Virtual Dyno

[thetyrant]

Not sure if anyone on here has played around Virtual Dyno (VD for short) on a PC but ive used it for years on different cars to compare modifications etc and if you put in accurate data it spits out pretty accurate information, proven by many people comparing what it generates against running car on chassis dyno however its critical to do road runs on flat as possible road and have all car data correct to get meaningful numbers of course, any rise or fall in road throws number way out as does having the weights and things incorrect, ive ran same road i use for this which looks nearly perfectly flat (must measure it at some point!) in both directions and data from car shows its not quite as flat as it looks but close enough for what im doing.

Link to website here for more info and download if you want to try it: https://barnhill.bitbucket.io/

Of course i would always take the actual power/torque figures as a guide only without comparing on an actual dyno that measures the output as there are so many variables especially doing it on road, but even chassis dynos can vary massively so with both methods to me its more about comparing shapes of graph it generates after changing something on the car rather than actual numbers, as above using VD you need to use same road (flat as possible to get most accurate numbers) and as close to same weather conditions as you can get to keep things consistent.

As im about to do a turbo install on my car ive done a few datalogs of stock n/a car to give me something to compare with in VD once turbo goes on, was also interested to see what my stock n/a mapping on Link Ecu generates compared to the Oem ecu on same spec car, its not a totally fair comparison really as ive generated the data 2 different ways and weather was different but it gives me an idea if ive made car worse or not!.

First off using stock car on OEM ecu i used the Torque App on my phone with an OBD adaptor to get engine data, which is ok if a bit slow on datalogging but it gives you something to work with at least so better than nothing, next same spec of car but now with my Link G4+ ecu installed and im able to pull much more accurate data from the logs it generates so will be using this going forward, im still able to pull data from Torque App as the stock ecu runs in piggyback mode to control dashboard but it doesnt collect much data now that Link has taken over running engine so not as much use.

Based on all this ive now got some data both using Oem ecu and same spec car mapped by me on Link G4 ecu, only engine mod between the 2 is car now on 470cc injectors ready for turbo but dialled in for n/a, runs were done using same piece of road near my house which is flat as i can find locally,  ambient temp was lower when i did the Link G4 run now that weather has turned so that will of affected the peak numbers as engines love cool air, but it shows at least ive not made car anyworse with my mapping it seems    still fine tuning the n/a mapping while i get ready for turbo install so some small improvements to be made im sure but im not really concentrating on extracting anymore power from car in n/a mode just want it running nicely before getting the turbo on there.

Below is screen shot from VD showing the 2 runs overlayed, the MR2 settings isnt in VD as a default car so i had to add as custom car using data found online, you need things like weight, CD, frontal area, gear ratios and tyre size etc. To get truly accurate numbers these need to be perfect but numbers ive used will be close enough for what im doing and car will go on dyno once turbo at some point so i can see how close it comes.

As you can see according to data ive picked up a little torque and power right through the rev range (12hp & 5ftlbs peak), colder air will probably contribute to about half of this i would guess but overall happy to see same sort of shape graph as i got with oem ecu to show im on right track with mapping, car does feel a touch more peppy on the Link but could just be the cooler air of course.

Edit to add my VD settings for reference.

[Petrus]

Thank you for sharing.

I have an app on my telephone. There are many.
Any smartphone will do and the app uses the accelerometer (which also senses which way you hold the phone).

Keeping the instrument horizontal gives the most accurate g-reading.

More importantly, make sure the position is always the same and use the same stretch of road.
Bar differences in relative air density ( temp./moisture/barmetric pressure) and wind conditions, the results are fine for comparison. Never mind the actual numbers, it is about the difference.
Véry handy tool indeed.

Again, thanks for the heads up.

[thetyrant]

Yes Phone apps are ok for a quick guide but generally not very accurate compared to proper datalogging and software, the sample rates on phones and the accelerometers are often not fast/accurate enough to get meaningful data, if you look at my attached picture and the dots on the graph lines (which are the sample data points you can see how many more i got from the Link ecu on red lines compared to those on the blue lines from my phone.

Im looking forward to trying this once the turbo goes on and hoping for a much larger area under the curve

[m1tch]

I think looking at some of the dyno figures for a tuned NA car I think around 10-15bhp is about right, think a lot of the gain with tuning NA is actually drivability and slightly more down low torque rather than top end.

[thetyrant]

I think looking at some of the dyno figures for a tuned NA car I think around 10-15bhp is about right, think a lot of the gain with tuning NA is actually drivability and slightly more down low torque rather than top end.

That's interesting to know as ive not really seen any tuned n/a 1zz graphs, there is certainly a little more to come and you can see on the graph above where it richens up and dips into 12's AFR low down and right at top end where the power/torque dips accordingly, getting this right wont make more power but will improve area under curve a little and help with response etc,  the low down dip around 3-4k is down to the VVT and im going to have a play with VVT zero'd out at full throttle see what we get, will make It easier to tune the fueling for sure.

I was reading on the HPA website about tuning VVT that when using a MAP sensor setup (like I currently have on the Link as opposed to a MAF like oem) getting fueling right during VVT is tricky, its seems best way time permitting is to tune the car with VVT zero'd out initially to get fuel/ign right and then add in VVT afterwards and tweak to suit, ive certainly found it a little tricky when doing the low speed cruise fuelling and ended up bringing in VVT a little higher up load wise and it helped with that a lot, otherwise had a very spiky looking fuel map which I don't like!

Once turbo'd VVT wont really be needed so much as its main use is to bring in some midrange low/midrange torque which it does well and helps with emissions as well, once you start forcing air into the engine its not so critical as the boost sorts the midrange, little bit of VVT low down maybe to help spool the turbo although with the tiny one ive got I don't expect any issue with that.

All interesting stuff and good to be back on tuning cars as its been too long!

[m1tch]

The NA graphs are in the performance section of the site:

https://www.mr2roc.org/forum/index.php?topic=27521.0

I have also collected together a few FI graphs as well (and there are others in the performance section of the site).

[m1tch]

Here are a few other dyno figures with the 1zz:

[shnazzle]

Just downloaded the mobile version. See how it goes. As it only really needs engine/throttle, I'm tempted to plug in the Emanage to get much more accurate logs instead of the OBD2 readings.


Only annoying thing so far in the app is that you can only enter one gear ratio. So you have to pick the gear you're going to do the pull in in advance.

Also, I have no idea how much my car weighs Guesttimated 2200 lbs

See what it comes up with.

[thetyrant]

Mobile version is a bit scaled down hence why ive got it PC, much more flexible and easier to see rather than squinting at my "normal" size mobile

Ive done a few 3rd gear pulls now turbo is on as per my build thread but the other day i did a 4th gear run to see how it compared and to check mapping etc, i will upload it here when i get a minute but shows pretty consistent with the 3rd gear pulls just more topend as extra load made more boost,  4th gear is best (on the 6speed anyhows) as nearest 1:1 ratio but the speeds at redline in 4th are too much for me these days, unless i find a nice long flat private road to max it out,  at least in 3rd speed isnt crazy at redline.

Ian

[thetyrant]

These are setting i have in my VD and driver weight i use 155lb which is right for me.

[JB21]

Used it when I had my tuned Mazda 6 MPS. It was pretty accurate compared to a RR session, both around 300hp/300 ft-lbs. I used it with my Cobb tuning device plugged into the OBD2 port where logs were saved to and then uploaded to VD.

[thetyrant]

Another one below with JB21's 2ZZ (red lines) compared to my 1ZZ turbo (blue lines), different roads etc so not that accurate but gives you an idea on difference in way they deliver the power,  the data sample rate was slow on the 2zz run as well which flattens the lines and makes the lift less noticeable, although as JB21 says on another thread it doesnt feel as strong in lift as previous 2zz engine he had,  will add my stock 1ZZ to it as well later...