Pages: [1]
Author Topic: Simos 18 Load and Torque Datalogging/Monitoring  (Read 7499 times)
eliotroyano
Hero Member
*****

Karma: +47/-7
Offline Offline

Posts: 796


« on: October 07, 2021, 03:47:40 PM »

Hi friends I have some doubts about Simos 18 cars and high hp/torque/load values measured.
1) It is well known that Simos 18 ECUs have a diagnostic/datalogging/scaling limit of 1389mg/stk (at least using VCDS right now) that I imagine that is not the real limit that ECU can see. Then my first question is exist any way to read or measure the real load (mg/stk) from ECU? maybe a RAM datalogger can do that?
2) Something similar scene occurs with engine torque values. The are some channels like the "Engine Torque (Nm)" that can show in theory that info. But compararing different cars, different dynos, the taken values, are generally bigger at the dyno than what ECUs shows in the datalogging channels like previously one commented. Then the second question is there is an specific channel that can show more precise information? or it is the sum of other channels like the previous one + "Drag Torque (Nm)?
I appreciate your comments and tips.
Thanks,
 
« Last Edit: October 09, 2021, 03:43:47 PM by eliotroyano » Logged
prj
Hero Member
*****

Karma: +903/-420
Offline Offline

Posts: 5788


« Reply #1 on: October 07, 2021, 11:26:05 PM »

https://docs.vehical.net/logger
Logged

PM's will not be answered, so don't even try.
Log your car properly.
d3irb
Full Member
***

Karma: +131/-1
Offline Offline

Posts: 185


« Reply #2 on: October 08, 2021, 09:20:52 AM »

For RAM logging, like MAF above 1389mg/stk, you are right that it's no longer overflowing or capped internally, just in the ReadIdentifier value.

I agree with prj  Wink , just buy VehiCAL. It's a great value for what you need.

For the reversed/indicated torque model, I'd recommend you read the FR. It is just based around a simple airmass to torque table, purely modeled data, so I wouldn't spend too much time expecting it to match closely with measured values from a dyno. The most important thing about the reverse torque model is making sure it is not underscaled by too much on a DSG car, as it drives the feedforward in the DSG clamping controller and therefore the DSG clutch pressures.
Logged
eliotroyano
Hero Member
*****

Karma: +47/-7
Offline Offline

Posts: 796


« Reply #3 on: October 08, 2021, 10:44:55 AM »

For RAM logging, like MAF above 1389mg/stk, you are right that it's no longer overflowing or capped internally, just in the ReadIdentifier value.
I agree with prj  Wink , just buy VehiCAL. It's a great value for what you need.
I have saw it. It is a really nice option.

For the reversed/indicated torque model, I'd recommend you read the FR. It is just based around a simple airmass to torque table, purely modeled data, so I wouldn't spend too much time expecting it to match closely with measured values from a dyno. The most important thing about the reverse torque model is making sure it is not underscaled by too much on a DSG car, as it drives the feedforward in the DSG clamping controller and therefore the DSG clutch pressures.
You got my point, specially for DSG cars. Then Load (mg/stk) and Torque (Nm) defines the basic model of that ECUs.
I understand your comment but how do we avoid to be totally blind about torque numbers and torque modelled maps?
And how without some at least rounded numbers, can we tweak it to avoid torque/boost/throttle monitoring issues?
« Last Edit: October 08, 2021, 10:56:54 AM by eliotroyano » Logged
prj
Hero Member
*****

Karma: +903/-420
Offline Offline

Posts: 5788


« Reply #4 on: October 09, 2021, 01:48:22 AM »

You need to understand difference between indicated and clutch tq.
Also there's no issue just doing linear extrapolation and you can go close to even 1000nm and it's still more or less accurate.
There's actually a lot more than those two maps that goes into the torque calculation, there is ignition angle efficiency and so on.

That said, I don't understand your point here. Don't underscale anything, extend the maps and there's no problems, so I am not exactly sure why this discussion is needed?
Logged

PM's will not be answered, so don't even try.
Log your car properly.
CarHood
Newbie
*

Karma: +0/-0
Offline Offline

Posts: 10


« Reply #5 on: October 09, 2021, 08:06:48 AM »

Hi friends I have some doubts about Simos 18 cars and high hp/torque/load values measured.
1) It is well known that Simos 18 ECUs have a diagnostic/datalogging/scaling limit of 1389mg/stk (at least using VCDS right now) that I imagine that is not the real limit that ECU can see. Then my first question is exist any way to read or measure the real load (mg/stk) from ECU? maybe a RAM datalogger can do that?
2) Something similar scene occurs with engine torque values. The are some channels like the "Engine Torque (Nm)" that can show in theory that info. But compararing different cars, different dynos, the taken values, are generally bigger at the dyno than what ECUs shows in the datalogging channels like previously one commented. Then the second question is there is an specific channel that can show more precise information? or it is the sum of other channels like the previous one + "Drag Torque (Nm)?
I appreciate your comments and tips.
Thanks,
 

I use Vehical and highly recommend it.

However one workaround I've found for the 1389 airmass limit is to instead log mass airflow (g/s).  I don't have the specific PID info here on my phone.  But you can convert that to mg/stk and it will read well above 1389 and works fine for most intents and purposes.  I did this on HP tuners prior to moving to a ram logger.
Logged
eliotroyano
Hero Member
*****

Karma: +47/-7
Offline Offline

Posts: 796


« Reply #6 on: October 09, 2021, 03:43:20 PM »

I use Vehical and highly recommend it.
However one workaround I've found for the 1389 airmass limit is to instead log mass airflow (g/s).  I don't have the specific PID info here on my phone.  But you can convert that to mg/stk and it will read well above 1389 and works fine for most intents and purposes.  I did this on HP tuners prior to moving to a ram logger.

I have saw that option too but I need to test it.
Logged
eliotroyano
Hero Member
*****

Karma: +47/-7
Offline Offline

Posts: 796


« Reply #7 on: October 09, 2021, 04:17:55 PM »

You need to understand difference between indicated and clutch tq.
Also there's no issue just doing linear extrapolation and you can go close to even 1000nm and it's still more or less accurate.
There's actually a lot more than those two maps that goes into the torque calculation, there is ignition angle efficiency and so on.
That said, I don't understand your point here. Don't underscale anything, extend the maps and there's no problems, so I am not exactly sure why this discussion is needed?

Thank for you comments. Please correct me if I am wrong. From my research, in a really simplicated view the Simos 18 Torque Modelled Control is based basically by load/torque maps, limiters and control threshold maps, of course adapted by many other controls like ignition/lambda/pressure/etc..etc. In a Bosch/Siemens style manner there are 2 big maps that "still should match (at least last 2 rows)", the IRL(Bosch) / Mass airflow(Siemens) and IOP(Bosch) / Reference Torque (Siemens). Of course if some deviation occurs, some kind of torque monitoring or control behaviour starts to be applied.

Then to finish, I have seen how ECU applies some engine controls around, for example at 450-460Nm (VCDS datalogging) / 500-520Nm(dyno) (+1389mg/stk), using Audi S3 8V OEM maps that "should" be valid around those load/torque values. Please know that ECU torque limiters and TCU software are updated. Of course something is missing / doing wrong, that is the point, then some tips will be highly appreciated.
« Last Edit: October 09, 2021, 04:22:59 PM by eliotroyano » Logged
prj
Hero Member
*****

Karma: +903/-420
Offline Offline

Posts: 5788


« Reply #8 on: October 10, 2021, 11:14:49 AM »

You can continue writing walls of text or you can log the ECU. One is going to help the other not.
Logged

PM's will not be answered, so don't even try.
Log your car properly.
eliotroyano
Hero Member
*****

Karma: +47/-7
Offline Offline

Posts: 796


« Reply #9 on: October 11, 2021, 01:32:21 PM »

You can continue writing walls of text or you can log the ECU. One is going to help the other not.

Sorry. Log attached.
Logged
eliotroyano
Hero Member
*****

Karma: +47/-7
Offline Offline

Posts: 796


« Reply #10 on: October 14, 2021, 05:35:53 AM »

Sorry. Log attached.

Hi, any comment?
Logged
IamwhoIam
Hero Member
*****

Karma: +43/-99
Offline Offline

Posts: 1030


« Reply #11 on: October 14, 2021, 05:38:54 AM »

Hi, any comment?

yeah, use vehical, 7 samples every 10 seconds won't get you very far at all, EVER.
Logged

I have no logs because I have a boost gauge (makes things easier)
eliotroyano
Hero Member
*****

Karma: +47/-7
Offline Offline

Posts: 796


« Reply #12 on: October 14, 2021, 05:46:40 AM »

yeah, use vehical, 7 samples every 10 seconds won't get you very far at all, EVER.

Thanks for the idea, that is planned now. But I ask about load/torque issues in log. Around 450Nm(log) 500Nm(dyno) car start to produce some weird behaviour, slow down and once time generate a DBW Monitoring error (not a torque monitoring issue???).
Logged
IamwhoIam
Hero Member
*****

Karma: +43/-99
Offline Offline

Posts: 1030


« Reply #13 on: October 14, 2021, 07:50:18 AM »

I'm sorry but I can't figure out shit from fuck in a log where each sample happens a few thousand RPM after the previous. Nope, not even gonna bother to look at it.
Logged

I have no logs because I have a boost gauge (makes things easier)
eliotroyano
Hero Member
*****

Karma: +47/-7
Offline Offline

Posts: 796


« Reply #14 on: October 14, 2021, 09:23:02 AM »

I'm sorry but I can't figure out shit from fuck in a log where each sample happens a few thousand RPM after the previous. Nope, not even gonna bother to look at it.
Sorry. Ok thanks.
« Last Edit: October 14, 2021, 09:26:40 AM by eliotroyano » Logged
Pages: [1]
  Print  
 
Jump to:  

Powered by SMF 1.1.21 | SMF © 2015, Simple Machines Page created in 0.105 seconds with 17 queries. (Pretty URLs adds 0s, 0q)