VEMS Home Products Documentation Wiki Webstore
WikiHome ProjectPage DocsPage UnderDevelopment MembersPage SiteIndex RecentChanges Preferences
MembersPage/MichaelRichards/Projects/NicksGTX (2005-03-26 02:12:20)

This engine is a Mazda BP-26. It has the following:

The customer is looking to have 300bhp but I aim to provide whatever the turbo will support with all the bells and whistles to make it a nice clean streetable install.

Interface

Display using a Noritake CU20025ECPB VFD display. This was chosen for it's bright display, wide temp range and 5V interface. It's the only VFD I've been able to find that's completely LCD compatible.

The customer will also have the option to plug in a PS/2 keyboard. This is important as he may need to make cold startup adjustments and is 2500km away in Northern Canada.

I am also considering providing several switches to control the setup.

One switch may switch to a regular fuel (87 octane) map that runs boost of 8PSI from the wastegate. I am fairly confident I can make this work with the 10 PSI wastegate and 8.65:1 compression ratio.

Since the keyboard is not likely to be plugged in all the time I'd like to put a pushbutton beside the LCD that allows the driver to change the screen that's being displayed. This is a 2 line VFD display but I had an idea... If 2 rows is defined then it should just double the number of available screens and the next screen should be the bottom 2 lines of the 4 line screen. This would require some firmware mods.

Another switch may control a street or race boost profile. I have a proprietary boost control scheme that I was exposed to while doing some consulting work. It makes for a very nice street driven car that your mom could use. This scheme will be good in the winter as well.

The clutch will be wired to it's own switch so the ECU will know when to engage an auxiliary rev limiter. I would like to hook up a VSS sensor but it uses a mechanical speedometer so I don't think this is possible. Given the lack of gearbox strength full throttle shifting is probably a bad idea anyway!

Engine Control

The plan is to have a fully sequential wasted spark fuel management system. Fully sequential injection is necessary as the injectors (550cc) are quite large for a good stoich idle. Wasted spark was chosen because I will be using standard GM 2 post coils and I am not planning to implement any per cylinder ignition trims.

Detonation detection is a possibility as the engine came equipped with a knock sensor. While this will be wired in I'm not sure that I will be using it yet.

I will be using one of the wideband O2 channels and providing on-screen mixture display. The O2 is mounted in the turbo elbow just before the downpipe. If this does not prove to be accurate enough then I will weld a bung into the downpipe.

The car has a standard resistive TPS sensor so I will be supplying +5V to one side and GND to the other.

+5V can come directly off an onboard L7805 (EC36-pin28 HALLsupply) or after a 270 Ohm protection resistor (EC36-pin29) TPS5V output. When using the same source for different units (eg. HALL + TPS), it's a good idea to provide the weakest, smallest current-limit one to the TPS (TPS can be easily calibrated anyway).

The car is also equipped with a hall effect distributor. MembersPage/Fero has been successfully using this ford/mazda distributor with success so I will be using his scheme for the primary and trigger. I'm not too parial on using cam based triggers but the alternative isn't much better considering it's harsh environment.

CLT sensor will be using the mazda sensor. This differs from the domestic temp sensors but most imports seem to have this in common. Subaru, Honda and Mazda all use the same resistive profile. (link or some data?)

IAT sensor is mounted after the intercooler and just before the throttle body. It is using a standard GM IAT sensor.

Boost control is accomplished via a 3 way solenoid valve. In the rest (and failure) state this valve routes boost directly to the wastegate. In the powered state it plugs the boost supply and diverts the wastegate to the atmosphere.

The MAP sensor uses the supplied (WebShop) MPX4250AP motorola sensor. I'm not partial to having a long vacuum hose into the passenger compartment. In my experience this does not lead to a very responsive engine. I will machine a protective case for this sensor (sounds nice. more details / photos?) and mount it close to the throttle body.

Dash mounted "turbo" light. In factory form there was a little turbo icon that came on whenever the turbo started to spool. This is not critical but I like adding extra touches and I'd like to have this bulb operate with the VEMS. Not sure where to plug it but any MAP triggerd output could be used with a 1bar abs switch point.

The serial interface for tuning will be wired through the EconoSeal connector and mounted in the glovebox. Nobody wants to see wires coiled up near the passenger's feet so this will provide a clean access point. Maybe someday I can make it talk the OBD2 protocol and just use factory wiring (or use the factory wiring with non-OBD2 protocol).

I would like to provide electronic fan control via a standard Bosch relay. I have the option to use the factory radiator installed thermoswitch but with 1.8l 4 cylinder engines switching a fan on can sometimes cause them to stall. With electronic fan control I can cause the ECU to increase the IAC duty cycle to maintain a constant idle speed.

The econoseal plug is wired via 24" of cable to a panel mount milspec marine 24 position connector. This billet piece provides very clean hookup between the engine and the firewall. The O2 sensor will pass via it's own separate chassis wire and be plugged in near the front of the engine. I'm not sure if this will need to be shielded or not, but the nernst signal can be infected with ignition noise and the heater control ruined (=>sensor too hot). I have never had a problem with using unshielded wire for this (but LSU4 WBO2 heater control is very different from NarrowBand - you challange your luck with unshielded, the location of wires will decide if it becomes good or bad).

I suggest to add mdh.. / mdn.. commands for the lines below (where appropriate). Helps a lot when reviewing subsystems.