Tuning Wiki

This is a historical revision. Read the current page or return to its history.

Page organisation in this revision
Page type
Technical article
Subject
Diagnostics
Parent
Section level
Applies to
No ECU-specific applicability
Tool groups
None
Manufacturers
Volkswagen

Changes from revision 1 to revision 1

Removed text is struck through; added text is highlighted.

Bosch meters on the TDI

Volkswagen TDIs of this generation use a five-pin Bosch hot-film meter. 0 281 002 216, Volkswagen 028 906 461, and 0 280 217 121, Volkswagen 06A 906 461, are the same meter apart from the length of the housing. The transfer curve drawn for them matches Bosch 0 280 218 018. Wire colours and control-unit pins below are from a 66 kW ALH.

PinFunctionColourGoes to 1Temperature output. Not used on the ALH. 212 V supplyYellow/blackRelay 109 3GroundBrownControl-unit pin 4 45 V referenceRed/greenControl-unit pin 50 5Output voltageBrown/blueControl-unit pin 52

The signal wire is pin 5. Colours on a particular car can differ from the ALH colours. Output voltage rises with mass flow. At idle with the EGR closed, a Bosch meter on a 1.9 reads about 2.2 V.

Pierburg meter

The Pierburg meter, Volkswagen 074 906 461, has six pins. Colours and pins below are from a 66 kW AGR and an 81 kW AHF.

PinFunctionColourGoes to 15 V referenceRed/greenControl-unit pin 50 2Signal groundBrown/blueControl-unit pin 25 312 V supplyYellow/blackRelay 109 4Not connected 5Power groundBrownGround 6Output voltagePurple/brownControl-unit pin 52

At the same idle, EGR closed, the Pierburg meter reads about 1.6 V. A control unit calibrated for the Bosch 2.2 V will under-read the air if it is given the Pierburg voltage unaltered, and the other way round. Matching one family to a control unit calibrated for the other is a gain and an offset on the signal wire: the Pierburg voltage has to be raised and shifted to look like the Bosch curve, and the Bosch voltage has to be attenuated to look like the Pierburg curve. The output of these meters will not drive a heavy load, so any network on the signal wire has to buffer it.

Milligrams per stroke and kilograms per hour

The control unit reports mass per intake stroke, not kilograms per hour. A four-stroke four-cylinder engine takes two induction strokes per revolution. At 900 rpm that is 900 × 2 × 60 = 108 000 strokes per hour. A reading of 480 mg/stroke is 108 000 × 480 × 10⁻⁶ = 52 kg/h.

480 mg/stroke is the idle reading on a 1.9 with the EGR closed. A full cylinder of 0.48 litre, at about 1290 mg of air per litre, would be 0.48 × 1290 = 619 mg/stroke if the cylinder filled completely. It does not: the valve is open for a limited time, so the real idle mass is lower. A turbo that was actually making boost at idle could push the reading above 619 mg/stroke. With the EGR about 78 % open the same idle should fall to about 290 mg/stroke, because exhaust gas is displacing fresh air through the meter. A reading far below those figures is a blocked air path or a tired meter. The air path includes the filter, the intercooler, a partly closed anti-shudder valve, an intake full of soot and oil from EGR and the crankcase breather, and an EGR valve stuck open. Valves that no longer seal belong on the same list.

A meter that has gone low

On an 81 kW AFN with 180 000 km, wide-open throttle still made 2000 mbar of boost and the intake was clean, and the control unit stored no air-mass fault. Specified mass at 3000 rpm, wide open, is 850 mg/stroke. The meter in the car was well below that, around 600 mg/stroke on the bench for a meter bad enough to feel flat. A new meter on the same engine brought actual mass back onto the specified trace and the response came back. Actual and specified are not identical on a healthy meter at every moment. They meet at the 3000 rpm point. An AHF, which uses the same meter, showed the same 850 mg/stroke actual and specified through a full-throttle run in first to fourth, with the gear changes at 47, 53 and 60 seconds.

Raising the output of that tired meter by about 15 %, starting from its 1 V zero, put the logged mass back to 850 mg/stroke at 3000 rpm on that car. The signal was still the shape of a worn element. It was enough for the control unit to deliver the fuel the boost was asking for. Small-signal transistors of the BC548B and BC557B family, or 2N2222 and 2N2907, are what that gain stage was built with.

A 2.0 litre petrol meter on a TDI

06A 906 461A is the 2.0 litre petrol meter. The TDI meter is 06A 906 461, without the A suffix. The petrol meter’s voltage sits higher for the same mass. The control unit then believes there is more air and allows more fuel, which is the smoke some of these swaps produce. An attenuator in the cut signal wire, between the meter and pin 52, brings that voltage back down. People who tried the swap reported both a clean result and heavy smoke, and the smoke tracks the higher voltage.

Mercedes CDI meter in the same tube

The Pierburg meter in the Mercedes 200 and 220 CDI drops into the Volkswagen Bosch tube and uses the same connector. Against a healthy TDI meter, 06A 906 461, the CDI meter reads a little lower. At 350 kg/h the curves are closer than at 200 kg/h, and the TDI control unit still calculates about 25 kg/h less, about 7 %. Readings above 200 kg/h in that comparison were taken on the road, so they are less repeatable than the bench comparison from 0 to 200 kg/h. During ordinary driving the control unit trims the difference: at 120 km/h it was observed to close the EGR by an extra 10 %, which raises the fresh-air reading back toward the value it expected and leaves less recirculated exhaust. That raises NOx. On an 81 kW TDI the difference was not obvious to drive, and it was closer than fitting the 2.0 litre petrol meter, whose high reading is what produces the smoke.

On a scope the CDI output falls all the way to 1 V at idle while EGR is active. One volt is the meter’s way of saying the fresh-air flow is zero. The waveform is stepped, which is a digital-to-analogue reconstruction: the element is digitised, trimmed, and turned back into a voltage. The steps themselves are not a problem for the control unit. The drop to 1 V is a likely reason some cars store a fault on this swap. A healthy meter is not a flat line either. Flow rises while a cylinder is drawing and falls while the valves are shut. Two meters in the same duct also do not read identically at high flow, because the sample each one takes depends on where it sits in a flow that is not uniform. Above about 250 kg/h the CDI and TDI curves may cross again.

Manifold pressure and temperature

The combined sensor sits just after the intercooler. Volkswagen 038 906 051 and Bosch 0 281 002 177 cover the 66 kW and 85 kW TDIs. Colours are from an ALH, an AGR and an AHF.

PinFunctionColourGoes to 1Signal groundBrown/blueControl-unit pin 25 2Temperature voltageGrey/greenControl-unit pin 13 35 V supply and referencePurple/redControl-unit pin 39 4Pressure voltageYellow/blackControl-unit pin 40

The temperature element is an NTC between pins 1 and 2, 2000 Ω at 25 °C. Inside the control unit a 730 Ω resistor sits between the 5 V supply and this NTC.

Coolant temperature

The coolant sensor is two NTCs. One is read continuously by the engine control unit. The other is sampled for the instrument cluster. Colours vary. The alternatives in brackets were recorded on other cars.

PinFunctionColourGoes to 1Ground, a low resistance to groundBrown/green (brown/blue)Control-unit pin 54 2GroundBrown/white (brown)Ground 3Engine-control temperatureBrown/blue (brown/yellow)Control-unit pin 70 4Instrument temperaturePurpleInstrument control-unit pin 8

A 730 Ω resistor in the engine control unit sits between 5 V and pin 70. The instruments read their sensor 40 times a second, and the voltage is high for only 4 to 6 ms on each sample, so a normal voltmeter cannot see it. A scope, or a meter with a peak-hold function, can. The instrument side uses about 230 Ω between 5 V and pin 8. Resistance is the practical check. At about 90 °C the instrument NTC is about 115 Ω and the engine NTC is about 250 Ω. The instrument curve is not reported below 30 °C. Extended back to 0 °C, the instrument NTC is about 2200 Ω. Both curves are logarithmic.

Rendered revisions

Open each version to inspect formatting, tables and photographs.

Revision 1: Volkswagen TDI air-mass measurement

Bosch meters on the TDI

Volkswagen TDIs of this generation use a five-pin Bosch hot-film meter. 0 281 002 216, Volkswagen 028 906 461, and 0 280 217 121, Volkswagen 06A 906 461, are the same meter apart from the length of the housing. The transfer curve drawn for them matches Bosch 0 280 218 018. Wire colours and control-unit pins below are from a 66 kW ALH.

PinFunctionColourGoes to
1Temperature output. Not used on the ALH.
212 V supplyYellow/blackRelay 109
3GroundBrownControl-unit pin 4
45 V referenceRed/greenControl-unit pin 50
5Output voltageBrown/blueControl-unit pin 52

The signal wire is pin 5. Colours on a particular car can differ from the ALH colours. Output voltage rises with mass flow. At idle with the EGR closed, a Bosch meter on a 1.9 reads about 2.2 V.

Pierburg meter

The Pierburg meter, Volkswagen 074 906 461, has six pins. Colours and pins below are from a 66 kW AGR and an 81 kW AHF.

PinFunctionColourGoes to
15 V referenceRed/greenControl-unit pin 50
2Signal groundBrown/blueControl-unit pin 25
312 V supplyYellow/blackRelay 109
4Not connected
5Power groundBrownGround
6Output voltagePurple/brownControl-unit pin 52

At the same idle, EGR closed, the Pierburg meter reads about 1.6 V. A control unit calibrated for the Bosch 2.2 V will under-read the air if it is given the Pierburg voltage unaltered, and the other way round. Matching one family to a control unit calibrated for the other is a gain and an offset on the signal wire: the Pierburg voltage has to be raised and shifted to look like the Bosch curve, and the Bosch voltage has to be attenuated to look like the Pierburg curve. The output of these meters will not drive a heavy load, so any network on the signal wire has to buffer it.

Milligrams per stroke and kilograms per hour

The control unit reports mass per intake stroke, not kilograms per hour. A four-stroke four-cylinder engine takes two induction strokes per revolution. At 900 rpm that is 900 × 2 × 60 = 108 000 strokes per hour. A reading of 480 mg/stroke is 108 000 × 480 × 10⁻⁶ = 52 kg/h.

480 mg/stroke is the idle reading on a 1.9 with the EGR closed. A full cylinder of 0.48 litre, at about 1290 mg of air per litre, would be 0.48 × 1290 = 619 mg/stroke if the cylinder filled completely. It does not: the valve is open for a limited time, so the real idle mass is lower. A turbo that was actually making boost at idle could push the reading above 619 mg/stroke. With the EGR about 78 % open the same idle should fall to about 290 mg/stroke, because exhaust gas is displacing fresh air through the meter. A reading far below those figures is a blocked air path or a tired meter. The air path includes the filter, the intercooler, a partly closed anti-shudder valve, an intake full of soot and oil from EGR and the crankcase breather, and an EGR valve stuck open. Valves that no longer seal belong on the same list.

A meter that has gone low

On an 81 kW AFN with 180 000 km, wide-open throttle still made 2000 mbar of boost and the intake was clean, and the control unit stored no air-mass fault. Specified mass at 3000 rpm, wide open, is 850 mg/stroke. The meter in the car was well below that, around 600 mg/stroke on the bench for a meter bad enough to feel flat. A new meter on the same engine brought actual mass back onto the specified trace and the response came back. Actual and specified are not identical on a healthy meter at every moment. They meet at the 3000 rpm point. An AHF, which uses the same meter, showed the same 850 mg/stroke actual and specified through a full-throttle run in first to fourth, with the gear changes at 47, 53 and 60 seconds.

Raising the output of that tired meter by about 15 %, starting from its 1 V zero, put the logged mass back to 850 mg/stroke at 3000 rpm on that car. The signal was still the shape of a worn element. It was enough for the control unit to deliver the fuel the boost was asking for. Small-signal transistors of the BC548B and BC557B family, or 2N2222 and 2N2907, are what that gain stage was built with.

A 2.0 litre petrol meter on a TDI

06A 906 461A is the 2.0 litre petrol meter. The TDI meter is 06A 906 461, without the A suffix. The petrol meter’s voltage sits higher for the same mass. The control unit then believes there is more air and allows more fuel, which is the smoke some of these swaps produce. An attenuator in the cut signal wire, between the meter and pin 52, brings that voltage back down. People who tried the swap reported both a clean result and heavy smoke, and the smoke tracks the higher voltage.

Mercedes CDI meter in the same tube

The Pierburg meter in the Mercedes 200 and 220 CDI drops into the Volkswagen Bosch tube and uses the same connector. Against a healthy TDI meter, 06A 906 461, the CDI meter reads a little lower. At 350 kg/h the curves are closer than at 200 kg/h, and the TDI control unit still calculates about 25 kg/h less, about 7 %. Readings above 200 kg/h in that comparison were taken on the road, so they are less repeatable than the bench comparison from 0 to 200 kg/h. During ordinary driving the control unit trims the difference: at 120 km/h it was observed to close the EGR by an extra 10 %, which raises the fresh-air reading back toward the value it expected and leaves less recirculated exhaust. That raises NOx. On an 81 kW TDI the difference was not obvious to drive, and it was closer than fitting the 2.0 litre petrol meter, whose high reading is what produces the smoke.

On a scope the CDI output falls all the way to 1 V at idle while EGR is active. One volt is the meter’s way of saying the fresh-air flow is zero. The waveform is stepped, which is a digital-to-analogue reconstruction: the element is digitised, trimmed, and turned back into a voltage. The steps themselves are not a problem for the control unit. The drop to 1 V is a likely reason some cars store a fault on this swap. A healthy meter is not a flat line either. Flow rises while a cylinder is drawing and falls while the valves are shut. Two meters in the same duct also do not read identically at high flow, because the sample each one takes depends on where it sits in a flow that is not uniform. Above about 250 kg/h the CDI and TDI curves may cross again.

Manifold pressure and temperature

The combined sensor sits just after the intercooler. Volkswagen 038 906 051 and Bosch 0 281 002 177 cover the 66 kW and 85 kW TDIs. Colours are from an ALH, an AGR and an AHF.

PinFunctionColourGoes to
1Signal groundBrown/blueControl-unit pin 25
2Temperature voltageGrey/greenControl-unit pin 13
35 V supply and referencePurple/redControl-unit pin 39
4Pressure voltageYellow/blackControl-unit pin 40

The temperature element is an NTC between pins 1 and 2, 2000 Ω at 25 °C. Inside the control unit a 730 Ω resistor sits between the 5 V supply and this NTC.

Coolant temperature

The coolant sensor is two NTCs. One is read continuously by the engine control unit. The other is sampled for the instrument cluster. Colours vary. The alternatives in brackets were recorded on other cars.

PinFunctionColourGoes to
1Ground, a low resistance to groundBrown/green (brown/blue)Control-unit pin 54
2GroundBrown/white (brown)Ground
3Engine-control temperatureBrown/blue (brown/yellow)Control-unit pin 70
4Instrument temperaturePurpleInstrument control-unit pin 8

A 730 Ω resistor in the engine control unit sits between 5 V and pin 70. The instruments read their sensor 40 times a second, and the voltage is high for only 4 to 6 ms on each sample, so a normal voltmeter cannot see it. A scope, or a meter with a peak-hold function, can. The instrument side uses about 230 Ω between 5 V and pin 8. Resistance is the practical check. At about 90 °C the instrument NTC is about 115 Ω and the engine NTC is about 250 Ω. The instrument curve is not reported below 30 °C. Extended back to 0 °C, the instrument NTC is about 2200 Ω. Both curves are logarithmic.

Revision 1: Volkswagen TDI air-mass measurement

Bosch meters on the TDI

Volkswagen TDIs of this generation use a five-pin Bosch hot-film meter. 0 281 002 216, Volkswagen 028 906 461, and 0 280 217 121, Volkswagen 06A 906 461, are the same meter apart from the length of the housing. The transfer curve drawn for them matches Bosch 0 280 218 018. Wire colours and control-unit pins below are from a 66 kW ALH.

PinFunctionColourGoes to
1Temperature output. Not used on the ALH.
212 V supplyYellow/blackRelay 109
3GroundBrownControl-unit pin 4
45 V referenceRed/greenControl-unit pin 50
5Output voltageBrown/blueControl-unit pin 52

The signal wire is pin 5. Colours on a particular car can differ from the ALH colours. Output voltage rises with mass flow. At idle with the EGR closed, a Bosch meter on a 1.9 reads about 2.2 V.

Pierburg meter

The Pierburg meter, Volkswagen 074 906 461, has six pins. Colours and pins below are from a 66 kW AGR and an 81 kW AHF.

PinFunctionColourGoes to
15 V referenceRed/greenControl-unit pin 50
2Signal groundBrown/blueControl-unit pin 25
312 V supplyYellow/blackRelay 109
4Not connected
5Power groundBrownGround
6Output voltagePurple/brownControl-unit pin 52

At the same idle, EGR closed, the Pierburg meter reads about 1.6 V. A control unit calibrated for the Bosch 2.2 V will under-read the air if it is given the Pierburg voltage unaltered, and the other way round. Matching one family to a control unit calibrated for the other is a gain and an offset on the signal wire: the Pierburg voltage has to be raised and shifted to look like the Bosch curve, and the Bosch voltage has to be attenuated to look like the Pierburg curve. The output of these meters will not drive a heavy load, so any network on the signal wire has to buffer it.

Milligrams per stroke and kilograms per hour

The control unit reports mass per intake stroke, not kilograms per hour. A four-stroke four-cylinder engine takes two induction strokes per revolution. At 900 rpm that is 900 × 2 × 60 = 108 000 strokes per hour. A reading of 480 mg/stroke is 108 000 × 480 × 10⁻⁶ = 52 kg/h.

480 mg/stroke is the idle reading on a 1.9 with the EGR closed. A full cylinder of 0.48 litre, at about 1290 mg of air per litre, would be 0.48 × 1290 = 619 mg/stroke if the cylinder filled completely. It does not: the valve is open for a limited time, so the real idle mass is lower. A turbo that was actually making boost at idle could push the reading above 619 mg/stroke. With the EGR about 78 % open the same idle should fall to about 290 mg/stroke, because exhaust gas is displacing fresh air through the meter. A reading far below those figures is a blocked air path or a tired meter. The air path includes the filter, the intercooler, a partly closed anti-shudder valve, an intake full of soot and oil from EGR and the crankcase breather, and an EGR valve stuck open. Valves that no longer seal belong on the same list.

A meter that has gone low

On an 81 kW AFN with 180 000 km, wide-open throttle still made 2000 mbar of boost and the intake was clean, and the control unit stored no air-mass fault. Specified mass at 3000 rpm, wide open, is 850 mg/stroke. The meter in the car was well below that, around 600 mg/stroke on the bench for a meter bad enough to feel flat. A new meter on the same engine brought actual mass back onto the specified trace and the response came back. Actual and specified are not identical on a healthy meter at every moment. They meet at the 3000 rpm point. An AHF, which uses the same meter, showed the same 850 mg/stroke actual and specified through a full-throttle run in first to fourth, with the gear changes at 47, 53 and 60 seconds.

Raising the output of that tired meter by about 15 %, starting from its 1 V zero, put the logged mass back to 850 mg/stroke at 3000 rpm on that car. The signal was still the shape of a worn element. It was enough for the control unit to deliver the fuel the boost was asking for. Small-signal transistors of the BC548B and BC557B family, or 2N2222 and 2N2907, are what that gain stage was built with.

A 2.0 litre petrol meter on a TDI

06A 906 461A is the 2.0 litre petrol meter. The TDI meter is 06A 906 461, without the A suffix. The petrol meter’s voltage sits higher for the same mass. The control unit then believes there is more air and allows more fuel, which is the smoke some of these swaps produce. An attenuator in the cut signal wire, between the meter and pin 52, brings that voltage back down. People who tried the swap reported both a clean result and heavy smoke, and the smoke tracks the higher voltage.

Mercedes CDI meter in the same tube

The Pierburg meter in the Mercedes 200 and 220 CDI drops into the Volkswagen Bosch tube and uses the same connector. Against a healthy TDI meter, 06A 906 461, the CDI meter reads a little lower. At 350 kg/h the curves are closer than at 200 kg/h, and the TDI control unit still calculates about 25 kg/h less, about 7 %. Readings above 200 kg/h in that comparison were taken on the road, so they are less repeatable than the bench comparison from 0 to 200 kg/h. During ordinary driving the control unit trims the difference: at 120 km/h it was observed to close the EGR by an extra 10 %, which raises the fresh-air reading back toward the value it expected and leaves less recirculated exhaust. That raises NOx. On an 81 kW TDI the difference was not obvious to drive, and it was closer than fitting the 2.0 litre petrol meter, whose high reading is what produces the smoke.

On a scope the CDI output falls all the way to 1 V at idle while EGR is active. One volt is the meter’s way of saying the fresh-air flow is zero. The waveform is stepped, which is a digital-to-analogue reconstruction: the element is digitised, trimmed, and turned back into a voltage. The steps themselves are not a problem for the control unit. The drop to 1 V is a likely reason some cars store a fault on this swap. A healthy meter is not a flat line either. Flow rises while a cylinder is drawing and falls while the valves are shut. Two meters in the same duct also do not read identically at high flow, because the sample each one takes depends on where it sits in a flow that is not uniform. Above about 250 kg/h the CDI and TDI curves may cross again.

Manifold pressure and temperature

The combined sensor sits just after the intercooler. Volkswagen 038 906 051 and Bosch 0 281 002 177 cover the 66 kW and 85 kW TDIs. Colours are from an ALH, an AGR and an AHF.

PinFunctionColourGoes to
1Signal groundBrown/blueControl-unit pin 25
2Temperature voltageGrey/greenControl-unit pin 13
35 V supply and referencePurple/redControl-unit pin 39
4Pressure voltageYellow/blackControl-unit pin 40

The temperature element is an NTC between pins 1 and 2, 2000 Ω at 25 °C. Inside the control unit a 730 Ω resistor sits between the 5 V supply and this NTC.

Coolant temperature

The coolant sensor is two NTCs. One is read continuously by the engine control unit. The other is sampled for the instrument cluster. Colours vary. The alternatives in brackets were recorded on other cars.

PinFunctionColourGoes to
1Ground, a low resistance to groundBrown/green (brown/blue)Control-unit pin 54
2GroundBrown/white (brown)Ground
3Engine-control temperatureBrown/blue (brown/yellow)Control-unit pin 70
4Instrument temperaturePurpleInstrument control-unit pin 8

A 730 Ω resistor in the engine control unit sits between 5 V and pin 70. The instruments read their sensor 40 times a second, and the voltage is high for only 4 to 6 ms on each sample, so a normal voltmeter cannot see it. A scope, or a meter with a peak-hold function, can. The instrument side uses about 230 Ω between 5 V and pin 8. Resistance is the practical check. At about 90 °C the instrument NTC is about 115 Ω and the engine NTC is about 250 Ω. The instrument curve is not reported below 30 °C. Extended back to 0 °C, the instrument NTC is about 2200 Ω. Both curves are logarithmic.