TVR Regeneration
I have 'misplaced' a few parts of the car over the last few years. One of them happens to be it's original VEMS. (It's now doing service in another of my cars...)
I have some parts of an old VEMS where the main board is beyond repair, so my cunning plan is to purchase a new v3.8 main board and build up an ECU using the spare case and other parts. I note that a few changes have happened in the last few years so I will sketch out a plan and include some questions to make sure I've got all the changes covered.
Outputs
Fuelling Outputs
- 8 large Siemens injectors driven from the 8 injector outputs. I used to use IGBTs for these because they proved more reliable.
Q: Is it still the case that there are reliability issues with FETs on these outputs - especially if I play with the flyback?
A: There are no current known hardware issues of Siemens injectors on 30v flyback rail, however if you increase the flyback rail considerably there is a point where you will go over the rated FET voltage; this is therefore not recommended.
Q: I used to use a 30v transient suppressor diode in the flyback because it worked well. Is there a new scheme I could employ to force the injectors closed faster? (the highZ Siemens injectors do seem slow)
A: No 30v flyback rail is the fastest available; If you are using LowZ injectors, active flyback your best option.
Q: If I do this, does it adversely affect the flyback of other output types, meaning I need to configure them differently?
A: For outputs controlled by injector FET which do not benefit from heightened flyrail, like: IAC, boost control valves etc. mounting a local flyback diode across the solenoid connector is recommended.
- P259 output to drive the fuel pump relay.
Q: Has the resilliance of the P259 been improved? I had one or two problems with it in the past (usually my fault!)
A: p259 outputs have been outfitted with additional on board flyback diodes, this does not help against over-current or ground/powerground connection problems but its does make it more resilient against flyback spikes
Q: Is there a convenient way to include within the ECU case some N channel FETs (with suitable flyback protection - relays are inductive after all) to provide higher current outputs? 1A to 2A would be sufficient.
A: yes, see https://shop.vems.hu/catalog/vems-p-48.html extra NFETS options.
- 2 Wideband lambda sensors, connected with decent shielded cable, apart from the power. They will be LSU4.2 because I have a box of these to use up!
- Evaporative emissions control valve. Driven from P259, however it is right on the limit of drive capability. (See previous question).
Q: Is there a good way of PWMing this output now? I used to use the boost control table for this, but it seems like a waste! (Note that I need the fan and water pump controls for fans, but these won't be PWM driven.)
A: Secondary PWM control is your best option for this, this works independently of boost control.
Ignition Outputs
- 8 Coil-near-plug ignition channels driven from the 8 IGBT outputs. I have both crank and cam sensors so can use them as fully sequential ignition. It looks like nothing much has changed here, and this bit always worked well.
Idle control Outputs
- I have retained the standard 4 pin stepper motor. This is driven using the stepper outputs on the EC18 connector. I see no need to change this.
Misc Outputs
- 2 cooling fans on separate outputs so that they can react to different temperatures. P259 drives relays for these. A FET drive would be nicer if possible.
- Tacho output on P259.
- Shift lamp currently on P259, but may be replaced by more intelligent instruments.
- Check engine / Fault lamp
Summary of Outputs
It would appear I've used all the available outputs on the EC18 and EC36 connectors and still might have one feature that can't be realised. So I need to decide between shift lamp or fault lamp. (It's a TVR, so it will probably need to be the fault lamp!)
Inputs
Sensor Inputs
- All the usual sensors from the EC36 connector are unchanged as far as I can see. There are some minor changes in triggering, but nothing more than resistor values.
- For reference the crank trigger is a VR 36-1 tooth wheel and the cam is a hall based single hole wheel mounted on the cam pulley. Both have been shown to work robustly in the past.
- One channel of knock sensing. (Two would be nice, but so far it seems a little unnecessary)
Q: This was always connected as a "green wire". The v3.8 board looks like it has dedicated tracks. However, Knock #1 now conflicts with WBO2 #2 Nernst cell on EC18 pin #1. Is this correct?
My choice would therefore appear to be to use knock #2 as the main sensor input and limit myself to one sensor, or bring the two sensors out on a separate connector mounted on one of the case ends. The former keeps the wiring simple, but the latter frees an EC18 pin to be a spare analog input, and these are always useful - see later.
- Fuel temperature. A useful input to detect heat soak! The car had this from new, so it seems only fair to keep using it on an "anytrim" input. It is to be connected to an analog input on EC18 pin #3.
- Wheel speed input. Again I would like to get at least one speed input into the ECU.
Q: This is on EC18 pin#8 and conflicts with the WBO2 #2 pump + signal. Is there an alternative pin (for either function) that doesn't require too much rework?
Q: Additionally, is the second wheel speed input tracked to a particular EC18 pin? If I want two wheel speed inputs then it might make more sense to output them on another separate connector.
- Oil pressure input - mainly for complete logging, but has a benefit for engine safety.
- Oil temperature input - mainly for complete logging, but has a benefit for engine safety.
Control Inputs
- Config switch. Currently on EC18 pin #16.
- launch control enable. Currently on EC18 pin #12
- Shift cut cluch switch. Currently on EC18 pin #6
- RS232. I am led to belive I can have two of these now. #1 is now tracked to EC18 pin #14 and #15.
Q: Does RS232 #2 go only to the pads near the MAX232 chip?
Summary of Inputs
Again, it seems I've used every pin at least once, and will need to introduce some more connectors too. The trick will be to work out what I don't need and what can be done more efficiently. I can then write down a good specification for my spare main board.