Oops - forgot photo. It's def a ducted fan, and a helluva blowjob if you ask me... :)

http://forums.pelicanparts.com/uploa...1505838532.jpg

I will see if I can get an actual photo of the blades in mine - they have had quite a bit of usage.

"Brushless" on air stock is a good sign! Motor size and blade diameter is key. Be interesting to see blade wear if there is---not that I'm wanting it! Have to ask... what's your take on noise it makes? Thanks for diving in Dave :)

Turbine config decided

Single turbine positioned just ahead of bucket is path being taken. Set this up now to confirm yesterday's speed readings and measured 50 mph today where yesterday it was 70. Small delta would be overlooked---20 mph is too significant to let slide. If you're interested in details and discoveries, see "Edits" in yesterday's post (#138 on page 7.)

Current thoughts: Could take major detour to fine tune understanding of system's air speeds & air flow dynamics. To get such hard data... more sophisticated measuring gear is needed than what's in hand. While this is tempting inquisition detour, am inclined to focus on main priority at this stage, that being: Whatever air speed measures are... what's important is air flow through condenser---and evening that flow out to a reasonable degree. Am satisfied with flow. Evening it out has yet to be sorted. (Deflector is done but untested.)

Went ahead with adjusting air valve's spring torsion. Existing set up used single 180 x 0.183 x 0.15 (4 coil) spring with horns bent in. That no longer works---valve's open without any main intake obstruction (given new, increased air speed.)


http://forums.pelicanparts.com/uploa...1505927307.JPG

0.015 wire diameter spring with mandrel compared to heavier 0.021 (with 5 coils.) Buying a variety of springs in one shot was good idea. Allows instantly playing around with different combinations.


http://forums.pelicanparts.com/uploa...1505927307.JPG

System now needs 2 thicker wire springs (where it used to have one thinner.) Higher horn bend (top springs in pic) was too much. Lower springs have bend that works. (Bend is to pre-tension doors to keep them closed under normal op conditions.)


http://forums.pelicanparts.com/uploa...1505927307.JPG

Where there were two rubber and two nylon spacers next to spring... there's now one of each next to 2 springs. Valve now opens as it did before with exception being it not opening as fully as thinner single spring set up. Is still enough air passage so turbine burning up from air-starvation is off menu. Change in system's sound when main intake is obstructed = distinct. Torsion spring source: Gardner Spring.

Playing around with springs and different ways to install doors... too much messing around with doors---I broke a hinge off. Hinge welds were done with 350dF media which can be reheated/reflowed. I did that to both doors (with them on pivot post.) My impression is 750 dF media does not reheat---it gets brittle. Am no expert with this stuff so don't quote me.

Getting close to doing install-in-car to check if turbine with rubber reducers will make slight turn as needed. One way or another... they'll turn. CEET will be fit and final-cut. Cradle bracket---to mount intake stack to car---is already riveted to tub. Will apply thin rubber cushion on stack where it contacts cradle to isolate stack from car. 3 weeks till AC system is charged---am optimist.

Whipping post

Had private input concerning my apparent fook'd-up air speed measuring capabilities. Point made to me: “It’s annoying to look at threads with bad data.” OK. On possibility there’s more than one PorscheBro thinking this… my public apology and perspective: This is not an engine build where factory specs are being reported in error. This is experimental build with new ideas, materials, and techniques applied around almost every corner. Screw ups are inevitable---a given when sticking nose where it's not been before... a natural part of the discovery process IMHO. In nutshell: I share info---the good and the fook-ups---so you know what’s going on with the challenge in hand… perhaps my trials & errors will benefit you in some way. Hope you’re with me in this.


Wiring at brushless - now with rubber coupling

http://forums.pelicanparts.com/uploa...1506358207.JPG

This coupling setup is going to take minutes where hours were invested on the housing-inside-housing program. After footers and other appendages have been cut off (done a few days ago) is now a matter of adjusting wires so connecters can be connected on outside of couplings. Being unfamiliar with this type motor, am not inclined to apply heat to terminals to swap-in longer wires. Will splice in instead.


http://forums.pelicanparts.com/uploa...1506358207.JPG

Heat shrink will slip over splices.


http://forums.pelicanparts.com/uploa...1506358207.JPG

http://forums.pelicanparts.com/uploa...1506358207.JPG

Bullit connectors. Have used spade connectors for years. Liking bullits better--- they take less space. Will bury this connection under insulation (that will exist around ducting.) When system’s cable is run, 3-wire waterproof junction will install as system’s electrical disconnect in wheel well.



Final fit

http://forums.pelicanparts.com/uploa...1506358207.JPG

Parts going in car. Have not weighed it yet but will. Articulating duct joints at top… have them just in case. (Didn’t need them.)


http://forums.pelicanparts.com/uploa...1506358207.JPG

http://forums.pelicanparts.com/uploa...1506358207.JPG

http://forums.pelicanparts.com/uploa...1506358207.JPG

http://forums.pelicanparts.com/uploa...1506358207.JPG

Stack installed. Extension added to Coastr’s air-valve allows 90 degree elbow to turn into wheel well’s set back. Intake’s tilted down for drainage. Would go with short cobra instead of elbow here if choice was made again. Reason = less complicated fitting.


http://forums.pelicanparts.com/uploa...1506358207.JPG

Clearance between air-valve and tub is controlled in part by rotation of stack in cradle. Can also move rubber elbow forward and back. Moving it back gets it closer to tire. Did not mount tire to confirm clearance today---am relying on measures taken a few days ago.

http://forums.pelicanparts.com/uploa...1506359453.JPG

Turbine’s new position = attached to air-bucket. Slight bend necessary (to compensate for air-bucket’s inlet pointing a touch too far towards outside of car) is managed by angling rubber couplers on turbine’s housing. Red wires are “testers” connected to ESC/Arduino/motorcycle 12v batt. (Is first time system is run in car.)


http://forums.pelicanparts.com/uploa...1506359453.JPG

Extra CEET’s been trimmed. Installed length = 26.5”. Another 4” can be eliminated by placing aluminum duct where turbine used to be located (at end of intake stack.) Attempting to slip bare CEET ends onto receiving ends… PITA. Have mounted aluminum duct-sleeves to each end of CEET as “males.” Much easier making connections with this config.


http://forums.pelicanparts.com/uploa...1506359453.JPG

iPhone “panorama” pic.


http://forums.pelicanparts.com/uploa...1506359453.JPG

CEET has heavy duty coil laminated between interior and exterior duct material. Coil is rigid but will deform (and hold new form) given adequate force. Deformed section here is what wedges into crevice. Weggie is done to maximize clearance between CEET and tire below. Will insulate entire duct and bucket prevent air in route to condenser from absorbing engine heat.


In-cabin sound

Windows closed, no music, engine ON at idle… if you know what stock 911 sounds like at idle, only maybe would you sense what faintly sounds like an extra fan running. For anyone not familiar with 911 sounds… they wouldn’t notice it. With musac ON… zero fan-sound. Of the multiple discoveries needed for this idea tohave wings… this is a big one. Could add noise barrier to fender and/or reduce turbine’s speed… with neither mod having much appeal.

Turbine’s high pitch whine experienced with CEET duct (and not with metal duct)? No whine now. Stuffing CEET into crevice is considered probable cause.

When outside of car… fan/motor sound and air wooshing through condenser is distinctly heard. Comparison to racket Spal doesn’t make… turbine is choice for noise lovers. (Sound is actually fine. Were it not, I’d say so… and prepare for more critical whipping.)

Block main intake and system’s sound changes pitch. Is built in alarm. TY Coastr.

If you’re thinking… why not test sound in cabin much sooner during project? Why wait till now? I did think early-on about doing so. Conclusion back then = all aspects of system config might be influential on acoustics. Not knowing what final config would be… early sound test seemed waste of time. Noise levels of system on bench were pondered over. Found turbine’s whine kicking in when running brushless motor above 80% of motor’s capacity. That discovery led to programming Arduino to ramp turbine up to that 80% capacity as a max. New rubber hose was had to test as an alternative to CEET but product did not bend as promised---was returned to Race Part Soultions (who took hose back providing refund including my return shipping---admirable.) With CEET in hand and clock ticking… idea was to install CEET and see what happened. Maybe installed config would effect sound for the better. It did.

Would insulating fender area completely eliminate fan-sound heard in cabin? Maybe. Work involved for an inconsequential gain---not worth it.

Removal of system = 10 minutes (does not include disconnect from refrigerant hoses.)

Sorting out how to insulate duct and air-bucket comes next. Condenser's ports also need re-positioning.

http://forums.pelicanparts.com/uploa...1507290037.JPG

Phase2 clamps. (Sizes on their website are not correct. Talk to CS before ordering.)


http://forums.pelicanparts.com/uploa...1507290037.JPG

Sleeves are on each end of CEET. Peel-and-stick on insulation = not used---insulation floats… all of it.


http://forums.pelicanparts.com/uploa...1507290037.JPG

Slight turn needed at entry to bucket is made with couplings.


http://forums.pelicanparts.com/uploa...1507290037.JPG

Depression between couplings is filled with 2 insulation layers.


http://forums.pelicanparts.com/uploa...1507290037.JPG

Floating insulation on CEET when duct is bent.


http://forums.pelicanparts.com/uploa...1507290037.JPG

Locator notes. Each is covered with clear mylar tape. Durability of mylar here is in question.


http://forums.pelicanparts.com/uploa...1507290037.JPG
http://forums.pelicanparts.com/uploa...1507290037.JPG

Turbine duct section and bucket are yet to be insulated, ports on condenser still need to be moved, and screen needs to be installed on exhaust side of condenser.

Running electrical cable

http://forums.pelicanparts.com/uploa...1507397961.jpg

Rather than put hole in tub and run elec cable along floor boards, plan is to run it through tunnel. Not sure what year 911 reference above is from but am assuming it's close to '80. Cable has to pass through one of these holes excluding obvious no-goes.

Thanks to DJThom and dentist90 for pic and info.

Well your air valve is a thing of beauty - a far cry from the original idea I put forward of nicking an air box pop off valve and installing it.

How much negative pressure is in the intake? Is it a faint pull or somewhat like a small vacuum?

Credit is yours Coast for idea--TY. Neg pressure (vacuum) is noticeable 2-3" from intake. Start to obstruct air flow and air valve compensates. Was great idea on your part http://forums.pelicanparts.com/support/smileys/wat6.gif Without air valve, suction at intake increased intensity up to point of full obstruction. As soon as obstruction is slightly removed... pull increases. Is different dynamic with valve---suction at intake is equalized with valve.

Running cable for brushless/turbine

3-wire 16awg has taken place of 4-wire. Mod is result of choosing to serve brushless dedicated power from fuse panel (rather than risk existing wiring---that served pre-existing Spal---being inadequate.)

http://forums.pelicanparts.com/uploa...1508166484.JPG

Waterproof junction for cable's end in wheel well.


http://forums.pelicanparts.com/uploa...1508166484.JPG

Waterproof wire seals.


http://forums.pelicanparts.com/uploa...1508166484.JPG

http://forums.pelicanparts.com/uploa...1508166484.JPG

Electric sealant/insulation applied on junction's seams for added waterproofing and to prevent water from wick'ing into cable. (Cable has paper winding.)


http://forums.pelicanparts.com/uploa...1508166484.JPG

Hole serving single wire to pre-existing Spal is enlarged to 5/8” for 3-wire cable & grommet. Grommet’s collar-well is gel-epoxied to chassis (idea being rust prevention.)


http://forums.pelicanparts.com/uploa...1508166484.JPG

Entry to engine bay (just aft of fuel filter.)


http://forums.pelicanparts.com/uploa...1508166484.JPG

Run is forward along sill through collar... then along bulkhead...


http://forums.pelicanparts.com/uploa...1508166484.JPG

Drop is on right side down through gap between suspension crossbow & bulkhead.


http://forums.pelicanparts.com/uploa...1508166484.JPG

Tub ahead of right axle.


http://forums.pelicanparts.com/uploa...1508166484.JPG

Using speedo port for passage through tub.

http://forums.pelicanparts.com/uploa...1508167590.JPG

http://forums.pelicanparts.com/uploa...1508167590.JPG

http://forums.pelicanparts.com/uploa...1508167590.JPG

Heavy fish is fed up right side of tunnel (to keep cable from laying on top of shifter rod. Looks like fish is on left side but is not.)


http://forums.pelicanparts.com/uploa...1508167590.JPG

Forward exit from tunnel is up through existing access hole on top. Is possible to exit on side of tunnel but upwards avoids acell linkage area. After cable is pulled... shifter, clutch and acell are checked for interference. None.

Cool project. Loving your documentation and build skills.

So by the dead of winter your AC system will be blowing cold! ;)

Keep up the good work. Fascinating stuff.

Just avoid "heat sources" (heater boxes, heat exchangers, etc.)

Thanks for pos-notes Glen http://forums.pelicanparts.com/support/smileys/wat6.gif. Yeah... my timing sucks! (Perhaps I'm in wrong hemisphere.)


Since you're raisin that flag CG... is definitely worth a second looksee. TY ~~~

Electrical


http://forums.pelicanparts.com/uploa...1508255023.jpg

V.4 = final plan (maybe). 16awg is used throughout with exception of BEC between Arduino and ESC where 26awg exists (for 5v Arduino power, ground and signal. ESC has internal step down, 12v > 5v.)


http://forums.pelicanparts.com/uploa...1508255023.JPG

http://forums.pelicanparts.com/uploa...1508255023.JPG

http://forums.pelicanparts.com/uploa...1508255023.jpg

Bundle = Arduino, ESC, 40a Relay. Will place somewhere below dash. 30a fuses and inline spices. Philmore 30-10320 Rnd Paddle Lever Toggle Switch, SPST 3A@250vAC, ON-OFF.




Port Adjustments


http://forums.pelicanparts.com/uploa...1508255023.JPG

http://forums.pelicanparts.com/uploa...1508255023.jpg

http://forums.pelicanparts.com/uploa...1508255023.JPG

http://forums.pelicanparts.com/uploa...1508255023.JPG

I would probably go with 10awg wire over 16awg for 30A, particularly if the run is from the battery up front instead of the alternator. 16awg is usually fused to a max of 7.5A, so a 30A fuse in there makes the wire effectively unfused. 10awg will cost/weigh a little bit more, but you won't need any "special" connectors.

Random image I wish I'd seen this image before I did my amp install:

https://ceautoelectricsupply.com/wp-...Fuse_Panel.jpg

^^^
Tell me what is happening here. I see power and ground distribution. I see six relays. I see two groups of 6 outputs. What's it for?

I hope I don't add confusion to this thread...It's a custom board made for a specific automotive purpose, and has nothing to do with this fan setup, outside the fact that it is inspirational and informative.

https://ceautoelectricsupply.com/custom-relay-panels/

This particular board distributes (with fuses) power from the battery (giant red wire) to a series of relays (thick red wires) each of which has a trigger (thin colored wires) and an output (thick colored wires).

I really like the cleanliness of the layout, the ability to quickly disconnect it from the car, and the components themselves. It would be easy to add ground distribution or change the layout, etc. The board itself is just ABS or HDPE plastic, available from Amazon or any place that sells cutting boards...You could easily integrate all the Arduino hardware on a board like this, and have clean disconnects when you need to troubleshoot or modify.

Would like to think I have the fuse wrong rather than wire Treme---I'll go back and check specs. Hopefully I'm in the ballpark and a lower amp fuse will work. Input is much appreciated (as I'm an electrical-idiot.)

Image you posted... not showing up on my end.


Dru... perhaps I sketched wiring plan using "symbols" that mean one thing (to a professional electrician) when I'm casually intending another---There are 2 fuses, 3 splices, 1 relay. What is this for? It's an experimental project concerning a rear fender condenser (for AC system) in an '80 911. This project explores whether it's possible to intake air from ahead of the rear wheel and deliver it---with some measure of force---through a condenser located behind the rear wheel. Along the way, new ideas, materials, and techniques are tested to see what works and what doesn't. Specifically, wiring you're looking at exists to run a brushless turbine. This turbine serves as system's air pump. Premise to build is: cooler air should be more thermally productive for condensing than warmer air---and so this is a test of sorts. Knowing this, sensible next question is: Is this system necessary? No. Spals attached to fender condenser can deliver 33 dF at vent with a 50 dF delta. So, why bother? It's an exploration primarily but there's also interest in seeing how far the delta between ambient and cabin temps can be driven.


Those boards are nicely/cleanly setup. Makes my zip-tie-bundle look jungle-done. Perhaps I'll rethink "bundling" approach. Nice ref Treme... inspiring stuff---always good. TY.



Schedule

Plan was to have everything assembled and system charged by Friday this week before I leave town for 3 weeks. Urgency was to charge system in summer's final heat (so ambient-to-cabin-temps could be looked at with ambient still on broil setting here in Miami.) November... probably not that exciting to test AC in but one never knows. Problem on hand = Bombshell---not the good kind---has landed totally screwing my schedule for rest of week. I'll only manage to get a few things done now before departing.


.