I have just had the fuel distributor and WUR rebuilt on my November 1984 QV with K Lambda and installed a Tubi exhaust and a test pipe. I would like to know how to adjust the mixture. As I understand it, that is done by measuring the duty cycle of the frequency valve. I believe that I understand that procedure: measuring duty cycle by connecting the uty cycle meter to the frequency valve with modified connections which allow meter connection to the FV contacts, making adjustments as necessary to the mixture screw with the O2 sensor disconnected, testing again with the O2 sensor reconnected and the mixture screw access hole plugged etc., etc. However, I have read references to the notion that the Tubi exhaust and/or test pipe call for modified mixture adjustments. Some recommend that, with the Tubi or test pipe, the O2 sensor should simply be disconnected. Since my 308 is actually the newest of my collector vehicles, I am more a carburetor guy, but pretty ignorant when it comes to Bosch FI. Any advice with respect to test procedures and settings would be greatly appreciated. I do not have access to exhaust gas analysis equipment and would prefer not to go that route if not absolutely necessary.
I have an '83 so no lambda but I just pretty much did it by ear the old fashioned way. Adjust, drive, adjust, drive, rinse and repeat until it's right. You're probably not going to want to set it to the published emissions compliant specs anyway.
Doing it the old-fashioned way does sound appealing. I may well go that route but would like to understand the "modern" procedure. Meanwhile, to make things even more complicated, I was reading an interesting two-year-old post that says that connecting to the frequency valve does NOT give the duty cycle information you need. It said that is available on pin 17 of the Bosch ECU, but advises attaching an oscilloscope to the frequency valve connections to make mixture adjustments. Any further thoughts?
If you want to use the O2 sensor as your "gas analyzer" for the warm idle mixture adjustment: At warm idle with the O2 sensor single wire unplugged, you would target using the mixture screw to get a voltage on the unplugged wire of the O2 sensor to be maybe an average 0.6V~0.7V DC relative to ground (i.e., a little to the rich side -- this won't be rock solid and may drift around a little). Of course, you need to block the access hole to the mixture screw (like with a finger tip) so no air leaks in that way to get a meaningful reading (i.e., you can't read the voltage from the O2 sensor while making the actual adjustment of the mixture screw). The injection ECU sets the FV duty cycle to a constant 50% in this mode. At warm idle, with the O2 sensor single wire plugged in, you would make small adjustments of the mixture screw to get an average 50% duty cycle at the frequency valve (again, the access hole to the mixture screw must be blocked to get a meaningful measurement). During this process, the voltage on the plugged in wire of the O2 sensor will be 0.5V DC average, but will be "wandering" every second or two from something like 0.2V (lean) to 0.9V (rich). The attached document (Service Bulletin 10-13) is the best thing that I've seen in the F documentation to describe how their K-Jet with Lambda system works. Some FV duty cycle information is given around pages 9-10.
Many thanks, Steve!!! I'm going to have to study the Service Bulletin a few times to fully absorb it. I have continued to research this question. Is it true that the mixture screw only controls IDLE mixture? If so, why do so many people talk about monitoring the system (with oscilloscope or gas analyzer) under other than idle conditions to adjust that screw? Are they simply misguided? I understand that I have to block the access hole (and probably rev the engine after adjustments to "clear the system") and I understand that, with the O2 sensor disconnected, the ECU defaults to 50% duty cycle. As I mentioned, some have asserted that you cannot measure "meaningful," or sufficiently precise, duty cycle at the frequency valve but need to use pins 19 and 17 on the ECU. Is this just wrong? I have just had my fuel distributor rebuilt by Larry Fletcher. First, do I need to follow the procedure Larry has on his website for initial setting of the mixture adjustment screw? (Basically, turning the screw, presumably from lean to rich, until one of the injection ports begins to fill with fuel.) THEN, I like your idea of adjusting by O2 sensor voltage. So is it a matter of continuing to adjust the mixture screw until the DISCONNECTED O2 sensor gives a reading of 0.6V or .7 V DC and , with the 02 sensor connected, fluctuates between 0.2V and 0.9V but averages 0.5V? As I mentioned, I am a Neanderthal carburetor guy. Often people don't realize that some carburetor mixture judgments only affect IDLE mixture. What if I just set the mixture screw by ear and with a tach for a strong, steady and non-hunting idle that does not dip and then rebound when you take your foot off the gas?
As I recall it only adjusts idle mixture. Also, need to cover the adjustment hole to get correct measurement after adjustment..
Yes. If you set it super-rich at idle (very low airflow), it will only be a little rich at higher airflows. Not familiar with any of that (nor can recall it being discussed here). Not familiar with any of that, and see no reason why measuring the duty cycle at the FV would be invalid. Yes, that gets you in the initial ballpark. Yes, this part is OK (called open-loop mode) No for this part (called closed-loop mode). If you want to adjust with the O2 sensor plugged in, you want to target setting the FV duty cycle to 50%. The injection ECU will be changing the FV duty cycle in closed-loop mode to keep the O2 sensor output at 0.5V average. An adjustment done in open-loop mode should give a result very similar to an adjustment done in closed-loop mode. That certainly doesn't hurt, and to get in the initial ballpark of running at all that's basically what you need to do. The "problem" is that IC engines can really run quite well when set rich or even very rich -- however, this will cause the catalytic converter to overheat. This is one of the main reasons for needing "with Lambda" systems (to minimize the amount of unburnt fuel entering the cat).
↑ "Is it true that the mixture screw only controls IDLE mixture?..." Interested to read that. Many moons ago I found that the mixture on my car had been set very very rich. It happened because there was a split in a vacuum pipe that had gone unnoticed which was bleeding in unmetered air & someone had adjusted the mixture to compensate. When I fixed the vac pipe the CO reading at idle went up towards about 6. Adjusting it back to where it should be (at around 1.1-1.2) did two things (aside from it running a lot better): firstly the plugs had always previously been black when I pulled them, that changed to the more desireable biscuit colour & secondly average fuel consumption dropped from 16/17mpg to around 24. i.e. a change of around 30-35% . So I always assumed that the CIS mixture screw affected the thing under all conditions..... (UK , non cat car, UK mpg, not US)
It will affect the mixture at all conditions, but the burn rate is not perfectly linear like a modern car.
Didn't say that it had no effect at all at higher RPM -- just that the richness is reduced at higher airflow if set rich at idle. See the comments on Page 10, Chapter 7 of the Probst Book: "None of these adjustments is particularly well suited to fine tuning the fuel mixture at mid to high rpm, since any change which is effective there brings with it a more dramatic change at low rpm and idle. Thus, there are the usual side effects of hard starting, poor driveability, higher fuel consumption, and exhaust emissions. Competitors in the now defunct SCCA Rabbit/Blisten Cup racing series (for near-stock, Volkswagen Rabbit sedans equipped with K-basic) used to carefully adjust the idle mixture screw to a very rich setting - the series rules prohibited any real modifications. The result was that the large enrichment in fuel mixture at idle brought with it a smaller but meaningful enrichment at high rpm, where it would give a little more power. At idle and moderate engine speeds the cars ran very rich, but for racing, the slight power increase (at high rpm) made this trade-off acceptable." I'd add that your situation was not a case of a properly working K-Jet without Lambda system (aka K-basic) just set very rich at idle. You had an (uncontrolled) air leak compensated for by using the mixture screw to expose more of the slits when the deflection of the plunger was very small (because of less air deflecting the airflow plate at idle due to the air leak). The funnel shape is designed to give a very predictable change of airflow vs slit opening (when the airflow plate is starting from the correct idle position -- yours wasn't), and the behavior of your air leak might be something totally wacky and unknown vs engine RPM giving a very different overall result.
I can't thank you enough, Steve. I hope this information is useful to other owners. This is the article that stated both that duty cycle cannot be accurately measured at the frequency valve and that an oscilloscope should be used to take measurements out on the road. https://www.benzworld.org/threads/k-jetronic-with-lambda-overview-and-adjustment.2182425/ Granted, it does involve a Mercedes 450 SL, but it is talking about K Jetronic with Lambda of 308QV vintage. It sounds knowledgeable and refers to some other interesting articles, but I assure you I am not debating your judgment but am just passing this on for your information. I have also seen references to people testing out on the road with exhaust analyzer probes, but perhaps they just go to extremes. I myself was a fanatic, especially when it came to motorcycles where tiny adjustments can make a significant difference. As you suggest, and Iain concurs, if the mixture screw is set extremely lean or extremely rich it apparently can have some impact beyond idle mixture. I think I will go with the open-loop adjustment you outline and the old "ear method" and see where that takes me. My "final question": I noted that I am going to a Tubi exhaust with test pipe. I don't want to waste gasoline or excessively pollute the atmosphere, but with this set up and without a catalytic converter, will the car run BETTER if adjusted on the rich side or are you simply saying that it will run okay if set rich but still should be set as close to a "neutral" mixture as possible?
No, it doesn't -- I didn't read the whole thing as this is not how Ferrari implemented K-Jet with Lambda: "However, our frequency valve as used in the 1980 to 85 cars is just a modified fuel injector stuck I to the WUR control pressure loop. It's an engineering stop-gap to get cars to meet emissions regulations of the 1980s." This type of design (using a frequency valve to change the control pressure to move the plunger to change how much the slits are open for a given airflow) is how many K-Jet without Lambda Ferrari were modified to meet US importation rules, but it is what I've called "quasi-Lambda" in the past (and may use a different injection ECU). Ferrari's implementation of K-Jet with Lambda uses the frequency valve to change the lower chamber pressures in the fuel distributor which changes the pressure drop across the slits and changes the amount of fuel delivered for the same size slit opening. The reason this is a way better design is that it can respond much more quickly -- moving the fuel distributor plunger by changing the control pressure is a slow process with a lot of inertia; whereas, changing the lower chamber pressures near-instantly changes the pressure drop across the slits. Bottom line -- that article does not apply to Ferrari's K-Jet with Lambda system design. No, it won't run "better" intentionally set rich. It already has provision to increase the richness under high load (the throttle WOT microswitch closes at large throttle opening which "tells" the injection ECU to go open loop and increase the FV duty cycle to add enrichment for best power. Go for neutral or just barely above neutral IMO.
Thanks again, Steve. I'm not surprised there is a distinction and I appreciate you clarifying that for me. I'm learning.
