OK, I am going to build a little PCB to do this. But one question for the pro's what range accel do I need? I figured that nobody will ever crack 2g's is this right or do I need the +-5g part?

-pete

 

2G's will cover all my applications, even if I wan't to put it some serious cars (over 1000bhp @ 1100kg) I will just test from 125+kph. 2G will be heaps unless you wan't to test launch's and **** like that, I just wan't power curves.

Thanks mate

I will wait for you're pcb !

 

Ok Update !

No need to build one...After mucho searching I found this (exactly what I wanted) from a UK company.

Check it out http://www.race-technology.com/WebPa.../DL90Home.html

I will get a quote and place my order, this is the perfect item.

 

Has anyone seen the new G-tech coming out? It graphs on the screen (LCD) and memorizes your HP and tells you when your best shift is. It also hooks to a PC for better resolution. It also measures lateral G's. I saw it at SEMA in October, and he said it should be out sometime this spring. I just need to call him back to find out when. I think it is going to be under $300 or so.
Don.

 

Whoah way too much!

I'm still going to build one - there's no way the race-tech or Gtech for that matter can actually get the real parasitics (wind resistance, rolling friction) they must use rough guesses from Cd frontal area etc.

-pete

 

Race Technology do, that is why I am getting it ...they use GPS to get ultra accurate speed sensing, there accelerometer is the most accurate I have read specs on as well between the two they eleiminate the time accumulate inacuracy tha accelerometer alone are prone to (no matter how good they are).

They calc the actual CdA and rolling resistance by doing two basic runs and this combined with the exact weight will give you the power, the only issue I have with this method of CdA and rolling resistance was of accurate speed measure, but with GPS & accelerometer this is not an issue as the data is logged and is extreme in accuracy, HENCE THE PRICE !!!!!

I have been watching a 26min race file on their software and the unit is excellent, I just placed my order, will keep you informed on how it performs in my car when I recieve it.

 

thanks for that,

My only wonder now is how do they get accurate speed measurement out of GPS? I though it wasn't specific enough with position?

-pete

 

This is in the help file for the software that runs the DL90............."Speed

Vehicle speed, with units of either kilometres per hour (kph) or miles per hour (mph) is calculated from combining the information from either the GPS and the accelerometers, or if GPS is not available then the wheel speed pickup is used. If neither GPS nor wheel speed data is available then just the acceleration data is used, however unless the run is short or the road is very flat the speed data in this final case will become inaccurate due to accumulation of errors. The method used to combine GPS and accelerometer data is very complex and takes maximum advantage of the different data sources. In very general terms the accelerometers are very good at recording fast changes, but the accuracy of the speed estimate deteriorates with long runs. In contrast the values of speed from the GPS is only available every second, tends to have significant inaccuracies, but does not suffer from drift as it is a more direct measurement. Combining both data sources results in a speed trace that is both available every 100th of a second so each gear shift can be analysed, yet does not drift off irrespective of the length of the run.
"

I think that explains it ?

 

Speed measurement using GPS can be made much more accurately than what you get if you just use the position measurements over time (differentiation). Most GPS receivers have the capability to measure speed directly, by looking at the doppler frequency shift of the satellite carrier signal. This kind of speed measurement has some biases that prevent it from being used to track position, but it's far better than deriving speed from position.

 

Ok latest update !!!!!

For anyone interested here is what I FINALLY chose and why.

I have decided to go with the ROAD DYNO package for the following reasons. If you have accurate data this unit will operate EXACTLY like a stationary dyno, the ONLY difference is that it is using the car instead of a drum or inertia/mechanical/electrical brake to apply resistance load. From what I have determined unless you test a top fuel car where the tires change diameter ALOT this machine will be super accurate on a flat road. And for comparative tests NOTHING will match the repeatability of this unit as you are in control of the input variables, just test on same section of road and make sure it is not a really windy day. The ease of logging and transferring the data into a graph is excellent.

Now my system with an accelerometer is prone to much more inaccurate readings, in particular initial set up of the unit and needing to start from 0 speeds as well is a pain in the ****! It is only accurate over short time spans. I have not found a unit that in theory from my research can be as accurate as the road dyno set up.

Now the last one I was looking at from Race Technology is prone to allot more inaccuracy after playing with sample data files fro the last 24hours straight!!!! It when you put in "virtual" beacons to reference spots on a data log can give varying speeds from the GPS due to the inbuilt inaccuracy in this, you will find this if you download the software and "trim" the data logs to say a section for an acceleration sample and the data is out on average by up to 5% in the speed readings I analyzed. It's G readings and other data all would have to be manipulated to achieve a power graph like you get at a dyno shop.....THIS IS A LOT OF ******* AROUND FOR THIS PURPOSE!!!!!!!!!
It is excellent for collecting lap time data and split times as they stay and after much playing with their software it is VERY impressive in this, It uses data smoothing to generate a actual circuit map, I found you need to do at least 5 laps to get a smooth map that does overlay one corner on top of another, THIS HOWEVER IS USELESS FOR GETTING A HORSPOWER GRAPH ! Like I want to do.

In summary the ROAD DYNO system will be the most accurate (only relies on rpm input & everything else is pure physics!) and repeatable thing by theory and practice, assuming you're input "car" variables are right, you have no wheel spin, you are on a pretty good road (400m flat for most cars, ours around 200m flat for 2000rpm to 8000rpm+ pull in 3rd). Even if you **** up with the input variables so long as you are consistent and use the same test section of road ANY mods you do to the car (good or bad) will show up on the power graph........I will trust this MUCH more than a dyno shop where there variances are LEGENDARY

I just order the ROAD DYNO, I will keep you posted of how I go with it and load up some results ASAP.

 

You're sure you don't want to revisit this analysis when you haven't been up for 24 hours straight?

Seriously, is tire "slip" an issue with this Road Dyno? I'm not talking about tire "spin"; I thought it was possible to get 10-15% slip under power and braking. That would tend to increase RPM by a similar factor, relative to the no-slip condition, right?

 

I ordered mine last night as well... Can't wait to see the HP and TQ curve on my Jeep Cherokee!!! That is, until I get my 20B FC running

 

We can compare graphs !

We should start up a data base of different cars for variables like, Frontal area, Mass, Cd etc etc.....

The details for my car are 1070kg + 70kg = 1140Kg (2513lb) driver, Frontal area 2.1m^2 or 22.6ft^2, Cd 0.32, 3.909 Final drive, I will test in 3rd gear 1.391 ratio and tires are 205/50/15....now all I need is a power/torque graph ! I am going to use the same piece of road that I have been using for the last 5 years for my Autronic data logs, it is perfectly flat and is sheltered from wind etc.

Can't wait.........

 

I hope I dont offend anyone but I think both systems have major flaws. I read a777's story at the time about the road dyno and can now see a few flaws:

1) They're badly made - he had plenty of failed runs
2) They don't directly measure force. This is rediculous because they are then prone to missing wheelspin, clutch slip etc.
3) They have no way of guessing counter-forces so you could gain HP by testing on a different road, with different tyres or if the wind was strong. Hell increases in tyre pressure (from heat alone) would screw your readings.
4) It's tacho based so there goes the roll-down idea

And as far as accelerometers go I dont agree with you. I've seen systems work that can get it spot on - and I think I know the trick. Either way we'll see when I build my system over the coming weeks. I would really like to compare data so let me know when you get your system.

 

It is impossible to do the test I want to do, and most of us want to do with an accelerometer, there is no system out there that will give you the graphs the the road dyno will give, not to mention in such a short quick time.

The road dyno does measure force ! Through the time it takes to accelerate a known mass & the aero load due to increased velocity through the Cd factor and frontal area, It is easy enough to find the frontal area and Cd factor of the car you own or test. It is easy to double check the final drive of a car, I will use my GPS (I use for camping) to get a real speed at 3000rpm in 3rd or 4th in the cars I do test to find out exactly what diff ratio the owner has in their car, the rest gets inputed into the program......it is very simple.

In regards to the "other" counter forces, yes you will need to test on a still enough day and on a flat road. Wheel spin is not an issue in 4th gear or even 3rd for most cars on the road. And if you have wheel spin then you will never know what power you are getting at the rear wheels on the road or on the dyno (I have seen many of my customers cars wheel spin on a dyno when strapped).

I hope the unit is built well enough, they give a 30day money back guarrantee, so if it's not up to scratch I will be sending it back. In theory I like their system the most from all I have seen the downfalls being flat road/still day/and speed verification of diff ratio if the car is not yours i.e. do not know what diff ratio you/customer has?

So for me with a GPS this is not an issue, all I need to do is buy some corner scales or go the weigh bridge and get exact data of car I am testing. I will post up data as soon as I get the unit & test it. In my case I know my engine makes around 500bhp (engine) in road trim, so see what it puts to the ground will be intersting

 

I still like the Griffin method of measuring the G's, how ever I want to map "actual" engine rpm via a pick up to do this you need to monitor the ignition or the rear wheel speed vai a sensor. If you do mapping just based on vehicle speed and calc back through ratio's you will not pick up wheel spin either ! That is why the way the road dyno does it is so simple and so right.

Any dyno Engine/Chassis etc will give false readings if ramped wrong by operator or if there is any slip in the clutch/tires etc. The only difference with the road dyno is being exact with the variables to give real results, build quality is a different issue, but the theory is sound, from my Mech Eng background anyway.

Just remeber to use the nicest (smooth/falt) raod possible (300m stretch ample) test on a still day, and verify variables such as weight/test gearing to give most accurate result and theory says it will give totaly accurate RWHP.....Much more reliable than many chassis dyno operators.

My only concern now is with the build quality/reliability from other users, I have contacted othe people who have used this product and they have not had any issues. I will find out soon enough.

 

That's why I'm taking a tacho input - best of both worlds! the difference between measured accel and engine accel gives you slip - sounds a lot like the traction control ferrari were cheating with doesn't it

You can calculate force (and therefore power) indirectly by measuring engine accel but the road dyno has to rely on guesses (who knows their frontal area?) an idealistic formulas (Cd isn't static in the real world IIRC)

Even if I never get the track mapping stuff working, I can at least dump 5 mins of data at 10ms splits giving RPM, accel vector and calculated speed. straight into an excel spreadsheet. surely that's better than any data a road dyno can give me?

-pete

 

Pete

I am gonna get the road dyno anyway (no choice now ! I orderd it already ) and I will test her, If it's crap I will send it back and get your system.

I still like the idea of the G's measurement it's just gonna be a little more complex test, get the decel forces, get the speed, get the accel forces and process the data.

One question, are you going to input speed via which way? If deriving it from the accelerometer you will need to start from a know point and that will need to be 0kph stationary start. Either that or you will need a calibrated speedo input i.e. GPS or other source if speedo is not correct or will you input speed of a wheel sensor?

With just accelerometer it can only be done from stationary start...right?

 

 

Simple - the system is powered by accecories (eg cig plug) so when you start the car I assume the car is stationary and that's your zero point. As you work through the gears the computer compares calculated speed vs speed calculated from the tacho input. The difference is wheelspin/clutch slip.

-pete

 

Welp, got the road dyno in the mail yesterday... everything was there.... but I was too dog tired to mess with it... We're gonna through it on my friends WRX and see what it does... We have a new Gtech meter as well, so that may help calibrate the drag and whatnot, who knows

We're building quite an inventory of gizmos... Gtech, Road Dyno, Wideband O2

 

Mines not here yet

Though I am in Australia, bit further away

 

Well, another update !

I got the road dyno, here are the results....

Unit works fine on mild powered cars, tries it on my friends Nissan, no problem result to with in 2% of chassis dyno, even picked up the difference in running with no air filter !

tried it on my car, got 632rwhp !!!!!! this was due to wheel spin/slip in third gear !!!! tured the boost up even more and the unit cannot process the data ! it comes up with a message saying data has to "many errors" try another run ! **** THAT I did about 10 runs from 1500rpm in third to 8400rpm (170kph!) and the mother ****** flies but the dyno will not handle it due to the slip on the tires !

I logged the sensors on my Autronic and you can see when it comes on full boost at around 6000rpm that the rmp spikes at 6700 and at 7400odd rpm (my torque peak) and this is what is ******* up the Road Dyno software hence not letting me process the data !

So after all the posts and research I have found that the RD does not work on a powerfull turbo rotary with not enough traction, I would try to play with it more but ten runs is pushing my luck with the LAW, I passed a cop just after my 9th run still did one more !

So I am sending the unit back tommorrow, back to the drawing board !

RPM-PWR have you built that unit yet

 

Hey Rice:

Do you think putting more rubber on your car would fix this problem?


-Zach

 

To a degree, yes.

But I would need like 345/40/17 or something like that to totally overcome the problem in a street tire.

Ferrari F40s used a tire this size to transfer 480bhp in a 1100kg car so I think that wouldbe a good guide as to the size required