NA8 air intakes

[manga_blue]

Interesting findings, droo.

I'd been starting to think more about temps after the club event at Wakefield a couple of weeks ago. High main straight speeds in the first couple of laps of the day melted away as the morning warmed quickly. Stockish motor, no turbo. Losing about 1 sec per lap on main straight by late morning, even with higher corner exit speeds on grippier hot tyres.

It's a bit concerning that your inlet temp is 3 deg above ambient. What's the condition of the strip seal between the engine bay and the HVAC collector box on yours? They're reasonably cheap from Mazda and only take a minute to fit (also does wonders for giving you cool air in the cabin). If the your seal is good then that throws a question mark around the whole perceived wisdom about drawing air from there.

I can't comment on your CAI. On mine there's no shielding between the headers and the Loch Stewart air box but the box itself is plastered with heat reflective tape wherever it faces the headers. That stuff seems to work well. It has protected coolant hoses within 3 cm of the headers for years. I'll take a pic tomorrow.

The NB8A cross pipe is a better option than the NA8 original. It is made from the same thick insulating plastic and fits straight in. It has wider bends and looks less restrictive. It doesn't have the resonator or that ugly squashed section below the bend into the throttle body. You just have to tap into it for the extra idle air supply pipe to the NA8 throttle body.

In total though the amount of heat soak is still pretty worrying, particularly with respect to what happens when stationary. When we're doing sprint sessions we usually warm up the car in the pits and dummy grid before heading out onto the track. What you're indicating is that by doing so we're robbing the car of power for a significant proportion of the session.

rxmx's might be the way to go, thanks for that rxmx. Nice to know that the water doesn't affect you there. I'll keep exploring around the cold side, maybe looking for something like rxmx's but with the much smaller NA8 MAF fitted into the downpipe from the throttle body and an enclosed pod in the wheel well drawing air from a pipe sticking into the side of the car's mouth, just in front of the off-side corner of the radiator.

[gslender]

Droo, what are you using to measure air temps, what sensor and do you have multiple sensors?

G

[manga_blue]

Correction. The NB8 crosspipe is a heavier more flexible rubberish material than the NA8 one. Here it is in an NA8. Check the heat tape on the heater hoses and around the bottom of the air box.

Yesterday was the first time I popped the bonnet after the last Wakefield event when I drifted through the new dirt they dumped inside turn 4. You can see why we need air filters lol.

[evil_weevil]

Wow! hasnt this turned into an awesome thread with better info than expected!

rmmx and droo - thank you for posting all that info, certainly very useful and I hope to take a lot out of it too (and tbro/Terry, thank you too!)
time to start working on this before next race meet hehe

[rxmx]

No worries, actually credit goes to Dweezle as he sent me a pic of something similar he wanted built before he sold his car. After doing some more research found that its a great source for a cold air setup and plus I wanted to remove all the plastic from the engine bay, just never really thought of routing it cold-side. Also if you put a sheet under the front bar towards the wheel like a mini splitter you will capture more pressure that is forced down by the wheel..

I just used a turbo intake pipe i had laying around, blocked off the BOV, IAT ports and used that.

[droo]

Ive just remembered on my suzuki gti i had tried to mount the afm so air passes up vertically - i remembered it wasnt getting readings properly and would hesitate on slow cruising speeds - just to note to keep the afm with air passing horizontally through it if you choose RXMX's method!
The GTI does have a different internal setup to the NA8's afm but as they're both hotwire based, i'm just putting that out there : )

Manga_blue -
The temp sensors are TC75Ax chips (x=0..7), TO-220 package components which give a digital output in celcius. Eight of these chips can be linked off the bus and each individually addressed, so you can get eight realtime readings essentially simultaneously. No calibration required for it's output. I have my arduino setup reading the temps in my dash. (I did work out how to address multiple groups of sensors, but figured i dont need to read more then 8 locations, so it was unnecessary to implement!) There is an error range for them at nominal environment temperatures however on my testing (freezing them and boiling them, they've been pretty accurate, knowing where zero and 100deg c are.. - as near as makes no difference for me if there is a 1-2 deg c error range.

ranting, sorry.


my cowlilng temps have always been a few degrees warmer then ambient.
3 degrees is the absolute minimum difference (once the car is warmed up and have done 10-50km driving)
On a hot day, the diff can be up to 5 degrees.

The temp readings from all the sensors when the car is cold are all pretty much the same when it's first started after leaving it over night.

In the sun & heat, Black car i figure, black plastics above the cowling. On the 40+ deg C days Bendigo had early Jan this year, the cowling would read well above 50 degrees as it was exposed to baking heat from direct sunlight and no place to park undercover at work : ( - not to mention a gear knob, steering wheel and handbrake i needed to wrap my hands in a few layers of cloth so my hands didnt burn on them.. (but the air con via it's insulated pipes worked amazingly through that cooling everything down pretty quick )

The cowling would take about 5 minutes to cool down (from the extreme temp) when moving, but what was really disturbing was having engine bay temps hit 89 degrees at the top rear of the engine (where the oem coil pack sits). This would drop down to a 10 degrees above ambient while driving on the highway at speed for some time but a big concern to me when doing urban traffic driving.


The cowling insulation strip I put there early last year and compared to what was there earlier, cabin temps reduced from the vents i'd say it's helping/working. I'm thinking of putting heat reflecting tape or some insulation on the inside of the cowling wall facing the engine. That metal seperating the firewall from the cowlling is pretty thin, right at the top of the bay where heat is rising to and essentially sitting there. Short of having some sort of venting to let that heat out, i'm thinking that that 3-4 degree temp increase in the cowling wont really go away.
I thought it may also be engine bay heat heating the bonnet as air passes over it, but with the expanse of bonnet/metal - and rate of airflow hitting it - i'm thinking thats not really a big factor..
Trapped engine bay heat i have no doubt is being soaked into the cowling via that thin metal wall tho..


For now, i need to source some heat reflecting tape!

[manga_blue]

I might take a run today and put a temp gun on the surfaces around the collector box. The tape on the firewall sounds like a promising idea.

I've just been thinking about the downsides of rxmx's solution. Track temps can be pretty high on track days. I often feel the track surface temp before a session and in doing that I've noticed that the air can be quite a bit hotter in the first few inches above the track. It might be the case that air drawn from the collector box, although 3 deg higher than ambient, might still be cooler than air flowing into the wheel well area.

Has anyone found a good inlet setup for near the top of the nose cone?

[hks_kansei]

Has anyone found a good inlet setup for near the top of the nose cone?

Note sure on your class regs, but have you considered using the indicator?

Another one I've seen is to cut a notch (or a round hole with bellmouth) out of the bumper between the headlight area, and then runa duct directly to the new opening.
Like this:

[93_Clubman]

Great posts (as usual) Droo.

[gslender]

The temp sensors are TC75Ax chips (x=0..7), TO-220 package components which give a digital output in celcius. Eight of these chips can be linked off the bus and each individually addressed, so you can get eight realtime readings essentially simultaneously. No calibration required for it's output. I have my arduino setup reading the temps in my dash. (I did work out how to address multiple groups of sensors, but figured i dont need to read more then 8 locations, so it was unnecessary to implement!) There is an error range for them at nominal environment temperatures however on my testing (freezing them and boiling them, they've been pretty accurate, knowing where zero and 100deg c are.. - as near as makes no difference for me if there is a 1-2 deg c error range.

droo, Googling for TC75A turns up nothing... and various combinations don't either. Can you remember more about them?

EDIT: Found it... http://www.futureelectronics.com/en/tec ... .3MUA.aspx

I wonder if they are susceptible to surface/heat soak or radiated heat more than just measuring air temperature (of which air isn't a good conductor of heat) ???

[droo]

CRAP! (my bad) TC74Ax
ranges from TC74A0 to TC74A7
3.3 and 5.0 versions (smd and TO-220 packages)

I endd up getting 5 or 6 of each value from a Thailand manufacturer - pretty cheap individually but postage was stupidly expensive..

here's one datasheet -

http://pdf1.alldatasheet.com/datasheet- ... 50VAT.html


I'm really sorry about my typo..

[droo]

another link this time for arduino code.

the chips use 4 wires - power, gnd, arduino, SDA pin and SCL pin (located differently on each different arduino board)

http://www.instructables.com/id/Thermos ... uino-code/

really simple to interface (which is why i chose it) and isnt affected by other arduino loads on other pins (unlike most other 3pin temp sensors)

[bruce]

I wonder what all this work nets you in horsepower or throttle response? I'd hazard not a lot.

[manga_blue]

FWIW these are surface temps averaged from 2 runs of 20kms in hilly country at 80-100kmh. Not spirited, just normal highway cruising. Ambient temp was 26.



The 120 is on the mounting bolt at the bottom corner of the filter box. Metalwork in front of the radiator was reading 30-33.

[gslender]

CRAP! (my bad) TC74Ax
ranges from TC74A0 to TC74A7
3.3 and 5.0 versions (smd and TO-220 packages)

I endd up getting 5 or 6 of each value from a Thailand manufacturer - pretty cheap individually but postage was stupidly expensive..

here's one datasheet -

http://pdf1.alldatasheet.com/datasheet- ... 50VAT.html


I'm really sorry about my typo..

droo, I'd be wondering if this sensor is really suitable for detecting changes in air temp in a surrounding enclosure - it may not lend itself well to detecting air temperature, and instead (due to heat soak of the sensor packaging itself) be telling you about the latent heat absorption differences of various materials and the radiant impact they have on the sensor - you may find the air itself isn't impacted at all by those differences you are seeing.

Is the sensor significantly in the air path? And does it have enough of its surface to be largely cooled by air temperature changes? Maybe a simple test of heating up the sensor (with a hot gun) and then getting a steady stream of air blown over the sensor and recording how quickly it can change and/or if it changes quick enough etc.

What is your thoughts on that?