[WIP] 1965 Ford Mustang

[chrskly]

Was interested to take a look at your code but can’t find it?

Here you go https://github.com/chrskly/electric-mus ... in/bms/src

[rstevens81]

There are some things to think about in allowing a BMS to control contactors and BMS code in general ....
1. Given that car is a very electrically and efm noisy environment a watchdog in the code is imperative because if the chip locks up you loose BMS, so therefore a watchdog to reboot it in that event is imperative.
2. The behaviour of the watchdog is also important...i.e. if it were to lock up and reboot what would happen would the contactors open and then close again? Under load this would substantially reduce the life of the contactors even the big ones. That's not all of the failure modes but a big one to illustrate the point.
3. Opening of contactors under load is always going to be bad for them and everything connected as electricity under load is like flowing water through a pipe if you suddenly stop it in a pipe you get hammershock the clunk noise you hear in pipes if you have an old house, in electrics the inductance of the circuit (wires) will try to raise voltage whilst you disconnect to make a bigger spark on the contactor, how big of an effect this is I don't know but point was also that a false flag will substantially reduce life of contactor.

I am aiming for a different approach it may not be right but is simpler in my mind as the simpler it is the less we have to think this may or not be useful...
1. Contactors controlled by VCU, when starting the switch on process just add a line of code the check the voltage dela...tbh it probably can be quite simple in that if your cell voltage delta is above X mV (of all 192 cells) you go into error or warning state and you just chech for error or warning before switching on.
2. If so etching goes wrong you enter a warning state a buzzer sounds if after 10 seconds (arbitrary to allow you to pull to side of road and switch off ignition) if after this X seconds the inverter is commanded to shut down and then commanded to open contactors.
Although this isn't perfect what situations would 10 seconds not be enough and if so could you skip the warning stage and go straight to powering down inverter?

Hope this is helpful in someway ... It all stems from the age old problem who is the biggest master BMS or VCU/inverter? You sorta have to choose 1 and live with the consequence and work round it to the best you can i.e. if you were to choose BMS as master it would tell the inverter to shutdown X seconds before it would open the contactors.

[chrskly]

There are some things to think about in allowing a BMS to control contactors and BMS code in general ....
1. Given that car is a very electrically and efm noisy environment a watchdog in the code is imperative because if the chip locks up you loose BMS, so therefore a watchdog to reboot it in that event is imperative.
2. The behaviour of the watchdog is also important...i.e. if it were to lock up and reboot what would happen would the contactors open and then close again? Under load this would substantially reduce the life of the contactors even the big ones. That's not all of the failure modes but a big one to illustrate the point.
3. Opening of contactors under load is always going to be bad for them and everything connected as electricity under load is like flowing water through a pipe if you suddenly stop it in a pipe you get hammershock the clunk noise you hear in pipes if you have an old house, in electrics the inductance of the circuit (wires) will try to raise voltage whilst you disconnect to make a bigger spark on the contactor, how big of an effect this is I don't know but point was also that a false flag will substantially reduce life of contactor.

I am aiming for a different approach it may not be right but is simpler in my mind as the simpler it is the less we have to think this may or not be useful...
1. Contactors controlled by VCU, when starting the switch on process just add a line of code the check the voltage dela...tbh it probably can be quite simple in that if your cell voltage delta is above X mV (of all 192 cells) you go into error or warning state and you just chech for error or warning before switching on.
2. If so etching goes wrong you enter a warning state a buzzer sounds if after 10 seconds (arbitrary to allow you to pull to side of road and switch off ignition) if after this X seconds the inverter is commanded to shut down and then commanded to open contactors.
Although this isn't perfect what situations would 10 seconds not be enough and if so could you skip the warning stage and go straight to powering down inverter?

Hope this is helpful in someway ... It all stems from the age old problem who is the biggest master BMS or VCU/inverter? You sorta have to choose 1 and live with the consequence and work round it to the best you can i.e. if you were to choose BMS as master it would tell the inverter to shutdown X seconds before it would open the contactors.

Hey,

thanks for the input. Definitely happy to hear all perspectives on this. I'm just figuring stuff out as I go.

1. Yeah this is a good point. The pi pico has support for a watchdog. I'll add this in. Thanks.

2 + 3. Random/unwanted open/closing of contactors is definitely something I want to avoid. When the contactors are already open, the BMS can block future closing of the contactors. So, for example, we can end up in drive mode with only one pack enabled. In this case, if the BMS reboots, then something unintended could happen. The question is, can the pico boot up fast enough that the contactors wouldn't notice a quick off+on? Probably not. But it's worth trying/testing.

Buzzer is a great idea also. I plan to have a 'dashboard vcu' for driving the original analogue gauges and lights. Will add some sort of buzzer to this also. The dash vcu will also have a logger built in and be a hub for monitoring the overall state of the car (rather than having one wifi ap per device). So, any errors from any part of the car will be surfaced on the dash and also a web interface provided by the dash.

[aaronmcg]

Excellent progress!

[Jacobsmess]

Hi Chris, how are you finding the is300H transmission? I'm wanting some assurance it can deliver motorway driving speeds comfortably without MG1 exploding but I've found no evidence as of yet and as it's early in development in the EV conversion world there isn't a wealth of knowledge from running projects.

[Jacobsmess]

Just for clarification, I was hoping you might be able to provide some information on the RPM from the gearbox output at different voltages.

[chrskly]

Just for clarification, I was hoping you might be able to provide some information on the RPM from the gearbox output at different voltages.

Hey, sorry, I missed this somehow. To be honest my gs300 has been sitting untouched on the garage floor for months. I have lots of other work to do on the car until I get to the stage of spinning the wheels. For example, replacing rusted out floors

[PatrcioEV-ATX]

Hey, sorry, I missed this somehow. To be honest my gs300 has been sitting untouched on the garage floor for months. I have lots of other work to do on the car until I get to the stage of spinning the wheels. For example, replacing rusted out floors

tempImagebUzNxr.gif

Keep at it. Someday you'll have a gorgeous classic to tool around in!

[chrskly]

Keep at it. Someday you'll have a gorgeous classic to tool around in!

That's the hope

[Huskte]

Been a while since an update. How're things going with the project?

[chrskly]

Been a while since an update. How're things going with the project?

I broke rule #1 of ev conversion club. I chose to convert a rust bucket. The bodywork continues...

[rstevens81]

I would say your in good company
It's the slight problem when you choose the car you want not just something that's randomly cheap.

[Huskte]

I broke rule #1 of ev conversion club. I chose to convert a rust bucket. The bodywork continues...

IMG_9264.jpeg

Oh wow, I't looks like it may have been easier to build a new car from scratch haha. At least you can honestly say that you've touched every single nut and bolt on the car.

Best of luck, I envy your strength of commitment.

[Huskte]

I would say your in good company
It's the slight problem when you choose the car you want not just something that's randomly cheap.

Some people are just cursed with good taste i guess.

[Gregski]

this is such a fantastic build, I hope it picks up again, and boy do I miss my '68 Rustang now

[chrskly]

Wow, my last update was August. Time has really flown by.

This project is still very much alive. I've not posted anything because, frankly, I'm still trying to get ahead of the rust. Here's some of the gore for those who enjoy.

Cowl tank replacement.

Front inner fender cleanup.

Rear wheel arch cleanup.

Front frame rails

It's interesting what you learn about yourself doing projects like this. Mostly what I've learned is that I like doing things The Hard Way™. See below for what I mean.

I don't seem to have posted anything around what I'm doing on the LV side here. The original loom is a write-off. Absolute junk. I won't use any of it. What I need will be very far from what was originally there, so it's a ground up redesign (The Hard Way™). I'm breaking the LV stuff up into three chunks : front, rear, and lighting. You can probably guess for yourselves what each of those does. The plan is to make a PCB with all of the fuses, relays, and diodes for each chunk on a single board. Why not just get generic fuse/relay box from the breakers yard or Ali? I know if I try and make one up using just spade connectors and wire, I will 100% guaranteed get lost and make a mess of it. But if I design it up in KiCAD (where I can leave little notes next to things) I'm much more likely to get it right.

Here's the current version of the 'front' box. I'll definitely do another version of this board. The right angle connectors are a waste of space. I should have the pcb take up the full area of the box and use vertical connectors instead.

I mentioned before that I'm implementing my own BMS (The Hard Way™). I'm using BMW PHEV modules so I've lifted the already-reversed communication protocol from SimpBMS (thanks to Tom DeBree!). The rest is a ground up rewrite. I want to be really, really sure that this works as I intend, so I'm going to do a lot of testing before putting it in the car. I'm building a testing rig to simluate all of the possible inputs (voltage, temperature, charge request, etc.) and all of the possible state transitions. After a bit of head scratching, I realised I could use one board to run the actual BMS code, and another board to run testing code which simluates all of the signals (CSC messages, ISA shunt messages, ignition on digital input, charge request digital input, etc.) and validates that the right thing happens. I hope to completely automate the testing process. I have a lot more work to do on the BMS code. So, when I make any changes to the BMS code I can validate that it still does everything it's supposed to do very easily.

As I'm writing this, I'm thinking that I should also test unexpected resets of the BMS itself ....

I'm working on my own CHAdeMO implementation (The Hard Way™). I got my hands on the standards docs for all of the CHAdeMO versions up to v3. I'm going to try to make it completely compliant with the standard. I've got the first version of the hardware back from JLC. There are no doubt errors with this and I'll need at least one more go around. I'll likely create a testing rig like I have for the BMS.

The car will have CCS as well as CHAdeMO, so I'll need somewhere to put those two charge ports. I don't fancy my chances of adding charge port doors into the fenders and not making a total mess of it. Besides that I think it would look ugly. So the plan is to put the charge sockets behind the two tail lights. I'll make a door that goes over the original rectangular hole where the whole tail light assembly sat. The tail light lens and chrome will bolt to the outside of the door and look more or less original. I can then make a small tub to go on the inside which will be the mounting for the charge port. There won't be room for a filament bulb any more of course, so I'm working on a PCB that will sit inside the lens which has a bunch of LEDs on it. It's all a bit complicated, but hopefully the end result will be nice. The charge ports will be totally hidden.

I will need to extend the opening about 5mm - 10mm on the top and bottom to make room for the CCS socket.

After lurking on some conversations in Damien's discord, I'm also rethinking the rear battery location. One thing that this car is notorious for is, when rear-ended, the fuel tank gets crushed. If I were to put the batteries where the fuel tank goes, there's not a lot of car to absorb an impact before getting to them. It's a shame, as the batteries fit so well into this space - it's like they're made for it. But there's only something like 300mm between the rear bumper and the edge of the battery box. So, instead I think I'll put them in the boot, behind the back seat, over the rear axle. When I looked at this before, I thought I wouldn't fit all 6 modules there. But, dumb dumb that I am, I didn't consider just putting them on their side. Then they fit easily. So, the plan then is to make up a battery box, probably out of aluminium, and crane it in whole. This also has the benefit of it being further into the car, and further away from wet.

I'll still finish and use the tub that slots into the old fuel tank spot. Instead it will now just house the OBC, rear HV junction box, charge controllers, BMS, and some other bits.

I got my hands on an iBooster. Seems like it will fit nicely. The old booster is somewhere in my garage, I'm just not sure where. I can compare push rod lengths at some point, but there's a lot of other things to do before I get to that.

The car will be 60 years old next year - in June if I remember right. I'd really, really like to get it back on the road by then. I've been able to block out a few hours a week to work on it, so hopefully that'll accelerate progress a bit.

[Alibro]

Looking good
Would it be easier to put a cage around the back of the car in the boot or behind the rear bumper to strengthen it rather than compromising the battery position?

[chrskly]

Would it be easier to put a cage around the back of the car in the boot or behind the rear bumper to strengthen it rather than compromising the battery position?

In the long run its probably just easier to relocate the batteries rather than trying to add strength to rear of the car. They will actually fit quite well above the rear axle too as it turns out.

[chrskly]

Correction, the rear-most point of the car is about 150mm from the edge of the fuel tank area. So, even worse than I thought.

[rstevens81]

From someone who has quite literally driven himself nuts worrying about battery placement .... Rember the batteries are not as explosive than than petrol. For what it worth either (a) put them above the rear axel bits obs quite a bit of weight high up or (b) place them in the fuel tank area but put them side ways in an arrangement of 2x2 that way the connection goes in a circle and the cables can't really short out in event of a big shunt.
The thing to rember is that even with relatively recent cars upto the early 2000s had fuel tanks located behind the rear axel on the crumple zone (vauxhall omega is a good choice) and also rember that r100 only specified that a crush test is required if it's less than 300mm from the furthermost rear (or forward edge)

edit: slept on it a bit ....and changed my advice...since you can fit them above the axel do so, i would get some threaded rod and use these.

Edit 2 this is how I would arrange batteries above the rear subframe frame rather than on their side stack them up...this gives you lots of space that you will need for fuses and contactors etc (the sbox is a bit weak to be reused really)

[chrskly]

Thanks for the input, it's a big help

Yeah, the more I think about it, the more the batteries-over-the-axle option seems like the better one. Like I say, I had sort of discarded the idea early on. They didn't seem to easily fit above the axle and the dimensions of the fuel tank lined up so well. It's your classic "if I fits I sits" situation.

I don't think I had considered arranging the batteries the way you're suggesting. Which is surprising as I have pages and pages and pages of scribbles of different configurations for the front box I have max 280mm of height above the axle. It might be a little less as I'm not confident in the accuracy of that measurement. The BMW PHEV modules are 105mm tall with the CSC clipped to the end and 125mm tall with the CSC clipped to the top.

CSC on top => two modules tall = 250mm
CSC on end => two modules tall = 210mm

Both options are fairly tight, but possibly doable.

Not that it makes any difference, but there is a little error on your drawing. The battery terminals are always on the same side as each other. So, I just drew it out again, just to double-check. Basically just the 4<=>5 link (on your diagram) is on the other side.

Min dimensions of CSC on top => 753mm wide, 360mm deep, 250mm tall
Min dimensions of CSC on end => 800mm wide, 360mm deep, 210mm tall

I'll have to factor in enough gaps to fit everything together, plus the battery box material itself.

..... time to double check that height measurement ....