Quote:
When I first mentioned a resistor I was thinking to put it in series in order to reduce the voltage between the receiver input and ground. I can see how putting a resistor in parallel will reduce the current flow through the receiver circuit but I don't understand how the amount of current relates to a digital circuit where the signalling is in terms of presence, or absence, of current. With due respect to your expertise in this area, I am planning to try Calum's solution because it has been shown to work on a JECS Nissan ECU and therefore it should work on a JECS Subaru ECU. Phil. |
Quote:
Quote:
Quote:
|
Quote:
|
Quote:
The converter is powered by the car battery and it does include a voltage regulator. |
I got the ASCII mode working !!!!!
As Calum pointed out there is a flag in ROM at 8029 that determines whether the ECU starts up in binary or ASCII mode. This flag gets copied to RAM at 12CF. If the flag is 0, the ECU will run in binary mode. If it is 1, ASCII mode. The subroutine at C005 switches the flag to 1, but I don't know what set of circumstances will make the ECU call this subroutine. Maybe it's some sort of test mode? Anyway, the simple solution is to set the flag manually. Set the laptop to 1953 8E1 as normal. Send 7812CF00 to read location 12CF. Should get 12CF00 back repeatedly. Then send AA12CF01 to write 01 into location 12CF. The output stops immediately because the ECU comms parameters have changed. Then set the laptop to 4808 8N1 and send 7F313233343030 to read location 1234. You should get back 3132333430300a0d which is ASCII for 123400 + line feed + carriage return. It does not repeat like the binary mode. I'll have to do some tests and see how the speed compares to binary mode. I don't see it being very much faster because it's sending more data. But it's nice to see it working. |
Quote:
Overload of an input will upset the output, such that there is not a true virtual on/off output signal. The input arrangement included in the circuit you are going to try, overcomes such problems and suggested adjustments are provided. The "comparator" circuitry which follows must surely provide a clean output. You are doing the right thing and will have both belt and braces. ;) |
I didn't see something on Kashima-san's site about how to engage the ASCII mode... there was something about a header triggering it. I'll read through again and see if I can figure out why it won't change to ASCII mode by itself. Oddly enough, Kashima-san suggests that binary mode will work under the conditions of ASCII mode... I really do have to read through again, but I swear... :confused:
I'll test that mode out when I get home. |
Phil, courtesy of Nomake, I have only now checked your circuitry and have been negligent in not doing so earlier.
The open ended floating input will have a high impedance and it is little wonder that overload has been a problem. C3 is included specifically to bypass high frequency noise not controlled by smoothing capacitor C1. “Capacitor C3 should be best physically connected directly between pins 16 and 15 on the IC, and C4 likewise 16 to 2. I do not know how you have made up the circuit, but short current paths in these areas are important. “Sensitive” circuitry is involved and earth loops can present problems. It is important that all ground/earth points are physically close. A dual IC is being used and I would suggest that the second unused input (pin 10), as a precaution should be grounded rather than be left open. However this is probably not that critical. You have certainly set a pot boiling and I would suspect that you may by now have become frustrated as a result of the number of cooks invading your kitchen. I commiserate accordingly. :D |
Quote:
If I have to rearrange the board to add the comparator then I will incorporate these suggestions at the same time. I don't really fancy it though because I've trimmed the legs short on all the components. I think as a first (half-assed) attempt I'll add the comparator on a separate board and see if the problem goes away. Quote:
|
Quote:
For this type of project amateur radio guys (hams) often use what is referred to as "ugly wiring" and this is very effective. The idea is simply to assemble the components in a logical three dimensional framework, all being supported by legs, pigtails and the odd added single strand interconnection. The resulting ugly concoction is in practice very rigid and can be stuffed into an enclosure and retained with suitable pliable packing. For absolute permanence a filling of expanding foam can be injected. When it comes to VHF circuitry this ugly method can be very efficient electrically and hard to beat on all counts. For experiments the advantages are obvious. Components can be reused without problem and most important, alterations and additions become a simple procedure. The idea is to make connections prior to soldering, using simple "U' bends rather than twists. . Give it a try as this would be an excellent method for the applications you have on hand. If you think about it, the assembly can be neat and tidy, even if "ugly". PC board, move over.:p :D |
I still can't quite understand what Kashima-san is saying on ASCII. However, the other Japanese site you linked is much clearer.
From the paragraph under the Binary Mode table: Quote:
But I did give it some thought on my own. So if you overwrite the address, it changes over to ASCII mode. Cool. Now is there a way to change modes back without changing the UART speed? Or is that all hard-coded so that when you change the mode flag it automatically changes the UART back to normal? Oh yeah. And another question I came up with. The select monitor interface... does it go through the ECU before the other computers? Or does it interface with the ECU, TCU, ACU, etc separately? The reason I ask is because we all know the TCU works fine. But if the interface has to pass through the ECU before reaching the select monitor plug... then what gives? We get a perfect signal from the TCU and an imperfect one from the ECU. But if the TCU signal is going through the ECU's transmitter... Grah. I don't know... |
Quote:
Quote:
Do you get a response from querying the ACU? The JDM and USDM aircons units are different and I was wondering if yours responds in the same way as mine. Try sending command 89 to query address 000F. If it starts sending data then press buttons on the A/C unit to see if the data changes. |
1 Attachment(s)
Time to answer some questions.
First off, I already did a test (long time ago, forgot to mention it!) where I had the ECU and TCU talking at the same time. It was a mess. But it made me wonder if there was a way to separate the inputs. Hm. Secondly, ASCII mode confirmed operational... and confirmed broken. I've uploaded a log of my test of ASCII mode using the command "7F313233343030". Adding a load to the electrical system makes ASCII mode hopelessly useless. Strangely enough, under the same conditions Binary mode only falters the instant I turn on the headlights. Once they're on, the errors vanish. Is ASCII just too complex a language? Third, ACU test initiated. Confirmed operational. Results: Off Button: 0A Auto: 09 Econ: 06 Def: 02 Vent: 0C Circ: 01 Out.Temp: 0B LO: 05 MED: 04 HI: 03 Interesting. Have you found the memory addresses for ACU error codes? According to the SVX User's Manual DVD, there is a debug mode for the ACU which causes the LCD to display numeric error codes for various possible faults. If it can display error codes, I assume it would also send them through the select monitor. |
Quote:
Quote:
Quote:
01=Recirc, 02=Demist, 03=Fan, 06=A/C, 07=Temp+, 08=Temp-, 09=Auto, 0A=Off, 0B=Out_Temp, 0C=Mode. I haven't looked for the error codes. I guess the way to find them would be to unplug the A/C clutch or something like that to trigger a code and then look for it in the ACU memory. As far as I can tell, the only addresses that return anything are 0000 to 001F, so it should be easy to find. I was thinking about the PM you sent me. Do you have a Knoppix CD? It's a bootable version of linux that can run off the CD without having to be installed on your hard drive. If not you can download it or I could send you one in the mail. http://www.knopper.net/knoppix/index-en.html With Knoppix you could run my software which will let you a query whole range of data rather than one address at a time. It makes it a whole lot easier to investigate stuff. |
Forgot to press TEMP+ and TEMP- while testing, hahaha. I bet the commands are the same. It's too bad that it merely reports what's been pressed rather than creating a "status" for it. I think it would be really freaky to change modes on someone from the laptop instead of the front panel, hahahaha.
My laptop is pretty old, and as such it has no internal CDROM drive. I have an external though and I think I can get it working again. I don't have a disc like that but I'm familiar with it since my dad uses them. I'll try that out later. Crossing my fingers as well, Nomake |
All times are GMT -6. The time now is 09:47 AM. |
Powered by vBulletin® Version 3.8.7
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
© 2001-2015 SVX World Network
(208)-906-1122