Problem reading some Sensors with OBDII Scanner+Torque

[hotpocketdeath]

Does anyone else have a problem reading some sensors?

Sensors I can read
Boost
RPM
EGT1
Speed
MAF
Ambient Temps
Engine Load
many more.......


Sensors I can not get to work
Coolent temperature
Fuel Flow
MPG
CatB1S1

Now I'm on an MK6 Golf TDI, but this guy has a MK6 Jetta TDI and he can obviously read the sensors and I'm pretty sure has the same OBDII reader.

I've tried 2 OBDII readers, both the same model and both can't read the sensors I'm looking for. Connecting the OBD readers to my brother's van and I can see the coolent and CatB1S1 sensors, but no fuel flow.

I do have a VCDS and when I connect it to my car and check the Generic OBDII section, and I can see all the data right where it should be.

Anyone have an opinion about this?

 

[Diesl]

My car is a different generation, but the torque behavior should be similar: when you define a new display, only values in green are available. You could also download and run torquescan, a free Torque plugin; that should also tell what is available. And I think the paid version has an option (from the main circular selector) to list in text form what it thinks is available. (Or maybe that is torquescan.)
EDIT: The list is under 'OBD Adapter status', from the main circular menu.

If the variables you are looking for are not listed in the Torque default set, all is not lost, as long as your car makes them available via OBD. Defining custom PIDs for Torque is even easier than defining Xgauges for Scangauge.

 

[hotpocketdeath]

I have the paid version of Torque w/ Torquescan. Those specific sensors are not showing up as even being available. But it seems like they should be.

I've even compared the PID to what Torque should be reading to what VCDS reads, and it's the same. VCDS reads that sensor, the OBDII reader I have does not. I've even tried making the custom PID reading, and still get nothing. It's like the reader just isn't able to read that data.

I just ordered another OBDII reader from amazon. Different design, so I hope it might work better. But it's cheap, so not a big deal.

 

[Diesl]

My car is a different generation, ....

That of course is complete nonsense. For some reason (not-so-early-onset dementia probably) I read your 2010 as a 2006. So our car electronics are the same generation.
Cat b1s1 should work (works on mine, see http://forums.tdiclub.com/showthread.php?p=4146263#post4146263), and I think also coolant temp. MPG is something the app computes itself, so I don't know what's needed for that (car setup in torque, length of connection, and things of that nature - to quote Aanold).
What does torque have to say about your adapter? Any errors? Mine was really cheap, reports a fake Bluetooth ID (11:22:33 etc.), and needs to re-pair every damn' time I start the app. But it does work.

 

[hotpocketdeath]

Cat b1s1 should work (works on mine, see http://forums.tdiclub.com/showthread.php?p=4146263#post4146263), and I think also coolant temp. MPG is something the app computes itself, so I don't know what's needed for that (car setup in torque, length of connection, and things of that nature - to quote Aanold).

I believe MPG relies on certain sensors before it can be calculated, such as Fuel Flow. So naturally, if I can't get the data......

The two sensors I would really like to see are the coolant temps and fuel flow rate.

What does torque have to say about your adapter? Any errors?

I'll have to get back to you on that. So far, I haven't seen any errors, but I'll recheck.

 

[hotpocketdeath]

Well, I did some more testing. No errors are appearing from the OBDII adaptor. And the ECU definitely shows those sensors as being available.

I tried another OBD reader "Dash Command", and it's able to read the Fuel Flow sensor, so it appears to be an issue with the Torque app.

 

[hotpocketdeath]

Well, after messing with some settings that were suggested by the Torgue developer, I got my Fuel Rate to work.

Still don't have my coolant temps working yet.

 

[APT]

Are you using Torque Pro or just Torque?

 

[hotpocketdeath]

Are you using Torque Pro or just Torque?

I have Torque Pro.

I just got my Coolant temps to read, but only if I use PID 67. But even though the temps appear correct, I feel it's still not quite right. So I'm still playing with it. I think PID 05 is the more accurate data.

Although I believe I got CACT working (Charge air cooler temperature). Air temperature after the intercooler. I'll post the information after I double check the results to make sure the equations are right.

 

[Diesl]

[...] Although I believe I got CACT working (Charge air cooler temperature). Air temperature after the intercooler. I'll post the information after I double check the results to make sure the equations are right.

All the scangauge Xgauges I posted here should work under torque as well.

 

[Lightflyer1]

Diesl can you describe how to enter these into the custom PID's in Torque for those of us who are oblivious to how this works? Thanks in advance!

 

[hotpocketdeath]

All the scangauge Xgauges I posted here should work under torque as well.

Yep, saw those. I'm slowly building a number of custom guages for Torque. Some gauges work better than the ones built into Torque. Like Fuel Flow.

Edit: Hmmm, I've been going over some of my results compared to the information you posted in that link and some of the "equations" described are "off".

Take for instance this item:

BP2 70 05414670 4810 0001001E0000 bar     requested manifold pressure

When you request PID 70, the response with this data is bytes B and C. So put them together as "(B*256)+C)".

I've already pulled these bytes from my car at idle. I've also checked the OBD2 results in VCDS which gives 101kPa (or 1.01Bar). This is what the end result should match. Pulled from my car, I get B=12, C=169. This results in a number 3241.

Now the MTH number 0001001E0000 using this guide results in a formula of 1/30+0 or simply 1/30.

Next, according to the same guide, the RXF number 05414670 indicates that the results should be divided by 100

So the final equation is (((B*256)+C)*(1/30))/100. Plug into my calculator and I get 1.08Bar or 108kPa.

It's close, but it does not equal the results I read in VCDS when it was able to show me what the results should be.

I found that using a simpler equation ((B*256)+C)*.0312 give a far more accurate result that better matches what VCDS gives.

 

[Diesl]

Interesting! So maybe the divisor should be 32? I just guessed the 30. There goes my boost at idle theory...

Comparing to VCDS readings is definitely the way to go. Could you also have a look at the DPF pressure and the 7A variable? I could never get one of the other VCDS guys interested enough in the lowly OBD stuff to actually bother. Maybe you can figure out what these variables actually correspond to on the VCDS side.

 

[Diesl]

Diesl can you describe how to enter these into the custom PID's in Torque for those of us who are oblivious to how this works? Thanks in advance!

Hotpocket just posted a link to a recipe. I don't think I could explain it any better. Actually, I'm pretty sure I could do a much worse job...

 

[hotpocketdeath]

Interesting! So maybe the divisor should be 32? I just guessed the 30. There goes my boost at idle theory...

Comparing to VCDS readings is definitely the way to go. Could you also have a look at the DPF pressure and the 7A variable? I could never get one of the other VCDS guys interested enough in the lowly OBD stuff to actually bother. Maybe you can figure out what these variables actually correspond to on the VCDS side.

I'll check it out when I can. I remember seeing several measuring blocks dedicated to nothing but DPF information. I'll have to look for those blocks again. I was going to work on the fuel pressure controller next.

And I'd agree with the divisor being 32. It fits.
New Equation to get boost is even simpler: ((D*256)+E)/32
Bytes B and C are for requested, bytes D and E are actual pressure readings. I found bytes F-I returns null data.

Currently, the problem with this formula is it includes barometric pressure. My current formula subtract 97 from the results to compensate for barometric pressure. Ian (Torque Developer) got back with me about doing some advanced stuff dealing with assignable virtual PIDs. This will allow us to create a boost guage that automatically subtracts barometric pressure as it changes. So I'm trying it out.

Hi!

This is something you can already do - simply create a virtual PID (this is your variable) then you can reference it from other PIDs you make

You can look up the ID of the virtual PID by using the 'TorqueScan' plugin (it will be visible in the list of PIDs) - it will be something like [fe1234]

Leave the 'PID' field itself blank for virtual PIDs

You can then reference that PID in the equation of another PID: A+B+[fe1234]

Regards

 

[Diesl]

I'll check it out when I can. I remember seeing several measuring blocks dedicated to nothing but DPF information. I'll have to look for those blocks again. I was going to work on the fuel pressure controller next.

Thanks! I just don't see a VCDS in my future anytime soon, unfortunately.

And I'd agree with the divisor being 32. It fits.
New Equation to get boost is even simpler: ((D*256)+E)/32
Bytes B and C are for requested, bytes D and E are actual pressure readings. I found bytes F-I returns null data.

Mhmm. For some reason I thought it was the other way around. The values for D&E looked suspiciously 'even' in hex, and B/C would overshoot D/E when going full throttle.

Currently, the problem with this formula is it includes barometric pressure. My current formula subtract 97 from the results to compensate for barometric pressure. Ian (Torque Developer) got back with me about doing some advanced stuff dealing with assignable virtual PIDs. This will allow us to create a boost guage that automatically subtracts barometric pressure as it changes. So I'm trying it out.

I like having the 1 atmosphere offset in there. I view manifold absolute pressure as the 'volume (torque, power) multiplication factor' that the turbo creates. For people using PSI having to mentally subtract 14.7 is of course pretty annoying.
But thanks for sharing the tip on combining values using virtual PIDs. I think I'll try computing real power, and maybe something resembling DPF soot load from DPF differential pressure and expected pressure (assuming my wild guess that PID 7A is expected DPF pressure drop from ash is right).

 

[hotpocketdeath]

Mhmm. For some reason I thought it was the other way around. The values for D&E looked suspiciously 'even' in hex, and B/C would overshoot D/E when going full throttle.

It took checking the values in VCDS to determine which set of bytes belonged to which reading. At idle VCDS reported requested as 101kPa and actual at 97kPa. When I took the bytes from Torque, B+C was 3241 and D+E was 3103. So it looks pretty clear which is which.

I like having the 1 atmosphere offset in there. I view manifold absolute pressure as the 'volume (torque, power) multiplication factor' that the turbo creates. For people using PSI having to mentally subtract 14.7 is of course pretty annoying.
But thanks for sharing the tip on combining values using virtual PIDs. I think I'll try computing real power, and maybe something resembling DPF soot load from DPF differential pressure and expected pressure (assuming my wild guess that PID 7A is expected DPF pressure drop from ash is right).

The problem is that barometric pressure isn't always 14.7. Where I live, it's around 14.1-14.2. But it's nice to know that if I take a trip to the Florida coast, I won't need to make any changes this area.

Anyways. I tried out my new formula on the way home from work.

The results feel much better than the built in boost guage. Information I'm seeing is more what I would expect to see and response seems much quicker.

My settings:
Mode/PID = 0170
Long Name = VW Boost
Short Name = Boost
Minimum = 0
Maximum = 24
Scale = 1
Unit Type= kPa
Equation = (((D*256)+E)/32)-([ff1270]*0.1)
Header = Auto

I went with [ff1270] for barometric pressure which is the PID for the Android Phone. Since it was being read in mBar, I needed to convert it to kPa, which is what the 0.1 does. I didn't want to pull this info from the car since it would just add one more thing the adapter would need to read.

 

[hotpocketdeath]

Comparing to VCDS readings is definitely the way to go. Could you also have a look at the DPF pressure and the 7A variable? I could never get one of the other VCDS guys interested enough in the lowly OBD stuff to actually bother. Maybe you can figure out what these variables actually correspond to on the VCDS side.

Ok, I took a look at this PID and some others and here's what I got worked out for them.

PID 7A - Returns only one set of data in bytes B&C. Lists this information as DPF Delta and is in kPa. Formula appears to be ((B*256)+C)/100

PID 73 - Labeled Exhaust pressure, measured in kPa, same formula as PID 7A

PID 78 - Exhaust temp 1-4 (Bytes are 1-B&C, 2-D&E, 3-F&G, 4-H&I). Measured in degrees C. Formula works out to be (((B*256)+C)-40)/10

PID 6D - Fuel Pressure. Requested is bytes B&C, Actual is D&E. Measured in kPa. Formula appears to be ((D*256)+E)*10

 

[Diesl]

So the names and the formulas/lae are confirmed by or coming from VCDS?

 

[hotpocketdeath]

So the names and the formulas/lae are confirmed by or coming from VCDS?

Yep.

 

[hotpocketdeath]

PID 78 - Exhaust temp 1-4 (Bytes are 1-B&C, 2-D&E, 3-F&G, 4-H&I). Measured in degrees C. Formula works out to be (((B*256)+C)-40)/10

Update:

I have been talking with Ross Tech and the developer of Torque to talk about why I am seeing two different sets of temperature readings coming from the same sensors.

Without going into too much details, it looks like Ross-Tech is not using the correct forumla to calculate the exhaust temps in the Generic OBDII section.

This was confirmed by using VCDS to go to 01-Engine measuring block 99, which gave readings more inline with what Torque read.

But more importantly, I checked the temps when my car had been sitting for 14 hours. When all other readings showed close to ambient (around 60f), the generic OBDII under VCDS showed around 115f. Obviously not right.

I've already told Ross-Tech of my findings so that they can fix it in the next update.

So the proper formula should be (((B*256)+C)/10)-40. Although, it seems like my readings are a few degrees high, so I'm using "-43" to compensate. That's the nice thing about custom PID guages.

 

[Diesl]

Hey HPD, on my car actual manifold pressure and actual fuel pressure only use the first byte. From one of your posts (#17) it seems you are really seeing 16 bit values?

 

[hotpocketdeath]

Hey HPD, on my car actual manifold pressure and actual fuel pressure only use the first byte. From one of your posts (#17) it seems you are really seeing 16 bit values?

Well, unfortunately, I can't guarantee all PIDs will work. My car won't give me a reading on PID 05, which is the normal engine coolant PID. I have to use a diffferent PID. I also don't get Cat B1S1 (PID 3C).

I do know that the fuel pressure control (PID 6D) returns a 6byte value.
Byte 1 = Sensor info (which sensors are used)
Byte 2-3 = Requested fuel pressure
Byte 4-5 = Actual fuel pressure
Byte 6 = Fuel Temperature

I get all data from the first 5 bytes, but byte 6 returns null data on my car. So I only get partial information. Watching this guage, I've seen my fuel pressure go all the way up to ~24,000psi. Which seems to be inline with what the fuel system can do.

How are you checking for data? All I was doing was loading the PID, setting the formula only to the byte I was looking for data on and hit test.

 

[Diesl]

Well, unfortunately, I can't guarantee all PIDs will work. My car won't give me a reading on PID 05, which is the normal engine coolant PID. I have to use a diffferent PID. I also don't get Cat B1S1 (PID 3C).

Same here (not all pids work). But the bitmap pids x0,x20 etc tell me which ones work.

I do know that the fuel pressure control (PID 6D) returns a 6byte value.
Byte 1 = Sensor info (which sensors are used)
Byte 2-3 = Requested fuel pressure
Byte 4-5 = Actual fuel pressure
Byte 6 = Fuel Temperature
I get all data from the first 5 bytes, but byte 6 returns null data on my car. So I only get partial information. Watching this guage, I've seen my fuel pressure go all the way up to ~24,000psi. Which seems to be inline with what the fuel system can do.

Interesting, both the fuel pressure observation, and the number and meaning of bytes. So the rail pressure really does vary with rpm ( and maybe load). For pid 6D I have to look at my notes in the car. For pid 70, I know that the byte at 5008 (second byte of requested) is always zero.

How are you checking for data? All I was doing was loading the PID, setting the formula only to the byte I was looking for data on and hit test.

Same procedure, I check it byte by byte.

 

[hotpocketdeath]

Interesting, both the fuel pressure observation, and the number and meaning of bytes. So the rail pressure really does vary with rpm ( and maybe load). For pid 6D I have to look at my notes in the car. For pid 70, I know that the byte at 5008 (second byte of requested) is always zero.

I wouldn't say the fuel pressure varies with RPM, but rather with throttle load.

PID 70 should operate very similarly as PID 6D.
Byte 1 = Sensor info (which sensors are used)
Byte 2-3 = Requested boost pressure
Byte 4-5 = Actual boost pressure

Except there are 4 more bytes associated with this PID (9 bytes total). I believe these bytes are for additional sensors or cars with dual turbos. These bytes only return 0 for me.


And it's weird watching fuel pressure. If I'm driving along at a moderate throttle, my fuel pressure will be at around ~12-14,000psi. But when I lift my foot off the pedal, it will jump up to ~20,000psi for a few seconds before coming back down to ~4,000psi. I think this is caused because the injectors are no longer giving the fuel a pressure release. At full throttle, I see it jump up to ~24,000psi, which looks normal to me.

It's amazing just how much the pressure in the fuel system jumps around while driving.

 

[Diesl]

PID 70 should operate very similarly as PID 6D.
Byte 1 = Sensor info (which sensors are used)
Byte 2-3 = Requested boost pressure
Byte 4-5 = Actual boost pressure

On my car (2012 JSW TDI w/DSG) byte 1 is 3, and byte 5 is always zero.

Except there are 4 more bytes associated with this PID (9 bytes total). I believe these bytes are for additional sensors or cars with dual turbos. These bytes only return 0 for me.

Zero for me as well.

I haven't watched 6D as much, mostly because I wasn't sure it really is fuel pressure. Why would they (VW) vary fuel pressure, instead of just injection duration (and number of pulses)? Is the range in injection quantity not large enough without varying the pressure?

But anyways, if VCDS believes it's pressure then that should be good enough for me as well I guess I'll remove the question marks from my table in the other thread.


As far as values go, with the conversion/MTH as given in my table (other thread), I've seen values up to 1700 bar, which more or less matches your numbers.

 

[Redrocketdi]

It took checking the values in VCDS to determine which set of bytes belonged to which reading. At idle VCDS reported requested as 101kPa and actual at 97kPa. When I took the bytes from Torque, B+C was 3241 and D+E was 3103. So it looks pretty clear which is which.
The problem is that barometric pressure isn't always 14.7. Where I live, it's around 14.1-14.2. But it's nice to know that if I take a trip to the Florida coast, I won't need to make any changes this area.
Anyways. I tried out my new formula on the way home from work.
The results feel much better than the built in boost guage. Information I'm seeing is more what I would expect to see and response seems much quicker.
My settings:
Mode/PID = 0170
Long Name = VW Boost
Short Name = Boost
Minimum = 0
Maximum = 24
Scale = 1
Unit Type= kPa
Equation = (((D*256)+E)/32)-([ff1270]*0.1)
Header = Auto
I went with [ff1270] for barometric pressure which is the PID for the Android Phone. Since it was being read in mBar, I needed to convert it to kPa, which is what the 0.1 does. I didn't want to pull this info from the car since it would just add one more thing the adapter would need to read.

I have tried 3 different formulas and the gauge doesn't seem to work for me. I'm using torque pro app and it doesnt give me the option to enter a PID number?

 

[Diesl]

If you can't enter a PID, how are you getting any measurement?