Talk to me about automotive A/C service and modern compressors

In dealing with the trials and tribulations of my Saab's A/C, I finally broke down and bought a manifold and vacuum pump so I could evacuate the system and recharge it by weight, which I did and it worked properly (well, it worked until I realized the compressor clutch was still Tango Uniform, but the charging part worked).

My question is regarding topping up a system that's low rather than empty. I know that in theory one is supposed to evacuate the system, pull vacuum, and then recharge, but that seems silly and wasteful unless you have a way to recover the remaining refrigerant, which I don't. I've found charts with high and low side pressures based on ambient temps, but I have also found some discussion of "controlled compressors," something my BMW 128i apparently has. According to BMW, the pressure readings don't apply to these compressors - they give list no alternative method of determining the state of charge that I can find.

That said, I put the gauges on the system - both sides were low, which I expected. There's a very slow leak somewhere in there that causes it to lose a bit of refrigerant, enough that it starts to struggle to cool effectively a year after it's charged. I want to top it up, but I'm not entirely sure how to proceed. My gut feeling is that BMW is being overly cautious and somewhat pedantic in their instructions, and that applying the standard pressures as a guide should get me back to where I should be. Am I off base on this? Is there something I'm missing in my understanding? And what exactly is a "controlled compressor"?

This is very relevant to my current situation. 

Ive got 3 systems that need work, and im about to buy a manifold and vacuum pump myself. 

Completely new at this level, as ive always had success just adding freon.

So  watching,  and hopefully learning what the non-hacks do!

Your antique Saab probably has a sight glass somewhere.

Modern stuff really should be evacuated and charged properly, but if you add a few ounces at a time, it's good when it gets cold inside...

In reply to Streetwiseguy :

Yeah, the Saab was no problem. The system was empty, I recharged by weight (a massive 1kg), and it was cold for like 5 minutes before the clutch started slipping (there's another thread on that problem - no need to go into it further here).

For the BMW, that's kind of what I'm figuring - add some, watch the pressures, and check the temps in the cabin. It's a puny system, 5?? grams, so it shouldn't take a whole lot based on the pressures I saw when I checked it, maybe half a 340g can or so.

This article explains how the BMW variable-capacity compressors work. Seems like voodoo to me, but it does list some ballpark figures for high and low side pressures (which seem a little lower than what I've seen for standard compressors), so I think I'll go with those and hope for the best.

EDIT: Here's another article that deals with the whole IHKA system and talks a bit more about using a scanner to read information electronically. I have to see if my scan tool will do this.

And this is one of the more comprehensive general A/C articles I've found. Might be of use to those not dealing with BMW's black magic.

When the system is off, you have what's called static pressure which should be the same on both high and low side. Static pressure is a decent indicator of 'charge' if you don't have a better way like evacuating and recharging by weight. As far as variable displacement compressors, the pressures still determine what's actually going to happen when it's in use. I.e. you cant have a 134a system running 90psi on the low side and still putting out 40 degree air. So when it's in use you can still look at those pressures and determine whether the cause of a cooling issue is refrigerant related or not, it just makes it harder to judge CHARGE by that because the system is adding variables.  This is not really a new thing, because there were lots of old 'dual units' (ie front and rear evaporators in big suvs and vans) that would have pressures roughly in the ballpark even though they were super low on charge.  I helped a guy working on an ~89 Suburban a week or two back and he couldn't understand why it wasn't cooling after he put ~1.5 12oz cans of refrigerant in it. He never looked at the capacity of the system which was more like SEVEN cans and once we added that it worked. So running pressures being a poor indicator of charge is not a new problem, it just works a lot better in some cases than others.

In reply to Vigo (Forum Supporter) :

Interesting explanation - I admit I'm not strong when it comes to understanding systems like this even in the best of circumstances, and the whole variable compressor thing just makes it even harder for me to wrap my brain around.

I did check and my scanner does provide live data for many, though not all, of the parameters shown in the second article I linked. Based on that, taking a conservative approach, and using both the scanner and the manifold gauges, I put in ~150g of refrigerant, or about 25% of capacity. That got things to the low end of where everything suggests they should be, and the air is blowing cold again (well, as cold as it ever gets in this car - it is not a strong system). The one thing I didn't check is the static charge - I'll do that once the car cools off.

Streetwiseguy said:

Your antique Saab probably has a sight glass somewhere.

Modern stuff really should be evacuated and charged properly, but if you add a few ounces at a time, it's good when it gets cold inside...

My old school AC guy sticks a thermometer in the vent and adds refrigerant until it cools down the air discharge enough.  I understand R-134a is more sensitive to quantity and needs to be evacuated and teh correct amount installed. 

In reply to spitfirebill :

I think there's still something to be said for results. The BMW scanner allows me to see live temp data for outside, inside, evaporator, and vent output - I'm sure it's using that data to control the system, but the latter two also provide a pretty direct indication of whether the system is within operating parameters, at least generally.

134 is dirt cheap. Evacuate and charge by weight is ideal.   Typical can is 12 oz. Say the system takes 1.25 lbs.  1 full can plus half another one and that's close enough.  System wont care if it's over or under by a bit.   that's about 8 bucks in refrigerant assuming you throw the half can away. 

In reply to TGMF :

If it leaks down over the winter to the point that it's not cooling at all in the spring, that's likely what I'll do, since I've got the equipment to do it now. At the moment, however, it was just not cooling as well as it had been, and it seemed, I don't know...unnecessary, to evacuate it when I figured a small boost would keep it going for the next two months. It's a really small leak, so it takes a good long time before it's noticeable.

Just adding is honestly fine like 99% of the time. The thing about just constantly adding is that each time you interface with the system you end up with a tiny amount of atmosphere 'air' in the system (like whatever is stuck between your fittings and the service ports on the car, or  if your hoses on your fill apparatus are 'empty' all the air in there too) and that lets in moisture. IF enough moisture accumulates in the system it can cause localized freezing right after the orifice that causes the pressure drop in the system (orifice tube or thermal expansion valve) and cause performance issues that way. However.. I've done a whole whole lot of just adding to systems over the years, and i've personally NEVER encountered a negative side effect. 

Variable compressor volume isn't new.  In the late 80's Nissan/Caloric had a valve on the suction side of the compressor that varied flow and the compressor clutch wouldn't normally release unless the cold air outlet was <38f.   Sounds like the IHKA setup "back valve" might be a compressor bypass/recirc valve that varies compressor outlet pressure as needed based on inside temperature.  

So, since this seems to be the right thread. Can i ask for some basic 101 course stuff?

Was given a set of gauges from a contractor friend,  but it seems ill need some sort of adapter to meet up with the ports on cars. Ideas? 

How do we do leak detection? Is there a simple way that isnt "stuff full of freon and dye and wait?" As that gets expensive quick. 

As ive only ever done"add a bit", whats the basic process for doing it right?

In reply to Dusterbd13-michael (Forum Supporter) :

The third of the links I posted is a pretty decent basic walk-through, at least to my untrained eye. The quick release fittings are universal, so they shouldn't be hard to find - my gauges came with them. They make sniffers for leak detection, but once you pull a vacuum on an empty system and turn off the pump you'll know pretty quickly if it's holding or not (minimum 30 minutes, should be no drop).

In reply to djsilver (Forum Supporter) :

If you look at the second of the three links I posted above you'll see that the compressor actually has an eccentric that allows it to vary the output. Honestly it seems to be another one of BMW's modern "answer to a question no one asked" solutions, but probably also produces some miniscule fuel efficiency gain.