T3--Under dash aftermarket AC unit review//install

[Piledriver]

Very nice to know Pile!....I will back read to see what compressor you are using. I have also though about making the primary four shroud bolts 10mm....just to make sure load is no issue.

The overhead fan unit I am working with will probabaly use an evap unit from a mk 4....good and easy to come by.

The overhead unit I am designing will have welded tabs to sit on the internal rails under the headliner....bolted in place. The 412 is ideal for overhead....huge overhead space...rails/flanges to mount on.....and I have had underdasb units from the factory. There is too much under the dash that needs regular service to use them on this car. Ray

GM used 3 long 8mm bolts to hold the compressor on.
My mount is total overkill, but I don't have to worry about it.

It would probably be better to use a modern spring loaded idler for belt tension.
I think the 4 rib I used is for a Dodge Neon or something, you'd want to match tension to belt width.
If I ever redo the exhaust to be a forward exit, it may grow a dry sump pump under the AC compressor.

Currently I'm more concerned the centered 10mm jack bolt will dig a hole in the mag.
(seemed like a good idea at teh time, and the end is radiused and polished)
I could ~easily make the lower bolts the jack bolts, with a pair of locknuts (inner/outer)...
MUCH easier access too.(will do that next weekend if I remember)

I considered putting 8mm timeserts in the mag fan housing, but will defer that until its an actual issue.
I used bolts with plenty of length/good engagement.

I figure if the T4s came with the big old York single lungs, the modern 5/7 cylinder compressor kicking in would be like a bug hitting the windshield by comparison.
When I turn it on/clutch kicks in/out at full AC load, you can hear/feel something happened, but it's not really enough to disturb the idle speed.

[raygreenwood]

Yes...thinking of some of the same comparisons. The original cooling manifold mount is pretty sturdy. I may simply put steel bushings around the the 8mm bolts coming out of the case to tighten up the fit between the cooling shroud and the four 8mm bolts.

I am figuring on using a pivot and jackbolt set up as well. If the jack bolt ends up against the mag shroud I will put a small steel plate under the jack bolt point on the shroud to spread the load.

The pulley and belt "may" have to do double duty. I am exploring the possibility of using either a Ford or VW water cooled power steering pump (both are about the size of my fist)...and removing every other sliding vane inside to to cut the volume and load way down...and use it for a gear oil pump system for the 004 trans in this car...which has some of the same gear oil distribution issues as the vanagons.

Ray

[Piledriver]

I was processing on how this setup works (and could work) driving home.

You (Ray) would probably probably want to start with an orifice tube setup vs. a TXV, with a clutch.(although it works great the way I implemented it with the TXV)

Basically, the compressors automagic function trys to keep the evap temp just above freezing at all times, replacing the same function the TXV does by varying the capacity to "hit" lowside pressure/temp target. (it is controlled ~entirely by the lowside pressure, as refrigerant has a very well defined temperature vs. pressure profile)

The problem is you probably don't need 35F air coming out the vents, so modern climate control systems will also mix hot air in to set the temp if you want it above a certain point. You could probably rig up an evap bypass path for recirc air with a mix door to do the same on our cars, pickup near the heater for defrost.

The compressors automatic throttling only gets you so far by itself, but bypassing a % of air out of the blowers around the evap (and letting the evap temp/control loop take care of itself, as it was designed) would get the job done.

The in dash tstat cycles the clutch for higher setpoints, just like on a txv... mine is set at about 25% (way warm) as it will otherwise freeze me out driving home at night when its only 85 out, even set with the fan turned down.

To control it fully electronically you need something with at least some programmabilty and a temp sensor as well as a "control" temp input slaved to the in dash thermostat. MS3 has a couple generic closed loop programmable pwm output channels that look quite capable, may do, you can set the inputs to look at ~anything and it has a preset table for initial conditions, so you could say throttle the compressor back based on rpm and setpoint temp (as example that may work)
You can also use the other generic PID loops output as an input...

Later model GM setups pretty much all have a clutch as well AFAICT, apparently the inability to not be able to turn the compressor completely off causes issues worth spending the $$ on.

I have my own freon recovery gear etc, so I may pop in an electric control solenoid and try some sort of compressor throttling under ECU control while keeping the TXV. (I suspect GM does this on more recent models)

I think I would need to add an accumulator on the suction side if converting to orifice tube... Its meant to keep liquid freon out of the compressor, much more likely with the orifice tube design, the TXV regulates the freon flow into the evap.

[raygreenwood]

Thanks Pile!

Yes....looking at a, few systems out there.....I just recently replaced a compressor and dryer on a 2005 sable......most of the newer systems that still use an orifice tube system with an electronic climate control system ....and no variable stroke......seem to use one or more accumulaters around the system.
They also tend to have short clutch and bearing lives if you either cant turn the system off....or dont. The constant declutching especially if as you note the system is used as temp maintenance without using outside air mix.....meaning outside air is closed off and the system is maintaining your less than frigid temp setting on an otherwise not so hot day......by compressor cycling alone.

I say short life......but not of the compressor......but of the bearing system in the clutch pulley. I have seen this a bit here and there.....and when shopping parts....various reliable parts people I know state that its hideously common ...every day in fact......that customers with cars with constant climate control systems seem to come in at relatively moderate miles and lifespans........between 5 and 7 years old and 70k miles or higher....with burned up clutch bearings and seized pulley.

The sad part is that getting the clutch, bearing and pulley off of many of these systems without trashing an otherwise excellent compressor. .....is usually a pain. Most buy a reman or new pump complete with pulley, clutch and bearing.

So...yes......many of the variable stroke systems seem to be constant run. Some things I need to ponder. Ray

[Piledriver]

Put about 6 oz of freon back in based on temp measurements, there really has to be a way to calculate the charge weight based on line length and components...


After it was working to my satisfaction, went to tighten up the belt...

I considered putting 8mm timeserts in the mag fan housing, but will defer that until its an actual issue.
I used bolts with plenty of length/good engagement.

It's not a potential issue amymore, the bolts were plenty long but it looks like most of the threads were krunchy from a previous strip-out event with much shorter bolts. One "popped" just as I finished up of course.

I would prefer to put in solid inserts but I'm about to hit Oreallys for a 12mm long 8x1.25mm insert kit against my better judgment.

Whats the OD thread size on 8mm case savers? Probably Whole Lot Cheaper than a Timesert kit.
(designed the mount so i could use a box end on the bolts, prefer to stay 8mm)

[raygreenwood]

Put about 6 oz of freon back in based on temp measurements, there really has to be a way to calculate the charge weight based on line length and components...


After it was working to my satisfaction, went to tighten up the belt...

I considered putting 8mm timeserts in the mag fan housing, but will defer that until its an actual issue.
I used bolts with plenty of length/good engagement.

It's not a potential issue amymore, the bolts were plenty long but it looks like most of the threads were krunchy from a previous strip-out event with much shorter bolts. One "popped" just as I finished up of course.

I would prefer to put in solid inserts but I'm about to hit Oreallys for a 12mm long 8x1.25mm insert kit against my better judgment.

Whats the OD thread size on 8mm case savers? Probably Whole Lot Cheaper than a Timesert kit.
(designed the mount so i could use a box end on the bolts, prefer to stay 8mm)

True value hardware stores carry metric thick wall Easylok inserts. They probably have that at Elliots in Dallas. Ray

[Piledriver]

True value hardware stores carry metric thick wall Easylok inserts. They probably have that at Elliots in Dallas. Ray

That would have been a better deal if it comes with the case (and I was going to Dallas today)
Heli-coil kits are carded so cutting up the plastic just so is required to provide a hokey storage container you can hang on the wall,, the ez-lok kits case comes with extra "holes" for the different inserts.

http://www.threadtoolsupply.com/m8x125- ... r-kit.html

I'm also not sure a couple of the holes in the Vanagon fan housing would take an M12 tap without seeing threads externally.
Helicoils are in and she's back together.

I was able to drop the compressor side of the rear bar down a couple inches and swing the compressor/hoses out of the way in about 2 minutes. Removed 3 bolts from crossbar so it could pivot.

(rant)
Why doesn't anyone make spark plug solid inserts from silicon aluminum bronze (or other flavors?)?
Its not expensive, very machinable, about the right expansion, and conducts heat and wears far better than aluminum.

You can get the plug inserts in anodized aluminum but the kit costs you a kidney, the insert costs are only ridiculous.
Steel (carbon) not a great plan, SS not a good plan at ALL as it conducts heat poorly.

I'm thinking silicon (something) bronze would be a much better solution.

[Piledriver]

Ended up taking one more pass at charging it, forcing the condenser fan on by grounding the control relay, and setting charge by high side only based on ambient temp/pressure chart just on the high side of 134a range @ 85F.

I'd rather risk a small overcharge than any undercharge due to compressor oiling concerns.
(5% undercharge=50% reduction in oil flow, supposedly, although the PAO is likely to stay in the compressor for the most part)

The TXV setup and long return line w/uphill bit by the trans should greatly reduce slugging concerns from a mild overcharge.

AC performance is truly excellent now, but there has got to be a better methodology for calculating the refrigerant charge of a variable displacement compressor system.

Interesting tidbit, temp readings indicate that a lot of cooling is happening in the high side long line to the condenser under the car.

[raygreenwood]

Why doesn't anyone make spark plug solid inserts from silicon aluminum bronze (or other flavors?)?
Its not expensive, very machinable, about the right expansion, and conducts heat and wears far better than aluminum.

You can get the plug inserts in anodized aluminum but the kit costs you a kidney, the insert costs are only ridiculous.
Steel (carbon) not a great plan, SS not a good plan at ALL as it conducts heat poorly.

I'm thinking silicon (something) bronze would be a much better solution.

Have you tried the copper plated steel spark plug inserts from TimeSert? Jake noted years ago that these are what he uses.

They work well if you PROPERLY install them...which may require slight machining to the spark plug recess to get the nearest size/length to be flush with the opening in the combustion chamber. This "can" be thought about while using the cutter toll that resurfaces the gasket seat area....by purposely selecting the next length down and cutting the seat area until teh depth is right.

http://www.timesert.com/html/sparkplug.html

Ray

[Piledriver]

Addendum as to determining correct freon charge with variable displacement internally regulated compressors:
(automotive)

//Big Fat Disclaimer:
This is MY opinion. I am not a professional HVAC guy. I am mostly self taught, but I do have a bad tool fetish and have the right equipment, and have been working on and off on AC systems for years incidentally (not for a living) from back when you didn't need a freon recovery system to work on cars, at least.

I only work on my own and familial AC junk.(not for pay)

I have also worked on helium compressors and cryo pumps, and vacuum pumps and systems incl. UHV of most descriptions. I do THAT incidentally as part of my real job.

Warranty:
If this method blow up your junk you get to keep all the pieces.
Do not even attempt without the correct equipment and knowledge, and at least some experience.
//
There is no manual that covers this subject for custom installations, and asking around on HVAC/auto AC forums got ~nothing definitive (except for learning what tools worked best, plenty of opinions there)

...but I suspect the freon load OE mfr recommend for each vehicle is determined this way, and it works.

The variable displacement compressor systems seem to like being filled to the point they cannot quite maintain 25PSI lowside, then take out an ounce or two until it just can. (typicaly 25 PSI-ish)

(~as soon as you begin to charge it/make min pressure so it can start, compressor will destroke to maintain lowside setpoint. Feed system freon just until it starts to go up, then recover a bit just until it hits regulated value again)

Note: Regulation pressure can vary by vehicle/compressor model, you'll find out what it is by observing response, or can probably find spec it by the control valve part# or vehicles tech data)

If it doesn't seem to be working like that, STOP.
... and do some more research. Something is not set up right or broken, or you don't have a internally regulated compressor. An ECU controlled setup requires special considerations/tools and not discussed here.


If you overcharge, the compressor cannot do its low-side regulation to maintain 25 PSI and it never gets to max stroke/capacity/evap temp unless its cold out, which is kinda the opposite of the point..

Charging just to upper capacity limit of regulation is the largest (probably correct) freon charge (best cooling and lubrication distribution) allows them to run at max capacity, and thinking how the system works makes the most sense.

I ended up with about an 18 oz load (allowing for typical gas losses in manifold/hoses etc) and given that the (compressors) original app took 15 oz and the condenser is same type/size and high side hose lengths to the evap are similar, its ~sane.

I have 2 freon scales, the one I like best cost me $8 on ebay, bought to see if I could fix it....and just needed new batteries.

The fancy inficon scale works to 100kg though, has come in handy as a scale (+/- 0.1 oz, can be NIST traceable calibrated) You could probably use it as a food scale... Nice looking oval metal brushed aluminum plate with a remote.

It came with some other nice toys (the nicest, an NLA 3 channel thermometer was free) from a nice retired commercial HVAC guy I bought a dead recovery system from for $100.(works great, now) The recovery tank cert expired even though it was brand new...

Safety tip of the day:
Don't fall off the roof of a Target. Probably Not how you want to retire.(if lucky)

The original dealer-installed freon (R12) load on a T3 was huge, partially because it had serpentine condensers with huge internal volume, but also due to the fact the condenser was usually back by the transmission... VERY LONG fluid refrigerant filled lines condenser to evap.

My 12"x12" condenser is a parallel microchannel flow (very low internal volume/very high internal surface area and efficiency) and lives hidden behind the R front headlight. (There's room on a T3, late models, anyway)
Typical high side(liquid refrigerant) line length for a front mounted evap, probably shorter than many.

Now, it blows 38F air now when told to, and lots of it.
Bit more load on the engine when it kicks in.

[Piledriver]

As the pictures in this thread have all been eaten by bitworm at the last server upgrade, reposting some, starting with the condenser sealed in the front fender (late T3, a lot more room there)