VCDS to Usable Data???

When pulling fueling info from VCDS in a log, it always has a range of 0 to 51 mg/str no matter if you car is stock or tuned; kinda like scaling a 4 bar sensor to the diagnostic limit in Group 11 for boost. Can this be converted to usable data, converting volume per stroke to volume per unit of time? The conversion is multi-step but simple, however the VCDS numbers are somewhat arbitrary according to many members on here.

Does anyone have helpful information we can coalesce in one spot regarding the flow data of each nozzle type we have on the market? We've had threads analyzing the flow performance between the nozzles, but only in relative terms and nothing is specific, abosolute units with tolerences. Also, nozzles are usually refered to as what is the horsepower they support so it kinda adds mud to the water.

I know we can use the voltage going to the pump to figure out where we lie on fuelling too, but I wouldn't expect the transfer function of that voltage to be linear with respect to the maximum flow rate of a nozzle.

Also, as a side note, does anyone have access to add to the following page, or make a sister page with the current, larger nozzles on the market?
http://pics2.tdiclub.com/gwillie/VW/nozzlesize/index.htm

We have flow maps for turbos, just not flow maps for nozzles

I go by QA Voltage instead of IQ from the same group

And that's what I understand is a better way to do it, but how can we translate/convert that to usable flow data? I ask because its good information to know, but it also helps identify how to setup a water/meth kit or even calculate AFR...

A given QA collar voltage equates to a given effective pump displacement. Since we primarily deal with either 10mm or 11mm pumps, a given QA voltage should be more/less same fuel delivery to the nozzle for a stock 10mm or 11mm pump.

Where it gets complicated is what each nozzle does with this slug of fuel. A smaller nozzle will take the same QA position and deliver a higher pressure in the lines and less fuel in the cylinder all other things being the same.

And that is something that is partially reported in Group 8 or 13 (fuzzy memory) where we can see the each nozzles relative IQ. Though I'd imagine we'd have to hook up flow meters on the pump's inlet and outlet. Though I don't know how useful those would be in the pump has any sort of a reservoir. Pardon my ignorance on the topic.

This can be altered to allow a larger diagnostic limit and the pump voltage map re calibrated to show correct quantities.

However, 99% of tuners don't have the equipment or time to do this, costs to do it properly would be alot more than anyone is willing to pay for a tune....

I think the proper term for the QA should be Duration Adjuster. It effectively controls the injection duration. Combine duration with the injection pressure and you can calculate quantity knowing how a nozzle flows.

What would the base unit be? milliseconds, micro-?

I know this isn't an easy question, but how close can one get in estimating these values?

According to the FAQ:

Manual transmission A4 TDI's have 800 bar (11,760 PSI) pump side pressure with 1100 bar (16,170 PSI) injection nozzle pressure. Automatic transmission A4 engines have 950 bar (13,965 PSI) pump side pressure with 1350 bar (19,845 PSI) injection nozzle pressure

Click to expand...

in your ecu maps there are limiters (map addresses) that you can raise from '51'mg to 'whatever you like' so that VCDS will display it properly and your MPG readout will be correct.

Whether it's for fuel economy, AFR, or any other statistical means, how does one figure out, like what JFettig mentioned, what value to set it to? How do we go from arbitrary to something meaningful?

Measure how much fuel and air is being used at every speed and load site.....

Digital Corpus said:

What would the base unit be? milliseconds, micro-?

I know this isn't an easy question, but how close can one get in estimating these values?

According to the FAQ:

Click to expand...

The fundamental units are degrees of crank rotation as a time base is only relative for a given engine RPM. The pump, lines, nozzle are all a system that work together to deliver a mass of fuel to the cylinder.

The pressures given in the FAQ are singular operating points (and how the pressure at the pump can be lower than the pressure at the nozzle is curious...). Those pressures will change with fuel delivery and engine RPM.

What you're asking is technically possible, however very complicated with many interdependent factors to get repeatable results, not to mention accurate results that can be compared (not unlike dyno testing).

Fix_Until_Broke said:

The fundamental units are degrees of crank rotation as a time base is only relative for a given engine RPM. The pump, lines, nozzle are all a system that work together to deliver a mass of fuel to the cylinder.

Click to expand...

Gotcha, this makes more sense. Physical time woul then be related to RPM as the injection event lasts a certain number of degrees.

Fix_Until_Broke said:

The pressures given in the FAQ are singular operating points (and how the pressure at the pump can be lower than the pressure at the nozzle is curious...). Those pressures will change with fuel delivery and engine RPM.

Click to expand...

I've seen the simplified graphs on DBW's site. After the pilot pressure of the injector is met, ther is a sudden drop as the injector opens, pressure then builds back up and injects the remaining fuel for that window.

Fix_Until_Broke said:

What you're asking is technically possible, however very complicated with many interdependent factors to get repeatable results, not to mention accurate results that can be compared (not unlike dyno testing).

Click to expand...

So, the end result is being that it's not possible to know the mass of the fuel injected during the event based off of the IQ number or voltage (though voltage would be the better measure) unless someone has the time and money to invest in developing a system to monitor the variables such as pump voltage, needle lift sensor, and crank position with high accuracy and time resolution?

If that is all completely true, then how does the ECU manage accurate fueling in MAPs that are far less than 255 units by 255 units, which could also be considered "crude"?

Below is my own conjecture as I've never done the fully detailed testing/analysis etc...

I'm guessing that the OEM's use full on data acquisition with air/fuel sensors, mass flow information from the injector manufacturer (Bosch in this case) and spend a lot of time (months?) tuning the engine to know what it does so they can make it work in the 255 units of the ECU and meet emissions, which is their main goal. They're not dealing with all the variables that we are (intakes, nozzles, heads, cams, turbo, tune, exhaust, intercoolers, compression ratio, porting, etc) so they can instrument the crap out of it and make the ECU work with very limited resolution since they're dealing with the same hardware.

I'm guessing that if you have the same nozzles* on the same injection pump on different cars, that you can compare QA voltages and get similar fuel delivery (within say 10% or so). *this assumes that the nozzles are identical, same pressures, etc

OK - great, so the fuel delivery is the same, what does this mean? How do you prove it?
Are the IP's equal - same fuel temp, same piston/bore clearance, same collar absolute location vs QA voltage, etc. EGT's the same, Smoke the same, Power the same - all of these are very subjective and have tons of variability.

I'd take the simplified graphs from DBW with as much salt as they put on the roads up here (a lot). No units and not at all forthcoming with any information other than a picture of some unidentifiable data and a lot of "because I said so". I'm guessing that a pressure vs time trace would probably look like that in certain operating conditions, but it may look significantly different at other conditions as well depending on fuel delivery, RPM, etc.

To bump this thread with another possible means of measuring injection quantity, but in a more gross manner, and not specific to a nozzle:

Does the pump have a reservoir? If not, could we not measure the flow rate of the fuel going into the pump and the flow rate of the pump's return line and then just subtract it? I know there is a return line from the injectors as well so that may have to be measured.

I know this won't give specific IQ, but if there is no buffer in the pump, then this would give us accurate fuel consumption numbers, no? If so, anyone know of industrial flow meters that can deal with these low flow volumes and high temps of diesel?

I have been thinking of this as well, and we do a lot of this at work measuring low flows of various chemicals. I did a preliminary search and came up with some companies that make low flow diesel measuring devices.

http://www.jv-technoton.com/fuel_flow_meter/

Differential volume measurement will give a rough estimate of net fuel flow, however you'll find that at low to medium power that you're returning more volume than you're consuming due to the thermal expansion of the fuel and the very low percentage of the gross fuel flow that is actually used.

If you put a temp sensor in each line and account for the thermal expansion it will be more accurate, but you will be much better off if you measure mass instead of volum
e.

Want to know the relative flow of each nozzle in a car? (where it matters)

Easy

Get ye to a dyno, with a bunch of different injectors. Test each one at the same load and engine speed, (*NOT* WOT) see what VCDS reports for IQ (which you can then translate back to the positioner.)

Easier/cheaper, do it on the road at a fixed speed, let's say 75 mph, over a few miles, install new nozzles, rinse and repeat, and use some statistical tools to analyze the data. It is important to have a significant load on the engine, let's say at least 60% or maybe between 20-25 mg/r. In my experience, 20 mg is a good place to start with OEM injectors. The results do become less linear/reliable as you get farther from stock, though. OEM calibration volts vs IQ is usually pretty close to accurate over the OEM operating region for that particular car with OEM equipment, but the farther you get from this baseline, the less correct the correlation within the factory's Volts/IQ map calibration. Also a host of potential errors introduced by things that change in-cylinder pressure conditions when the fuel is injected, like boost and SOI for example. That's why I say start with 20 mg when testing stock. For example, with a normal 10mm IP, R520 will flow something like ~10-13 mg compared to ~20 mg on stock injectors at the same load/road speed in an otherwise 100% stock 5-speed car. (Another hint: it's not a strictly linear relationship between volts/IQ, either!) Group 20 for a ve pump TDI will show you throttle position and IP volts with the associated IQ.

Alternately, you could hook up the injection pump to a bosch test stand, run the pump at a set rpm, and set the positioner volts. Measure into a graduated cylinder, change one variable, rinse and repeat.

Or some combination of the above if you want to design a more rigorous test protocol. But the idea here is to predict the results in a car, so why not just test them in the car, using the measuring tools that Bosch and VW so kindly supplied with each car?

The IQ (volume) numbers you get by road testing will vary based on fuel temperature and the fuel grade, both of which affect density. The number you see in the VCDS display has the specified IQ adjusted for fuel temperature- the ecu injects longer at higher fuel temps so there's consistent performance over a wide range of climactic conditions. So for example, if the reference amount is 20 mg, the ecu may specify 21 or 22 mg instead (meaning longer duration for the same torque output) if the fuel is hotter. So it's important to look at both the IQ value and volts compared to throttle position- And this is why I like group 20 for these types of measurements.

But where it really counts is at WOT, and the in-cylinder conditions are different than at part-load. And this varies even more depending on the tuning. Then you have to consider, more worn injectors (or poor quality ones) will have a higher leakby/return flow, so need to inject more more duration for the same in-cylinder delivery. But if you are after a more or less acedemic discussion of how much an injector flows relative to another, this is a decent enough place to start.

interesting, and wouldn t it be possible to fix a device or two on the diesel lines around the pump which would evaluate flow/fuel quantity effectively circulating threw them, at the entry and exit of the pump lines, add some temperature gauges to be able to calculate more accurately volume and do roads/dyno tests at whatever rpm/load/acceleration/speed it s been decided to compare consumption...?

Easiest way in my opinion would be to use a wideband O2 sensor on the downpipe and the MAF (if in good condition) for estimating the amount of fuel going into the engine.

Then, you can do the needed corrections on the pump voltage map of your ECU and that would enable you to display correct numbers (or at least not too far) for Injection Quantity (if you know how to enlarge diagnosis limit of 51mg but it's really easy)

cruiserboy said:

Injection Quantity (if you know how to enlarge diagnosis limit of 51mg but it's really easy)

Click to expand...

Hi Cruiserboy
Please tell me how

cruiserboy said:

Easiest way in my opinion would be to use a wideband O2 sensor on the downpipe and the MAF (if in good condition) for estimating the amount of fuel going into the engine.

Then, you can do the needed corrections on the pump voltage map of your ECU and that would enable you to display correct numbers (or at least not too far) for Injection Quantity

Click to expand...

I m not running on diesel only so i can t do that