Weber and Coulomb's MX-5

Post up a thread for your EV. Progress pics, description and assorted alliteration
User avatar
weber
Site Admin
Posts: 2936
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Weber and Coulomb's MX-5

Post by weber »

Now I'd like to tell you about our efforts to eliminate problem number
3. A torsional body oscillation (shudder).

This shudder happens, at something like 10 Hz, when one rear wheel hits a bump or pothole. Not sure if it's only rear wheels, but it seems that way. Newton (Jeff Owen) and I both did a lot of web research on this problem and it sounded like what the MX-5 community call cowl-shake, which can sometimes be excited merely by travelling at a particular speed, where a less-than-perfectly-balanced wheel might generate the appropriate resonant frequency.

The theory is that, due to the relative flexibility of the central part of the chassis, being a convertible, without a rigid roof, it twists easily. And I supposed that I had made things worse by putting cells out far from the roll axis and thereby increased the roll-moment-of-inertia at both ends of the car, so front and rear were like two dumbells with a torsional spring between them, and no significant damping.

The most popular solution was to stiffen up the central section of the car with this pair of top-hat chassis rail reinforcements with a butterfly brace between them, from Flyin' Miata in the US. This is bolted underneath with many bolts through the cabin floor under the carpet.

Image

This would have cost about AU$800 landed in Australia, and it appeared to be compromised by having to fit around an exhaust system and having to be relatively light and in pieces for shipping. So Newton and I agonised for many days over the best way to build something ourselves and where to attach it. One idea is to effectively close the transmission tunnel to make it a rigid tube between the fore and aft bulkheads.

As you can see below, I did some pretty sophisticated FEA on a model of the open tunnel with bulkheads, to see how it moved when twisted, and tried it with braces in various positions.

Image

FEA? Fibrous Element Analysis, i.e. paper, string, staples and drinking straws. Image [Edit: What you see is that when you twist it, one side of the tunnel goes forward while the other goes back, and the middle of the tunnel closes up.]

For quite some time I fooled myself into thinking that threaded rod on the diagonals was a good idea, preloaded in tension, like diagonal wind braces between tall house stumps. Newton knew in his guts (or was it his bones) that this was a bad idea, but couldn't come up with an explanation that dissuaded me. But when he said "So why doesn't anyone else do it on cars?" I went searching and found a few chassis braces that did have threaded preload adjustment. But they were always to be preloaded in compression.

So I went back to my FEA models (including sticky-tape this time) and soon realised that diagonal braces in tension cause twisting (unless the rectangle is constrained to be planar by other means, as it is in the house case). You might think, as I did, that two crossed diagonals would cancel each other out in that regard, but they don't. They just give two stable twisted states, with the ability to pop from one to the other if enough force is applied.

In the end I went to Bunnings and bought two 2 metre lengths of 25 x 25 x 1.6 mm square steel tube for $33 and cut and welded it to make the 6-ended star you can see below.

Image

MX-5s with an exhaust system can only dream of such a simple bracing arrangement. At the front the diagonal struts bolt with some existing bolts that attach the front subframe to the chassis. At the rear, I found points that were spaced the same distance apart (600 mm) and had three layers of metal overlapping. They were 1100 mm from the front bolts. For the lateral struts, in the middle, I made the holes 540 mm apart and they were 4 mm above the plane of the crossed diagonals. They came up into a top-hat section above the floor, to which the front of the seats are bolted. The only complexity came with having to allow a hydraulic brake-line to pass through one strut, which required reinforcement around the cutaway. On a conventional MX-5 there would have been two fuel-lines there as well.

Image

Of course you cut and drill these things precisely to within +-0.5 mm then you weld them up and no matter how evenly you do the welding the cooling welds bend the tubes in all sorts of directions. Some holes end up being 10 mm from where they are supposed to be and it's no longer anywhere near planar. But it's nothing that can't be fixed by some careful overbending in the other direction. When chocking it up off the concrete and jumping on it doesn't work, you use a jack and a prop under something heavy, like your house.

Image

Here's where the brake-line had to pass thru, and not straight thru either, but going around a bend.

Image

Eventually it's all back in spec and you paint it (penetrol and grey primer on the cuts and welds, flat black all over), mark and drill the 10 mm holes in the floor of the car from underneath, bolt it up and weld the nuts on top of the floor and paint the welds (penetrol and red oxide primer). The nuts in the middle were also welded to both sides of the vertical part of the top hat section, as you can see below. The M10x1.25 (fine) high tensile (class 8.8) bolts have to be very tight as we are relying on friction, not the accuracy of the holes, to stop the ends of the braces moving relative to the floor.

Image    Image

Image

So how did it go? It didn't do a damn thing for the shudder. No reduction whatsoever! So I'm sure glad I only spent $48 and some hours on it, instead of $800. [Edit: I eventually got it to work. See viewtopic.php?title=weber-and-coulombs- ... 462#p51462] Although Newton has a theory that with the Flyin' Miata brace, most of the work is done by the chassis rail reinforcements resisting bending, rather than the butterfly closing the tunnel and resisting twisting.

But we both agree that another possibility is that the shudder is happening not between the front and rear bulkheads but between the rearmost battery box and the rear bulkhead. The next experiment will be to drop that under-boot battery box out and go for a bumpy drive (on one half-pack). But not today. I have to leave her working and looking pretty for our guest speakers at the Brisbane AEVA meeting tonight - a delegation from Sinopoly Battery Limited.

There is some visible effect of the brace. Previously, as a passenger, I had witnessed the back of my door moving up and down by about 3 millimetres relative to the jamb as we drove. Now it only moves about a millimetre.
Last edited by weber on Tue, 17 Jun 2014, 10:27, edited 1 time in total.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).
User avatar
weber
Site Admin
Posts: 2936
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Weber and Coulomb's MX-5

Post by weber »

I had a surprise helper last Friday. Bladecar. He let me take his white Blade Electron Mk 6 for a drive, It has the T4 motor and controller. It was very smooth and responsive. Quite the polished product.

He also helped me pull the under-boot battery box out of the MX-5 in record time. He was a mechanic in a former life. The verdict on its effect on the torsional shudder?: No detectable effect. It still shuddered just the same on the same bumps.

Given that taking the under-boot battery box out made no detectable difference to the shudder, I can now decide that we're keeping the 36 cells in it. I need to put it back in again soon, since I can't drive very far or use very much power with it out. But I decided to take the opportunity while it's out, to reduce its size. It was built to take 60 cells in 3 rows, but because of weight distribution issues the rearmost row is completely empty and no longer needed.

I don't like being able to see the bottom of the rack for the rear row, showing below the bumper. It's indecent for a refined lady to let her slip show like that. As shown in the photos below, I have cut away the rear row from the base after replacing it with two struts (made of 30x30x3 angle) that bridge from the existing rear mounting brackets to the back of what was the middle row. I will leave the lid of the box as it is.

[Edit: I should explain that, in the photos below, the battery box is rotated so that its rear end is facing upwards.]

Image   Image

Image   Image

Image
Last edited by weber on Tue, 03 Jun 2014, 19:44, edited 1 time in total.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).
Damnthematrix
Noobie
Posts: 9
Joined: Sat, 31 Mar 2012, 14:55
Real Name: Mike Stasse
Location: Cooran Qld
Contact:

Weber and Coulomb's MX-5

Post by Damnthematrix »

In a previous life, I owned a Triumph TR4. Whatever structural problems your MX5 might have, I am certain would not compare to what we used to call "scuttle shake"!

Image

My TR eventually sported what was called a Surrey Top, an optional extra glass rear screen on which either a soft roof or hard top could be fitted to bridge the gap to the windscreen in inclement weather. I never owned a hard top, but we once drove the TR to the Snowies for a holiday and I borrowed one..... the resulting increase in torsional stiffness had to be experienced to be believed, it was like a different car!

Hard tops are available for MX5s...... Personally, I could never again drive an open car, the sun is so much harder now than it was when I was 18. Maybe I'm just getting 'soft'!

Mike
Last edited by Damnthematrix on Wed, 04 Jun 2014, 01:47, edited 1 time in total.
User avatar
offgridQLD
Senior Member
Posts: 1827
Joined: Tue, 23 Jul 2013, 16:05
Real Name: Kurt
Location: Fleurieu Peninsula, SA

Weber and Coulomb's MX-5

Post by offgridQLD »

I have to agree with the above. Personally I think the MX5 looks great with the hard top fitted.

+ it gives you some place to put solar panels so you can drive around for free all day....(joke)

Kurt
User avatar
weber
Site Admin
Posts: 2936
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Weber and Coulomb's MX-5

Post by weber »

Funny that, of the 10 battery boxes in the MX-5, the one we built first is the last to be finished. Here's a stereo pair of the finished product.
How to view cross eyed 3D images.

Image
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).
User avatar
weber
Site Admin
Posts: 2936
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Weber and Coulomb's MX-5

Post by weber »

I never thought I'd be complaining that a road had been sealed. Image I've just been for a very pleasant drive out to Lake Manchester in the MX-5. Google Maps: Bardon to Lake Manchester

I was intending to test the effectiveness of an upgrade to my DIY butterfly brace, on a bumpy dirt road. But when I passed this amusing sign

Image

I instead hit fresh bitumen extending beyond it as far as eye could see. In fact it was so fresh I could count the wheel tracks of the cars that had gone over it before me, and the tar between the wheel tracks was still glistening. You could hear the stones sticking to the tyres and being thrown off inside the mudguards as I crawled the MX-5 along it. Yuk!

I eventually got my wish as it turned back into gravel and I accelerated away. Potholes, corrugations, patches of loose gravel, uphill and downhill bends, yahoo! Great stuff. The verdict? ... I think it's Good Enough (TM). Put it this way: If I'd never driven a Nissan Leaf (whose middle name is "smooooooth"), I'd probably declare the electric MX-5 now much the same in the shudder and shake department as many other cars I've driven. Just occasionally you hit a certain shaped bump in a certain way and get a shudder, but it decays in about half a second and you're back to sailing over the tops of the corrugations (or at least that's what it feels like).

I've concluded that the worsening of what is a known problem with convertibles, was not particularly caused by the increase in roll-moment-of-inertia due to putting batteries out wide, but was more a consequence of using much stiffer springs (and hence shockies) to deal with the 30% extra weight. I could have got away with less stiff springs if I had had them custom made to be longer than they would be on a conventional MX-5. But it can only improve when I upgrade to wider tyres and can therefore run them at lower pressures.

I think you're right Damnthematrix and OffgridQLD, the consensus on the web seems to be that, even on their first pass, the Mazda engineers managed to make the MX-5 stiffer than most of the British sports cars it was based on. I agree that a hardtop would probably make a big improvement, but I really enjoy being able to easily put the top down, particularly in the beautiful weather we've been having in Brisbane lately. Janelle and I had a lovely drive out to Cedar Creek, via Samford, weekend before last. And I'm not sure I want to spend $1000 on a hardtop.

There are two more low cost options for further stiffening. These are:
1. Reinforcing the chassis rails under the floor by bolting over them some 75 x 50 x 20 top hat sections folded from 2 mm sheet.
2. Welding the seams around the doors and windscreen frame (under the weather strips), and related gussets, to make up for spot welds that may be getting a little tired after 24 years.

But I think I'll leave this alone for now, and move onto some of the other finishing touches required. I've got an exhaust pipe to install Image, and the voice synthesizer just arrived Image. But I'll explain the changes to the butterfly brace, with some photos, in my next post.
Last edited by weber on Thu, 12 Jun 2014, 12:41, edited 1 time in total.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).
User avatar
Johny
Senior Member
Posts: 3749
Joined: Mon, 23 Jun 2008, 16:26
Real Name: John Wright
Location: Melbourne
Contact:

Weber and Coulomb's MX-5

Post by Johny »

Great to hear. That problem has been a pain for you.
User avatar
weber
Site Admin
Posts: 2936
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Weber and Coulomb's MX-5

Post by weber »

Thanks Johny.

Below I describe the changes I made to the chassis brace under the transmission tunnel of the MX-5. It was initially constructed as a 6 pointed star (two diagonals and one lateral) which you can see here:
viewtopic.php?title=weber-and-coulombs- ... 0&start=61

From the start, Newton (Jeff Owen) had recommended including front and rear laterals as well, but I noted that these were not included in the Flyin'Miata butterfly brace, which is shown at the start of the post linked above. And my paper model hadn't indicated much benefit from them, so I thought I'd avoid the extra work.

However when Mazda themselves decided to add tunnel bracing to a later model of MX-5, here's what they did.

Image

As you can see, there is very little in the way of diagonals but lots of laterals. The lack of diagonals might only be because the exhaust system makes them difficult. But in any case, it seems unlikely they would include those laterals at the front and rear of the tunnel if they didn't have a benefit. So, given that I had declared the star brace a failure, I decided to try adding these extra laterals, turning the star into a butterfly, but at 90 degrees to the Flyin' Miata butterfly.

After cutting and test-fitting the two new laterals, I finally pulled the star brace out of the car and found something very interesting, which was quite sufficient to explain why it hadn't been working. I found that at the mating surfaces of all 6 mounting points the paint had been rubbed off the brace and the underbody sealer and paint rubbed off the floor. This was despite me having re-torqued all the bolts after driving it for a while, as suggested by Newton.

Image   Image

As mentioned in the earlier post, this brace totally depends on the Class 8.8 high tensile M10 bolts and nuts (torqued to 50 Nm) producing enough friction between the brace and the floor to prevent relative movement. The degree of wear at all 6 points in such a short time, indicated that there was a lot of movement, and so if this could be resisted there was a chance the brace would have a significant stiffening effect. But clearly the paint was just lubricating the joints.

I did some searching on how to increase the friction between two metal surfaces. It turns out that when the two surfaces are of the same metal, the best thing to do is polish them smooth and leave them bare. So I got to work with a scraper and then sandpaper on both the floor and the brace.

But I had already gone to considerable trouble making the front and rear laterals, so I decided to weld them in anyway. So I don't know if the star brace alone might have been just as good, with just the paint removed from the mating surfaces. In any case, the complete "Weber & Newton" butterfly brace weighs only 4.7 kg. The Flyin' Miata brace weighs 8.6 kg (and that's not including the required frame rails, which add another 5.9 kg).

The difficulties in making the front lateral strut were avoiding the brakeline and working around the front subframe whose bolts are reused by the brace. It ended up positioned 12 mm lower than the other struts, but still higher than other parts of the car in that vicinity.

The difficulty in making the rear lateral strut was that the PPF (Power Plant Frame), which is the aluminium truss beam that connects the gearbox to the diff, is 10 mm lower than the floor there, requiring a cutaway in the 25 x 25 x 1.6 mm square tube the struts are made of. I gave it 7 mm of clearance to allow for the diff and motor bouncing up and down on their mounts, and so I cut away all but 8 mm of the square tube. This is because I found a scrap of 1/4" (6.4 mm) bar left over from Augusta Marie which, when welded against the inside of the 1.6 mm tube wall, made up 8 x 25 mm of solid steel to hopefully match the strength of the uncut tube.

Image

In the photo below, you can see that I made a nice hot cup of tea, hooked it up to certain equipment, and made all the molecules in Mexy's undergarments leap simultaneously one metre to the right. [HHGTTG reference].

Image

As mentioned in my previous post about the test drive, this chassis brace works well enough that I can now move onto other things. Yay!
Last edited by weber on Mon, 16 Jun 2014, 12:58, edited 1 time in total.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).
User avatar
Adverse Effects
Senior Member
Posts: 1172
Joined: Sat, 01 Jan 2011, 03:30
Real Name: Adverse Effects
Location: Brisbane

Weber and Coulomb's MX-5

Post by Adverse Effects »

i have an idea about making the bracing lock more rigidly with the under body

do you know about dimpling tools?

panel dimpling tool

if you did that to both the brace and the under body i think they would "lock" to each other
User avatar
weber
Site Admin
Posts: 2936
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Weber and Coulomb's MX-5

Post by weber »

Adverse Effects wrote: i have an idea about making the bracing lock more rigidly with the under body

do you know about dimpling tools? ...

Thanks Adverse. I didn't know about dimpling tools. That's a great idea. Except that by the time I'd torqued the bolts up to 50 Nm, the dimples would be flattened out again. It seems such dimples are intended for countersunk-head screws and rivets in low stress applications. So what I need is for the "dimpling tool" to stay in place permanently. In other words I need concave and convex washers. Do such things exist? I asked google. Yes they do! Also called "spherical washers". They are designed for a different purpose -- bolts that are not quite at right angles to the surface -- but they could be used for this locating purpose too.

Image
Click on the image to go to the seller's page.

Unfortunately I've already welded the nuts in place above the floor, so that would make installing concave washers there difficult. But I may still get a benefit from using convex washers on the bolts against the six brace-ends from underneath. Now to find somewhere I can buy such a small quantity.

In case anyone ever tries to duplicate this brace, or do something similar, I should mention that the bolts I used are somewhat special, in that, although they are M10, they only have a 14 mm head (flanged). These are common enough on vehicles, but the ones you can readily buy have a 16 or 17 mm head (unflanged). A 14 mm socket just fits between the sides of the 25 mm square tube, so the tube can simply be cut at 45 degrees to give access, without significantly weakening the tube, as you can see in some of the photos above.

I still don't know where to buy small quantities of such bolts as I reused those from the front tie-downs of the MX-5, which were removed to save weight and to repurpose their holes to mount battery boxes. These have an "8" on their head (property class 8). Don't be fooled by the ones that are used to hold various brackets on the old internal combustion engine, which appear to have a "9" on them. After stripping one and having to clean the remains of its thread out of the class 8 nut with a tap, I can tell you it is not a "9" but a "6". Image
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).
User avatar
Adverse Effects
Senior Member
Posts: 1172
Joined: Sat, 01 Jan 2011, 03:30
Real Name: Adverse Effects
Location: Brisbane

Weber and Coulomb's MX-5

Post by Adverse Effects »

not the correct size BUT something like this could help

Bolt, Lug 16mm x 1.5P x 23mm Tapered Head

User avatar
weber
Site Admin
Posts: 2936
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Weber and Coulomb's MX-5

Post by weber »

I fear the cone on the wheel bolt head meets the surface at too steep an angle, about 45 degrees, while the convex washer meets it at about 20 degrees. Unless a significant area of the thin (1.6 mm) tube wall is clamped, it will simply deform around the bolt as the chassis flexes, elongating the hole and failing to resist the movement. In fact I think there is a chance of this happening even with the convex washer, if it is not used with a matching concave washer, so I won't be using them this time.

But the concave/convex washer combination looks like a winner for future DIY chassis braces of this type. Thanks for the idea.

I should mention that I sprayed inside all the tubes of the brace with Fishoilene to resist corrosion. I put masking tape on all the bare steel mounting surfaces before I did so and will clean them all with acetone before bolting the brace back in.

After it has been driven a bit and retorqued I will apply some underbody sealer around each mounting point to exclude air and water from any bare metal that remains exposed.
Last edited by weber on Tue, 17 Jun 2014, 09:03, edited 1 time in total.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).
User avatar
weber
Site Admin
Posts: 2936
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Weber and Coulomb's MX-5

Post by weber »

I've made a few other improvements to the MX-5 over the past 3 weeks:

1. Debounce. I implemented debounce in software for all the inputs to the EVCU (Electric Vehicle Control Unit) (one of Tritium's "EV Driver Controls" units). This debounce filter also functions as noise spike rejection. If a change only appears for one 10 ms sample, it is ignored. It's kind of surprising this wasn't implemented in the standard software, since it boils down to one line of code:
outputs = (inputs & prev_inputs) | ((inputs | prev_inputs) & prev_outputs);
This has solved a problem where switch inputs that were normally open circuit (with a 10k pullup or pulldown) would pick up motor EM noise which could be interpreted as a signal, from the crash switch or a charge-cable-detect switch, and cause us to drop out of drive mode. This typically happened just after starting off from the traffic lights, with other cars having just started moving behind you. Not good. The workaround until now had been to disable the crash switch and charge-cable-detect switch. Not good either. All fixed now.

2. Auto 12 V battery charge. I added EVCU software to automatically maintain the 12 V battery with occasional charges as required (via the DC-DCs off the traction pack). It turns on when the battery falls below 12.5 V and turns off when it reaches 14.2 V. This occurs even when the key is off and we're not charging the traction pack (but not if the crash switch is operated).

This gets around the fact that our DC-DC converters cannot keep up with the maximum 12 volt load (power steering, headlights and demister fan, all at once). The 12 V battery can make up the shortfall for an hour or so, but it's no good if you come back from one trip with the 12 V battery nearly empty and expect to use it again later without charging. This change also prevents the battery from being flattened by the consumption of the EVCUs if it isn't driven for a week or two. But this does mean that now, if you want to work on the HazV wiring you must disconnect the 12 V battery, since otherwise the battery breakup contactors could come on at any time. And if you leave the headlights on they can flatten the traction battery, given long enough.

3. Accelerator cable oiled. I pulled out the pot-box, disconnected the accelerator cable, made a duct-tape funnel around the end of it, taped it to the underside of the bonnet to keep it pointing upward, primed it with some WD40, followed with chain oil (viscous and sticky), jiggled the cable back and forth and waited until it started to run out the end inside the car, with a rag under it. Then I cleaned up the mess and put it all back together. I also oiled the accelerator pedal hinge and spring, and the pot-box hinge and spring. Then I logged the pedal values at the two extremes and changed the constants in the EVCU software accordingly, so we now get 0 to 100% with negligible wasted travel (but noise clipped) at each end, and without the steppy response we used to get due to stick-slip when trying to move the pedal slowly.

4. Rear shocky damping. I increased the damping on the rear shockies by 4 clicks of the adjuster, to match the new stronger springs. So both front and rear are now set 12 clicks clockwise from the softest setting. I didn't have to remove the contactor boxes. I only had to unbolt the chargers and then loosen the bolts holding the contactor box's aluminium brackets enough to get my fingers under them. While I was at it I eliminated some possible rattles by sticking some neoprene foam to the top of each shocky adjuster (compressed against underside of contactor box) so they can't rattle in their holes or against the underside of the contactor boxes. I also put neoprene foam between each battery charger and the rear fuel-tank battery box.

Then I put it all back together and went for a test drive over the speed bumps. It's just perfect! No detectable overshoot or undershoot. Very nice!
Last edited by weber on Wed, 18 Jun 2014, 14:19, edited 1 time in total.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).
GRMarks
Groupie
Posts: 140
Joined: Sat, 13 Feb 2010, 18:14
Real Name: Glenn Marks
Location: Dandenong,Vic

Weber and Coulomb's MX-5

Post by GRMarks »

Hey weber early in your post you played with elcon chargers (now TC chargers), you opened one up and had a peek inside. Mine has just died 1/2 way through charging. Is there a fuse inside (wishful thinking thats its something simple)? I am not an electrical engineer so I may have to take it to someone to get it fixed. But if there are some simple potential problem points I could check out those things first before taking it to someone.

Any ideas?

regards

Glenn
Regards

Glenn Marks
User avatar
weber
Site Admin
Posts: 2936
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Weber and Coulomb's MX-5

Post by weber »

GRMarks wrote: Hey weber early in your post you played with elcon chargers (now TC chargers), you opened one up and had a peek inside. Mine has just died 1/2 way through charging. Is there a fuse inside (wishful thinking thats its something simple)? I am not an electrical engineer so I may have to take it to someone to get it fixed. But if there are some simple potential problem points I could check out those things first before taking it to someone.

Any ideas?   regards    Glenn

Hi Glenn,

I just checked kennybobby's TCCH Elcon 1.5kw charger schematics and yes, there is a fuse inside, on the active input from the mains. In the photo it looks like it might be soldered to the PCB.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).
User avatar
coulomb
Site Admin
Posts: 6357
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Weber and Coulomb's MX-5

Post by coulomb »

GRMarks wrote: ... early in your post you played with elcon chargers (now TC chargers), you opened one up and had a peek inside. Mine has just died 1/2 way through charging. Is there a fuse inside (wishful thinking thats its something simple)?

This seems like a charger topic, so I've started a thread here: Can I fix my TC/Elcon charger myself?. Hopefully others will add more ideas to that thread.
MG ZS EV 2021 April 2021. Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
Patching PIP-4048/5048 inverter-chargers.
If you appreciate my work, you can buy me a coffee.
7circle
Senior Member
Posts: 540
Joined: Thu, 04 Feb 2010, 02:18
Real Name: Ken
Location: Melbourne

Weber and Coulomb's MX-5

Post by 7circle »

.... lost my reply... try again but brief...

Would you like more longnitudinal strength in chassis?
Like the kits you linked to showed.
Your Butterfly Brace looks sweet. Not too heavy.

erhh .. debounce .. why not just beef up RC filter on pins?
A port pin read ... then a second port pin read ... doesn't effectively stop erradic signals on 10kohm input signals. If it gets two consective logic states doesn't mean the signal is stable.

And why are you OR masking the output with input states and etc.
If the inputs have triggered internal processes and CAN bus anouncements.
I guess I should go entertain myself with code browsing on source forge.

I guess too, that when you up the Volts with new Wavesculpter Model lower currents on DC side will lessen EMI ...or perhaps not.

Sounds like your really enjoying the drive now with tuned in rear shocks and stonger body.
User avatar
weber
Site Admin
Posts: 2936
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Weber and Coulomb's MX-5

Post by weber »

7circle wrote:Would you like more longnitudinal strength in chassis?
Like the kits you linked to showed.
Your Butterfly Brace looks sweet. Not too heavy.
Thanks. I don't think we need more longitudinal bending stiffness. I think the rocker panels handle that fairly well. They have 3 components, so they are like a tube with a vertical partition, like this "(|)". I think the main problem is torsional.
erhh .. debounce .. why not just beef up RC filter on pins?
A port pin read ... then a second port pin read ... doesn't effectively stop erradic signals on 10kohm input signals. If it gets two consective logic states doesn't mean the signal is stable.
That's true. And I fixed one earlier by just changing the pullup to 1k. And I will do that eventually to these others. But I got lazy, and it seems to be working so perhaps the noise consists only of isolated spikes (more than 20 ms apart).
And why are you OR masking the output with input states and etc.
If the inputs have triggered internal processes and CAN bus anouncements.
I guess I should go entertain myself with code browsing on source forge.
The names "inputs" and "outputs" are merely relative to the debounce filter itself. Only the filter reads the inputs and the rest of the code takes the filter's outputs as its inputs. The filter has to remember not only what the previous inputs actually were, but also what it previously passed on to the rest of the code.

But I freely admit that Boolean expression makes no sense to me in terms of ANDs and ORs. I simply drew up a truth table for all 8 states of inputs, prev_inputs and prev_outputs and decided what the outputs needed to be, then used a Karnaugh map and DeMorgan's law etc. to mimimise the number of instructions required, and that's what fell out. And it works.
I guess too, that when you up the Volts with new Wavesculpter Model lower currents on DC side will lessen EMI ...or perhaps not.
Dunno.
Sounds like your really enjoying the drive now with tuned in rear shocks and stonger body.

Oooh yeah. Image
All I want now is that 900 V WaveSculptor. But Tritium_James seems to have my email going direct to his spam bin or something.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).
User avatar
weber
Site Admin
Posts: 2936
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Weber and Coulomb's MX-5

Post by weber »

This is a bit like "Mexy -- The Lost Episodes", but Coulomb's recent post on TC/Elcon charger serial connectors made me realise I'd never posted about the optic fibre conversion I did on the second charger in early October 2013, between when we got her registered and when we took her to Canberra for the AEVA Festival.

I'll let the photos tell the story ...

Image

Image

Image

Image

Image

... except to say that's a 470R between charger serial output and fibre LED (blue wire, blue connector), a 1k5 pullup to +11 V from fibre phototransistor (red wire, black connector), and a 5k6 current limiting resistor to charger serial input. The black wire is the common 0 V.

Image
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).
User avatar
weber
Site Admin
Posts: 2936
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Weber and Coulomb's MX-5

Post by weber »

Coulomb & Weber's MX-5 -- The Stealth EV


One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).
User avatar
Adverse Effects
Senior Member
Posts: 1172
Joined: Sat, 01 Jan 2011, 03:30
Real Name: Adverse Effects
Location: Brisbane

Weber and Coulomb's MX-5

Post by Adverse Effects »

LOVE IT
User avatar
4Springs
Site Admin
Posts: 926
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Weber and Coulomb's MX-5

Post by 4Springs »

I want one!
User avatar
woody
Senior Member
Posts: 1716
Joined: Sat, 21 Jun 2008, 02:03
Real Name: Anthony Wood
Location: Mt Colah

Weber and Coulomb's MX-5

Post by woody »

At first I thought it was a video of a hamster wheel...
Planned EV: '63 Cortina using AC and LiFePO4 Battery Pack
User avatar
Johny
Senior Member
Posts: 3749
Joined: Mon, 23 Jun 2008, 16:26
Real Name: John Wright
Location: Melbourne
Contact:

Weber and Coulomb's MX-5

Post by Johny »

Ha!
User avatar
weber
Site Admin
Posts: 2936
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Weber and Coulomb's MX-5

Post by weber »

I plan to eventually take a video at night from another car following Mexy's Plasma-Exhaust (TM) Image.

According to the Star Trek wiki, the ignition of plasma exhaust will not cause damage to a vehicle provided its hull plating is polarised, and it is travelling at half-impulse. Image

Here's how the plasma exhaust went together: I started with this 80 mm USB-powered Plasma Ball from Jaycar for $20.

Image http://www.jaycar.com.au/productView.asp?ID=GE4089

Then I cut a 300 mm length of 90 mm diameter PVC stormwater pipe, available in 1 m lengths from Bunnings. I painted the inside of the pipe (and the first 90 mm of the outside) flat black. The last 210 mm of the outside I wrapped in this chrome-like self-adhesive plastic film.

Image http://www.ebay.co.uk/itm/A4-Chrome-Sil ... 1523914414?

I put the overlap at the bottom where you don't normally see it. I painted the end of the pipe with White Knight Super Chrome paint using a small brush.

After temporarily removing the electronics from the plasma ball's base, I hack-sawed the bottom level off the base so the remainder would fit inside the pipe.

I wrapped some dark grey polyethylene foam tape from Clark Rubber around the glass ball to wedge it inside the pipe and protect it from mechanical shock. I put the small gap in the foam at the bottom so water can drain through it.

Image

Image

I powered it via a $6 DC-DC converter. You can see it here, siliconed inside the base of the upside-down plasma ball.

For some bizarre historical reasons, coming from the world of radio controlled hobbies, it's not called a DC-DC but a UBEC (Ultimate Battery Elimination Circuit). I'm told "Ultimate" means it is a switching circuit, not a linear one.

I experimented with running the Plasma Ball off more than the 5 volts you get from a USB port. It got a bit brighter, but its TO-126 package transistor got hot, so I replaced it with a TO-220 device with a piece of copper busbar as a heatsink. So it's actually running off a 6 V UBEC, but I can't recommend it.

Image

You can see above, that bulk water is excluded from the front of the pipe (which of course does not connect to a muffler) by closing it off with a disk of sleeping-mat foam, cut from a sheet by filing a sharp edge on another piece of 90 mm storm-water pipe and using it like a cookie cutter. The foam disk was sprayed flat black and siliconed in place. A small notch was left at the bottom of the disk to drain any water that might enter from the rear when parked.

Image

It's mechanically mounted to two tabs that are welded to the rear battery box, with box and tabs all painted flat black. Greg Breslin (BladeCar) made the tabs for me when he was helping me a few weeks back. Thanks Greg.

There's also a third support, which is a long screw through the top rear of the pipe, up into the inside of the plastic bumper. The heads of all 3 mounting screws are painted flat black and the long one going upward has some flat-black plastic tubing over it (unshrunk heatshrink).

I didn't want to run a separate wire from the front of the car to power it, so I had to decide whether to power it off the tail lights, the brake lights, or the coil drive for the traction battery breakup contactors. In the end I went with the contactors, so Plasma is on whenever she's being driven or charged.

Image

Atomic batteries to power. Turbines to speed.

One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).
Post Reply