Krade

After a few teething probs I have got my AVC-r installed and working, now come the confusing part for a real non techie, setting it up! My car used to peak at at 16.5-17 psi which by my workings out is about 1.14 bar or 1.16 Kg/cm2.

When I have the boost control off and monitor the boost it peaks at around 0.75 kg/cm2.

When it comes to setting the boost and duty cycle, I'm not too clear on exactly how to set it up.
I guess I would set the Target boost to 1.12kg/cm2 but what do I set the Valve duty to? The book isn't to clear on it as the examples given doesn't really explain much, Is there a rule of thumb for or is it just an approximation/guess and the controller will work it out in the self learning mode?


Also the book shows the self learning at 4000 rpm and mine shows it as 3000 rpm? Do I need to change smething to get it to 4000rpm?


Any help/advice or tips ?
Cheers

Krade

btt

nom

If you do a search (on 'Apexi AVC'), there's lots of bits of info scattered about the place. They don't appear too much fun to set up, no... what MY & turbo does you car have? Might be able to have a stab at a starting wastegate setting then, but it would be a bit of a wild stab
I think the 4,000 on the instructions are just as an example?

JamesS

OK....

1) Make sure you have set the sensor type, gear n/v and knee points in the `etc` screen.

2) Go to the 1st boost and duty cycle screen. Set desired boost eg 1.2bar. Then set duty cycle to 70%. If you then go to the next screen you will see that the rpm*boost table and the rpm*duty cycle table have been set to your (flat line) requirments.

3) Find a quiet bit of road where you can run from idle to max rpm in 4th gear. Start at idle and go WOT until the red line. The screen where you initially but the boost and duty cycle will now show *** - This means self learning has begun. Everytime you go full throttle the AVCR will modulate the duty cycle required to delivery the required boost target. Basically job done!

4) If boost is oscillating then your initial duty cycle setting is too far away from the `correct` setting and the AVC will not be able to converge ie the initial setting has to be about right.

5) You will see that you do not achieve full boost in 2nd, now you can go the `start duty` screen and increase the start duty in 2nd gear. Note - anything other than 0 will disable the self-learn mode in that gear.

6) Don`t bother with the feedback speed setting - does nothing!

7) You may want to shape the boost profile ie have it dropping at high rpm. You can do this in the rpm*boost table screen........

Krade

Cheers James,

I must be doing something wrong here as I tried setting it up a few times this weekend. I have tried it with the duty set to 30, 50,40,70 in both 4th and 3rd and it hasn't gone in to the self learning mode at all. I have the boost set to 1.1 do do you reckon this may be too high, even tho the ecu boosted it to 1.2??

---=CeeDee=---

I did it also this WE...boost set to 1,25b, duty to 55% (my car was boosting to 1,3b before by itself) and it hone in on it and went to learn mode. I did it in 4th, but couldnt do it from idle (couse i was on highway) so i did it from 3k rpm...well, scarry things happened on first run couse car didn't run smootly on high rpms (backfire/hesitation) but after selflearn was done it is all ok (duty cycle bar looks akward though..it has a drop @ 4k rpm to cca 40% and than jump to 80% @ 5500rpm...but if i switch to monitor mode and select boost and then go to graph mode everything looks good, no spikes, fast rise of boost and than flat line)...i adjusted 5th by -8% and 1st for +9% and 2nd +4%...3rd and 4th are still in learn mode(+-0%)...imho it's kind of stupid thing that start duty cancel learn mode (what is the point of it if you can actually learn only one gear) if you edit (which you must if you want to achive target boost in lower gears) start duty...

JamesS

If boost is oscillating, then the duty cycle you set is not close enough to the `correct` level.

A TD04 running 1.1 bar should be about 70% at 3000rpm and may go as high as 80-90% if you try to maintain to the rev-limit.

The duty cycle should not have any big `holes` across the rpm range.

Self learn mode will only activate if full throttle is held for a period of time.....



[Edited by JamesS - 12/15/2003 1:02:29 PM]

---=CeeDee=---

2 JamesS :

boost is ok after selflearn is done...

I do have TD04 and if i set duty to 70% as you suggested than it overshoots to 1,5b..so for me it is 55% droping down at 4k and going up and than down again...

according to my boost gauge and avc-r monitor no spikes...

i managed to get it to learn mode with 55% duty...

if i leave 4th & 5th gear in learn mode it overshoots in 5th gear for .15b...i think 5th is -8% but will look at it and post it...whan i think of it now maybe it's -4%..will check

forgot to mention...car is decated..supersprint downpipe...fujitsubo exhaust...Blitz SUS...HKS AFR +8

Krade

I am on the verg of re-installing my old solenoid taking the apexi one out and just using the display for a fancy boost guage if I cant get it sorted soon

each timeI ran mine full throttle in 4th or 3rd to the redline and held it for aslong as I dared with out lifting (keeping an eye on the knock link tho) went from almost stand still to 120mph before I backed off so surley it should have triggerd the self learning.

I have a 95 wrx with HKS induction, a full decat, and some remapping done to the car before I got it.

[Edited by Krade - 12/15/2003 2:05:30 PM]

JamesS

Are you sure you have set-up the sensor type and the gear selector correctly, also that the start duty is about right?

Also how have you wired the AVC-R - You have two options, one allows to monitor throttle position, the other allows you to monitor injector duty cycle. If it has been wired for injector duty cycle, the self learn function is deactivated. The AVC-R has one signal wire that you either connect to throttle postion or injector pin at the ECU. Maybe this is worth checking?

CEEDEE; If it works on your car at 55% then it is correct for your car!

I`ve found on mine, with a couple of different turbo`s that 65-70% was about right, to start!

Krade

Set for 4 cly, throttle was reading as 100% on close so i changed the direction of the arrow. How do I know what sensors to set it up for?

---=CeeDee=---

2 James: I assume duty cycle is dependant to how much was your car boosting before...my car boosted on a cold day to 1,3b so it must be lower than yours...with all cats in place i never exedeed 0,95b and if that was the case i think 65% would be a good starting point for duty cycle...i will take a picture of my duty cycle after learn and post it

nom

AFAIK...
Typically, duty cycle should be around 70%, but this is only if the car's been set up right at the factory (I believe that they fiddle around with restrictor sizes to try to get it in this region, restrictor size being the main dictator of duty cycle given the same turbo). Mucking about with the engine's VE/flow will make quite a difference as well (why overboosting can be such a problem).

Andy.F

James - This is incorrect, the FB speed is the rate of correction application with respect to offset from setpoint.
This should be set higher in the lower gears due to the greater acceleration rate requiring faster correction.

Andy

---=CeeDee=---

2 nom : well...don't know why but it overshooted a lot whan i used 70% so i tryed with 60% and it was still too much than i went with 55% which went well regarding self learn mode...

2 Andy : i don't have problem with spikes or oscillating boost so my FB is default 5/5/5/5/5

2 James : checked out my 5th gear Start Duty and it was -6%

maybe difference is couse i'm driving standard eu turbo 155kw/208bhp version...maybe difference is couse it was fogy cold morning...what bothers me most is that learned duty curve have huge oscillation (almost 40% between min/max values)...

well, maybe it's time to do one more trip and make it learn again (i'm not that exited about driving 5th gear from 1,8rpm to the redline though)

JamesS

Andy, I know what the feedback speed is supposed to do, but it doesn`t make any noticable difference IMHO.

The sensor set-up is detailed in the manual (somewhere!), its something like 4,4, up-right arrow (?) - I`ll check later.

You do not use a restrictor with the AVC-R!

Ceedee, I spend alot of time on the M-way hence leave the start duty off for 4th & 5th ie I drive in these gears so it learns in these gears. If you do mostly A and B roads then let it self learn in 3rd gear and apply a negative offset to 4th & 5th - makes sense! I would `smooth` the duty cycle though.

Basically you should leave the learn function active in the gears you most use and then offset the others......

nom

Don't worry too much that the wastegate duty cycle should be 70%, as it may be very different. This point is set by your particular set-up's requirements, not by the AVC-R (all that needs to know is what point is the right point!) so there's likely to be a bit of guesswork involved
Normally, if it overshoots, it means the initial duty cycle is set too high & should be reduced. If it doesn't quite reach the boost target, the initial duty cycle is too low & should be raised. The AVC-R might work differently, though, but the throry will be the same
Also, even if it says no restrictor should be used in the manual, personally I wouldn't believe it - if there's no restrictor, it may well mean there's too little available adjustment for the ECU/solenoid, which would result in an oscillating boost. The restrictor is there to do 90% of getting the correct boost in the first place - the ECU controller is there to fine tune. So restrictors should really be 'played with' until the final stable duty cycle is somwhere around 70% - certainly at least 50% or you may have problems with oscillations

Krade

This could be the problem with mine. Will have to look in to it.

---=CeeDee=---

2 James & nom : Hmmmm...restrictor thingy...it was placed (before AVC-R) in hose that goes from turbo to T...as i read here on scoobynet everyone suggested that restrictor should be placed in new hose that goes from AVC-R NO port to turbo...i'm confused now..

regarding selflearn...i will try one more time and see how curve will look like...btw one more thing that confuses me in whole this AVC-R thing...Duty cycle learned curve is same for all gears (u cant see different curves for different gears) and it doesnt change in real time as you drive so whats the point of gear judge and self learn if it same for 2nd and 5th...beats me

nom

Can't remember which is the NO port, but IMO it should be back where it started - basically immediately after the compressor outlet nipple, whatever tube that is
But I may be wayyy off here - I must admit that boost profiling is not one of my stong points! Come to think of it, I'm not sure if I have any strong points...

JamesS

The restrictor should NOT be used with the AVC-R!

It is only there on the OEM set-up as the boost control hardware / software is relatively crude. The restrictor acts as a damper on the boost signal.eg. gives the OEM set-up `time` to catch the signal.

One reason the AVC-R is so pricey is that you are paying for a much better control algorithm in the software and a higher quality solenoid (speed of responce etc).

With the AVC-R you bin all the OEM hoses and solenoid. NOM - not sure what your talking about with regard to needing the restrictor to give the solenoid/ECU enough range - see above!!

Also ensure the pressure sensor that comes with the kit is installed upright and not at an angle. Also make sure it has a decent pressure signal that is not being used elsewhere.

With regard to the duty cycle screen - you are correct in that it only shows `one curve` I guess the unit is able to modulate around the base curve by a couple of % either way?

Also keep pipe lengths as short as possible. I have installed the solenoid on the same base plate as the OEM one.......phew


[Edited by JamesS - 12/16/2003 1:40:55 PM]

nom

The control range has nothing to do with the control system, it's simply dependant on the volume of passing air. Opening and shutting a flapper very accurately doesn't help if there's simply too much air about in the first place!
Hence the wonder of the restrictor.
I know that my restictor is too large at the moment - changing between 23% duty cycle and 24% - atmittedly with a Link controlling it which isn't the most impressive boost controller - will change the system from underboosting (1.3 bar, creeping extreemly slowly to 1.35, target 1.4) to overboosting (1.8 bar in half a second & slamming into fuel cut). This isn't the boost device getting it wrong or reacting too slowly, it's simply the amount of air passing through the system renders it far to granular to control. The restrictor was changed (not enough), the base duty cycle was rasied to 35% and it still isn't good, but much better (34% & 36% are now under/overboost points).
The AVC-R won't be able to cope with this.
A smaller turbo is far less likely to cause problems simply because so much more air needs to be dumped to keep the system stable, hence more granularity given the same restrictor...

JamesS

So how does a restrictor reduce the volume of air in a closed system? It simply introduces a time lag into the system by reducing FLOW rate of boost signal to wastegate diaphram.

AVC-R copes fine with no restrictor as the solenoid is able to respond faster than the OEM........and it is on a `bigish` turbo.

I believe the BRD 3-port solenoid does without the restrictor - why? Cos it can!

[Edited by JamesS - 12/17/2003 8:36:43 AM]

JamesS

OOps!

[Edited by JamesS - 12/17/2003 8:35:36 AM]

EMS

This statement would be correct if it actually was a closed system. It is NOT, the boost solenoid acts like a bleed with variable flow. The restrictor causes a pressure drop when air is bleeded by the boost solenoid, this allows the boost control system to achieve a higher boost pressure as the wastegate opening pressure. The smaller the restrictor the higher the pressure drop, the higher the possible boost.

AND a smaller restrictior will slow down the response time also, you will have bigger boostspikes with smaller restrictor.

Edited to mention this is for the standard boost control system with 2 port solenoid!

Mark.

[Edited by EMS - 12/17/2003 9:08:41 AM]

JamesS

I know it is not a closed system over time....just phrasing as example....

nom

The 3-port solenoid can work without a restrictor with some turbos. Others it can't, for the same reason as described before (I use a 3-port, and it workded very nicely indeed with a MD254 & no restrictor; however, a TMI-141 proved too much for it so a restrictor is now used ).
Obviously the use of a restrictor is also dependant on the solenoid's control system - but it seems a little silly to purposefully make the control system's job harder!

JamesS

....unless using a system that does not require a restrictor due to better control and can therefore give a faster responce and quicker boost rise rate.......