It is currently Wed Jul 18, 2018 7:53 pm



Forum locked This topic is locked, you cannot edit posts or make further replies.  [ 132 posts ]  Go to page 1, 2, 3, 4, 5, 6  Next
Author Message
 Post subject: o5e board project
PostPosted: Mon Jun 01, 2015 1:28 pm 
Offline
User avatar

Joined: Thu Mar 26, 2015 11:55 am
Posts: 112
Location: Massachusetts, USA
Hi all, I've been working on making some hardware targeted for this project. So it's time to show some stuff and introduce myself here on the forum.

First much thanks to Mark for helping me get started and answering a bunch of questions in email. I've been loaned a Freescale starter trak board and I hope to get an engine running on it soon. I feel it's important to get experience on actual hardware for this project early on, so I've been learning about parts for ECUs and tools for PCB design. Here is an initial take on actual hardware for o5e.

This is a screenshot of an MC5634m CPU board. To start I want to build a general CPU board and a ECU support board. I hope this divide and conquer approach minimizes some of the risks and lets updates and improvements happen without having to throw everything away every time there is a change.

This CPU board will take a partner card that fits on top and provides the actual drivers for the fuel injectors and also the IGBT's for the coils. Mark has warned me that it might be difficult to combine these in one box, but since they are already different boards, they can be different boxes easily. They could even be different boxes in one box, if you want to be sneaky...

The work has been done in KiCad, so it will be shareable - I need to learn a bit more about copyrights and licenses for hardware. It's a rather confusing subject.

I'm presenting this work early, it is by no means finished, so contributions will be welcome.

The board pictured is about 75mm x 125mm ( 3" x 5" ) and the partner board would be the same size. When I was looking at this a week or two ago I thought no way can I fit all that stuff on the board, but it looks like there is plenty of space now. It still needs a couple of parts put on, some ferrite beads a couple of resistors and then some stuff to provide more power protection for "load dumps".


Attachments:
C9_ECU_cpupcb_early2.jpg
C9_ECU_cpupcb_early2.jpg [ 243.45 KiB | Viewed 25372 times ]
Top
 Profile  
 
 Post subject: Re: o5e board project
PostPosted: Mon Jun 01, 2015 2:46 pm 
Offline
User avatar

Joined: Sat May 11, 2013 9:45 am
Posts: 729
Location: PA, USA
Looking good!


Top
 Profile  
 
 Post subject: Re: o5e board project
PostPosted: Wed Jun 03, 2015 3:02 am 
Offline
User avatar

Joined: Thu Mar 26, 2015 11:55 am
Posts: 112
Location: Massachusetts, USA
Here's another shot, this one without the rat's nest. I have been doing a little practice with putting down copper traces, but mostly still optimizing the placement. It seemed like there was room so I added another 2 VR channels for a total of 4 and another 2 user configurable analog channels. These have vias that can take leaded components in a vertical orientation. So you can set up your own pull ups and passive filter.

I should give a little more explanation of what you can see on the board. There are three major connectors. The one at the top of the board is a 40 pin header that carries the analog inputs. My thought here is to use a ribbon cable while the prototype testing is going on. Basically punting on connector choice there for the moment. I would expect to use a colored and maybe twisted pair ribbon cable.

The vertical connector on the right side of the board carries power and SPI for the partner card with the digital drivers for injectors and coils and other things. The connector at the bottom of the card carries 16 digital outputs to drive the 8 injectors and coils directly.

I think the partner card will also have an SPI chip with 8 digital outpus ( some PWM ) and also an SD card socket.

FInding space for the labels on all the chips is starting to become a challenge..


Attachments:
cpupcd_early3.jpg
cpupcd_early3.jpg [ 247.83 KiB | Viewed 25355 times ]
Top
 Profile  
 
 Post subject: Re: o5e board project
PostPosted: Sat Jun 06, 2015 4:27 pm 
Offline
User avatar

Joined: Wed Jan 21, 2015 1:53 pm
Posts: 49
Location: WxSW U.S.
Horizenjob wrote:
Here's another shot, this one without the rat's nest.

The "rat's nest" was interesting. What does it depict? It would also be interesting to know how these boards are envisioned to be used and by whom.


Top
 Profile  
 
 Post subject: Re: o5e board project
PostPosted: Sat Jun 06, 2015 4:44 pm 
Offline
User avatar

Joined: Sat May 11, 2013 9:52 am
Posts: 304
Location: Over here, doing 'over here' things.
US56 wrote:
Horizenjob wrote:
Here's another shot, this one without the rat's nest.

The "rat's nest" was interesting. What does it depict? It would also be interesting to know how these boards are envisioned to be used and by whom.

"Rat's nest" refers to the connections between components, a.k.a. "air wires" since they haven't been physically routed around the board yet.

_________________
/me goes off to the corner feeling like Jerry Springer with a mullet.

My O5E candidate: 1982 Honda CX500TC motorcycle.


Top
 Profile  
 
 Post subject: Re: o5e board project
PostPosted: Sun Jun 07, 2015 7:11 am 
Offline
User avatar

Joined: Sat May 11, 2013 9:45 am
Posts: 729
Location: PA, USA
US56 wrote:
It would also be interesting to know how these boards are envisioned to be used and by whom.


anyone needing an ECU. Way back the goal was to get some o5e HW put together so people can actually install and try it and this board is a continuation of that plan...a general purpose ECU with lots of I/O and a very low price tag.

Also i mensioned on my frankenFerrari thread I bought an EngineLab.net setup which I've been playing. Thids setup lets you take an AEM infinity ECU and replace the code with the EngineLAB stuff and program the control /logic layer in it anyway you please. So... I've been busy building the control logic in EngineLab and running on their simulator which is much faster than working in C, compiling, loading, etc. But I have structured it very much like a C program so once I'm happy it it will be a pretty simple matter to transfer it back into o5e. Long story short o5e is getting a much more complex control strategy very much like whats in the M1 motecs but better :)

So, progress on the HW and progress on the FW.


Top
 Profile  
 
 Post subject: Re: o5e board project
PostPosted: Mon Jun 08, 2015 12:21 pm 
Offline
User avatar

Joined: Wed Jan 21, 2015 1:53 pm
Posts: 49
Location: WxSW U.S.
mk e wrote:
...a general purpose ECU with lots of I/O and a very low price tag.

So it could be used to run an engine from on board a vehicle and not just for tinkering on the bench.

mk e wrote:
I've been busy building the control logic in EngineLab and running on their simulator which is much faster than working in C, compiling, loading, etc.

And the board(s) will also be used as a target platform to develop C code from the EngineLab framework in a "real world" development environment.

Given the above, it's a very good idea to make it relatively easy to replace the MCU. There are several good reasons but the one of which most people seem to be ignorant or forget is that flash memory has a limited number of write operations before it wears out. In a heavy dev environment it's not a matter of "if" but when. If the boards are hand populated, at least initially, saving the board and the labor invested will pay for the extra effort and cost the first time an MCU dies.

The usual method to make a biggish QFP device easily replaceable is to put it on a QFP to pin-row adapter which plugs into a base board. There are off-the-shelf adapters readily available for 144-lead LQFP. A clever design might be able to accommodate either 144- or 176-lead devices.

Horizenjob wrote:
Mark has warned me that it might be difficult to combine these in one box

It would be a mistake not to put a lot more thought into the issue of enclosure(s) and harness connector(s) up front. Those determine the form factor of the board(s). #rusefi is still struggling with these issues and there was never anything resolved here or on the old forum. Going straight to a PCB design in a vacuum now might be more fun but could lead to a lot more work and expense later in the project.

The first impression of this was that it was "that guy" again but on a different tack this time. Sans "rat's nest" it looks like a Microsquirt Module on steroids. Not being a fan of anything MS with the possible exception of the MS3 Pro which, begrudgingly, is apparently not a bad design for what it is, an MS3 Pro on steroids might not be a bad design model.


Top
 Profile  
 
 Post subject: Re: o5e board project
PostPosted: Mon Jun 08, 2015 1:57 pm 
Offline
User avatar

Joined: Sat May 11, 2013 9:45 am
Posts: 729
Location: PA, USA
US56 wrote:
So it could be used to run an engine from on board a vehicle and not just for tinkering on the bench.


Well.....most anything that works on the bench would meet that general description.

This will be finished ECU intended to be installed permanently into a vehicle...that can also be used for development if you choose. That is probably a more accurate description.

It will have a case. It will have a connector. For the most part neither will be up for public debate, they will simply be what they are...because when given options most everyone wants something different for VERY important reasons!..... but when there are no options there are also no longer any reasons why it NEEDS to be different :lol:

Very basic development will almost certainly continue to be done on dev boards of one kind or another simply because they are cheaper and easier to work with. I don't see replaceable processors as viable on a low cost production unit...if you kill the flash you'll need to buy a new unit, but they will be pretty cheap so that shouldn't be an issue.


Top
 Profile  
 
 Post subject: Re: o5e board project
PostPosted: Tue Jun 09, 2015 4:33 pm 
Offline
User avatar

Joined: Thu Mar 26, 2015 11:55 am
Posts: 112
Location: Massachusetts, USA
Yes, the rat's nest shows where you need to either put in copper traces for signals or make a connection with a via to a ground or power plane. It's pretty intimidating looking at first when it's just a pile of random parts with random organization which is what the board design tool shows when you first read the schematic data. You can make a lot of progress just placing parts and looking for places where lines cross to help you decide how to place wires on the connectors or orient the components.

It's a good suggestion to put a socket or tiny carrier card in for the CPU - it's true that there is a lot of flash burning during development. I'm not really a very good "batch" style programer so I tend to burn a lot of images and realize I made a simple mistake and then burn it again. I think though that I will live with this issue and let this compromise require buying new boards periodically. It will be a sign of success if this ever becomes irritating.

I'm trying to make decisions that will get this project over the goal line. Which is a real world ECU that people will want to use. If someone on this forum is actually contributing and manages to wear out their flash - they will get another board for free or very low cost. So the price for anything more complicated than just getting another board is too high, I think.

In the same vein we will probably land up producing this board with different processors at some point. I think the way forward is the newer multi cpu chips without the eTPU unit. So making a few boards with a different chip is pretty simple. I did spend time trying to decide between the 144 and 176 pin versions. The new parts would be 176 pin, but the current parts do not seem reliably available in that pin count....

I just put the crystal on the board and that has taken up some room. So I think the board is going to grow a few mm wider. I'm trying hard to actually get to the point of laying out the traces, but I'm cleaning up straggling issues to get there. So I spent a day trying to pick ferrite beads, wondering how to do the layout for the crystal and doing placement for all the caps around the cpu.

In order to keep this simple I am choosing to keep all the parts on the top of the board. Maybe that doesn't really save any money in assembly though - I need to talk to someone about this.

Any questions or more suggestions, just fire away!


Top
 Profile  
 
 Post subject: Re: o5e board project
PostPosted: Thu Jun 11, 2015 1:12 am 
Offline
User avatar

Joined: Sat May 11, 2013 9:52 am
Posts: 304
Location: Over here, doing 'over here' things.
Mark is going to cringe. ;)

Whilst I'm not offering these as a suggestions, but more like just a pause or two for thought:
Raspberry PI Foundation offers the RPI as a "compute module" which contains the basic guts of the RPI without all the interface sockets and such. It sits on a DDR2 SODIMM type board and is meant for integration into other products where USB, video outputs and such locations need customized, and where other peripherals might be integrated on the main board.
Intel Edison is in a similar vein, being a compute module in itself, using a very low profile 70-pin SMT connector and is about 25x35mm. Intel provides an Edison to Arduino Uno-type breakout board, for instance.


So what is really lost having a separable MCU daughter card?

You gain:
- more board space on the main board.
- possibly reduced main board size.
- change or even upgrade the MCU if necessary.
- provide functionality the main board doesn't offer- Bluetooth, et cetera.
- substitute MCU's from other vendors with different voltage requirements.

;)

_________________
/me goes off to the corner feeling like Jerry Springer with a mullet.

My O5E candidate: 1982 Honda CX500TC motorcycle.


Top
 Profile  
 
 Post subject: Re: o5e board project
PostPosted: Thu Jun 11, 2015 9:22 am 
Offline
User avatar

Joined: Sat May 11, 2013 9:45 am
Posts: 729
Location: PA, USA
abecedarian wrote:
Mark is going to cringe. ;)


It doesn't make me cringe.....I literally begged several people on several occasions to design me something modular for the development and DIY users so you can change whatever to do whatever.

I think for most people though COST, simplicity, reliability, flexibility in that order are all that really matter....so the right answer depends a WHOLE lot on the the question which in this case is "who is the unit for exactly?"

The unit Marcus is designing will feed the masses and provide SOME flexibility for development/DIY projects so I think its a good starting point I think. Additional flexibility will come from external CAN based expansion options so you can complicate things to your hearts content :)


Top
 Profile  
 
 Post subject: Re: o5e board project
PostPosted: Thu Jun 11, 2015 12:59 pm 
Offline
User avatar

Joined: Thu Mar 26, 2015 11:55 am
Posts: 112
Location: Massachusetts, USA
I considered doing what you are saying. It might have been possible to even use the $25 Starter Trak module from Freescale.

This board is close to what you are saying though. The difference is it has some analog filters and the 2 VR chips. I did split the design into 2 boards with the drivers on a second board. What you're asking for would have taken 3 boards. Connectors are surprisingly expensive and they also cause their own issues.

So, for better or worse, at this point I think we will go down the slightly more integrated route. There is a lot of Flash on the chip, if the wear on the flash becomes an issue perhaps we can work out ways to use the different blocks available for leveling.


Top
 Profile  
 
 Post subject: Re: o5e board project
PostPosted: Thu Jun 11, 2015 1:43 pm 
Offline
User avatar

Joined: Wed Jan 21, 2015 1:53 pm
Posts: 49
Location: WxSW U.S.
mk e wrote:
It might have been possible to even use the $25 Starter Trak module from Freescale.

What "$25 Start Trak module"?


Top
 Profile  
 
 Post subject: Re: o5e board project
PostPosted: Thu Jun 11, 2015 2:39 pm 
Offline
User avatar

Joined: Thu Mar 26, 2015 11:55 am
Posts: 112
Location: Massachusetts, USA
They make several, but not one with the 5634 at that price. I sort of think the 5643 dual cpu model would work though. You would need to write code that performs the eTPU function to run in one of the two PPC processors.

http://www.freescale.com/webapp/sps/site/overview.jsp?code=STARTERTRAK_HOME


Top
 Profile  
 
 Post subject: Re: o5e board project
PostPosted: Fri Jun 12, 2015 11:15 pm 
Offline
User avatar

Joined: Sat May 11, 2013 9:52 am
Posts: 304
Location: Over here, doing 'over here' things.
Funny they have a Kinetis EA kit listed in powertrain.

_________________
/me goes off to the corner feeling like Jerry Springer with a mullet.

My O5E candidate: 1982 Honda CX500TC motorcycle.


Top
 Profile  
 
 Post subject: Re: o5e board project
PostPosted: Sat Jun 13, 2015 10:43 pm 
Offline
User avatar

Joined: Thu Mar 26, 2015 11:55 am
Posts: 112
Location: Massachusetts, USA
[url]Funny they have a Kinetis EA kit listed in powertrain.[/url]

I'm not sure what you mean? I think about 10 or 15 years ago traditional board design moved onto the chip. The new SOCs are very complex and include most or all of what was in a traditional computer.

So you can look even at the top end of the Freescale PPC line and look at the extremly powerful chips they use in cell tower base stations. These do real time signal processing at 70 billion samples per second. They have 16 to 24 10 G serial lines to move data in and out. There talking about swapping out the PPC cores and offering versions that also have ARM cores.

There is tremendous amounts of work invested into the infrastructure on these chips and I think that is an area that Freescale is way ahead of the other folks.


Top
 Profile  
 
 Post subject: Re: o5e board project
PostPosted: Sun Jun 14, 2015 12:20 am 
Offline
User avatar

Joined: Sat May 11, 2013 9:52 am
Posts: 304
Location: Over here, doing 'over here' things.
Horizenjob wrote:
[url]Funny they have a Kinetis EA kit listed in powertrain.[/url]

I'm not sure what you mean? I think about 10 or 15 years ago traditional board design moved onto the chip. The new SOCs are very complex and include most or all of what was in a traditional computer.
I meant it's "funny" an ARM Cortex M0 is listed in powertrain MCU's.

SoC has been common for well over a decade, so ....
Quote:
So you can look even at the top end of the Freescale PPC line and look at the extremly powerful chips they use in cell tower base stations. These do real time signal processing at 70 billion samples per second. They have 16 to 24 10 G serial lines to move data in and out. There talking about swapping out the PPC cores and offering versions that also have ARM cores.

There is tremendous amounts of work invested into the infrastructure on these chips and I think that is an area that Freescale is way ahead of the other folks.
June 12, 2015, I was doing ETTCS and do not recall seeing any FS chips. I'd like to learn more about this.

_________________
/me goes off to the corner feeling like Jerry Springer with a mullet.

My O5E candidate: 1982 Honda CX500TC motorcycle.


Top
 Profile  
 
 Post subject: Re: o5e board project
PostPosted: Sun Jun 14, 2015 1:30 pm 
Offline
User avatar

Joined: Thu Mar 26, 2015 11:55 am
Posts: 112
Location: Massachusetts, USA
There are two branches to the high end PPC chips from Freescale. The B4860 is at the top of the cell base station line and there is a similar line that swaps out the signal processing stuff for just more PPC cores. They come with up to 12 PPC cores, I think, and even more 10G serial lines ( 24 now, I think ).

The emphasis now is towards the pico and femto cell size units. The way that cell phones are going to get more bandwidth is by drastically reducing the size of the cells. Expect these chips to start showing up on every floor of every building and every other phone pole in your town. Last year I started seeing ads for Comcast guys installing them in your conference rooms to improve your internet...


Top
 Profile  
 
 Post subject: Re: o5e board project
PostPosted: Mon Jul 27, 2015 7:20 pm 
Offline

Joined: Tue Sep 16, 2014 2:10 am
Posts: 45
Location: Australia :)
AN4156_MPC563xM_Engine_ReferenceAug2011_updated.pdf

*edit by abecedarian to keep forum presentation consistent. Link still works as advertised.

_________________
Steve
Bosch K-Jetronic Injection with FrankenCIS
Yes! mechanical/hydraulic constant flow injection can be managed by a modern EMS


Top
 Profile  
 
 Post subject: Re: o5e board project
PostPosted: Mon Sep 07, 2015 3:30 pm 
Offline
User avatar

Joined: Thu Mar 26, 2015 11:55 am
Posts: 112
Location: Massachusetts, USA
Thanks Steve, I don't think I had that ap note and a quick read shows some helpful stuff. Also sorry it took me so long to notice your post.

I've completed a first pass on my layout for this prototype. This has been slower than I would like but I've had to learn some things and the changeover to KiCad has been a little painful. I still have more to learn but things are going quicker with the design and the CAD now that I have some more experience.

Any review or comments are appreciated. Probably what makes sense here is to give an little overview of where this stands and a picture of the board. I don't have the 3D stuff for KiCad working and I wish I did because a real 3D picture of the board would be more interesting and also useful since it's time to put it in an enclosure. I'd like to see how the board and partner driver board fit together.

I'll update this list with more specifics or corrections. Sorry it's not in a final state, but if I don't start pushing this stuff out I'll just keep doing other things...

The board dimensions:
- 127mm (5"), 69mm (2.72"), ~13mm (~0.5")
http://www.alibaba.com/product-detail/beautiful-strong-74-29-110mm-Custom_60224919325.html
Who doesn't love a beautiful box?

Analog
- 8 inputs, 2 without pullups for oxygen sensors
These have ESD TVS protection at the input and Schottky diodes before the CPU. I suspect the TVS may be a little extra because the analog filter are probably strong enough - but that's the current state of these. They should also get weak pullups for fault detection.

Position sensors
- 4 VR differential sensors inputs.
Haven't done the work to figure a simple way to convert these to Hall, but it seems doable.

Digital
- 1 CanBus
- 2 Digital on the analog connector with TVS and diode protection.

There are several more available on board. I need to route 2-4 of these to the driver partner board for direct control. The driver partner board also has an SPI connection so we can choose 1 or 2 chips to provide digital signals that are driven by SPI.

Driver Partner Board
- 8 high impedance injector drivers
- 8 coil drivers for 4.5mH coils
- more to be decided....

The injectors and coils are driven by Freescale 33810 and maybe the IRGS14C40L for coil driver.

I started the process of getting quotes for getting the boards built and it looks like the PCB will be around $4 in a high temp fiberglass with 2 oz copper. Cheaper fiberglass and thinner copper only saved around $0.50, so they may make sense. I don't understand yet how they quote the copper thickness. Do they include the copper plating or not? I need to know so I can understand the current capacity of the vias. The vias I am using between the copper planes and also the power transistors will carry some current and I would like a long lifetime.


Attachments:
C9_ECU8a_r0-1_small.jpg
C9_ECU8a_r0-1_small.jpg [ 247.43 KiB | Viewed 24575 times ]
Top
 Profile  
 
 Post subject: Re: o5e board project
PostPosted: Tue Sep 08, 2015 2:53 pm 
Offline
User avatar

Joined: Sat May 11, 2013 9:45 am
Posts: 729
Location: PA, USA
Nice work Marcus!

Is this a double sided board with the CPU on the back or are you stacking?

Can you post a pic of the case...the link just brings up a sign-in page.


Top
 Profile  
 
 Post subject: Re: o5e board project
PostPosted: Tue Sep 08, 2015 4:22 pm 
Offline
User avatar

Joined: Thu Mar 26, 2015 11:55 am
Posts: 112
Location: Massachusetts, USA
This is a double sided board with just little pieces on the back. The CPU is on the top and the partner board plugs onto the 2 shorter sets of pins. The long set of pins on the right side of the board carries all the analog stuff. It's a little hard to decide where to draw lines, but I tried to make the CPU board be a bit more generic. In theory you can use it for other things, with or without some other type of partner board.

That link works for me without a sign-in, so they are probably doing something with cookies or whatever to track people. Here is a link to the factory, but using what I thought was the part number brings up a very similar but different product. This one would take the same board I think, but it is taller and has fins so slightly bigger outside. Possibly a good alternative, but I lost track of that one earlier and couldn't find it again. It's frustrating.

I wish I had 3D images for my board, maybe I should whip up something very simple in SKetchUp.

The enclosure in the above link is 74mm X 29 MM.

http://yonggu-enclosure.en.alibaba.com/product/1899002544-218701521/80_45_115mm_Best_sheet_metal_extruded_aluminum_electronic_enclosure.html

This one may work and it has side flanges which would maybe make better mounting options. I'm trying to fit 6 boards onto a 13"x9" panel. This enclosure would require to add 6mm or so to the width which isn't really much of a change and would still probably fit in a panel.

http://yonggu-enclosure.en.alibaba.com/product/1900516397-218701521/Customized_for_you_104_28_95mmaluminum_industrial_electrical_enclosures_Junction_box.html

Sigh, they have 38 pages of descriptions of these boxes... I would like a nice one. When I worked on a friends megasquirt the other day he commentated that he thought their boxes had a really cheap feel. It was the type they offer here where the top and bottom are identical halves that mate. I think I am wishing they offered these extrusions with about 1mm more thickness.


Top
 Profile  
 
 Post subject: Re: o5e board project
PostPosted: Sat Sep 12, 2015 2:57 pm 
Offline
User avatar

Joined: Fri Aug 16, 2013 12:04 pm
Posts: 57
Location: Jersey City, NJ
US56 wrote:
#rusefi is still struggling with these issues

Nope, it does not. 64 pin + Mazda OEM case now, 121 pin waterproof + Nissan OEM case one day.

_________________
another open ecu


Top
 Profile  
 
 Post subject: Re: o5e board project
PostPosted: Sun Sep 13, 2015 5:12 pm 
Offline
User avatar

Joined: Thu Mar 26, 2015 11:55 am
Posts: 112
Location: Massachusetts, USA
I've been going thru the application note that Steve provided and it's been helpful.No big changes but a couple of differences and some food for thought.

They mention that the system basis chip dissipates an unexpected amount of power in their design, so on this board i have more than doubled the size of the ground pad under the chip and included thermal vias to a matching and larger pad on the back of the board. I have also segmented the ground pad under the IC pad with solder mask. This should ensure a good solder bond without air bubbles.

These changes were made by changing the shape of the ground pad and then drawing solder mask on top of the segmented area and the increased region to "tent" the thermal vias and prevent them from drawing solder to the other side of the board. I hope this works, I don't know if there is a better way to do this with KiCad.

I put TVS diodes and Schottky diodes on the inputs to the CPU. They argue this is not required and I should have read the hardware manual or data sheet in more detail. I don't remember how well this was covered.

Not using the TVS diodes requires using high voltage and also larger caps in the front end which also cost more too. The high voltage caps also lose a large percentage, like %75, of their capacitance with a high voltage bias so it's not clear they offer strong protection. So far I'm thinking I'll keep the TVS units in. There is a reality here that an OEM ECU spends it's whole life installed in a car being ignored. An aftermarket ECU is probably subjected to being played with during winter months as people install and remove engines or their parts. So potentially a lot more static exposure.

I'm less sure about the Scottky parts. They provide much more protection compared to the CPU diodes, 50+ mA vs. 1-3 mA. You could probably hit a signal line with a welder or plug a sender into a wall outlet.

The cost of these parts is about $3 a board, so not much money. They do take up board space though and I could upgrade from 8 A/D channels to more by taking those parts off. I may defer the decision to remove them until after the proto boards are made.


Top
 Profile  
 
 Post subject: Re: o5e board project
PostPosted: Mon Sep 14, 2015 2:36 pm 
Offline
User avatar

Joined: Sat May 11, 2013 9:45 am
Posts: 729
Location: PA, USA
Way more robust for just $3 seems like an easy choice.

Overall it'a light on I/O for a 6 or 8 cyl so CAN expansion will need be the next project


Top
 Profile  
 
Display posts from previous:  Sort by  
Forum locked This topic is locked, you cannot edit posts or make further replies.  [ 132 posts ]  Go to page 1, 2, 3, 4, 5, 6  Next

Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
cron
Theme designed by stylerbb.net © 2008
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group
All times are UTC - 5 hours [ DST ]