3D printing ECU case - cost?

[AndreyB]

What is the state of 3D printing cost? If we CAD the shape, how much does it currently cost to print something like an ECU case? Not to buy a 3D printer but order a print online?

Say 6 by 6 by 0.75 inches like ?

[puff]

if it's made of plastic then it's pointless - it won't filter any emi noise.
still too expensive. and probably much more expensive if its printed out of metal. besides, I guess it won't be that rigid comparing to conventional cases.
one could cut out the case of aluminum bars (5mm thickness, 50mm wide. for me the most interesting way would probably be to get access to milling equipment.
i didn't get what's wrong with aluminum extrusion cases? or with those taken from the junk yard?

[kb1gtt]

These guys seem cool http://www.shapeways.com/about/how_does_it_work?li=nav Lots of options, not horribly expensive either from what I recall. However also not cheap. As a ball park price, I see this metalic plastic IPhone case cost less than $10 http://www.shapeways.com/model/720311/iphone-5-bumper-case-prototype.html?modelId=720311&materialId=28 I'd bet we could get a case for between $50 and $100. Perhaps we could keep the costs low be using plate aluminium covered or enhanced with 3D parts.

[Sudo]

There are games you can play to filter EMI with a 3D printed plastic case design. One thought is to design a multi layer case that allows the user to sandwich aluminum foil in the walls of the case. Cost would be cheap to print. But I think the better point for this idea is that it will be totally DIYable.

[AndreyB]

if it's made of plastic then it's pointless - it won't filter any emi noise

I have that WBO controller under the hood - does not it have plastic case? So how comes the WBO is OK - what's the difference between an ECU and a WBO controller, why would one tolerate plastic and the other one would not?

Here is a test print for one of the boards I had in my pile:

[puff]

okay, it pretty much depends upon the density of printing and the weight of the object. When I wanted to print my tps enclosure, some service wanted 300RUB per 12g. in bulk quantities the cheapest you could get was 5-7 RUB per g (at 38RUB per USD). There are chinese guys who would cast custom enclosures out of aluminum. Probably, that is much better option? According to aliexpress, that shouldn't be any more expensive than printing it out of plastic.

[kb1gtt]

Concerns with EMI are based on if you expect to emiter radiate radio frequencies, of if you expect your board to be effected by possible near by radio frequencies. You can use plastic enclosures if you expect the RF is not a problem. If it is going to be a problem, you typically need the metal enclosure for shielding.

[abecedarian]

Concerns with EMI are based on if you expect to emiter radiate radio frequencies, of if you expect your board to be effected by possible near by radio frequencies. You can use plastic enclosures if you expect the RF is not a problem. If it is going to be a problem, you typically need the metal enclosure for shielding.

I'd argue that if you're generating injector and ignition currents within the ECU, you really need to be concerned about EMI. It'd suck to have an injector trigger an ignition event, right?
And if the 3rd/4th order harmonics keep your iPuff from sending music to your stereo...?

If you ever hope to sell a unit, with the WiFi and BT things I've heard mentioned as 'options', EMI / EMF are paramount concerns.


But what do I know? I only work in the wireless communications industry and have to deal with external interference issues all the time.

[kb1gtt]

Correct, if you expect to emit RF like an injector driver or ignition driver could do, you should be concerned with EMI. As well if you have sensative electronics you should like an iPuff, you should be concerned with EMI. I think the above WBO is dealing with signals below 1kHz, and probably has a low pass filter on it's sensitive inputs. The chances of coupling in energy below 1kHz via EMI means is not very likely and is not very likely to effect it's operation, so it's probably OK to be plastic non-protected from EMI. However I don't know what it is or what it does, so it might just be cheap, or it might not be required.

[kb1gtt]

I recently obtained this 3D printer for about $200
http://www.aliexpress.com/snapshot/7440999179.html?orderId=73668081914971

Then I obtaining this multi-color pack of ABS filament that's costs about $5 per 1 lbs of the plastic,
http://www.ebay.com/itm/PLA-ABS-filament-for-3D-printer-only-5-spool-18-colors-Box-/222075800126?hash=item33b4bfca3e:g:zj0AAOSwNSxVbfk0

Then a program like Repetier, makes the slicing and such reasonably easy, as well it provides filament length, and run times for making an item.
http://www.repetier.com/downloads/

Then plugging the numbers from the above into this web page, we get a predicted cost for a 3D case.
http://www.ic3dprinters.com/3d-printer-job-cost-calculator/

It looks like a 3D case could be made for about $25-ish. I'd kind of like to find a program similar to Repetier but one that includes the cost predictions natively . For example, Repetier claims it needs 30,000mm of filament, but I don't know if that's the 0.4mm diameter that's extruded, or the 1.75mm diameter that's feeding the extruder. I suspect there is a chance the cost is mangled as they might consider themselves to be different from each other. I'd also like to find a slicing software that allows me input the basic extruder dimensions, and validate if the G code could collide or not. The collision tools I've found so far are very primitive and basic.

Any how, it appears that materiel costs are less than $1 for a printed case.

Does anyone want to take a stab at making a 3D solid model of a case? I'd like to get more real numbers for this, but time is a resource I have in sort supply. Perhaps someone could make a 3D case model, I see FreeCAD and Sketchup are free and could make such a case.

[AndreyB]

you've lost me. $25 or $1?

[kb1gtt]

$1 for the plastic, but for electricity and time to run the equipment, it's closer to $25.

[puff]

From my undertanding, its the initial filament. Imho, the easiest way to check is to weight the case (fiy, people say, sometimes, up to 10% of filament is somehow evaporated)
I can hardly imagine how expensive your electricity is. Time to run is almost free, isn't it?

[kb1gtt]

10% you say, that seems a bit high to me, as it barely makes any odder. Perhaps that includes the bit of scrap that you need to do things like get the filament flowing, it has to print a minimum before it starts to work well. As well if you change color there is a certain amount of scrap as you purge the old color. However that's pretty minimal. Perhaps the 10% is a difference in density. The stuff I have seems it was extruded as fast as possible, you can see some fatigue cracks in the initial filament, so perhaps it's a bit less dense and after it's heated and put under mild pressure, perhaps the molecules are allows to re-arrage and get packed a bit more tightly.

The web page calculator asked for a $/hr number, I grabbed a number from my arse and put in $5/hr. I have a Kill-A-Watt EZ Power Meter so I can measure the energy used, then make a real prediction of how much the energy costs. I'm sure you are correct that the cost for the electrcity is also close to $0.
http://www.powermeterstore.com/product/kill-a-watt-ez-power-meter?gclid=CNKm9NH8pswCFUgehgodRxEJqA

The key problem I have is that I don't have a 3D model to work from, so all of my numbers are estimates, and not real numbers.

[Abricos]

VW plastic ECU case ...
tin sheet metal interference protection ...

[Abricos]

a quick and easy method of making plastic elements ... case ... in small production ...


[video][/video]
[video] [/video]
[video] [/video]

[kb1gtt]

A good way to make low cost parts is to make an model out of some hard materiel, you could 3D print, CNC mill, or whittle it out of a block of wood. Once you have this original form, you use RTV from your local automotive store to make a soft mold. These soft molds are the same thing shown in the above video, and as you damage the mold that he shows above, you can simply make a new mold by pouring more RTV onto the original hard mold.

[AndreyB]

General 3D question: what is the most common file to share 3D models? For instance we create stuff with Autodesk www.tinkercad.com and here are download options:

Is it .STL?

[puff]

yep.

[AndreyB]

All rusEfi 3D models on one public account - https://www.tinkercad.com/users/gJWaVCdbGlc

.stl files added into appropriate repository folders, linkes added into first messages of appropriate topics.

[kb1gtt]

Most low cost 3D printers want the STL file, which the slicer knows how to use. However if you want to modify the file, a STL is not a good format. It's good for printing, but not good for modification. So if can save some other format for the source file, that would be good.

[RustyGargoyle]

Another way to obtain "cheap" prints.

http://dirtypcbs.com/store/print3d

SLA prints at 0.95 per gram
Density: 1.3 grams per cubic centimeter

Am assuming it comes from china so the shipping should be free but you will have to wait an eternity.

Maybe you can laser cut a case?
http://dirtypcbs.com/store/lasercut

But again RF noise is an issue and possible moisture ingress.

In either case I don't think these are suitable for use in the engine bay. Would be a suicide mission.


side note: maybe you can consider moving the files to "https://www.thingiverse.com/jumpstart/thingiverse"
feels like the github of the 3d world.

[AndreyB]

Wow great links thanks for sharing!

[kb1gtt]

A hole in the ECU for the SD card is pretty much a no go for under the bonnet / hood. If you want SD card logging you should expect that you'll need to seal this some how from the environment. For now it's best to consider this as an in the cabin option. There are many issues including the connector is not water tight, etc.

[AndreyB]

A hole in the ECU for the SD card is pretty much a no go for under the bonnet / hood.

Well, the Denso case we currently use is probably a no go for under the hood - there are gaps around the connector. The not water-proof connector.

So SD card slow is not making it much worse, and the SD card hole is defenitely targeted at 1sr generation Miata (inside cabin) use-case.

[AndreyB]

Progress

[AndreyB]

Not much to see - getting the shape right, waiting for new revision of brain board and tall reset button from China.

We kind of should have matched the discovery reset button locaion on the brain board but guess it's a bit too late.

[RustyGargoyle]

Not much to see - getting the shape right, waiting for new revision of brain board and tall reset button from China.

We kind of should have matched the discovery reset button locaion on the brain board but guess it's a bit too late.

You could just drill a hole and print an extension piece for the current button. Basically a plastic rod maybe throw a pen spring onto it so its not flapping around.

[AndreyB]

next iteration

[AndreyB]

RC1