Have you got a boost surge after a re-map?

[urbanfox]

Hi Ben

Still waiting for you to get back to me. PM'd and emailed you....

Thanks

Why dont you just call him?

[morgan349]

Why dont you just call him?

Doubt there will be much point calling the office at 8pm on a Sunday mate

[urbanfox]

Doubt there will be much point calling the office at 8pm on a Sunday mate

Yeah, except it looks like you've been waiting for a reply since mid may lol, so why not pick up the phone and call him tomorrow?

[morgan349]

Yeah, except it looks like you've been waiting for a reply since mid may lol, so why not pick up the phone and call him tomorrow?

With all due respect (in the interest of keeping this thread on track), unless you either are suffering the same issue with your Shark remap, or you have some useful technical input on the subject, please resist pressing that 'add reply' button. Thanks.

[shark_90]

Hi Ben

Still waiting for you to get back to me. PM'd and emailed you....

Thanks

Got an email from 20:01 on Sunday night but can't see any others? PM box is massively behind with 69 unread messages.. Will deal with it today.

[morgan349]

Got an email from 20:01 on Sunday night but can't see any others? PM box is massively behind with 69 unread messages.. Will deal with it today.

no worries Ben. Flat out today so hard to chat to you about it but just let me know. Thanks

[PaulMoVRS]

Thought you had it sorted ollie must be very frustrateing :-(,

My map seems to be fine no surgeing no flatspots. The only issue i might have is i don't think i am getting enough torque between 2.5k and 3.5k not as fast as what i was expecting, ben has told me he will try my map on another car to see if i have a issue or if i am talking crap. I just expected there to be a bit more grunt in 3rd and 4th gear in that rev band.

But i am happy as i have no annoying surgeing :-)

Paul

[shark_90]

I still have a flat spot :( In the same place at 3.5k most noticeable in third gear. I have just changed my route to work so doing a lot of b road driving which has given me a constant test of the car. Its difficult to detect when fully accelerating which is why I held off sighting it as still being a problem. Quite frustrating as I thought it was sorted. Dropped Shark and email to make them aware.

Email received, if you can keep in touch with Cheryl about coming back down then we'll get you booked in. I have to say I'm surprised, so hopefully it will be something simple.

[OllieH]

I am going to do a bit more testing to make sure I can reproduce it before I come back down. I am still not sure, with all the issues its hard to relax so every time I hear a dip in boost noise its making me a bit paranoid. I suppose unless your on full power you will get dips in boost. :think:

[morgan349]

I am going to do a bit more testing to make sure I can reproduce it before I come back down. I am still not sure, with all the issues its hard to relax so every time I hear a dip in boost noise its making me a bit paranoid. I suppose unless your on full power you will get dips in boost. :think:

Ollie, Ben is away on Holiday until Wednesday so you probably will have a good opportunity to test the car thoroughly before then....

[RickRST]

Hi All,

I stumbled on this topic through Google and thought I would add my input.

I am an instrumentation engineer for large company, have a degree in Physics and in Electronic & Electrical engineering. In my spare time i build and map engines on different ECU’s including motronic.

The problems being described seem to be in two different issues – Surging and holding back, which are slightly different to each other.

Firstly, surging is where the car is pulling hard but the boost isn’t holding steady. This can be both hardware and software based. The N75 valves used to control boost aren’t the best, they wear, stick and fail, and often won’t give any fault codes. If yours is a few years old, or you are having a remap or any change of tune then I strongly suggest you fit a new one, they aren’t expensive. They can work at one moment, and then not the next. To complicate things, ME7 adapts it’s boost control to compensate for things like wear. You can fit a new N75 and things won’t be right because it has adapted to the characteristic of the old N75!

The N275 valve is a bit of a pain, and should be removed or bypassed on a tuned car. It controls how the DV’s react. It is connected to a vacuum reservoir which is charged by the engine during part load. As a safety feature the N275 can be triggered by the ECU which will cause the DV to open, and the engine to lose boost. It is overkill though, there are that many other safety nets on an E-Gas system it’s not needed. It has other downsides, especially on part load and trailing throttle conditions – especially on remapped cars running more boost. You tend to get sudden drops in power as you modulate the throttle while on boost, as the N275 makes the DV overly sensitive and dump boost when you don’t want to.

Software wise, a lot of tuners request too much load from the small turbo. When you run the turbo on the ragged edge like this, small changes in load can lead to swings in the boost pressure as N75 fights to make the turbo keep up with demand. The whole system becomes unstable and leads to wild swings in boost. There are two ways to cure this. Firstly, you can reduce the boost in certain areas but that can make the car flat. The correct way of doing it is to tune the boost control routine – few people seem to be able to do this properly though! You can control how quickly the N75 reacts to changes in load and requested load so that you don’t get these oscillations. On a custom map I would spend a lot of time on this as you are essentially able to reduce lag/time to reduce peak boost. Another thing tuners do when they can’t get the boost to do what they want, is bypass the closed loop control all together. This has its upsides and downsides. The biggest downsides are that you gain lag in certain areas of the rev range as you are relying on a fixed duty cycle, and you lose certain boost related safety nets, although there are others which will usually catch you. The upside is that you rarely get oscillation due to incorrectly tuned PID and load requests, and that you can run more boost than your map sensor is capable of reading. If you are experiencing surging problems, ask your tuner what their boost control strategy is: closed loop, or using the N75 linearization map to give a semi fixed duty cycle.

Hesitation is usually the result of the ECU cutting load/boost by restricting the throttle opening and N75 operation. This is due to the ECU not liking something, and this something can be very subtle – a slight overboost, a slight underboost, too much knock retard etc. The annoying thing is that this can often be intermittent, depending on the air temperature and how the car is being driven.

OllieH – bypass your N275. Leave it wired/plugged in, but re-route the vacuum piping so it doesn’t do anything. You are welcome to bring the car to me in Stockport one evening/weekend and I will take a look at your map to see if it is that causing this issue.

Rick.

[Oet]

top explanation Rick

a really egood explanation for those of us (myself included) who don't understand how or why these things occur!!

welcome too!

[OllieH]

Hi Rick, great explanation. How you have explained does translate to how the car feels at times you obviously know your stuff

[shark_90]

We made some more tweaks to the map for PaulMoVRS since you were in Ollie, so we'd appreciate the chance to try those modifications on yours before you take it elsewhere.

Rick, have you come across this surging issue with 363908 and other softwares of a similar vintage?

[shark_tech]

The correct way of doing it is to tune the boost control routine – few people seem to be able to do this properly though! You can control how quickly the N75 reacts to changes in load and requested load so that you don’t get these oscillations. On a custom map I would spend a lot of time on this as you are essentially able to reduce lag/time to reduce peak boost. Another thing tuners do when they can’t get the boost to do what they want, is bypass the closed loop control all together. This has its upsides and downsides. The biggest downsides are that you gain lag in certain areas of the rev range as you are relying on a fixed duty cycle, and you lose certain boost related safety nets, although there are others which will usually catch you. The upside is that you rarely get oscillation due to incorrectly tuned PID and load requests, and that you can run more boost than your map sensor is capable of reading. If you are experiencing surging problems, ask your tuner what their boost control strategy is: closed loop, or using the N75 linearization map to give a semi fixed duty cycle.

Good explanation, unfortunately the surge problem isn't that simple.

Our software still uses the full LDRPID algo, its running on the I limiter with slightly tweaked Q0 and Q1 parameters. Linearisation is standard, no reason to change that with a standard turbo. LDOB is disabled via TABLDOBN.

Whatever you do to the PID algo, you will still get the original surge discussed in this topic.

Problem is not on the boost PID, its torque intervention from the very early (and stupid!) "torque model" on these late ME7's. That problem is now solved, we'll look into the suggested flat spot (which is completely different issue than the original surge) after Ben comes back.

[RickRST]

A friend of mine bought an ex Police VRS recently and complained the boost was all over the place. Went out with him and it was surging, on the std map. This was down to a duff N75 though. It's since been replaced. I have written him a map which I'm hoping to flash tomorrow/Saturday. He's not local but i'm mapping a car on megasquirt in Liverpool and he's due to to come along. I'll see what it's like when flashed.

On a KO4'd 450bhp S4 I had an issue where the boost was incredibly lazy to build if you used the std boost contol. As a trial I used the maps from an RS4 which has the same turbo but different heads and cams. With the RS4 contol, there was horrendous surging. Logically this makes sense as the RS4 large port heads etc will be less responseive and so the control loop more aggressive. It took me a while but I tuned in a loop that I'm really happy with, boost comes in hard and stays flat. On a custom map this is what the customer is paying for it takes the most time along with the ignition and knock tuning.

Rick

[RickRST]

Hi Shark tech.

When you say it's running on the I limiter, have you FF'd KFLDIMX?

When you say it's a torque model issue, are you getting a corresponding change in requested load which follows the surge?

Rick

[shark_tech]

On a KO4'd 450bhp S4 I had an issue where the boost was incredibly lazy to build if you used the std boost contol. As a trial I used the maps from an RS4 which has the same turbo but different heads and cams. With the RS4 contol, there was horrendous surging. Logically this makes sense as the RS4 large port heads etc will be less responseive and so the control loop more aggressive. It took me a while but I tuned in a loop that I'm really happy with, boost comes in hard and stays flat. On a custom map this is what the customer is paying for it takes the most time along with the ignition and knock tuning.

Rick

Writing PID for K04/S4 (I've done over 20 of those here in Finland) is easy. Getting the hidden fuel adaptations sorted is the tricky bit on them, especially if running with 750cc+ injectors.

We're drifting slightly off topic here. Like I said, the surge issue on those cars is not a PID control problem. Thats what I originally thought, and have spent alot of time before finding out the actual reason for it.

[shark_tech]

Hi Shark tech.

When you say it's running on the I limiter, have you FF'd KFLDIMX?

When you say it's a torque model issue, are you getting a corresponding change in requested load which follows the surge?

Rick

No I meant I haven't, thats why its "running" on that. If you FF it you just use the last column of KFLDRL which is not the right way to tune ME7.

I also haven't done the typical mistakes of increasing KFLDIMX too much etc.

You can see the symptom with VCDS, but not the cause. Its an internal "bug" in the torque model algo. Its not a PID problem, its not over/underboost, nor is it deviation between specified and actual load.

-Mikko

Edited 10 June, 2011 by shark_tech

[RickRST]

Cool, I'm glad you have it sorted. ME can certainly be fussy!

Rick

[RickRST]

No I meant I haven't, thats why its "running" on that. If you FF it you just use the last column of KFLDRL which is not the right way to tune ME7.

I also haven't done the typical mistakes of increasing KFLDIMX too much etc.

You can see the symptom with VCDS, but not the cause. Its an internal "bug" in the torque model algo. Its not a PID problem, its not over/underboost, nor is it deviation between specified and actual load.

-Mikko

I'm a post behind here! Agreed regarding KFLDIMX, i see far too many FF's in peoples tunes!

Rick

[rk696]

That's all gone over my head

[shark_tech]

I'm a post behind here! Agreed regarding KFLDIMX, i see far too many FF's in peoples tunes!

Rick

KFLDIMX is a 16bit map, so you must mean FFFF

[rk696]

Hi guys, I'm back.

After having the map tweaked by Ben and Mikko, and replacing the N75, the map is really smooth now, but it doesn't seem to have the same power it used to have. I've been trying to convince myself that it's because I'm not getting overboost spikes from the faulty N75, but I'm not so sure.

Does anyone with a Shark stage 1 map have a boost gauge? I've done a logging session today (on my private runway of course ), and the ECU is only requesting 17 PSI. Is that right? I thought the Shark map was more like 19 PSI? It also tails off to 11 PSI at redline, so I may have a boost leak too.

MAF readings are 171 g/s (215 HP roughly), so the sensor looks OK, and the AFR ratio looks good too.

Any ideas?

Put the logs here:

http://briskoda.net/forums/topic/204671-done-some-logging-what-do-you-think/

[lee1972]

i have the same problem the 1st remap i had done was awful realy lumpy and running realy bad also knocked my mpg down to 21mpg not the 35-40 i was used to so he changed it and put another on worked lot better but changed the standard filter to a green improved a lot but still have few flat spots at high revs

i still have the standard dump valve should i change it for forge 007 or baileys?

Edited 16 June, 2011 by lee1972