Apexi AVC-R
#3
Thread Starter
Junior Member
Joined: Jan 2002
Posts: 45
Likes: 0
From: Australia
on the manual, said, it start self-learning when stable boost achiev. so can i go straight to self learning mod without any duty been set??
at moment, we have overshoot and unstable problem.
when we set the target boost at 1, actual boost goes to 1.2
also boost is not stable when push hard.
at moment, we have overshoot and unstable problem.
when we set the target boost at 1, actual boost goes to 1.2
also boost is not stable when push hard.
#4
Originally posted by Rotor Reaction
on the manual, said, it start self-learning when stable boost achiev. so can i go straight to self learning mod without any duty been set??
at moment, we have overshoot and unstable problem.
when we set the target boost at 1, actual boost goes to 1.2
also boost is not stable when push hard.
on the manual, said, it start self-learning when stable boost achiev. so can i go straight to self learning mod without any duty been set??
at moment, we have overshoot and unstable problem.
when we set the target boost at 1, actual boost goes to 1.2
also boost is not stable when push hard.
You can set the duty cycle to undershoot or overshoot your desired boost level. If you set it really high, for instance, you will overshoot your target boost. Set it too low and it will rise more slowly and not overshoot. There is a graph in the manual that shows this effect.
You can also set the learning cycle thingy (a technical term ). There is a setup screen where you can set this for each different gear (assuming you setup the gear judge already). This setting affects stability.
I put my AVC-R in and ran the car for a while before reading the manual carefully. The manual has a flowchart that takes you through the setup process that you might want to go through, or double-check if you went through it once already.
Also, you might want to start by double-checking your installation. I had to switch one of the nipples on the solenoid to a different port to get it installed correctly for external wastegate (two ports) control. And I still think my install needs some tweaks to ensure an unrestricted flow path between manifold boost and the port on the side of the wastegate. If that flow path is too small or has too much restriction, you'll get boost spikes and generally less than ideal operation. If you can't get the boost low enough (at a reasonable duty cycle level), you might need to add restriction to the hose that goes from the manifold to the solenoid. In monitor mode, you can watch it learn/adjust the solenoid duty in reponse to boost.
-Max
#6
I've been giving AVC-R advice out, and I've had mine hooked up wrong for months! My boost control lately has been my right foot, because my car just boosted out of control. My wastegate stuck shut and killed my motor when I was running with no boost controller (just a hose to the WG), so I figured that it finally just died and was stuck shut for good.
Tonight I took a little of my own advice and double-checked my connections and it turns out I had the hoses to the solenoid swapped. So basically, I was just totally bleeding the boost signal off and the wastegate wouldn't open. I also pulled one of the wires out of the MAP sensor connector somehow (I think it got snagged by the steering column). Anyway, I fixed the connection and moved the wire to safety so it wouldn't happen again. And then I went for a drive...
This thing works pretty great. I set my boost to 1.00 and the left the duty at 20%. It spooled up slowly with that setup, so I bumped the duty up to 60%. Much better, but the boost seemed to creep up at high RPM for some reason. Turned the duty back down to 45% and it seems to be working pretty well now.
The learning is pretty aggressive, so you should be able to get stable boost as long as you are anywhere near the "right" value. They have a little test procedure and a formula you can use to guess a good value.
In watching the duty cycle, I noticed that it goes to 90% as soon as you start to see some boost, presumably to hold the WG shut for fastest spooling. Then it quickly drops to about what I set it to and then learns from there. Watch what it does on graph mode showing boost and solenoid duty cycle.
I am happy with my boost control setup now. I even got to stomp on an Eclipse while I was out testing. I killed him on the first two races and I think he turned up the boost for the third because it was much closer (though I still pulled away as the revs climbed). This Eclipse was a sleeper -- totally stock looks but pretty fast. He asked me if I had nitrous when we got to the next stop light.
-Max
Tonight I took a little of my own advice and double-checked my connections and it turns out I had the hoses to the solenoid swapped. So basically, I was just totally bleeding the boost signal off and the wastegate wouldn't open. I also pulled one of the wires out of the MAP sensor connector somehow (I think it got snagged by the steering column). Anyway, I fixed the connection and moved the wire to safety so it wouldn't happen again. And then I went for a drive...
This thing works pretty great. I set my boost to 1.00 and the left the duty at 20%. It spooled up slowly with that setup, so I bumped the duty up to 60%. Much better, but the boost seemed to creep up at high RPM for some reason. Turned the duty back down to 45% and it seems to be working pretty well now.
The learning is pretty aggressive, so you should be able to get stable boost as long as you are anywhere near the "right" value. They have a little test procedure and a formula you can use to guess a good value.
In watching the duty cycle, I noticed that it goes to 90% as soon as you start to see some boost, presumably to hold the WG shut for fastest spooling. Then it quickly drops to about what I set it to and then learns from there. Watch what it does on graph mode showing boost and solenoid duty cycle.
I am happy with my boost control setup now. I even got to stomp on an Eclipse while I was out testing. I killed him on the first two races and I think he turned up the boost for the third because it was much closer (though I still pulled away as the revs climbed). This Eclipse was a sleeper -- totally stock looks but pretty fast. He asked me if I had nitrous when we got to the next stop light.
-Max
#7
I find that my AVC-R takes at LEAST 3 to 4 "cycles" to stabilize the boost.  The manual should descibe what is the proper way for it to learn - read the manual.  There should be one WOT run from a relatively slow speed, then several 5th gear pulls...
-Ted
-Ted