Dyno Preperations

[Jellybean]

Any Preperations you should take before a Dyno, what UK fuel would you normally use ?

[johnny g]

VPower fuel, full tank.

Full service before hand (or very recently).

Double check the ignition components.

Swap Iridiums for Copper plugs (in some cases).

 

In general just make sure is in the best health it can be.

[David Reid]

Personally I always like to have my cars mapped on 95 then I run them on 98 or V power. The reason is just to have margin in the ingition calibration. Even do this with my race car(s).

 

If you then ever get a dodgy tank of 98 thats actually 97 you dont melt a piston.

 

To the list above I would also add

 

Check coolant system no leaks and full.

If using a non hub dyno good quality rear tyres so that you can get the power down.

Edited June 29, 2010 by David Reid (see edit history)

[johnny g]

If you then ever get a dodgy tank of 98 thats actually 97 you dont melt a piston.

 

Maybe worth investing in an ECU with knock-control?

[wojtrek]

VPower fuel, full tank.

Full service before hand (or very recently).

Double check the ignition components.

Swap Iridiums for Copper plugs (in some cases).

 

In general just make sure is in the best health it can be.

 

these sound like tips on how to maintain your car. not preparing for a dyno - if you have a supra that you're afraid to put on a dyno, and that's a 15second full load pull, how can you not be afraid to drive it anywhere?

a 4/5th gear pull on the road is a lot harder on the car than a pull on a dyno, so don't worry

[dude]

Maybe worth investing in an ECU with knock-control?

 

Or a mapper with enough sense to back the timing back slightly from optimum:blink: Listening for det helps too;). the tires are good advice though, no 888's or drag radials and no cheapies or 'willpops' as we call them.

[dude]

a 4/5th gear pull on the road is a lot harder on the car than a pull on a dyno, so don't worry

 

So how come so many cars blow up on dyno's??? Don't listen to that advice the dyno is a LOT harder on components than road pulls, load etc can be varied , you do not have airflow thru the rad or cooling the tyres like on the road + many other things.

[wojtrek]

I own a dyno - nothing ever exploded on it

I have a huge blower (1meter diameter), cooling not an issue.

I understand that heat dissipation might be a problem on tyres or the radiator - but that's only an issue while tuning, doing pulls one after another and maybe even with some huge load. But going to a dyno JUST to check power, doing one or two pulls in inertia mode - not a problem.

[dude]

Where does it say he is just going to do a few pulls???? Even a power run puts more strain on the car than the road, does your supra take 15-20 secs to pull 4th gear??? If so you have a problem;)

[wojtrek]

OK, I agree I assumed it was just a normal trip to a dyno - but I think this is the case, otherwise he'd say he's got a dyno session before him or sth. and I imagine the guy doing the tuning should know how to prepare a car for that.

 

Anyway, I will not agree that a inertia pull on a dyno puts more strain on a car than a 4th gear pull on the road.

And no, I don't have a problem with my supra - put a 400bhp supra in 4th, slow down to 1000-1200rpms (which is probably 40km/h or 25mph) and floor it, hold it to rev limit which equals around 195km/h or 120-125mph and time it

if a 400bhp does this in less than 15s then I do have a problem with my car

I currently have around 420bhp, it takes 5seconds to hit 62mph and it takes 10seconds to get from 100 to 200km/h (62-125mph) which is BMW M5 (the 507bhp one) territory = fast. so a 0-125mph is 15seconds flat out. And that is through gears (1st to 4th), which is faster that doing this in 4th gear only - obviously. add to that the fact that at 1500rpm there's not much going on and your time of a full rpm scale pull in 4th is in the 15-20s range

 

If a dyno pull is done with a load, then of course it can put more strain on the car than a 4th gear road pull. But still - if you're afraid to do go to a dyno - you shouldn't be driving your supra like it is designed to drive

Edited June 29, 2010 by wojtrek (see edit history)

[dude]

OK, I agree I assumed it was just a normal trip to a dyno - but I think this is the case, otherwise he'd say he's got a dyno session before him or sth. and I imagine the guy doing the tuning should know how to prepare a car for that.

 

Anyway, I will not agree that a inertia pull on a dyno puts more strain on a car than a 4th gear pull on the road.

And no, I don't have a problem with my supra - put a 400bhp supra in 4th, slow down to 1000-1200rpms (which is probably 40km/h or 25mph) and floor it, hold it to rev limit which equals around 195km/h or 120-125mph and time it

if a 400bhp does this in less than 15s then I do have a problem with my car

I currently have around 420bhp, it takes 5seconds to hit 62mph and it takes 10seconds to get from 100 to 200km/h (62-125mph) which is BMW M5 (the 507bhp one) territory = fast. so a 0-125mph is 15seconds flat out. And that is through gears (1st to 4th), which is faster that doing this in 4th gear only - obviously. add to that the fact that at 1500rpm there's not much going on and your time of a full rpm scale pull in 4th is in the 15-20s range

 

If a dyno pull is done with a load, then of course it can put more strain on the car than a 4th gear road pull. But still - if you're afraid to do go to a dyno - you shouldn't be driving your supra like it is designed to drive

 

You have just backed up my argument, on the road the car will not take that long to pull redline in 4th gear, so that proves the dyno puts a diff strain on everything, you pull 4th gear for 15-20 secs JP pulls 4 gears in half that time!!!

[wojtrek]

dude, I don't what how you came up with this conclusion. I wrote exactly the opposite.

 

a power test for a 400-450bhp supra, in inertia mode, on very heavy rollers, in 5th gear (direct gear for a 6sp supra) takes around 15seconds (1200rpm to redline)

this is a few seconds less than what it would take to do this on the road but in 4th gear in that same car.

how is this more strain on the rollers?

 

if JP does 0-125mph in 10seconds through gears, than his 5th gear pull on a dyno would be probably in 10s region, maybe 9 or even 8. there is no way he does 25mph to 125mph in 4th gear only, on the road, in less than that. so again, less strain on the rollers.

 

and to make things clear again, if that power test is done with an extra load (with a brake on a dyno) it might be still shorter than a 4th gear pull on the road or longer if the power applied to the dyno brake is larger.

 

a rolling road simulates road, right? the point of putting extra load on the rollers is to simulate road conditions - if you have succeeded to create the same load conditions (strain) as a 4th gear pull would on the road you're more than OK. putting more load is pointless unless you're tuning the car of course and want to see what exactly is going on with the car, or test extreme situations etc.

[David Reid]

Maybe worth investing in an ECU with knock-control?

 

None of the current after market ECU's have in my opinion good enough knock control to deal with this as they are far to simple.

 

I do know what I am talking about as I calibrated the OEM knock control system on most of the current JLR v8's that allows them to run a 100 Ron base map but be safe down to 90Ron.

[dude]

None of the current after market ECU's have in my opinion good enough knock control to deal with this as they are far to simple.

 

I do know what I am talking about as I calibrated the OEM knock control system on most of the current JLR v8's that allows them to run a 100 Ron base map but be safe down to 90Ron.

 

What would it take to have a safe enough anti knock system??? Is it a case of faster processing or is it the knock sensor itself, I use the HKS knock amp with my F-Con and it seems pretty good.

[David Reid]

What would it take to have a safe enough anti knock system??? Is it a case of faster processing or is it the knock sensor itself, I use the HKS knock amp with my F-Con and it seems pretty good.

 

Its the way in which they work and in relation to processing the knock signal. The problem is the output from a knock sensor is very difficult to process it’s not just on /off.

 

To give you the basics to get a good operating system you need:-

 

• Selection of knock sensor for each cylinder i.e on a supra cylinder 1,2,3 to use sensor 1, 4,5,6 use sensor 2.
   
  
• Sampling of the signal, only around the firing event say 20CA for each cylinder, this way any valve noise is not included in the knock signal.
   
  
• I my experience a minimum of 2 calibratable filters to process the signal through one high speed one low speed arround 3KHz and 8 Khz. (To give you an idea I think the latest OEM ECU's have 16 calibratable digital filters)
   
  
• Then you need a high speed knock IC as you say, to process the signals for each cylinder. You then need to add a clever algorithm, normally Logarithmic or differential and compare your current individual cylinder signal to your preprocessed threshold (derived from previous samples). Then make a decision on knock event or not for each cylinder.
   
  
• Once you have done that you then need quite a complex piece of ignition software to allow you to apply fast retard, followed by slow advance. This allows the cylinder to cool and as it cools ignition can be reapplied to keep ignition optimum. Basically the modern systems use a fast say 2-3 degree retard step which is advanced back by 0.25CA over 30 seconds, there is then an adapation strategy on top of this to lean the mean advanced value.
  

 

That’s a very brief overview

 

My understanding of current system like Motec etc is they dont sample to make an individual cylinder knock judgement, therefore valve closing event ect are caught in the signal and you could miss an individual cylinder knocking and are prone to falsely detecting knock.

 

They also don’t have filters user calibratable filters (and every engine is different knock frequency is basically a function of bore size) or the ability to swap filters with increasing engine speed.

 

The after market systems I have seen also only have a global retard function, no advance other than just resetting at key off.

 

Overall they are just too basic to be worth while in my opinion and overall your better off setting the base map slightly retarded or mapping on a slightly lower grade fuel (this also has some advantage in giving head room for evaluated coolant and IAT temperatures)

 

If an OEM type system was ever implemented in an aftermarket ECU, I don’t know if anyone would ever be able to map it correctly as its just too complex. For example it takes me approx 3 weeks of engine testbed work just to be able to determine the filter frequency for each cylinder and I have to use hugely expensive quick sampling kit measuring cylinder pressure as well as knock sensor outputs. So it just wouldn’t be possible or economic to map it.

[Jellybean]

Hi Dave

 

Waht about the SARD Rom chips for the OEM ECU?

 

http://takakaira.com/Main.aspx

 

Analyse

 

 

 

Analyse is an ECU set to provide alteration of high boost & rev control at high speed driving, and provides accurate engine control in various conditions, such as partial range, street driving, and engine idling.

 

 

 

Resetting items of Analyse:

 

*

Fuel map resetting.

*

Ignition map resetting.

*

Rev limiter resetting.

*

Speed limiter cancellation.

*

Boost limiter cancellation.

 

It's strongly required to use high-octane gasoline after the replacement of ECU (in Japan "high-octane gasoline" values is 100~101%.

[Ryan.G]

Its the way in which they work and in relation to processing the knock signal. The problem is the output from a knock sensor is very difficult to process it’s not just on /off.

 

To give you the basics to get a good operating system you need:-

 

• Selection of knock sensor for each cylinder i.e on a supra cylinder 1,2,3 to use sensor 1, 4,5,6 use sensor 2.
   
  
• Sampling of the signal, only around the firing event say 20CA for each cylinder, this way any valve noise is not included in the knock signal.
   
  
• I my experience a minimum of 2 calibratable filters to process the signal through one high speed one low speed arround 3KHz and 8 Khz. (To give you an idea I think the latest OEM ECU's have 16 calibratable digital filters)
   
  
• Then you need a high speed knock IC as you say, to process the signals for each cylinder. You then need to add a clever algorithm, normally Logarithmic or differential and compare your current individual cylinder signal to your preprocessed threshold (derived from previous samples). Then make a decision on knock event or not for each cylinder.
   
  
• Once you have done that you then need quite a complex piece of ignition software to allow you to apply fast retard, followed by slow advance. This allows the cylinder to cool and as it cools ignition can be reapplied to keep ignition optimum. Basically the modern systems use a fast say 2-3 degree retard step which is advanced back by 0.25CA over 30 seconds, there is then an adapation strategy on top of this to lean the mean advanced value.
  

 

That’s a very brief overview

 

My understanding of current system like Motec etc is they dont sample to make an individual cylinder knock judgement, therefore valve closing event ect are caught in the signal and you could miss an individual cylinder knocking and are prone to falsely detecting knock.

 

They also don’t have filters user calibratable filters (and every engine is different knock frequency is basically a function of bore size) or the ability to swap filters with increasing engine speed.

 

The after market systems I have seen also only have a global retard function, no advance other than just resetting at key off.

 

Overall they are just too basic to be worth while in my opinion and overall your better off setting the base map slightly retarded or mapping on a slightly lower grade fuel (this also has some advantage in giving head room for evaluated coolant and IAT temperatures)

 

If an OEM type system was ever implemented in an aftermarket ECU, I don’t know if anyone would ever be able to map it correctly as its just too complex. For example it takes me approx 3 weeks of engine testbed work just to be able to determine the filter frequency for each cylinder and I have to use hugely expensive quick sampling kit measuring cylinder pressure as well as knock sensor outputs. So it just wouldn’t be possible or economic to map it.

 

Syvecs does all that above with the full knock enabled and even has indiv cyl shut down thresholds so if the knock if that bad on one cylinder (ie faulty injector) it will shut it down.

 

The Very same knock control is what keeps half the teams engines going round leman for 24hours

 

Ryan

[David Reid]

Syvecs does all that above with the full knock enabled and even has indiv cyl shut down thresholds so if the knock if that bad on one cylinder (ie faulty injector) it will shut it down.

 

The Very same knock control is what keeps half the teams engines going round leman for 24hours

 

Ryan

 

Hello

 

They must have moved on e quite alot since I looked last. Are you saying Syvecs has individual KCS driven igniton control? How many filters do you have to choose from?

 

Finally can you actually get it to work? having a race team spend weeks setting it up on an engine dyno I would have thought is very different to the hours your normally get to completley map a car.

 

By "get it to work" I mean for example have you been able to map a 2JZ on 100 RON, then run it on 95 with no excessive audible knock on 95 and apply sensible igniton for 95 i.e not way retarded ?

 

If not whats the most extreme you have run and it and how much ignition do you allow it to control?

[Ryan.G]

Hello

 

They must have moved on e quite alot since I looked last. Are you saying Syvecs has individual KCS driven igniton control? How many filters do you have to choose from?

 

Finally can you actually get it to work? having a race team spend weeks setting it up on an engine dyno I would have thought is very different to the hours your normally get to completley map a car.

 

By "get it to work" I mean for example have you been able to map a 2JZ on 100 RON, then run it on 95 with no excessive audible knock on 95 and apply sensible igniton for 95 i.e not way retarded ?

 

If not whats the most extreme you have run and it and how much ignition do you allow it to control?

 

2 Filter options with adjustable CA listening area.

 

I would be very confident to say that the way i set it on the supra it would be able to cope with that no problem. Spent alot of time on it with my own car with is abused everytime it starts up pretty much. Normally allow it to pull up to 8 degrees and if it pulls more than that i ask it to shut the cylinder down and flag an ecu light with the full knock feature. Certainly saved my engine with a fuel problem i had at silverstone early this year.

 

No other aftermarket ecu will come close to the Syvecs Knock Control i know for sure and its one of the reasons i brought shares into the company.

 

Just take Jamiep car on here which had a stock engine and over 750+ for a good few years, ran 10's 1/4 miles, 190+ runs and still the engine is perfect. I would be confident to take a stock engine further tbh with the syvecs as it picks up knock so fast and deals with it that the cylinder pressures never spike severly. On my own car i dont even bother with det cans now!

 

All the safety features on the syvecs is what makes it the best ecu out as all the features like the engine trips on oil pressure, oil temp, knock control, egt control, input scopes up to 1000hz and much more make engines even more reliable.