Coast... Car I'd never finish. Air bucket will be done this year. http://forums.pelicanparts.com/support/smileys/wat6.gif

http://forums.pelicanparts.com/uploa...1517608438.JPG

Starting reference... 11 1/16".


http://forums.pelicanparts.com/uploa...1517608438.JPG

http://forums.pelicanparts.com/uploa...1517608438.JPG

Adjustments... Channel for engine release cable has been pulled down slightly by crevice area. Duct area that fits into crevice has been wedged-formed more. Duct area just ahead of crevice has been squashed. Above is inside view of duct is looking away from intake end. 3" OD duct would now seem a better solution in terms of fitment above wheel. How air flow would react to narrower passageway is questionable. (Am staying with 4" OD.)


http://forums.pelicanparts.com/uploa...1517609635.JPG

13/16” raise accomplished. (Realize now I did not place tape measure in same position for second reference. Tape measure is in a more vertical position second time. Might make a 1/16" difference. So 3/4" riase.) Can raise more but need to lengthen duct so it can arch up further. Have extended as much as possible now by resetting existing couplings. Got maybe 1" more duct length. Would need to replace a coupling with longer version to get length needed. Will run this position and see what happens.

1000 mile marker

Am looking at parts to check install is staying installed and for damage. Turbine blades are not checked at this stage due to the challenge that exists to remove and reinstall unit---problem being the insulation. Taking turbine out is easy. Putting it back in watertight condition is less so. For now am relying on sound to say turbine is working (sounding) as it should. Have been constantly running it when driving car with bypass switch. Insulation near turbine is not a good design. Needs rethink.

http://forums.pelicanparts.com/uploa...1526424559.JPG

After squashing duct further into crevice between fender and body (the first adjustment), there’s far less contact between tire and duct. Squashing duct further into crevice is not happening. Can see internal deformation of CEET duct in a post above. Contact as seen is acceptable for now. Were ride height of car a bit higher… would be no problem here. (Fender to ground = 23”.) Anticipated wear from debris being slung from tire onto system is absent---rubber "protection" looks as good as when it was installed but for water marks.


http://forums.pelicanparts.com/uploa...1526424559.jpg

Intake duct is where it was last seen. Bypass valve is “exercised” on occasion by closing intake off with hand. Pitch of turbine returning as it was prior to blocking intake suggests valve closing fully. Filter’s clean.


http://forums.pelicanparts.com/uploa...1526424559.JPG

Condenser’s outlet surface is undamaged. So too are surface areas facing tire.

Summer temps are coming on (in Miami.) Plan is to charge system towards end of this month---end of May (‘18). Looking for 90+ degree ambient to work in.

Head on over here and charge it, Karl - Friday's high in North TexAss is 98 :eek: (this place suuuuxxxxxxx !!!!)

PS - good to see you back and posting Kinky K! :)

Thanks for offer RonO! :) Were I less occupied I'd be knocking on your door.

Yeah, been absent a bit. Getting back in saddle with AC.

Having waited to charge system has proven fruitful---its allowed a refinement which would otherwise have been passed up now only to have been done later. Credit here to Charlie for tabling Spal idea a while back. Exactly how to blend a Spal into air-bucket concept has taken some time to sort out (because air-bucket remained an obstacle---in more ways than one.) Solution was simple once seen… get rid of air-bucket. Intake, filter, air-valve, duct and EDF (correct name for “brushless turbine”)... those parts remain installed. Premise now is to serve ahead-of-wheel air to Spal's intake side. Nozzle at end of EDF serves this air.


http://forums.pelicanparts.com/uploa...1536006564.jpg

From outset when EDF was planned, blade survive-ability has been questionable. These blades now have 2,000 car miles on them. (By-pass switch allows EDF to be tested independently of AC working. Have run EDF all the time when car is driven since its install.) Thanks to Glen for these blades looking as well as they do---intake filter was his idea. Unit was “choke tested”---small leaves & twigs being fed intake with NO FILTER installed---pics lost. (I did this with back up EDFs on hand.) Is capable of producing fine mulch for a while. Eventually there’s too much material to process and stuff gets caught between tip of blades & wall... Motor stops/fuse blows. Clean out and reset fuse and motor's back. Rain/wet has been no problem. (Is brushless motor.)


http://forums.pelicanparts.com/uploa...1536006564.jpg

http://forums.pelicanparts.com/uploa...1536006564.jpg

12” pusher Spal going in. 1,292 cfm @ 0.0 static pressure. Input = 12.6 A. Edge depth = 1.10”. Center depth = 3.44”. Fit between this 14” x 12” condenser and 12" Spal in hand couldn't be better. Only mod here = drill 1/4” bolt holes in condenser's flanges. As this is going together, am not sure how unit will fit in fender space.


http://forums.pelicanparts.com/uploa...1536006564.jpg

http://forums.pelicanparts.com/uploa...1536006564.jpg

Spal’s blades... ringed---makes sense. Junction... will cut off and place bullets.


http://forums.pelicanparts.com/uploa...1536006564.jpg

Another hole goes in chassis to run Spal's cable. Never fond of cutting into car like this. Visions of structure rusting (at these holes) and my son---who shall inherit this 911---... I see him cursing me for such tub/chassis violations. Am hopefully avoiding that distant angst in my Son... idea = put gel epoxy on exposed metal edges inside hole and into grommet’s seat before installing grommet. Surly better than nothing. Engine bay end of cable will get power from fuse panel near CDI (‘80 911 with ‘85 3.2). Believe top fuse is one to connect to. Don’t quote me on that---need to make sure. Relay to switch Spall ON/OFF is signaled from power turning compressor ON/OFF. 2nd cable running through bulkhead is for EDF.


http://forums.pelicanparts.com/uploa...1536006564.jpg

Reason for bullets... heat shrinking. Have seen HS done with silicone put on electrical connections and heat shrink tightened from center outward to both ends---squeezing silicone out. Seems fine idea for more permanent junctions than this. Did not go that route here in order to maintain easy disconnect/reconnect.


http://forums.pelicanparts.com/uploa...1536006564.jpg

New footers. Beef’d up (due to Spal weighing measurably more than bucket.)


http://forums.pelicanparts.com/uploa...1536006564.jpg

Spal-to-fender’s skirt clearance. Am usually guilty of too-tight vs too-loose. Flex of car while driving will likely see this clearance go negative on occasion. Will keep eye on it to decide if mod is called for.


http://forums.pelicanparts.com/uploa...1536007286.jpg

Clearance between Spal’s motor-house and fender-to-skirt seam. This measure is controlled by combination of unit’s tilt, footer height, lateral positioning, and front-to-rear positioning. Challenge here is parts are too large for space to start with---leaves very little adjustment making. Looking at this, Spal is 180 degrees wrong. Would have been better to place wires at bottom of housing instead of very near to fender's guillotine-seam. Is bad move. Return trip may be in order.

.

http://forums.pelicanparts.com/uploa...1536011053.jpg

Looking down at condenser’s exhaust side. (Engine bay is at left here.)


http://forums.pelicanparts.com/uploa...1536011053.jpg

Looking up. (Engine at left.)


http://forums.pelicanparts.com/uploa...1536011053.jpg

7”ish long nozzle is connected to EDF’s outlet. Bracket supporting nozzle bears little weight---most weight of duct & EDF combo is born by zip ties attached to rear lid release sheath. Connection of air-bucket to duct generated a number of complexities which are now gone with this config. This new direction genuinely qualifies as less is more program---a rare moment in this project.


http://forums.pelicanparts.com/uploa...1536011053.jpg

Rather than angle footer to accommodate unit’s tilt, nylon wedge-washers were made (by melting down one side of washer.) Installed washers is not an exact compensation for tilt. Final wedge has yet to be made & installed. Was planning to put rubber washer between footer and strut as a pad but… even that small dimension causes a fit-conflict. Rubber is out. Putting footer between strut and fender flange… that too creates conflict.


http://forums.pelicanparts.com/uploa...1536011053.jpg

Discharge extension pipe previously formed---to run parallel & behind strut… that run’s not happening now. Spal’s dimensions have forced entire unit forward so discharge port aligned with fender’s bracing strut. To move unit back required removing material from taillight assembly and from Spal. Those mods… too Frankenstein to bear for me. Was questionable they would even work. Path chosen is positioning unit so discharge port’s ahead of strut. Technically and visually this solution is not appealing. What is appealing here is it keeps things moving onward. (Abominable option = start over with smaller parts. Not.) This is way it is until next exploration. Discharge pipe… has rubber hose over it. If you look at bottom of vertical trim on mesh… bottom of unit has penetrated into wheel well’s arch. Azzuming I spent a good deal of time attempting to optimize new unit’s fit, here’s proof that 14” x 12” condenser with 12” pusher Spal (as used) is NOT an optimal fit into 911 SC’s fender cavity. Is not deplorable enough to demand a reinvention. Would this config fit 930 fender? Believe it would---everything could sit further back provided more "flare space."


http://forums.pelicanparts.com/uploa...1536011053.jpg

System is closed.

.

Keep GOING Karl!

Charged yesterday. Am traveling next 2 weeks. Ran out of time to post results before leaving. Will attend when returned. (Look out RonO!)

Done. Results are here: http://forums.pelicanparts.com/porsche-911-technical-forum/1009067-ac-pressure-temp-results-arduino-digital-control-air-bucket-builds.html

Update - 8.21.23...

Engine came out of car a while back so I'm late with this update compared to when things occurred---point being there was failure a while back with "the system" as opposed to just now. Two things happened. Air plumbing above the rear wheel suffered multiple attacks from the tire and that plumbing is now squashed to an unacceptable degree. (My car is set quite low.) Next, something hit the EDF (electronically ducted fan) and pushed in the housing enough so the house and blades made contact---the fan stopped.

If you've looked at the pressure & temp results you know what's happening in regard to AC performance. From all invested to date, some keys have been gleaned...

• Running an EDF is complex to set up. More so if one wants to collect system data---this must be done to make sense of what's going on. Is it all worth it? To climb the learning curve it surely is. From the point of being an "applied system," jury's out on this matter at this stage. I say this because of the complexity of it all. Mainly the wiring needed to be done.
• Rear wheel well is a torturous environment. Complex hardware in that space has to be well designed and robust. More so than the system is as it stands.
• From running the AC with a digital thermostat control, I'm inclined to believe the stock thermostat being a potentially limiting factor in how well stock AC systems cool. In a nutshell: stock thermo switch may not allow an evap to achieve its maximum cooling potential. This is an intuitive conclusion at this stage on my part. An interesting test would be to see exactly what temp that hardware genuinely operates at. If a system has been evacuated for 12 hours and no air whatsoever is introduced during the swap from evacuation to the charge, the charge is of a correct amount, oil level is correct, RD is fresh, compressor is good, and there are at least 3 condensors and a clean evap, the rear engine bay is sealed including the lid to the car and condenser surround to the lid, there are fans on the fender and front condensers, I think the evap should have the capability to reach a frosted exterior... and so cool quite well. Point is the evap may be able to do this BUT the stock thermostat doesn't let it.

AC performance in this "experiment"---as it stands---suggests the train runs well. It's just on the wrong track. The "wrong track" being the rear wheel well housing important parts. The "right train" being the digital thermostat control combined with the EDF.

On the mind's drawing board is moving the EDF from the rear wheel well to the front condenser. It would take the place of the stock fan. Great benefit in that location is the lack of torture the EDF would be subjected to. There would also be less parts called for to plumb it in. Downside might be a lack of available air. The potential problem being the size of the through hole in the tub that the stock fan uses to draw air in from behind the front bumper. The EDF used moves a lot of air---that is the entire reason for using it. I'm not keen on cutting the tub to enlarge that hole if needed. Don't have my car at the moment so can't eyeball this to draw any immediate conclusion. With the supporting system for the EDF already configured---thank you John if you happen upon this. (To John's credit, he brilliantly programed the Arduino that assimilates the data from multiple temp sensors, displays the data, and runs the compressor and EDF.) I'll get on with this new direction when I get my car back.
.

After recovering refrigerant (starting with an empty system) purge with nitrogen first.
How much for how long. Karl, I know you will enjoy figuring that one out.

Measure the low-side vacuum pressure with a micron gauge.

An alternative to using a thermostat in the evap core as the primary signal to cycle the compressor on and off: you could try a low-side pressure switch that turns off the compressor at a given PSI relative to the refrigerant temperature you are chasing.

Hope all is well Karl. Always good reading on your posts.

Griff

Karl, great to see you back on the project!

As you know I have Charlie's Mr. Ice project, 4 condensers, and his hurricane blower, variable speed fan switch and thermostat that has two LEDs to let me know when the compressor is cycling on or off.

Back in June I drove to Palm Springs, CA area. It was 107 degrees driving due west on I-10, with the afternoon sun shining right through the windshield. My dash was so hot I could not lay my hand on it for more than a few seconds. The passenger seat was hot, and I could feel the sun shining on my chest, through my t-shirt. I was getting 37 degree air blowing on me, and I was comfortable until I opened the door at the hotel and felt the 107 degrees.

The thermostat placement is tricky and I did not get it perfect. If I crank my temp to max, it will freeze the evaporator even when it is 107+ and the fan screaming at full speed. I have to dial the temp switch back about 1/4 turn to keep it from freeze up. Once I figured the real "max cold" setting, I was a cool dude.

On the way home, I took a full day detour to visit the Very Large Array radio telescope, Then down to Truth or Consequences, NM, over to White Sands Missile museum which I skipped as it was 109 I think and it is mostly outdoors. I did continue on to Alamogordo Space and hall of fame museum. I left my car parked outside in full 107 degree heat for 45 minutes or so.

I had driven 7 hours to get there, and it was 135 degrees inside the car when I got in, beyond toasty. The AC was blowing cold air in minutes, and inside of 10 minutes I was was comfortable and happy. On to Roswell, NM and it was still 105 and I stopped for dinner, then on to Clovis, NM for the night. The AC just kicked butt and just amazing.

The analog system will simply work for the condenser. Getting the sensor placed perfectly is evidently trickier than I can can get right. It is not worth trying a again, as I have near perfect AC in my 85 911.

I’ll go with zero amount and so zero duration... because I know you’re fond of goading me into proving my own stupidity—which I may have just done. But no worries. It’s all fodder for the learning curve. As I know it, N is used by HVAC pros to purge copper lines when & after brazing a line. Purpose being to rid the system of the oxide that would lead to internal scale in the system. Scale being an evil.

Speaking of scale, I know my system is rotting internally. I know this because I flushed it after the Black Death. Scale came out during that flush. Then I flushed it again when wanting to establish an unquestionable baseline for the Arduino/digital system (so there could be absolutely no doubt about the oil volume being right or wrong for the testing of the digital setup.) More scale came out during that flush….

http://forums.pelicanparts.com/uploa...1692810088.jpg

Above is what came from the rear lid condenser. (Flush fluid is being fed into the high pressure hose that connects to the compressor's outlet and is seen exiting at the end of the hose that connects to the entrance of the rear fender condenser.) Tapping on rear lid condenser released more scale than seen in photo. Am calling this “scale” because it's metal particulate. It had a corroded, old metal appearance & feel to it as opposed to being fresh metal from some damage (prior Black Death in my case.)

Why bother discussing scale? Point is I believe a system can run fine with scale so long as the scale releases in the system prior to the RD. Of course any scale’s not optimal but, how far does one have to take a system to make it perform reasonably well?...

http://forums.pelicanparts.com/uploa...1692810088.jpg

Results above took years of failures to accomplish so there's no brag here. I’ve also posted this record a number of times so it’s possibly annoying to some. Nevertheless, and why this is visually important here, these results are with a “scaling system.”

Having cut an RD open, seems to me that no particulate (scale) can make it past the RD—this is provided desiccant is always packed in every RD like the one I dissected. You and I discussed inline filters long ago—I said I wanted one just before the TXV. I recall you saying, in effect, “don’t waste your time.” Now I know why. RD is “the filter.”

To get painfully into detail, I don’t see the absolute need to N purge to preserve plumbing internals—am betting my AC pipes/condensers will likely outlive me absent N servicing. That said, I don’t know if there’s oxidation occurring in the gas side of the system. If there is scale in that section of plumbing—am referring to section of hard line that runs below passenger’s side of car—could scale make it up from the metal pipe to the compressor? Am speculating: not. If it does/can, those chunks could—possibly—get stuck in the reeds of the compressor. Perhaps this is overthinking it. You do goad me on to push the thinking envelope Charlie. Since I’m fond of operating outside the envelope, it’s all good.

Now for enlightenment from the Master Chef… What’s the bottom line with N purging Charlie?
.

Hey Glen. I know your system. Didn’t know about the LEDs. Interesting visual. You do get around! If your evap is freezing, BRAVO. That’s a fine measure of AC performance in my book. Keeping it from completely freezing and so having ice block the air channels is another measure of what I will conclude being your performance. Mentioned it in an earlier post that factory thermostat is a critical ingredient in regards to getting the most out of an evap. I didn’t say it that way but that’s the point. Whatever you have done, you’re there. Evap freezing—and not completely—is proof.
.

Karl, i honestly believe the biggest "secret" to getting the 911 AC system to work right, is a nitrogen purge. I have no scientific proof, and no data except my own experience with my 85 911 and a 86 El Camino.

Moisture is really bad pookie in any AC system. Our beloved G body 911s have a LOT of hoses, and my system has 4 condensers, an evaporator receiver-dryer and lots of places for moisture to hide.

The water turns into ice in a low pressure near vacuum. It can hide as ice for a long time, so a 20 minute vacuum pull is just not enough. I have a belts, suspenders, glue and staples, procedure to make sure my pants don't get pulled down with AC vacuum. I will freely admit, it is likely overkill.

Fact one: It is indisputable, one can't get "too good" or too much of a vacuum. Zero microns is impossible on a car AC but the lower the better. Always.

I start off with a 30 to 40 minute vacuum pull, then I pressurize with 125 to 150 PSI of nitrogen. I then go carefully over every fitting I can get to, and hose it down with soapy water looking for leaks. I let the nitrogen pressure stabilize for 20 minutes, and I photograph the gauge close up, from right in front of the gauge to have no parallax. Then I wait until the next day to see if the pressure has changed. Assuming no leaks, I pull a vacuum for 30 minutes, and then pressurize and purge with nitrogen, as it is 100% "dry" and has no moisture, but it will absorb any water. I let that sit for a while, then pull another vacuum. For my El Camino, with just a few feet of hoses, that is all I do before recharge. For the 911, I keep going into the crazy mode, of one more nitrogen purge to 150 PSI, and let that sit overnight.

Finally I pull a vacuum and remove the nitrogen rig and put in place my micron gauge. I change the oil in the vacuum pump, and let it run for a an hour or so, and check the pressure with the micron gauge. I want to get as low as it will go, but 200 microns is usually my goal. Micron are just a better way to measure really low pressures. 200 microns is 0.0038673555742632 PSI. It gets silly to use PSI at those levels.

Anyway, once it is to 200 or lower depending on the garage temperatures, I finally charge and rely on gauge pressures and garage temperatures.

As I said, I admit it is over the top, but it is just my time in my garage. I do not do AC work for anyone except myself, and I really like kick butt AC performance in my old antique cars.

Think you hit it Glen. In a nutsell: N can be used to pressurize the system to check for leaks BEFORE charging the system AND without introducing air/moisture to do so.

I recall AC discussion long ago about system o-rings moving inwards under vacuum and outwards when the system was pressurized. There was debate about this movement of the o-rings. Bottom line, a system could be sealed under vacuum but that did not guarantee a seal under pressure. Checking with N is indeed a fine idea.

What is the cost for the Nitrogen setup you're working?

.

I remember reading this too. And I think it makes sense. Also because once the AC system is running, when comparing absolute values, pressures reached on high side are a lot higher than the negative pressure of a deep vacuum.
So it is worth testing the system with the pressure values that it will have to face anyway.

I bought my nitrogen rig several years ago, and it was not really expensive, as I remember. No doubt prices have gone up.

I went to a local welding supply place that has a lot of the pressurized bottles of all the gasses. I told them I wanted an AC pressure rig and they said no problem. It is just a regulator much like any gas welding rig, of course set up for a nitrogen bottle. It has the standard fittings of a AC gauge setup, so all that is needed is the hose, and of course a nitrogen bottle.

The place I use lets you "buy" the nitrogen and the bottle is a cute little thing that is easy to move around. I just bring back the empty bottle, and they hand me a different bottle that is full, and it cost like 24 bucks for the nitrogen. Since the bottle just gets reused, I don't have to pay to have it retested, they just swap it like propane place.

http://forums.pelicanparts.com/uploa...1692897966.jpg

I also have a tire chuck for it, so it is easy to top off the pressure in a tire.

Glen... I counted your cycles: Vac - N - Vac - N - Vac - N - Vac - charge. That's wonderfully maniacal. Have to hand it to you for going beyond overkill... because you can. Am certainly not guilty of that sort of insanity myself ;). Questioned cost only because adding more gear to the AC program is unappealing (in my world.)

Am going to look into hardware and source for N and see where the chips fall. Appreciate the input / enlightenment.
.