Saab's APC (Automatic Performance Control) is an electronic turbo boost control system that adjusts maximum boost pressure as a function of engine knock (detonation) and RPM.
It consists of:
(1) Knock sensor
(2) APC control unit
(3) Pressure transducer
(4) Solenoid valve
(5) RPM signal from ignition distributor
Its working is as follows. Normally the maximum turbo boost (say, 0.7 bar) is set by the turbo wastegate actuator. Therefore the boost, in conjunction with the static compression ratio of the engine, must be set at a value at which no pre-detonation would occur even under adverse conditions (ambient temperature, engine's state of tune, fuel quality, etcetera). That in turns means that with turbo boost levels high enough to make an appreciable difference in relation to normally aspirated engines of the same size, the static c.r. must be quite low (in the case of non-APC Saab engines, 7.2:1).
With the APC system, the engine's static compression ratio was raised to at least 8.5 :1 and the wastegate was set to open at a low value (like 0.3 bar) so that even under worst-case conditions (low grade fuel, engine in need of tune up, high ambient and running temps) it would steer clear from pre-detonation. This is called 'basic boost'. However, when and as long as conditions allowed, the solenoid valve in the air line to the wastegate actuator would stay closed so that the pressure went right back into the intake instead of opening the wastegate, until the maximum boost pressure set by the APC controller was reached, or the knock sensor would detect detonation. In this case the solenoid valve would open and the wastegate actuator worked just like it was directly connected to the pressure side of the turbo.
The APC system was a major breakthrough in turbocharging and allows your engine to get the best performance from whatever fuel you're running at given conditions. That doesn't mean there is no scope for improvement. As long as there is no input from the engine knock sensor, APC will allow a maximum amount of boost following a programmed boost curve. This 'ceiling' was based upon the properties of the stock engine, fuel economy, and the driving requirements of the 'average' owner.
Do keep in mind however that the system is designed to 'bleed off' a certain amount of boost in a given timespan, i.e. from stock maximum boost to basic boost in a time short enough to prevent the engine from being subjected to harmful detonation. Raise the boost 'ceiling' and it might not return to the 'safe' basic boost level before pre-ignition occurs. So effectively, turning up maximum boost also alters the 'floor' level of the boost map, in which the APC system is able to fully protect the engine. Which in turn means cranking up the boost really high on a bone-stock engine might not be such a good idea. APC will cut back boost sooner and further than with stock maximum boost as cylinder temps rise under acceleration, but it might yet not bleed off enough boost soon enough to prevent your engine to blow its headgasket or melt a piston... Typically you want to 'tweak' the APC to make full use of other modifications that allow you to run more boost in the first place, like intake/exhaust, intercooling and fueling mods. Things that allow you to run more boost under optimum conditions will also raise the detonation threshold in worst-case scenario's, so the range of boost pressures in which the APC system has to work will stay about the same.
Which APC do I have?
A useful list of APC controller parts numbers for all APC-equipped Turbo models/years.
Basic boost only fault diagnosis
Although the APC system is very reliable, it is quite common for APC-equipped cars to have basic boost (the boost level that is determined by the wastegate actuator and is considered to be "safe" in a worst case scenario) only. If the boost gauge needle of your saab Turbo does not go any further than about 40-50 % of the yellow zone on full throttle, this is probably the case. It can be caused by something simple like a loose or porous vacuum hose, or by a failed or not properly connected electrical component of the APC system. Townsend Imports provide a handy troubleshooting guide.
Tweaking the APC controller
Basically there are two types of APC controller to be found in 900T's and pre-DI 9000T's: the under the back seat type (pre-'86 900 turbo's) and the ones located in the wheelwell area of the engine bay (9000, '86 on 900). The later ones are mounted in a bracket and open in two 'halves' using four screws to reach the trim potentiometers on the main circuit board, the ones under the backseat are bolted to the floor and have a plastic strip with a tamperproof seal on one side that covers the trim pots. There are three potentiometers on the ECU board. Om pre-'86 models you can reach them by removing the seal on the dark grey plastic strip which runs over the APC box; on '86 on cars you can remove the top half of the box with four small screws without removing the APC controller from the car. The pots are marked "P", "F" and "K" (from left to right, "P" and "F" being relatively close to each other). And that's where the debate starts, as you won't find the meaning of these markings in any (factory or aftermarket) manual...
...as published onto the Web in a TSN posting by Larry West back in 1992, when he tweaked the APC of his '84 900T to give 12.3 psi (0.85 bar) max boost. Remember this was at a time most turbo Saabers were shortening their wastegate rods to achieve higher boost on APC-equipped Turbo's, thereby effectively neutering the APC system...
According to Larry's original posting the function of the trim pots was as follows:
P = maximum boost pressure
K= knock sensor input sensitivity
F= frequency at which the system operates the solenoid valve (set at 12 Hz)
In practice you would only use the pot marked with 'P' to tailor maximum boost level. 1/4 turn clockwise usually gets you a noticeable increase without hitting the overpressure guard switch which is located under the dash left to the steering column (at least on LHD cars ;-) and set at about 0.2 bar over stock maximum boost. If anything, I would only use the 'K' pot to slightly increase the system's sensitivity to knock; you certainly don't want it to be less sensitive to the dreaded pre-detonation. Remember it's the high-frequency knock that is imperceptible to the human's ear that is most dangerous to your engine. The 'F' pot, finally, is best left alone unless you have an oscilloscope and really know what to look for...
An alternative view
It has since been pointed out by some, that the "F" pot does not have anything to do with 'frequency', as it's only a part of a voltage divider consisting of + --> R (resistor) 40 --> F --> pressure transducer. According to these sources, the reference voltage that the APC is working with is taken out between R40 and the transducer. This would mean the "F" pot setting also directly influences max boost. This view is supported by the results of comparisons between standard APC boxes and original Saab upgrades, like the 'Red Box' APC as described here. In all cases, settings of both "P" and "F" are significantly higher than standard.
Here's what Kevin Kelleher has to say about it in a TSN posting of May, 1999:
Most people/sites suggest that to crank up the boost you just turn in the 'P' screw in the ECU. Unfortunately this myth was supported by a Winkler article in Nines. based on extensive testin on my '83T over many years, I believe the proper way is to turn in the "F", and then turn in the adjacent "P" screw a little as a fine tune. The 16v's, up to M93, have essentially the same system. The screws (i.e. pots, potentiometers) are as follows:
P = Percent solenoid 'default' duty cycle
F= Full boost max boost vs. RPM map
K= Knock (knock) sensor sensitivity
Below about 5-6 psi and under 3,100 RPM the solenoid valve is 95% open, bleeding off the wastegate actuator to keep it closed. Once either of these limits is exceeded, the solenoid changes to a 'default' duty, set by the 'P' screw. This partly closes the valve, and reduces the rate at which pressurized air is bled off the actuator. This allows the wastegate to crack open and slow down how quickly the boost rises, all to avoid pressure spikes and severe overshoot.
The 'F' screw sets the max pressure allowed before the sloenoid valve quickly changes from the default 'P' duty cycle to a more closed condition in an 'active control' mode. This feeds more pressirized air into the actuator to open the wastegate more and reduce boost. The pressure map for my '83 T is flat up to 3,100 rpm and from there it tapers down 2.5 psi by 5,500 rpm. If I turn in 'F' fully, the controlled max boost limit is 12.5 psi up to 3,100 rpm, dropping to 10 psi by 5,500 rpm (the 16V max 'F' map should be higher).
The 'P' screw trims the rate at which boost rises. I found that turning in the 'P' too much results in excess overshoot, and an annoying run-away boost climb with just slight pedal depression in upper gears. However, if 'F' is turned in (increased) without increasing 'P', then the 'P-default' will open the wastegate before the higher 'F' limit is reached, and no significant boost increase is achieved. A relatively low 'P' setting limits higher RPM boost.
In my own experience, the "P" and "F" pots have about equal impact on the max boost levels actually reached, but in different ways. Turning in the "P" pot without increasing "F", as I first did on Saabine, does result in higher max boost levels, although this higher boost is not sustained for more than a few seconds. It does, however, provide a useful initial boost 'surge' on acceleration before cylinder temperatures rise and boost drops to pretty much the stock maximum boost or even a bit higher (given adequate fueling and intercooling as well as an in-tune engine). The effect is comparable to the 'overboost' Turning in 'F' on the Silver Arrow, with just a little adjustment on 'F' as Kevin Kelleher suggests, resulted in a 'flatter' boost curve with maybe 0.1-0.2 bar of extra boost towards the higher revs, but the boost does not build up as fast as it does on Saabine. If you look at the specs of the 'Red' APC box and the 'M'-type APC used for the Scandinavian saab 900 Turbo (8 valve) race series during the late Eighties, it's obvious that both the set values of 'F' and 'P'-pots are significantly higher than on standard APC boxes - which provides us with a logical answer. The combination of both potentiometers provides us with a big playground to tailor the APC boost behaviour in order to match the car's specification and desired characteristics.
DIY APC Controller mods
Paul Ellis has modified an APC control unit after comparing it to an 'APC upgrade' supplied by a well-known, but now defunct Saab aftermarket tuning company. Here is his guide to a DIY upgraded APC box:
Justin Vanabrahams' tweaking of the APC box innnards has intrigued me and I had the opportunity to compare a stock APC box with a Group 6 box, both from an '85 900T16v, and more importantly duplicate, or try to dupe the enhanced APC box, both owned by the same individual (not me).
Lots of disclaimers: Perform any and all of these mods AT YOUR OWN RISK. It would be an awfully good idea to mark and identify the parts substituted and/or replaced, in case one wants to sell their car and restore the stock functionality of the original APC.
I'm not certain whether these mods will work on 8v cars (the circuit board on 8v APC controllers is different- for instance it doesn't have a sub board. However an under-the-backseat APC box from a 16v will work on an 8 valve car with same. Eric), but who knows. Modifications have been tested on an initial basis, and the car they were guinea pigged in didn't grenade. Please don't make these mods commercially available. Furthermore, enjoy!
Here's the list of mods that I was able to detect and performed on the stock APC box, that seem to make a real difference in boost behaviours. I might have missed a component or three, so do take care.
APC box modified: P/N 75211826. 14-pin (or so) connector, main board P/N 5510779-01E. Sub-board P/N 55-11062-01A.
Main board: (all part placements made on component side, plug at 6 o' clock position):
-Add 1 megaohm resistor between bottom pin of R47 and RH pin of R83
-Add 1.5 megaohm resistor between top of R36 and jumper P03
-Over capacitor D24 (Thaddäus Delebinski just brought this to my attention. I think this should be a diode, not a capacitor - Eric), add 2.2 megaohm resistor (piggyback). I'm guessing this might be a dummy circuit to detect copying of the G6 mods.
-Substitute C1 with .22 microfarad mylar film capacitor
-Substitute C6 with disc capacitor (guessing .473 microfarad, number of cap was "473". I had in junkbox and used one with value of "472".
Important note: A '473' capacitor has a value of 47 mF, whereas one marked with '472' is 4.7 mF. Although Paul Ellis' DIY modified APC box seems to work very well as described, it is strongly advised that you use the same value as in the original, professionally modified box from which the component changes were copied.
-Adjust "P", "F" and "K" pots to taste. I turned the "P" pot 1/8 turn clockwise, the "F" just under 1/4 clockwise and chose to leave the K pot alone (chicken, yes).
Resistor values 1/4 or 1/2 watt, not critical.
The results were promising. Boost ran right up to the same mark as the base G6 box. Ran into the fuel cutoff, and didn't have a calibrated boost gauge to measure boost levels. Ran helfway into the red, good recovery patterns with F at new setting used. Car was an '85 5 speed 16vT, with rising rate fuel pressure regulator set at 3.0 bar, MSS sports exhaust.
Ladies and gentlemen, START YOUR SOLDERING IRONS!
DIY "Red Box" and "Type M" APC
Here you will find the specifications to replicate the "Red Box" APC. Apart from the settings of "P" and "F"-pots, the only differences between the two boxes are two resistors. The same goes for the type-M APC box for the 8 valve turbo.
Some useful comments and info from Finland
After reading the APC articles on this page Jari Madetoja from Finland wrote me with some highly interestingdsta and insight. It mostly supports keven Kelleher's view on the working of the APC trim pots, but goes more into detail about the relationship between the "P" and "F" pots, and the resistors R42 and R138, and give some more pointers about what results to expect from altering their values. Here it goes: