Ok, so let's go over two of my concerns in detail so that we can potentially cross them off the list.
1) It's important to maintain diameter as much as possible to preserve a true reading both at your wastegate and solenoid. The reason is because total pressure is a function of static pressure and velocity pressure. Pressure(total) = Pressure(static) + Pressure(velocity). In simple terms the actual pressure you will see is made up of a static pressure, plus velocity (also known as dynamic) pressure. Static pressure is a pressure like the pressure in a balloon or an oxygen take. The air isn't moving, but if you put a gauge on a balloon or pressure tank, you it will measure a pressure reading because the walls of the vessel is exerting force on the air. Velocity pressure is different. It is the pressure of the air actually moving. Imagine putting your hand out the window of a car moving fast. You will feel the air pressing against your hand, even though there are no boundaries/vessel, like in a balloon or oxygen tank. The boost pressure air coming off of a turbo is made up of a combination of velocity pressure and static pressure. The tubing is creating a "vessel" and creating a static pressure, but the movement of the air is also creating a velocity pressure. The sum of both is the total pressure, and in this case would be the measured "boost" pressure that we are trying to control. One potential problem I see is the "choking" down of different barb sizes at different components. Meaning, if the reference pressure barb size is a different diameter than the solenoid, there could be a potential for different measurement readings. That's because given a mass flow rate, a chance in cross section area will change static pressure to velocity pressure and vice versa. This diagram below demonstrates what I mean:
This is basically how a nozzle works. Given a certain flow, if I choke down the diameter I can convert my static pressure to velocity pressure. This could cause some discrepancies depending on how the each component is measuring pressure. Sorry if I'm getting too technical (I am a Mechanical Engineer :redface
. Basically what that means is that if you have various barb sizes and diameters, there is the potential you could be getting different readings from the same reference pressure. In other words, if you had the same 12 psi coming from the boost side of the turbo feeding two different devices with different diameter inlets, there is the potential that they would read two different values from that identical pressure (even if the tubes connecting to them were the same). So even if you used adapter barbs to maintain the tube size, if they must pass through a specific orifice size, that could throw off your pressure reading.
Also, you'll want the length of piping/tubing to be as similar in diameter and length as possible. If one run is significantly larger, that could also throw off your reading. Imagine a given pressure, but two different sized balls. One ball is the size of a basketball. The other ball is the size of the Goodyear blimp. If you blew into each ball the same amount of air, which ball would register a higher pressure reading? Obviously the smaller one, not the Goodyear blimp. So if the tubing to wastegate is significantly shorter than the tubing to the solenoid, that could also cause a mis-matched reading.
2) So having said that, many boost controllers can compensate for that using a feedback loop. When you mentioned "gain" that immediately indicates to me that you have a feedback loop (aka PID loop). There are two types of boost controllers: open and closed. In an open loop, the solenoid valve position is pre-fixed for given conditions. It's like if I charged a flat fee "per plate" at a buffet. The other type of control system is a closed loop. That's where the system looks at what is going on, then decides what to do at that point. In the analogy above, rather than charging a flat fee "per plate" it would be the equivalent of looking at the plate, adding up all the different things in the plate, and then deciding how much to charge. So a plate with more lobster tails and steak, would cost more than a plate with lettuce and tomatoes. Whereas in the open loop, both plates would have cost the same.
How that relates to controls in a solenoid is that in an open loop control system, the solenoid opens to certain pre-fixed positions based on a boost pressure. It is much more simple and straight forward, but it can't take into account variations (like restrictions, dynamic changes to the system, or unequal system flows described above). A closed system on the other hand, will constantly monitor these valves and "self correct" as the system changes. This is obviously better because it's like a guy watching over your system and should anything change, he automatically changes the "programming" to compensate for this change. There are two main variables connected with this "self analyzing" that are critical. They are sample rate and gain. The sample rate is how frequently the system looks for measurement. The gain is how "fast" the system reacts to those measurements. While it would seem intuitive to increase the sample rate and gain so that the system "checks" itself frequently and "reacts" quickly, there such a thing as too much. Too much sample rate, and too much gain will result in a system that overreacts and over shoots it's intended target. A perfect example of this is how people react to a car that is fishtailing. They tend to overcorrect and fishtail the car in the other direction. This is one possible cause of boost creep. The graph below is a very indicative explanation of too much gain and overcompensation:
If you are trying to get to a value of 1 (like 1 psi of boost) you ideally want to something that looks like the red line. The black line would be a case where you have too much gain (too much overcorrection) and it overshoots your target. Like in the car fishtailing example above, it would be someone overcorrecting the steering wheel position and pushing the back end of the car in the opposite direction. Then correcting back in the opposite direction to compensate for the overcorrection. Ultimately what that means is that you will get more boost than what your target boost is; aka boost creep.
A few ways to check to see if you have too much sampling/gain on your closed feedback loop is to adjust the values of sampling rate and gain. Rarely you'll see boost creep with too little gain, although there may be a very few cases where this could happen. However, mostly likely you have too high of a sampling rate and gain and are over shooting your intending boost level. The other option is to see if you can put your system on a open loop and see if you get boost creep then. If you stay conservative with your open loop values, there should be no way to get boost creep: at least from a programming situation.
Sorry I got overly technical (at work I call it nerding it up). While the probabilities are small that either one of these are the root of your problem, they are REAL, ACTUAL and PHYSICAL issues that can occur, so it would be prudent to rule them out before moving on to the next diagnostic step. So in summary:
1) Try to maintain diameters from start to end (turbo boost outlet to wastegate reference port, solenoid, to wastegate control port). If you can't for some reason, try to maintain the same pressure reading across each component.
2) Cross check your closed loop PID control on your solenoid by either playing with your sample rate and gains, or switching to an open loop. ALWAYS stay conservative with these values to ensure you do not go lean and blow up your engine.
If you can rule these to items out, we can start to look into other areas that might be causing issues.
Again, sorry for the lengthy post, redundant info you may already know, and "nerding" it up with technical mumbo jumbo. But at least if I break it down comprehensively, at least it will be recorded and searchable sometime in the future for others to potentially use. :smile:
