It's a nice trend and I also hope it continues.
Luke's 2001 NB MX-5
Re: Luke's 2001 NB MX-5
How will the cost / performance compare between the two packs? It sounds like there's quite a bit of fabrication work involved.
- jonescg
- Senior Member
- Posts: 4721
- Joined: Thu, 21 Jan 2010, 23:05
- Real Name: Chris Jones
- Location: Perth, WA.
- Contact:
Re: Luke's 2001 NB MX-5
Cost is actually slightly less than the original, and substantially higher energy density. The big pluses for the new pack will be the fact all 24 kWh will fit inside the fuel tank space, freeing up the whole boot. Negative is the lack of cooling (could be a positive considering how I got here) and the lower power of these cells. The DCIR is about 60 mOhm, or about 0.2 ohms for the whole system. This will mean a saggier pack but the motor is pretty weak anyway. 40 kW or so.
AEVA National President, WA branch director.
Re: Luke's 2001 NB MX-5
While saving for LG Chem modules from a Jaguar I-Pace (https://zero-ev.co.uk/product/lg-4p3s-module-2-6kwh/) I thought I'd get busy with the odds and ends that have been sitting for a while on the project.
Instead of using a mechanical sensor attached to the original throttle cable I opted to get a Toyota Prius throttle pedal with my motor package. This required a little bit fabrication work to get it to sit nicely.
I figured while I had the pedal box out that I would grind back all of the spot rust and give it a nice coat of enamel. It is a restoration + conversion project after all. All finished with a fresh coat of enamel. Perfect fit sitting in the exact spot the original throttle pedal sat. It has full range of movement and no flex. Though if you look closely the clutch pedal box is horrid so I thought I'd better take that off and give it the same treatment. Once off it was apparent that the clutch master cylinder had been leaking into the cabin, not great. Time for a replacement.
Instead of using a mechanical sensor attached to the original throttle cable I opted to get a Toyota Prius throttle pedal with my motor package. This required a little bit fabrication work to get it to sit nicely.
I figured while I had the pedal box out that I would grind back all of the spot rust and give it a nice coat of enamel. It is a restoration + conversion project after all. All finished with a fresh coat of enamel. Perfect fit sitting in the exact spot the original throttle pedal sat. It has full range of movement and no flex. Though if you look closely the clutch pedal box is horrid so I thought I'd better take that off and give it the same treatment. Once off it was apparent that the clutch master cylinder had been leaking into the cabin, not great. Time for a replacement.
Re: Luke's 2001 NB MX-5
Clutch pedal and pedal box refitted after a bit of a clean up.
And a new clutch master cylinder, slave cylinder and shorter braided line that doesn't go from the drivers to passenger to drivers side (WTF).
Some new rubber for the pedals too!
Re: Luke's 2001 NB MX-5
I really wish I'd taken some photos of the original engine bay but the washer bottle sat on the passenger side above the exhaust manifold. Problem is that whole area will now be batteries so I've had to relocate the washer bottle. Lucky being a MX-5 there's some pretty damn good aftermarket support. Like this awesome Denso washer bottle for god knows what that fits perfectly into the space between the firewall and windscreen.
Though the kit came with extra tubing for washer fluid but I didn't actually need any, plus it uses the original pump. All I need to do now is wire it up, which I'll do when I wire all of the other bits and pieces as the body loom is going to be pulled apart once everything has a home in the car.
If you look hard enough you can see the labelled washer
Though the kit came with extra tubing for washer fluid but I didn't actually need any, plus it uses the original pump. All I need to do now is wire it up, which I'll do when I wire all of the other bits and pieces as the body loom is going to be pulled apart once everything has a home in the car.
If you look hard enough you can see the labelled washer
- jonescg
- Senior Member
- Posts: 4721
- Joined: Thu, 21 Jan 2010, 23:05
- Real Name: Chris Jones
- Location: Perth, WA.
- Contact:
Re: Luke's 2001 NB MX-5
Wow, that footwell looks like it has substantially more room than a certain Prelude...necrogt4 wrote: ↑Mon, 25 Jan 2021, 14:00 While saving for LG Chem modules from a Jaguar I-Pace (https://zero-ev.co.uk/product/lg-4p3s-module-2-6kwh/) I thought I'd get busy with the odds and ends that have been sitting for a while on the project.
Instead of using a mechanical sensor attached to the original throttle cable I opted to get a Toyota Prius throttle pedal with my motor package. This required a little bit fabrication work to get it to sit nicely.
Pedal 1.jpeg
Pedal 2.jpeg
I figured while I had the pedal box out that I would grind back all of the spot rust and give it a nice coat of enamel. It is a restoration + conversion project after all.
Pedal 3.jpeg
All finished with a fresh coat of enamel.
Pedal 4.jpeg
Pedal 5.jpeg
Perfect fit sitting in the exact spot the original throttle pedal sat. It has full range of movement and no flex. Though if you look closely the clutch pedal box is horrid so I thought I'd better take that off and give it the same treatment. Once off it was apparent that the clutch master cylinder had been leaking into the cabin, not great. Time for a replacement.
Pedal 6.jpeg
AEVA National President, WA branch director.
Re: Luke's 2001 NB MX-5
Leg room is actually pretty decent. It's quite noticeable with the dash removed though the seats are in the normal positions.
What's really lacking is headroom. I took foam out of the drivers seat and still am looking directly at the black band at the top of the windscreen. Though some nice racing buckets will solve that.
What's really lacking is headroom. I took foam out of the drivers seat and still am looking directly at the black band at the top of the windscreen. Though some nice racing buckets will solve that.
Re: Luke's 2001 NB MX-5
So I've once again pivoted on batteries for the project. Somewhat.
The Jaguar i-Pace modules we're a perfect size, but rather expensive and halfway across the world. I managed to find 100 LG E63 cells at a really good price and those topped off with a further 24 found on eBay gives me a 31s4p pack @ 111v 260A pack. An extra ~20A (or more given that the Jag modules were second hand and these are new) so that's just shy of a 30kWh battery pack instead of 26kWh. Should make for an easy 200km+ range.
I've been working on a preliminary design based on the Chevy Bolt EV battery (https://www.youtube.com/watch?v=ssU2mjiNi_Q). The idea is to use acetal blocks at either end of each cell pack. The bolt via clamps to the frame ends and have a think alloy heat sink that wraps around the cells with thermal compound top and bottom (thin sides of each cell) and a thermal pad between each cell pack. The cells will then be held together with four long rods to create a bit of compression between each (the thermal pad at the back of each cell pack is also to take a bit of the cell expansion which is max 8% at 20% degradation. A bit over a mm). A thermal pad will sit in the bottom of the battery enclosure with a chill plate attached to the underside (keeping liquids away from the cells).
I've sent off the model of the end frames to be CNC machined out of acetal (I would need 62) and the cheapest quote I've got so far is a bit over $30ea. Which is um, a little expensive. So I need to rethink how I assemble these packs. After talking with Chris Jones, re-reading his technical battery build posts and watching his video (https://www.youtube.com/watch?v=LcLOT0m67Bo) I think I'm going to mock up some similar designs and see how that pans out.
That design of battery certainly looks like it would be cheaper and much lighter. My current design has each cell pack weighing 4.5kg while the cells only weigh ~940g each. That's almost an extra KG for each pack...
The Jaguar i-Pace modules we're a perfect size, but rather expensive and halfway across the world. I managed to find 100 LG E63 cells at a really good price and those topped off with a further 24 found on eBay gives me a 31s4p pack @ 111v 260A pack. An extra ~20A (or more given that the Jag modules were second hand and these are new) so that's just shy of a 30kWh battery pack instead of 26kWh. Should make for an easy 200km+ range.
I've been working on a preliminary design based on the Chevy Bolt EV battery (https://www.youtube.com/watch?v=ssU2mjiNi_Q). The idea is to use acetal blocks at either end of each cell pack. The bolt via clamps to the frame ends and have a think alloy heat sink that wraps around the cells with thermal compound top and bottom (thin sides of each cell) and a thermal pad between each cell pack. The cells will then be held together with four long rods to create a bit of compression between each (the thermal pad at the back of each cell pack is also to take a bit of the cell expansion which is max 8% at 20% degradation. A bit over a mm). A thermal pad will sit in the bottom of the battery enclosure with a chill plate attached to the underside (keeping liquids away from the cells).
I've sent off the model of the end frames to be CNC machined out of acetal (I would need 62) and the cheapest quote I've got so far is a bit over $30ea. Which is um, a little expensive. So I need to rethink how I assemble these packs. After talking with Chris Jones, re-reading his technical battery build posts and watching his video (https://www.youtube.com/watch?v=LcLOT0m67Bo) I think I'm going to mock up some similar designs and see how that pans out.
That design of battery certainly looks like it would be cheaper and much lighter. My current design has each cell pack weighing 4.5kg while the cells only weigh ~940g each. That's almost an extra KG for each pack...
- OzSpider
- Groupie
- Posts: 132
- Joined: Mon, 14 Dec 2020, 10:02
- Real Name: Derek
- Location: Brisbane, Australia
Re: Luke's 2001 NB MX-5
@necrogt4 what does the acetal block look like? Can you post an image of the model?
Is it possible to get it 3D printed in nylon or PET instead?
Come to think of it, when selecting a plastic for pack construction, what are you looking for in a plastic?
Strength, temp resilience, conductivity (lack of)...?
Is it possible to get it 3D printed in nylon or PET instead?
Come to think of it, when selecting a plastic for pack construction, what are you looking for in a plastic?
Strength, temp resilience, conductivity (lack of)...?
-
- Senior Member
- Posts: 826
- Joined: Mon, 17 Sep 2018, 16:30
- Real Name: Francisco
- Location: Brisbane, Strathpine
Re: Luke's 2001 NB MX-5
If you are going to make more than 2 or 3 it will probably be better to get them CNC machined. You will get a much better finish, the material strength will be better and you can use any material you like and it may turn out to be cheaper too.
- OzSpider
- Groupie
- Posts: 132
- Joined: Mon, 14 Dec 2020, 10:02
- Real Name: Derek
- Location: Brisbane, Australia
Re: Luke's 2001 NB MX-5
Actually @francisco.shi cost is where 3D printing can easily compete. $400 buys an acceptable Chinese made 3D printer and $30/kg for filament with no wastage.
-
- Senior Member
- Posts: 826
- Joined: Mon, 17 Sep 2018, 16:30
- Real Name: Francisco
- Location: Brisbane, Strathpine
Re: Luke's 2001 NB MX-5
$400 will get you 30kg of machined plastic of your choice. No need to spend time printing and the parts look better, are structurally better and can be made out of materials that do not melt or deform under pressure (like phenolic type materials)
3D printing is great for some circumstances but CNC machining is very fast and leaves a very nice and acurate finish and the parts are stronger, they can have press fit inserts and threads.
CNC machining will give you 0.1mm accuracy with very smooth finish.
For simple parts like the ones you are making as long as they can be cut can be made in a few minutes by a CNC and there is no post processing after the part comes out of the machine.
I have had plastic parts 3D printed and by the time I polish them to remove the lines I may as well have made them out of solid and I can never get tight enough tolerances and flat enough parts compared to machined parts.
Machined parts just look really nice.
3D printing is great for some circumstances but CNC machining is very fast and leaves a very nice and acurate finish and the parts are stronger, they can have press fit inserts and threads.
CNC machining will give you 0.1mm accuracy with very smooth finish.
For simple parts like the ones you are making as long as they can be cut can be made in a few minutes by a CNC and there is no post processing after the part comes out of the machine.
I have had plastic parts 3D printed and by the time I polish them to remove the lines I may as well have made them out of solid and I can never get tight enough tolerances and flat enough parts compared to machined parts.
Machined parts just look really nice.
-
- Senior Member
- Posts: 826
- Joined: Mon, 17 Sep 2018, 16:30
- Real Name: Francisco
- Location: Brisbane, Strathpine
Re: Luke's 2001 NB MX-5
I had to make 10.
Material was $20 total processing time was about 2hrs.
Not really sure of the weight but I am pretty sure you can not make a 3D printed one in 2hrs.
I assume you will need 30 or so of the black blocks for your pack?
- OzSpider
- Groupie
- Posts: 132
- Joined: Mon, 14 Dec 2020, 10:02
- Real Name: Derek
- Location: Brisbane, Australia
Re: Luke's 2001 NB MX-5
@francisco.shi The OP stated the best quote he could get was $30 ea and he needs 62 of them. Maybe you could pass on the details for your machine shop that can make these for a few hundred bucks.
Not disagreeing with of your statements about quality, materials etc, just cost. When the job is done you get to keep the printer as well. . Unlikely the machine shop will let you keep the CNC mill....
For low stress, low temp applications with complex shapes, 3D printing can compete.
Not disagreeing with of your statements about quality, materials etc, just cost. When the job is done you get to keep the printer as well. . Unlikely the machine shop will let you keep the CNC mill....
For low stress, low temp applications with complex shapes, 3D printing can compete.
-
- Senior Member
- Posts: 826
- Joined: Mon, 17 Sep 2018, 16:30
- Real Name: Francisco
- Location: Brisbane, Strathpine
Re: Luke's 2001 NB MX-5
If you send me a 3D model I can get you a quote.
Re: Luke's 2001 NB MX-5
I like acetal for is strength, ease of machining and the fact it can take a thread (not that this model has any). But what this model is meant to take is quite a bit of compression force. Oh, and the fact it is not thermally or electrically conductive also helps.OzSpider wrote: ↑Sat, 30 Jan 2021, 18:41 @necrogt4 what does the acetal block look like? Can you post an image of the model?
Is it possible to get it 3D printed in nylon or PET instead?
Come to think of it, when selecting a plastic for pack construction, what are you looking for in a plastic?
Strength, temp resilience, conductivity (lack of)...?
FWIW all I did compare the cost of 3D printing for prototypes and was able to get the cost down to around $16 or the CNC machining down to low $20 if using HDPE instead of acetal. It's also about 2/3's of the weight.
That said, I think the next iteration of design for the pack using PCB's for busbars (based on Chris Jones's method) is going to win out here. I've got some basics together and so far it's 10% lighter and that is what a shorter and thicker busbar. I still need to factor in compression, framing etc but it's promising so far. I'll put up some more complete designs shortly.
See below for some more renders of the model. And the next iteration, this has the battery tabs coming through the PCB and they're then bolted directly to straight busbars:
Re: Luke's 2001 NB MX-5
Here's the next iteration of design using the PCB cell connection method. It saves about 70mm in width being 360mm vs 430mm and is roughly the same length at 640mm.
I was with the first design planning on using a completely aluminium box but this one is a little more complicated in terms of materials and fixings but solves the issue of the walls of the case becoming a heat sink and soaking up both roads and motor heat (as one module will be sitting above the motor). The base of this module is an aluminium chill plate (yet to be completely modelled). The reinforced sides are acetal, the grey sides HDPE and the top will either be HDPE or polycarbonate which is about 15% heavier so not sure on that. Using composites instead of metal will both save weight and also insulate the batteries. While the metal base will cool them. Win win, I think.
The reinforced sides of the box will apply pressure to the cells but are fixed in relative position to one another so I'll probably need some spacers to compress the cells into the fixed space. The bottom rails inside the box will be welded in along with the square blocks which will both apply compression to the cells and help fix the sides to the base. The L brackets above the cells will prevent any vertical movement. I'll likely need something to prevent any horizontal movement too..
The BMS, contactor and fuses will sit on an insulated tray between the L brackets above the cells. Connections will come through the acetal side panel. I still need to models these in.
I was with the first design planning on using a completely aluminium box but this one is a little more complicated in terms of materials and fixings but solves the issue of the walls of the case becoming a heat sink and soaking up both roads and motor heat (as one module will be sitting above the motor). The base of this module is an aluminium chill plate (yet to be completely modelled). The reinforced sides are acetal, the grey sides HDPE and the top will either be HDPE or polycarbonate which is about 15% heavier so not sure on that. Using composites instead of metal will both save weight and also insulate the batteries. While the metal base will cool them. Win win, I think.
The reinforced sides of the box will apply pressure to the cells but are fixed in relative position to one another so I'll probably need some spacers to compress the cells into the fixed space. The bottom rails inside the box will be welded in along with the square blocks which will both apply compression to the cells and help fix the sides to the base. The L brackets above the cells will prevent any vertical movement. I'll likely need something to prevent any horizontal movement too..
The BMS, contactor and fuses will sit on an insulated tray between the L brackets above the cells. Connections will come through the acetal side panel. I still need to models these in.
Last edited by necrogt4 on Wed, 03 Feb 2021, 08:38, edited 1 time in total.
Re: Luke's 2001 NB MX-5
The new cell pack design does away with the large brackets either side of the cells in favour of a simple PCB busbar. What you're seeing here is half of the PCB and busbar. The busbar will be double the length to the left. The PCB in the opposite direction. This way two cell packs are attached via a PCB, with no conductivity between. On each side the PCBs will be connected via a busbar. This way the PCBs and busbars form rigid peice along each side of the cells.
The aluminium frame acts as a heatsink.
I'm not sure I need the red thermal pad if the cells are compressed together, if I ditch this it will save ~2-3mm per cell pack which does add up to a few cm per module which helps in such tight spaces in an MX-5.
The tabs on each cell are folded inwards then bolted to the PCB and busbar.
The aluminium frame acts as a heatsink.
I'm not sure I need the red thermal pad if the cells are compressed together, if I ditch this it will save ~2-3mm per cell pack which does add up to a few cm per module which helps in such tight spaces in an MX-5.
The tabs on each cell are folded inwards then bolted to the PCB and busbar.
Re: Luke's 2001 NB MX-5
Wow. I have been thinking of an EV. Your build is awesome. One part of me wants to rush out and pull the ICE out of my Benz the other part is trembling in the corner. Is there are there any good resources.
I sort of stumbled on the Hyper 9 via a gear reduction mounted to the rear diff of my Mercedes 190E.
Lost for words. You are brilliant.
Dan
I sort of stumbled on the Hyper 9 via a gear reduction mounted to the rear diff of my Mercedes 190E.
Lost for words. You are brilliant.
Dan
Re: Luke's 2001 NB MX-5
Thanks for the kind words Dan.
In terms of good resources, this whole forum is one to be honest. A couple of good and very lengthy threads are Francisco's Electric Pajero -> https://forums.aeva.asn.au/viewtopic.php?f=33&t=5721 and Prelude conversion project -> https://forums.aeva.asn.au/viewtopic.php?f=18&t=5502. http://evalbum.com/ is also pretty good for finding specific cars and sometimes the entries have decent blogs or forum posts linked in them.
If there's something that you don't know or can't seem to find a post in the technical discussion section here will usually turn up the information too.
In terms of good resources, this whole forum is one to be honest. A couple of good and very lengthy threads are Francisco's Electric Pajero -> https://forums.aeva.asn.au/viewtopic.php?f=33&t=5721 and Prelude conversion project -> https://forums.aeva.asn.au/viewtopic.php?f=18&t=5502. http://evalbum.com/ is also pretty good for finding specific cars and sometimes the entries have decent blogs or forum posts linked in them.
If there's something that you don't know or can't seem to find a post in the technical discussion section here will usually turn up the information too.
Re: Luke's 2001 NB MX-5
Among other things I've finalized my battery box designs. There will be three separate boxes. The largest will sit in the engine bay with 12 cell packs (4 cells per pack) and will be slightly taller by 20mm to house the main contactor. It also holds the Zeva EVMS v3, BMS unit, current sensor and fuse. The other two will be the same size as each other, house 10 cell packs with one sitting in the fuel tank cavity and one in the boot. All boxes will have chill plate cooling, more on that later.
Given the max current draw of 750A I decided to use custom connectors for the battery cables capable of that current rating as nothing with a sensible price exists to carry that current except perhaps cable glands but I want these boxes to be easily removed. The CANBus and other wises go through the two smaller plugs and there's also a charge cable connection too.
The main box: The rear box: From behind showing the busbars: And a better view of the connectors on the main box. I'll be using tapered head torx barrel nuts for the top two clamping bolts and tapered head torx bolts for the bottom two and the two bolts either side of the connectors. A little pricey but they will sit flush and offer higher torque than say hex nuts.
Given the max current draw of 750A I decided to use custom connectors for the battery cables capable of that current rating as nothing with a sensible price exists to carry that current except perhaps cable glands but I want these boxes to be easily removed. The CANBus and other wises go through the two smaller plugs and there's also a charge cable connection too.
The main box: The rear box: From behind showing the busbars: And a better view of the connectors on the main box. I'll be using tapered head torx barrel nuts for the top two clamping bolts and tapered head torx bolts for the bottom two and the two bolts either side of the connectors. A little pricey but they will sit flush and offer higher torque than say hex nuts.
Re: Luke's 2001 NB MX-5
So these arrived recently and I've been busy with my soldering iron.
In retrospect if I was to do this again I think I'd have the PCB manufacturer do it for me as there seems to be plenty of options for that kind of thing. So, first things first, I needed a jig. There was no way I'd get 256 4mm nuts perfectly aligned with the 4mm holes (yes I should have got for 4.5mm holes). So with a bit of grinding I made this: Which aligns the PCB's just right like so: Each board took about 10 minutes and I ended up running out of solder half way through. The Jaycar solder I bought (which was a little pricey) was sooo much easier to use and looks much better. Not that conductivity is important for this, just clamping force.
And the result after three nights watching trashy TV and chipping away at the PCBs:
Once they arrived I did a little bit of searching to find nickel plated nuts that I could actually solder to the boards. My local bolt shop only had stainless and Bunnings was no good either. But Altronics has quite a selection of different sizes and I made a drive down and bought several hundred nuts and bolts. I needed 256 nuts and bolts. Yep, I had to solder 256 nuts onto the boards.In retrospect if I was to do this again I think I'd have the PCB manufacturer do it for me as there seems to be plenty of options for that kind of thing. So, first things first, I needed a jig. There was no way I'd get 256 4mm nuts perfectly aligned with the 4mm holes (yes I should have got for 4.5mm holes). So with a bit of grinding I made this: Which aligns the PCB's just right like so: Each board took about 10 minutes and I ended up running out of solder half way through. The Jaycar solder I bought (which was a little pricey) was sooo much easier to use and looks much better. Not that conductivity is important for this, just clamping force.
And the result after three nights watching trashy TV and chipping away at the PCBs:
-
- Senior Member
- Posts: 826
- Joined: Mon, 17 Sep 2018, 16:30
- Real Name: Francisco
- Location: Brisbane, Strathpine
Re: Luke's 2001 NB MX-5
Should have mentioned this before. I have nuts designed to solder on PCBs.
They have a step to locate them into the hole. Then you use a solder paste stencil to apply solder paste place the nuts in the hole and put in the oven to reflow the solder.
It would have taken about 2min per board.
They have a step to locate them into the hole. Then you use a solder paste stencil to apply solder paste place the nuts in the hole and put in the oven to reflow the solder.
It would have taken about 2min per board.
Re: Luke's 2001 NB MX-5
LOL, yeah hindsight is quite the bitch. I also came across a catalog of PCB nuts after the boards arrived.
Well, this is the process of incremental improvement and learning.
At least my soldering skills have increase quite significantly
Well, this is the process of incremental improvement and learning.
At least my soldering skills have increase quite significantly