM interceptor or some way of cutting regen/assist if the pack gets too high/low?

If it works ok for a few days then I would make serious efforts to improve the general safety of the installation asap.

You can also get a cheap BMS from e-bay if you cannot afford LiBCM.
Something that beeps at least so you can throw the main IMA switch.

If you want a BCM Fooler PCB I'll send you one for $5 just to cover postage and help you out.

 

[1036Project]

Thanks, I appreciate that sentiment!

Actually 5ah is one of the few specs I do know. These are a bit smaller though, about the same width and height but only 2.5 in tall. After watching the video of Mudder disassemble G3 packs, I can tell the cells I have are compressed much tighter.

I have 58 good cells left. I currently have 52 in the car, but only 45 connected. most of them were the same voltage. They were all within .003 volts.

Do you have any particular recommendations for a cheap eBay BMS?

Thank you, I really appreciate the offer on the BCM Fooler PCB. I'm actually considering buying a few things from you. I definitely plan on current hacking after I get things buttoned up and a BMS. Do you have a list of all of your current products and prices somewhere?I'm debating on PCB route or completed products.

 

[retepsnikrep]

I don't have any cheap BMS recommendations, but they are out there.
I have an Orion 2 in my workshop (not cheap) and LiBCM of course.

E-mail me for details of my stuff.

@Bull Dog in the US can supply you a current hack pcb and 150A IMA fuse.

 

[mudder]

Thanks for sharing your prototype setup.
I strongly encourage you to add a BMS. Without it, you could easily burn your car to the ground. At the very least, please don't park your car in a garage attached to habitable space.

If price is an issue, I can sell you just the LiBCM PCB for a reduced cost (no grid charger, cables, adapters, etc). PM me if you're interested.

 

[1036Project]

I decided to go ahead and upgrade to a 60s setup. Partly because the DC-DC converter upper voltage limit has been raised. Mostly because I couldn't find any information on how much compression these cells need. I figured it'd be better to buy another set and leave them together as a factory unit.

I'm still saving up for a BMS so in the meantime I made a VMS

I would do a write-up on it if I thought it was worth someone's time. Based on the parts and labor cost, it's really not much cheaper than a Chinese BMS. It also doesn't offer any type of protection.

I have also added some RC charger balance leads and plugs to the batteries.

It's a plug & play unplug & play unplug & play.... system. It seems like it's going to be an exercise in spinning plates but that's the story of my life anyway.

I had quite a bit of trouble getting the voltage spoofing right. I didn't know if I had hooked something up wrong or if the car didn't like 230 volts. I decided to temporarily drop down to 42s to test voltage spoofing at a voltage I knew the car was good with. I was able to get it going so I hooked back up to 60s. I then used my NiMH discharger to drop the voltage to 220. I was able to get it going at this voltage.

In the process of voltage spoofing I inadvertently created something like a mini mini BCM interceptor.

I connected the shafts of two potentiometers together. This enables adjustment of the voltage on connector E and VPIN at the same time. This actually worked out surprisingly well. By turning them all the way down the SOC will slowly drop to four bars and force background charging. Turning them all the way up will slowly raise the SOC to full and cut out regen.

Using the interceptor I started slowly raising the SOC. When It reached about 230 volts the system stopped working again. Over the next hour or so the voltage slowly dropped from usage by the DC-DC converter. The system came back on again around 228 volts.

for the next few days the system worked pretty well in the mid 220 volt range. Unfortunately things started going downhill. I went to start the car the next morning and got the 12 volt chugga chug start. It had a p1444 code which I had seen several times while attempting to voltage spoof. This time though, it also had a p1647 code that I haden't seen before.

Did I say things went downhill, more like fell off a cliff. While I was checking my wiring, I accidentally dropped the mini mini BCM interceptor onto one of the balance leads.

This caused a direct short at 53s which created quite an impressive flash of light.

As I investigated further, I found that one of the bolts on my crudely made BCM Fooler was loose. I was hoping this was the cause of my P codes. This inspired me to build a BCM fooler directly into the BCM case.

I think it was successful because I don't have any new codes from it. Unfortunately it did not take care of the codes I already had.

After hooking everything back up, I noticed that I had no voltage on the battery. Who would have thought that directly shorting a battery would blow the main fuse. thankfully @Bull Dog had one on hand and was able to ship it out the next day. It came in today and I have everything renstalled. now I'm on to searching for the actual cause of my P1444 and p1647 codes. Any advice would be much appreciated.

All troubles aside I have to say 60s is a blast! I can't wait to see how it does with current hacking.

 

[retepsnikrep]

Please let me send you a proper BCM fooler pcb before you burn up your BCM/CAR.

IMO You are very unlikely to be able to keep voltages in step accurately enough with your double potentiometer like that.

The spoofed BCM Taps, MCM E and VPIN voltages have to be in sync.
You should work with at least 1% tolerance resistors and select them to match.
0.1% is better.

Work out you useable desired actual pack voltage range and select the resistors to bring down the pack voltage into the normal Nimh range that the car will see.

The + lead pre-resistor for the BCM fooler should match the VPIN resistor.

So you should end up with say a

10x10k 1% BCM fooler.

1x33k 1% BCM fooler + lead pre-resistor to shift the voltage down.
1x33k 1% VPIN resistor to bring the MDM voltage down.

If my maths is right 33k as above should bring 230V down to 172V for the car.


P1444... How is your VMS powered? How does it work? Schematic?
Is it isolated from the car 12V side? It should be if the HV is going into it.

 

[1036Project]

Peter, I really do appreciate the offer on the BCM Fooler PCB. This may sound weird and irrational, but I'm still hesitant to cut the BCM connector off of the bus board. I even emailed @Eli back in July of 2021 to see if he had an extra connector. I'm still waiting to hear back from him.

Please forgive my ignorance but I don't understand the problem with the internal BCM fooler. I was actually expecting a "burn down your car" comment for the VMS or interceptor. I was surprised to hear your concern with the BCM Fooler. Are the resistor leads too close together and at risk of arcing?

I originally attempted voltage spoofing with 1% tolerance fixed value resistors (I was only working with a kit of 60 values and have since purchased a 130 value kit). When I was unable to get it working, I decided to try a 5k linear potentiometer on the VPIN side. After a few attempts and readjustments, I was able to get it going. The potentiometer I was using was fairly loose. It was changing its set value due to vibrations from the car. This was causing the IMA system to randomly shut down. Eventually on one of the attempts, the IMA stayed active for a good while. I decided to probe the potentiometer to see what resistance the car was so happy with. To my surprise it was way under the value of which I had first set it. (side note I really need to learn to start documenting builds better. I wish I could give you exact numbers but I'm doing this from memory). I know I had a 27k ohm resistor on the PD for connector E. I believe the potentiometer on the PD for vpin was down to something like 2.61k ohm. This made it seem to me like they don't need to be so closely in step as long as the resistance on VPIN is lower. This is what inspired me to try the dual potentiometer setup. It seemed like a good way for me to find a sweet spot for the battery voltage. It's nice that it can somewhat function as a BCM interceptor. I'm sure it would be more practical to just use the two button interceptor instead.

I was originally going to use a 5K pot and a 50k pot. I figured it would be better to have more resolution. So I used a 2K pot in series with a 2K fixed value resistor, and a 20K pot in series with a 20K fixed value resistor. To keep the resistance on VPIN lower, I took advantage of the imperfections inherent to the potentiometers. With the potentiometer shafts connected. using an OHM meter on one side of the 20K pot. If you turn the pot from zero to 10K, the mirrored side of the 2K pot will read something like .970 ohm. If you turn the other side of the 20K pot from 0 to 10K, the mirrored side of the 2K pot will read something like 1.03 k ohm. Using the correct side of the pots will keep the resistance on VPIN lower than on connector E.

All that being said, I will concede that this dual potentiometer setup may still be the cause of my problem. I'm willing to try going back to a fixed value PD fooler to rule this out. Thank you for taking the time to do the maths on those values. I noticed you're using the same value now for both BCM Fooler and VPIN resistance. I take it for that 33k ohm VPIN value the other side of that PD would be 100k ohm. In my research of voltage spoofing, I have only seen your setups with the vpin PD being 1/10 of the PD for connector E. Was this change made out of necessity or convenience?

In regards to the VMS, I can guarantee you that it is 100% isolated from the 12 volt side of the car. At least it is now because I have disconnected it. You were spot on with that p1444 code. I delivered food for 104 miles last night and that code never returned.

The VMS isn't really anything special. It's actually just 10 cheap panel meters tapped every 6s and switched by relays.

In my shade tree engineering calculations, I figured as long as the relays could handle the sub pack voltage they would be sufficient. I like to learn by trial and error and this one is definitely the latter.

2hr 10min

 

[retepsnikrep]

I can send you a BCM connector and pcb, I have spares.

Yes inside the MCM is a 100k pull down resistor.

10 x 10k for the BCM fooler also = 100k

Hence 33k BCM fooler pre-resistor and VPIN reistor.

 

[1036Project]

^ You have a spare BCM connector, sweet! Email sent! I figured if Eli didn't have any no one would. Didn't think to message other battery suppliers.

Yes inside the MCM is a 100k pull down resistor.

I tried to find this information yesterday before I replied to your post. I learned something new about the search function on IC today

That's quite a big difference. I was finally able to find it!

MDM VPIN OUT >>> Resistor >>> MCM VPIN IN

 

[retepsnikrep]

Just use an isolated dc-dc to power your VMS

 

[1036Project]

^ I believe you are suggesting to power the VMS from the HV battery. That would be a great solution if this wasn't originally meant to be a temporary device. I'm already up to about $35 (not including parts that I already had) and about 15 hours of R&D. It's starting to look like this project was a fool's errand. After a preliminary search here is the closest one that looks like it would work.

TDK-Lambda PH50A280-12

Product Attribute	Attribute Value
Manufacturer:	TDK-Lambda
Product Category:	Isolated DC/DC Converters
RoHS:	Details
Number of Outputs:	1 Output
Output Power:	50.4 W
Input Voltage, Min:	200 V
Input Voltage, Max:	425 V
Output Voltage-Channel 1:	12 V
Output Current-Channel 1:	4.2 A
Isolation Voltage:	500 V

The list price on it is $126. I think at this point I'm just going to use a standalone 12v battery. I'll just turn it on periodically. It's not like I can constantly watch it while I'm driving anyway.

Thank you kindly for the suggestion though.

 

[retepsnikrep]

No i meant from the 12v with a little isolated converter.

 

[1036Project]

^ Oh, a little isolated converter probably would be the best way to go. You mean something like this?
PQP3-D12-S12-D

Output Power:	3 W
Input Voltage, Min:	9 V
Input Voltage, Max:	18 V
Output Voltage-Channel 1:	12 V
Output Current-Channel 1:	250 mA
Isolation Voltage:	1.5 kV

That's a much cheaper option as well. There is another free option though (for temporary use).
I could do some old-fashioned American over-engineering. I bought a meanwhell HRP-600-12, to go IMA-less, before I found cheap lithium. According to the data sheet...

Isolation Resistance
I/P-O/P, I/P-FG, O/P-FG:100M Ohms / 500VDC / 25℃/ 70% RH

I'm not sure the difference between isolation and isolation resistance.

 

[retepsnikrep]

Yes that's what i mean.