SDI Intake on ALH with Dyno Data

[vwmikel]

Did you take any boost logs when you did the dynos?

 

[Fix_Until_Broke]

Yes - Logs of channels 1 & 11 along with measurements of boost, exhaust and VNT actuator pressures. Pictures of the data can be found here. I can e-mail the raw data as well if you'd like it.

 

[vwmikel]

Yes - Logs of channels 1 & 11 along with measurements of boost, exhaust and VNT actuator pressures. Pictures of the data can be found here. I can e-mail the raw data as well if you'd like it.

That's quite a bit of data. It doesn't seem like the boost pressure is really affected by the volume of the intake much. I wouldn't be surprised if Stingray is right though and it could be an issue of volume where it just isn't that beneficial unless you're using a larger compressor.

You should send me your logs. I think we can dial in that tune a bit better.

 

[Fix_Until_Broke]

I'll try and get them sent to you (along with a handfull of other data I've gathered un-related to this) this week.

Is the 17/22 compressor not large enough to be considered a "larger compressor"?

I wish that VagCom was faster at sampling data or if I had the pinout for the IP so I could read in the QA command/feedback into my data acq.

 

[vwmikel]

I'll try and get them sent to you (along with a handfull of other data I've gathered un-related to this) this week.

Is the 17/22 compressor not large enough to be considered a "larger compressor"?

I wish that VagCom was faster at sampling data or if I had the pinout for the IP so I could read in the QA command/feedback into my data acq.

Well, a 1749 is a 49mm compressor and the 17/22 is a 1752, so it's really only 3mm larger. Which, is great, but the GTB2260VK is obviously a 60mm compressor. That's definitely a bigger step up and the reason we're seeing the numbers we are with them.

 

[turbo johan]

I had a SDI intake on my Lupo.
With 2+bar boost, ported head and big turbo it did really well.

Johan

 

[raydeee]

I have ordered a complete R32 Snorkle back TIP kit as well as a PD150 Intake From Ryanp. I will get it Dyno'ed with a before and after. Ill let you guys know.

Im interested on the results as well ,i will wait ,post it before i put mine on thnks

 

[Rub87]

This might be another trivial bit of information but I plotted the IQ and HP vs RPM on the very simple idea that more fuel makes more power.

Obviously that's not the case and/or there are other factors involved - Injection timing probably being a significant one (which I did not collect).

What I find very interesting is that the fuel delivery is more/less constant (per stroke) from 2000-4000 RPM, where as power is not. No big surprises here, probably just stating the obvious, just never seen the two compared to eachother.

The other interesting thing here is that for the same tune, the fuel delivery is higher earlier on the ALH than the SDI so that makes some sense as to why it makes more power at lower RPM's (there's more fuel), but that theory kind of falls apart at higher RPM's where the ALH intake maintains it's advantage even though fuel delivery is the same

51mg is just relative interal quatity, shape of torque cuirve depends probaly on the shape of the pump voltage map, was water temp always at 90 degC during all the runs?

 

[UFO]

Looks to me that the intake is not a significant flow restriction. Or maybe the MAF is saturated and that, or something else is limiting the fuel flow?

 

[JFettig]

What I take from these graphs and all the results from that dyno day is that there is something wrong with the dyno.

I think the max power someone saw that day was what, 140hp where that car previously came in over 200hp?

Jon

 

[UFO]

What I take from these graphs and all the results from that dyno day is that there is something wrong with the dyno.

I think the max power someone saw that day was what, 140hp where that car previously came in over 200hp?

Jon

That might explain a low number, but how does it explain a very clear difference over the rpm range?

 

[Fix_Until_Broke]

I updated the graph to include coolant temp - I don't see a correlation but maybe someone else does?

 

[Fix_Until_Broke]

Looks to me that the intake is not a significant flow restriction. Or maybe the MAF is saturated and that, or something else is limiting the fuel flow?

The tune that was run with the intake does not use the MAF

 

[Fix_Until_Broke]

Ok - so I re-plotted the data with the QA voltage instead of the IQ command per Rub87's comment and it shows a different story. More fuel = more power (more/less)

 

[VWBeamer]

I suspect a correction factor was changed on the dyno between runs....I'll be the first to admit this just my wishful thinking. I can not see how the SDI could make less HP.


I haver been to a lot of Dynodays and such with my aircooled car, and I can tell you that I've seen wildly different numbers put down by the same car on the same day because of different info put into the dyno's computer.

That might explain a low number, but how does it explain a very clear difference over the rpm range?

 

[m1ketdi]

I suspect a correction factor was changed on the dyno between runs....I'll be the first to admit this just my wishful thinking. I can not see how the SDI could make less HP.


I haver been to a lot of Dynodays and such with my aircooled car, and I can tell you that I've seen wildly different numbers put down by the same car on the same day because of different info put into the dyno's computer.

You can say what you like about correction factors, but as we can see the ALH manifold is flowing better at low rpm's meaning higher maf readings and the fuelling (QA voltage) is as shown higher in the low rpms.

 

[TDIMeister]

I have to say that I can think of no reason that the SDI manifold in and of itself would account for such a large magnitude loss of power vs. a stock manifold. If it were a bad-flowing or restrictive manifold, the difference in pumping work doesn't account for the ~20 lb.ft. torque deficit seen between around 2200-2500 RPM (~1.8 bar MEP). IQ doesn't show much as it's saturated at 51 mg/stroke. Past ~3300 RPM QA voltages are all the same, and doesn't explain the observed power difference.

More telling of the (in-)efficacy of airflow with the different manifolds are the MAF readings during the respective pulls. I also note that the SDI pull was done with the highest coolant temp as well. Plotting injection timing, VNT duty cycle and EMP might tell us something instead.

 

[ryanp]

The SDI manifold is designed to flow from the other side too, that wont be helping matters.

A calculated BHP figure should be quoted too if the car was removed the dyno, these TDI's have to be strapped down tightly due to the instant torque but it would have been easy second time around for the car to be strapped tighter or less than before.

I think a higher power output, 250+bhp, would have shown a better comparison

I wont be throwing my SDI inlet in the trash just yet

Thanks for sharing the info and putting the time and effort in to actually collect the data!

 

[VWBeamer]

Thanks for giving me permission to to say what I want on a public forum, but thankfully I do not need you permission to post here..

You can say what you like about correction factors, but as we can see the ALH manifold is flowing better at low rpm's meaning higher maf readings and the fuelling (QA voltage) is as shown higher in the low rpms.

 

[brum]

Haven't read the whole topic, but my two cents on the result.

With the SDI manifold you have better airflow. Better airflow means more air is required to achieve the same boost level so the ECU is closing the vanes on the geometry (can be seen on the results). More air in the cylinder = more air to be compressed. This is negative for the engine performance.

So the SDI is introducing two negative factors - requirement for more airflow (higher exhaust pressure, the geometry is closed more than with the stock manifold) and more air in the cylinder to be compressed. This two leads to the lower power of the car.

To confirm it measure the AFRs. I would bet that they will be higher with the SDI manifold.

So if all of the above is true - the SDI manifold is definitely the better option. But you need adequate tune for it. It would achieve more power with the same fuel cause it will require less boost for it. You need to adjust your AFR for optimal performance.

Edit: Actually the SDI manifold is introducing also third negative factor. The higher EMP. This means that the piston needs to do more work during exhaust phase.

 

[TDIMeister]

^ Emphasis that it's his 2 cents....... what it is, is finding any rationalization to explain a desired preconceived outcome without much sound physics basis. Let's let science and the data speak for themselves. Looking forward to F_U_B's update.

 

[TDIMeister]

There were 4 runs done total in this order

1) SDI Intake and TDTuning tune - MAF plugged in

came off dyno, intake swapped for ALH
back on dyno, warmed up

2) ALH Intake and same TDTuning tune - MAF NOT plugged in
Aborted run at 2000 RPM - forgot to plug in MAF after intake change and RC6 uses MAF
3) ALH Intake and RC6 - Different ECU
4) ALH Intake and RC5 - Different ECU

I don't know if the EDC15 ECU's "learn" driving style or not like the EDC16's do?

Here's an overlay with all the dyno runs numbered

I'm trying to make sense of the runs and their data. I interpret your statement as saying that run #2 was aborted at 2000 RPM due to your realizing that the MAF wire was disconnected but I see a complete run to 4000 RPM. I guess there were actually 5 pulls but the aborted one is not shown. OK.

Is it just me or is there a data alignment issue between power and QA voltage? Especially with the RC6 data, the QAV rises very quickly at low RPM without a corresponding torque increase (run #2 actually shows more torque with less QAV). If correct, I expect the RC6 to be comparatively smoky down low.

I'm increasingly of the suspicion that VNT duty cycle, EMP, IMP, MAF and injection timing (how about IAT and fuel temp?) will be quite telling. F_U_B, you're usually impeccable with data acquisition and I'm sure you had all of these logged.

 

[brum]

^ Emphasis that it's his 2 cents....... what it is, is finding any rationalization to explain a desired preconceived outcome without much sound physics basis. Let's let science and the data speak for themselves. Looking forward to F_U_B's update.

Actually I have some observation how much is the negative work of excessive boost. Tried to optimize my software for lower AFR in the whole IQ range and the fuel consumption dropped significantly...

Also this sounds as possible explanation . The SDI manifold should be optimized to deliver the air to the cylinders at atmospheric pressure. In the TDIs we've got the turbo to compensate for the lack of optimized manifold.

But even this is not scientific explanation it can easily be proved by the AFR data from the two runs. This would also provide some test data regarding the benefits of the SDI manifold over the other manifolds.

 

[m1ketdi]

If the sdi manifold really flowed better though we should see larger maf readings at low rpms.

But We can see from lower quantity adjuster voltages that MAF readings in the SDI run were lower below 2000rpm. Unless something significant had changed to limit fuelling such as fuel temperature extremely high etc, this is where logs of actual MAF would back up my thoughts.

 

[brum]

Agree. But looking at the graphs I can not say for sure if the IQ was more on some of the tests. If they are not the same, they are pretty close.

MAF data can also provide certain information about the manifold flow. Also the EGT can give some hints. But none of this is present here.

 

[TDIMeister]

But even this is not scientific explanation it can easily be proved by the AFR data from the two runs.

I agree that AFR will show something but in Diesel engines this is not commonly available. If F_U_B measured it, great, let's see it.

But We can see from lower quantity adjuster voltages that MAF readings in the SDI run were lower below 2000rpm. Unless something significant had changed to limit fuelling such as fuel temperature extremely high etc, this is where logs of actual MAF would back up my thoughts.

The QA voltage cannot be interpreted as an analog for air flow (QA = quantity adjuster = fuel)in isolation of the requested IQ, which is programmed in the tune. I'm not completely sure but I interpret the list of runs as saying that the TDTuning program does not rely on a MAF anyway.

 

[VWBeamer]

I should have mine on soon, and I will try to get some dyno numbers.

with my schedule, it could be a while though.

 

[Fix_Until_Broke]

Sorry for the delay in responding here...

The SDI manifold is designed to flow from the other side too, that wont be helping matters.

I'm no manifold expert, but I'd be hard pressed to see a difference between flowing in one side versus the other on this design. It's pretty symmetric inside.

 

[Fix_Until_Broke]

Haven't read the whole topic, but my two cents on the result.

With the SDI manifold you have better airflow. Better airflow means more air is required to achieve the same boost level so the ECU is closing the vanes on the geometry (can be seen on the results). More air in the cylinder = more air to be compressed. This is negative for the engine performance.

So the SDI is introducing two negative factors - requirement for more airflow (higher exhaust pressure, the geometry is closed more than with the stock manifold) and more air in the cylinder to be compressed. This two leads to the lower power of the car.

To confirm it measure the AFRs. I would bet that they will be higher with the SDI manifold.

So if all of the above is true - the SDI manifold is definitely the better option. But you need adequate tune for it. It would achieve more power with the same fuel cause it will require less boost for it. You need to adjust your AFR for optimal performance.

Edit: Actually the SDI manifold is introducing also third negative factor. The higher EMP. This means that the piston needs to do more work during exhaust phase.

So, if I follow your logic correctly, we should install a throttle in the intake tract for better performance?

 

[Fix_Until_Broke]

TDIMeister...

Yes, I was not clear about my 2nd dyno run (post edited now), The aborted run was actually at the beginning of the 3rd run with the RC tune that uses the MAF. I forgot to plug the sensor in - stopped the run, Plugged the sensor in and then re-started the 3rd run.

I don't think there was a data alignment issue with the QAV and dyno data - I was pretty careful to get the data synchronized. Yes, RC6 is smokey down low compared to RC5 or the TDT tune, however it is also much more powerful down low as well, particularly in day to day driving (not 100% throttle).

As to why the shape of the power curve and the QAvoltage curve don't seem to line up, I don't know. I'd love to know what makes the power curve change slope at ~2000 RPM.

The link in Post 32 takes you to the data (pictures) of the IMP/EMP/N75/etc like I've run in the past. Since the MAF tends to saturate quickly I did not record this channel. I only recorded Fuel and Boost (4 & 11?) to get a faster sample rate. In hindsight, I probably should have had injection timing, but the resultant sample rate is painfully slow (less than 1/second). One additional channel that I had this time is VNT actuator vacuum. I have a PD actuator with the position sensor, but have not taken the time to figure out the pinout. Watching the VNT vacuum and EMP's I can say with confidence that the vane position follows the command very quickly (not sure about accuracy/repeatibility). The vacuum gauge and EMP's move as quick as you can move your right foot.

No MAF on the TDT tune