it's fairly common mod in general, there just isn't a dedicated aftermarket solution for the RX8

pretty sure the appropriate model Diablo MAFia (1996-2004 Ford 6-wire) could be converted to work on the RX8 with some adapter wiring

 

If you want to do it properly, you could make a small opamp circuit, that subtracts a fixed voltage, for example 0.8V from MAF voltage, before ECU ADC converter. Now, you will keep your resolution, and get 0.8V extra headroom, and keeping MAF scaling, just with a reduced value. This is not fragile parts(IC and resistors), so reliability should be no issue.

If you want to do it extremely simple, you could most likely put a common collector amp in between, thats only one transistor and one resistor, and would subtract approx 0.7V.

Will ECU give a fault if MAF voltage is too low? If so, this will limit possibilities for a circuit like this, but still give approx 0.3V before you are at 5g/s.

I someone would like to try, I should be able to make a test setup and verify, then post schematic.

 

I believe it's approx. 0.87V or roughly 2 g/s on the factory MAF map. I'm not sure if it can be altered by adjusting other map values.

 

blah blah blah ............... somebody just do it and post some results !

 

Pull your head out from down under and get to it!

 

Had a suspicion about that. My map says 0.86V at 1.95g/s. From what I understand, changing this first value will give a Check engine light. Strange that this error cant be masked out.

Should be possible to get to more than 430g/s by subtracting 0.3V, MAF graph must be nearly flat until 5g/s, but that is OK, isn't it?

 

Ok, I know I am going to beat down over this but what exactly is the benefit of this? Is this a boosted application only or is it beneficial for a NA car as well?

 

wouldn't benefit an NA car as you would be hard pressed to move enough air to max out the maf. This is mostly for boosted cars that are moving so much g/s that they max the maf voltage out.

 

Appreciate it. I can understand the principle but I could not determine if it was for boosted cars.

 

OK ...... done

 

Lol ... I would do the larger MAF now rather than play the voltage trickery game

 

Seems to work ok

 

what's going on with the see-saw at the far top end?

 

I knew you would find something wrong .... lol
I don't know ... but It's certainly not unusual .

 

I know this is a late answer, but I did it with a NON-Stock MAF. The issue doesn't seem to be the low voltage, but instead what the low voltage converts to. So if you have a maf that reads .2 volts with no air flow and the ecu converts that to a negative number it will cause an error.

When I was doing it, the lowest voltage I had on the table was .56 at 1.88g/sec and there were no problems. I don't think going lower is a problem, but I had no need with my voltage divider setup.

 

Here is what I had in mind.

Input signal(MAFVolt) is offset by 0.8V to emulate original signal. So at 0.8V, you still have 0.8V, but in stead going to 4.6V, it's going to 4.0V. Offset and reduction is adjustable for what you need.

Should be a good start for you Brettus...?

 

Already did it and got the T shirt ..... See log above

previous highest log was 400g/s ... now will read to around 460 g/s . I only did 1k and 10k resistor and thought I would get around 500 g/s but reality is less ... might have to redo if i go that high.

 

Input to ADC(Analog to Digital Converter) is normally high impedance/resistance, so that should work fine, if you dont run into problems with voltage to low from MAF. If you do, you can get a local to make that circuit for you, should be 3$, a pricetag I know you kiwis like. If so, he must remember to use a opamp intended for single supply use, or else it wont work.

 

No problems so far . Although I do wonder what happens to those resistors when they get hot ?

Couldn't use Harlan's spreadsheet to re-calibrate because I couldn't open it . So did it by finding points on the table and joining the dots by eye . Worked well - took about 4-5 calibrations to get it honed in .

 

Resistors have a quite low temperature coefficient, close to none, you must almost put it into fire to see a significant change(have done this actually). On top of this, resistors can handle a lot of heat, 155°C(and more, depending on type) is max environmental temp, still being within tolerance.

And since its a voltage divider using most likely resistors of same type, drift by temp will be cancelled since both will have the same temp more or less. If your 1k resistor increase by 5%, so will your 10k. This means that unloaded, voltage will be more or less exactly the same. The load presented by ADC will offset this slightly, but most likely close to none.

 

It has nothing to do with any personalized BS like right or wrong.

I only look at and assess the data based on my own knowledge and experience. I would ask the *exact same question* if it was my own data rather than yours.

 

Oh ...ok . Past experience tells me it's all about finding the faults throwing in some sarcasm to boot . My bad.

 

Good to know ... thanks .

 

Glad to help out when I for once know what I'm talking about

 

The 500g/s estimate was actually about right as it turns out . The reason I thought i was running out of headroom was that I originally had a boost leak at the compressor housing and it was flowing more air than the engine was actually using.

This mod has worked well and given me no issues ........... FWIW .