Guest Ash Posted July 2, 2001 Share Posted July 2, 2001 I'll take the top off mine as soon as I get chance and have a root around to see if I can find any numbers that make sense. Yours, J Quote Link to comment Share on other sites More sharing options...
Phil Wall Posted July 2, 2001 Author Share Posted July 2, 2001 Let me know what numbers I should be looking for and I will look inside my spare. Quote Link to comment Share on other sites More sharing options...
GavinL Posted July 2, 2001 Share Posted July 2, 2001 Hi... can I make the tea ! Gavin Quote Link to comment Share on other sites More sharing options...
Guest Ash Posted July 2, 2001 Share Posted July 2, 2001 When you take the top off take care because mine was stuck down with a kind of strong glue I nicknamed "monster tack". Yours, J Quote Link to comment Share on other sites More sharing options...
Phil Wall Posted July 2, 2001 Author Share Posted July 2, 2001 Yes I found the monster tack. Sticks like ..... Must be a moisture sealant. However top off and large image 750K available if required. Most numbers clearly visible. No sugar please! Quote Link to comment Share on other sites More sharing options...
Guest Ash Posted July 2, 2001 Share Posted July 2, 2001 Phil can you whizz me an image please. Yours, J [email protected] Quote Link to comment Share on other sites More sharing options...
Phil Wall Posted July 2, 2001 Author Share Posted July 2, 2001 On its way to you sir. Quote Link to comment Share on other sites More sharing options...
Guest Ash Posted July 2, 2001 Share Posted July 2, 2001 Yep, just got it, thank you. Recognised it from when I was playing around, and reminded me of those surface-mount thingies that I just cannot get to grips with. When I was soldering components into circuit boards there was this little hole you poked things through. Paul, can you recognise any of the chip numbers??? Yours, J Quote Link to comment Share on other sites More sharing options...
Paul Booth Posted July 2, 2001 Share Posted July 2, 2001 Just returned from rebuilding ex-colleagues Tosh Portege, ground up. Looks like we've been busy bees haven't we (no sugar in mine). I haven't seen the piccy but I came to my PC specifically to ask if someone could take a digital picture, so, can I have one please? I've been giving it some thought and, so long as it doesn't use extensive boot ROM (as the SH1 MCU is inclined to do, unfortunately), we might get away with a ROMulator. It would mean some disassembling/decompiling of the code in order to achieve this but it's worth a shot. Obviously, I'd rather not have to read some compiled C output as unlabeled assembler, but if it saves me a G, I'll dig up my Epson sprocket printer. I was swapping email with Peter Betts from HMSE earlier (aka Hitachi Microelectronics Systems Europe - he's hovering around here somewhere) after he saw the postings. He made a comment which reminded me of something I knew but was not factoring in and made my blood run cold. Our friends in the orient are rather prone to embedding algorythms within the code rather than using mapping tables, i.e., tangents to characteristic curves are determined by in-line calculation rather than by mapping onto a 2D linearisation matrix. If this ECU has been developed in that manner, we might almost be as well taking the development version of the masked device used, with the piggy-back ROM socket (expensive but £10s-£100s expensive NOT £1000s expensive) and rewriting the software. By doing it that way, you can use a ROMulator onto the target device to emulate the masked component which is the end-game; only in our case we plug an EPROM in to development components, if it ever works. Quote Link to comment Share on other sites More sharing options...
Phil Wall Posted July 3, 2001 Author Share Posted July 3, 2001 Paul Pic on it's way to you now! Quote Link to comment Share on other sites More sharing options...
Paul Booth Posted July 6, 2001 Share Posted July 6, 2001 I just received an email. See if you can guess why I thought someone was taking the p155. please send me your up to date CV if you are interested in some contract work. I have a client who is looking for several testers to work on some embedded software. He is confident software engineers will be more than suitable for the role as well, if you are not interested please recommend anyone you think may be. The contract is located in the south east and will initially be a 3 month contract. All that is important is that you have worked with embedded systems before. The testing will be performed on an engine control system. The contract will start in a couple of weeks but he is selecting his contractors now so please send your details ASAP. Turns out it was genuine but, as usual, the automotive industry pays peanuts. Quote Link to comment Share on other sites More sharing options...
Paul Booth Posted July 6, 2001 Share Posted July 6, 2001 ECU issue not looking good guys. Discussed piccies with some heavy hardware friends I respect and they went (sound of inrush of air) "all custom bud", to a man. So, next job is to lean on Japanese friends in Tokyo and see if I can get honourable spec. sheets on devices. Mr Betts, what criteria did you use for your piggy-back box? Drop me a mail, by all means. Quote Link to comment Share on other sites More sharing options...
Guest Ash Posted July 7, 2001 Share Posted July 7, 2001 Then we'll have to go plan B. If we can get the spec sheets then it may be possible to retain the I/O devices and add a custom processor board. There must be some value in that. 1) We know the interface parts are all suited to drive the external sensors. 2) Everything already plugs in neatly. The Motec is reportedly about £1,500 but that doesn't take account of all the hassle in wiring it up and fashioning some kind of fixing brackets. Plus, interfacing it to work with the gearbox ECU. So the true cost of the Motec would be IRO £2,500. Device data sheets and a circuit diagram would be my idea of ECU heaven right now. As hardware is right up my street. All this talk of embedded code and 2D tangent tables is doing my head in... I haven't a flaming clue what you are on about. Yours, J Quote Link to comment Share on other sites More sharing options...
Paul Booth Posted July 7, 2001 Share Posted July 7, 2001 Unless you start doing PID operations with your outputs (now there's a thought... Mr Betts, where the H' are you?), it basically resolves itself down to interpolating intermediate values in a mapping table, e.g. very simply: input value x.ddd indexes output table y[x] and y[x+1]; op = y[x] + (y[x+1] - y[x]) * 0.ddd, i.e. simple interpolation. OK, you're outputs are modified by multiple inputs but you can either: calculate the output values against known curves (think this is what Toyota are doing) or use n dimensional mapping tables. The latter is easy to implement and easy to modify. Now if you applied a PID transform to your output driver you'd theoretically end up with a MUCH faster and smoother transition to the desired condition. Now this is where Mr Betts comes in, he's a DSP man. I'm going to send him an email. Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.