GTSBoy

Yes. Probably, given that there is only access from the bottom end of it, go with a drill bit. Don't start too small. 7 or 8mm is probably the right size. You want something that can make a big enough hole to do some damage, but not so bit that it clashes with the steel or binds up and breaks your wrist. A slow speed is probably a good idea too. Once the rubber is destroyed, you then have to get the crush tube off the stud, which will be the whole heat/oil/cutting exercise all over again, but this time with the need to strictly avoid damaging the stud (any further than the corrosion might already have done.

FFP is a step backwards. Would need to add a supercharger to offset the negatives. Hmmmm....

There is likely more action on FB groups (I wouldn't know, because I wouldn't go near a Meta business if I was on fire and they were the only way to put it out), but from what I gather it is mostly flogs. The real old knowledge is here. Many hundreds of thousands of posts covering almost everything you will want to know, and if you can't find what you're looking for or if the query is too current to be covered by historic posts, then just ask. Can't help you with Melbs workshops, but someone will be along to give some advice.

Not directly, but I measured the air temperature coming out the back (car running on dyno) and it was a small flow with some decent temperature in it compared to the temperature measured behind the radiator. Too many years ago to remember what the numbers were. Couldn't really quantify the flow rate though, so couldn't calculate a heat transfer rate, but my conclusion was that there was decent heat transfer to the air. Bro science, for sure. And granted, I will acknowledge that there are hot head/cam cover surfaces in there too, but they are only at ~100°C or less. I really should do it again, with my fine TC taped down onto the coils, with the cover on and off. But getting the cover off is a enough of a pain in the butt that I'm not going to do it just for fun.

The cooling is actually better with the cover on. The timing gears/belt does a good job of pumping some air in there. That's why the back of the timing cover has those louvres in it. With the cover off the air just goes away. With the cover on, it is forced to flow all the way to the back. It's still all very hot, but it is a little cooler with the cover on.

Didna think of putting some spacers under the cover? Lift it up 3-5 mm?

Yup, I'm on the bandwagon. That explains the excessively high boost and the slow control response.

You'd want to cap off the re-entry back into the dump too. Just something flat that can be held on by the Vband.

Yes, well, keep in mind that the air is intake air, which equal boost + possible oil. If there is a fine deposition of oil/scunge that then gets hot and carbonises, it could look just like that. Probably shouldn't be leaking. Might just be normal for that product. Hard to know if it is relevant.

Either the WG is reaching full opening, or it is not. The "it is not" case could only occur if there was not enough time available to swing the valve fully open during that boost event. I would consider that to be unlikely, as this is a commercial product that is in use elsewhere, so it really should work. But in your case, because there is definitely SOMETHING wrong, it should not be assumed that things like that are working as they should. You should put a video camera where it can see the actuator (if at all possible) during a run to see how far it is moving.

I think you're mostly on the ball there. With the straight gate, I suspect the weight of the spring will determine how quickly the gate can close, when not run with active pressure drive on both sides of the diaphragm. Otherwise, with drive on both sides of the diaphragm, you could almost go without a spring at all, only needing one to make sure that the thing was actually closed while completely off boost and not having pressure available to drive it closed. Butterfly valves have mostly symmetric loading when there is flow going through them, meaning that the gas hitting the upstream part of the blade is balanced by the gas hitting the downstream part of the blade, which means you don't need actuator torque to overcome any non-symmetric flow induced loads. But the gas flow does impart a purely normal load against the shaft, which transfers into the bush/bearing at each end of the shaft and does increase the torque required to make the shaft turn. Only a little, but it is there. I have no feeling for the amount of force involved in a WG application, but it certainly could make an argument for a decent spring weight being required. But all of this is just peripheral to the actual problem here.

This

You just need a datalogger of some sort. A handheld oscilloscope could do it, because it will make the trace visible on screen, so you can look at the peak, or whatever you need to look at. And there are cheap USB voltage loggers available too. You could get a 2 channel one and press a button to feed voltage to the second channel at points that you want to check the sensor voltage, when you knew what the guage was saying, for example.

While that sounds reasonable, this is definitely a boost control problem, but the real question is why are you having the boost control problem? Which is why I pondered the idea that there's a problem at ~4000rpm related to head flow. In that instance, you are not yet under boost control - it's still ramping up and the wastegate is yet to gain authority. So, I'm thinking that if the wastegate is not yet open enough to execute control, but the compressor has somehow managed ot make a lot of flow, and the intake side of the head doesn't flow as well as the exhaust side (more on that later), then presto, high MAP (read that as boost overshoot). I have a number of further thoughts. I use butterfly valves in industrial applications ALL THE TIME. They have a very non-linear flow curve. That is to say that there is a linear-ish region in the middle of their opening range, where a 1% change in opening will cause a reasonably similar change in flow rate, from one place to another. So, maybe between 30% open and 40% open, that 1% change in opening gives you a similar 2% change in flow. (That 2% is pulled out of my bum, and is 2% of the maximum flow capacity of the valve, not 2% of the flow that happens to be going through the valve at that moment). That means that at 30% open, a 1% change in opening will give you a larger relative flow increase (relative to the flow going through the valve right then) compared to the same increment in opening giving you the same increment in flow in outright flow units. But at 40% opening, that extra 2% of max flow is relatively less than 1/2 the increase at 30% opening. Does that make sense? It doesn't matter if it doesn't because it's not the main point anyway. Below and above the linear-ish range in the middle, the opening-flow curve becomes quite...curved. Here's a typical butterfy valve flow curve. Note that there is a very low slope at the bottom end, quite steep linear-ish slope in the middle, then it rolls off to a low slope at the top. This curve shows the "gain" that you get from a butterfly valve as a function of opening%. Note the massive spike in the curve at 30%. That's the point I was making above that could be hard to understand. So here's the point I'm trying to make. I don't know if a butterfly valve is actually a good candiate for a wastegate. A poppet valve of some sort has a very linear flow curve as a function of opening %. It can't be anything else but linear. It moves linearly and the flow area increases linearly with opening %. I can't find a useful enough CV curve for a poppet valve that you could compare against the one I showed for the butterfly, but you can pretty much imagine that it will not have that lazy, slow increase in flow as it comes off the seat. It will start flowing straight away and increase flow very noticeably with every increase in opening%. So, in your application, you're coming up onto boost, the wastegate is closed. Boost ramps up quite quickly, because that's really what we want, and all of a sudden it is approaching target boost and the thing needs to open. So it starts opening, and ... bugger all flow. And it opens some more, and bugger all more flow. And all the while time is passing, boost is overshooting further, and then finally the WG opens to the point where the curve starts to slope upwards and it gains authority amd the overshoot is brought under control and goes away, but now the bloody thing is too open and it has to go back the other way and that's why you get that bathtub curve in your boost plot. My position here is that the straight gate is perhaps not the good idea it looks like. It might work fine in some cases, and it might struggle in others. Now, back to the head flow. I worry that the pissy little NA Neo inlet ports, coupled with the not-very-aggressive Neo turbo cam, mean that the inlet side is simply not matched to the slightly ported exhaust side coupled with somewhat longer duration cam. And that is not even beginning to address the possibility that the overlap/relative timing of those two mismatched cams might make that all the worse at around 4000rpm, and not be quite so bad at high rpm. I would be dropping in at least a 260 cam in the inlet, if not larger, see what happens. I'd also be thinking very hard about pulling the straight gate off, banging a normal gate on there and letting it vent to the wild, just as an experiment.

Jason should have shown a real viscosity vs temp chart. All the grades have very little viscosity difference at full operating temperature.

That is a valid point.

This actually makes more sense than what I'm suggesting, which would make me feel happier anyway. I just reached for an idea on the cold side because the tests on the hot side theoretically ruled out WG problems. But now that you mention it, it did bother me that a 12 psi spring on what is supposed to be a good WG setup is on able to get down to 20 psi. That's bullshit. Read that way, that WG is not doing what it is supposed to.

My theory would revolve around the ~4000rpm point being peak torque territory on an RB, where you're supposed to be looking at max efficiency of the head, etc, and....just not having it because the skinny ports are doing something to f**k it up. I'll admit, it's a loose theory.

I'm thinking it is skinny NA Neo port sizing and cams. Somehow.

Photo of manifold showing gate location? I mean, it's 6Boost, so we probably shouldn't be worrying, but always wroth knowing what the layout is. Plumbed back to atmosphere? Or into the dump?

^ This is all good advice. I can imagine that there's some passive components in the HVAC controller that run that PWM output that could die, or suffer bad solder joints. It can be worth opening it up, taking a schmooze around looking for swollen electro caps, evidence of liquid escape anywhere, tracks that have been hot, lifted, cracked, etc. A DMM might not be suitable for seeing if the PWM output is pulsing. Might be too fast and too low voltage for a DMM to keep up. An analogue voltmeter might give a better hope. I use a handheld oscilloscope (<$100 from Aliexpress if you want something cheap). A DMM might see the voltage across the motor flicker. Otherwise, as above. If you can successfully see PWM action, then the control side should be good. If you can't see it with what you have, you might need to step up the instrumentation used, as above. Beyond that, and dbm7's advice on testing the motor directly, you're down to looking for broken wires, corroded connector pins, etc.

Ah. OK. I take it back. I hadn't looked closely at the R33/4 arms and presumed that GKTech did as GKTech do everywhere else, which is to use sphericals there. The poly bushings are made to be 100% interchangeable, should use the standard bolt just fine. Every other bush in every other place in pretty much every other car, does.

The J arm doesn't have bushes either. Assuming that by "J arm" you mean the part of the upright that runs down from the upper arm's outer bushes to the top of the hub. That has a kingpin style bearing in it. If you meant the lower control arm, it has 1 bush, at its inner end. If you have PU in there, that is superior to Nismo rubber. If you meant the caster/tension rod - it has 1x bush at the front end, and again PU is superior to Nismo rubber. But as I said above, I would definitely get the GKTech arms for that, as sphericals slay all other options there.

No. GKTech arms have spherical bearings in them. No bushes. You will not need bushes for those arms. The sphericals are a bit of a maintenance nightmare. I have replaced all of mine several times in the 5 or so years I've had them, and I have the arms out regularly to clean and lube the balls. Worth the pain on an R32, because the standard arm design is trash. If you need the camber adjustment, there are other options (than the GKTech ones), although I would still lean towards and prefer the GKTech ones, even with the maintenance load of the sphericals. The caster adjustment is also highly valuable, allowing for setting the car up to drive straight. There are a million options for these, including the GKTech ones. I've had Tein rods on mine for 20 years and the balls are much less trouble in that location. Never given me a moment's pain. All positives, no negatives. I consider them compulsory.

Most driving should* be done on one side of single lane divided roads. In the RHD world, you drive on the left side of the dividing line and the road is probably cambered equally on both sides. So your side of the road slopes away to the left. The same is true for the LHD world, just everything swapped to the other side and opposite slope. With a perfectly neutral, straight ahead wheel alignment designed to drive straight on a perfectly flat surface (or at least one that is level on the left-right axis, even if it has some slope in the fore-aft axis) you will not be able to drive on a cambered road without the car wanting to drift down the camber. You will need to add steering input in the opposite direction all the time. This is annoying. The solution has always been to set the camber and/or the caster to produce a continuous turning force in the opposite direction of the camber. The car will drive straight on the kind of camber for which it was set up, presumably as described in the top paragraph. But.... when the car is set up this way, as soon as you get into a lane, usually on a multi-lane surface road or highway, where the camber is not as presumed during setup, the car will usually pull to one side. In the RHD world, if you are in the fast lane on a big divided road, you are probably on the opposite camber compared to what the car was set up for (ie, sloping down to the right) and the combination of the setup and that camber will make the car want to go right pretty hard. Even a perfectly flat lane will tend to want to go right. There's no getting around it. Civil engineers who know their stuff (which is not an assumption that can always be made) will attempt to keep the variation in camber across a multi-lane road as small as possible, and if they can will attempt to make the fast lane as close to flat, or even cambered in the same direction as all the other lanes. This takes a lot of planning for drainage, control of levels, ability to deal with the elevation changes that occur at road junctions, etc etc. So it's not trivial to get it right. When they do make it work, then the annoyance is reduced, along with tyre wear, fuel consumption, etc. In theory, the civil engineers are supposed to worry about those aspects of road design also. * This used to be true, but now with very large highway systems, even just multi-lane surface roads running everywhere, it is less true now than it was, but the old assumption is the basis for describing the phenomenon, so let's just run with it for the moment.