When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
Tech / General EngineIs your car making a strange sound or won't start? Thinking of adding power with a new combination? Need other technical information or engine specific advice? Don't see another board for your problem? Post it here!
Well just picked up a Cardone maf. Now reading 5grams.
Good news, Now the diagnosis can begin. I still suspect the original MAF may be just fine, and should be stored as a good,working spare. You now know how difficult it can be to find one that works.
Cool!
Does the Cardone do the burn off cycle?
Asking because if it works, I'll considerer this one for replacement
Thanks
Depending upon which Cardone MAF was used, it could be a reman with new platinum wire and cleaning/reflowing solder, or could be a new unit with a TFT sensor in place of the hot wire. A reman should work exactly like the original. The TFT version would require some logic to "fool" the ECM into reading a 5+V signal when the burn-off cycle is initiated. The old Wells TFT units did that with a switching transistor on the "D" terminal and 5VDC regulator signaling the "C" terminal so the ECM thought the hot wire was at full current for the required duration. The scheme could vary, but I suspect all of the good ANALOG output TFT replacement MAF units do something similar.
If the air inlet uses the plastic resonator box between the TB and MAF sensor, the original straight convoluted rubber duct/adapter is GM10037612. If the original installation is the duct with an integral 90° bend (no resonator box) the duct/adapter is GM 10055895, and that could be a little tougher to scrounge up.
I think SbF is correct, especially with the Bosch logo and symbol so evident. That should be a good thing, since it will act just like an original (which it probably is).
I'm having the same high reading from my new Delphi MAF. I contacted Delphi and this was the response,
********
The AF10320 should be accurate down to 3.5 g/s but it appears we are having an issue with this part at low flows, you are now the second complaint on this. Thank you for bringing the issue to our attention. Our quality and engineering team are currently testing to find out where the problem is.
Thank you,
Delphi Technologies North America Light Duty Aftermarket
I'm having the same high reading from my new Delphi MAF. I contacted Delphi and this was the response,
********
The AF10320 should be accurate down to 3.5 g/s but it appears we are having an issue with this part at low flows, you are now the second complaint on this. Thank you for bringing the issue to our attention. Our quality and engineering team are currently testing to find out where the problem is.
Thank you,
Delphi Technologies North America Light Duty Aftermarket
A Brand of BorgWarner
**********************
Great! I just sent the following. So we are 3 complainants now.
I have installed a MAF Delphi #AF10320 on a completely original Camaro IROC-Z 1989 with the 305cid TPI engine. The AF10320 was obtained through RockAuto.com a reputable distributor. The original Bosch MAF has a reading of 40mv with zero airflow but the AF10320 has a reading of 800mv with zero airflow. This creates a problem at idle resulting in a very rich Air Fuel Ratio (AFR) condition. I have shared this information on Thirdgen.org web site where others have had the same issue. Delphi has replied the following to one of them:
"The AF10320 should be accurate down to 3.5 g/s but it appears we are having an issue with this part at low flows, you are now the second complaint on this. Thank you for bringing the issue to our attention. Our quality and engineering team are currently testing to find out where the problem is.
Thank you,
Delphi Technologies North America Light Duty Aftermarket"
With the the original Bosch MAF, normal warm idle at 750 rpm on my Camaro was recorded at 550mv, which corresponds to 5.80g/sec. Since the AF10320 cannot go under 800mv, that creates a higher reading for the Electronic Control Module (ECM). Consequence is an input of 6.43g/sec instead of 5.80g/sec which translates to 17% more fuel. At slightly higher rpm (800-1000rpm) the extra fuel gets over 25%. The ECM is incapable of compensating because it is out of range. When the engine is first started cold with the ECM in open loop operation, there is no fuel trim correction which leaves the engine running over rich, fouling the spark plugs with rough idle.
If you need more information, I will be happy to share it with you.
Last edited by SbFormula; Mar 22, 2023 at 10:09 AM.
So whats the conclusion on MAFs for our cars.
What brand can be bought and get correct readings?
Is BSE holding up and can deliver correct values or is it a hit or miss?
BSE 57854 (NEW) is that a Bosch unit
BSE 27854 (Reman) is that a fully working unit for 1989 ?
I need to order a new Mass Air Flow Sensor as well. Wondering the same thing. Should I buy the BSA New or the BSE Reman Unit ?
Any feedback would be great.
For an 86 Camaro Iroc 5.0 TPI
Last edited by pete2833; Mar 6, 2026 at 09:51 AM.
Reason: Update
I need to order a new Mass Air Flow Sensor as well. Wondering the same thing. Should I buy the BSA New or the BSE Reman Unit ?
Any feedback would be great.
For an 86 Camaro Iroc 5.0 TPI
I got a Remanufactured, BSE #27854 in 2023 for an '89 IROC-Z LB9 5 speed. It still works but low milage per year car.
Peace
I don't believe Delphi makes a MAF for V6 '86-'89. But I could always be wrong
Delphi makes one for V8 '85 #AF10328
Delphi might not actually MAKE any of these, but they DO list them as parts that they sell.
You are correct that the Delphi AF10328 is for the early 1985 F/M (not analog voltage) TPI installations. The OEM number for that was 14081249. That MAF has a design output range from 20-500 Hz (not KHz) and the no-flow number should be very close to 20 Hz.
Delphi also offers the AF10320 for the 1985-½ though 1989 analog TPI systems. The OEM number for that was 14094712. That MAF has a design output range from a theoretical 0.2 -4.5 VDC
That only indicates what the sensors were designed to do, but tells us nothing about their capacity to do that effectively or correctly, as many reports seem to indicate. Understanding HOW the hot surface mass air flow sensor functions should reveal that a zero output is a practical impossibility unless it is being calibrated and tested in a temperature-controlled oven that would melt the housing. I can describe that in greater detail if anyone wants to hear an old man rant later.
Delphi might not actually MAKE any of these, but they DO list them as parts that they sell.
You are correct that the Delphi AF10328 is for the early 1985 F/M (not analog voltage) TPI installations. The OEM number for that was 14081249. That MAF has a design output range from 20-500 Hz (not KHz) and the no-flow number should be very close to 20 Hz.
Delphi also offers the AF10320 for the 1985-½ though 1989 analog TPI systems. The OEM number for that was 14094712. That MAF has a design output range from a theoretical 0.2 -4.5 VDC
That only indicates what the sensors were designed to do, but tells us nothing about their capacity to do that effectively or correctly, as many reports seem to indicate. Understanding HOW the hot surface mass air flow sensor functions should reveal that a zero output is a practical impossibility unless it is being calibrated and tested in a temperature-controlled oven that would melt the housing. I can describe that in greater detail if anyone wants to hear an old man rant later.
I will listen to anything that @Vader has to say ! You are a wealth of knowledge. I’m a very good mechanic by experience. When I was in high school I was a gear head and always taking cars apart. When I had my 86 Iroc at 18 years old I wasn’t as knowledgeable as now and there was no internet in 1993 !!
I ordered a BSE Mass Air Flow Sensor from Rock Auto. Part # 57854
i updated my prom in the ECM. Was worried I had the infamous AUM prom. Turns out the ECM in the car is an AC Delco reman unit. It looks to have the original Prom in it.
It has an AMU 8958 prom with a date code of 8546. I was having all the symptoms of the AUM Prom. Really early lock up of TQ convertor and not running well. It runs much better now but still have Code 34 and a rough idle. I changed spark plugs today and cap and rotor is next.
New Mass Air Flow will be here Wednesday. Hope it cures the issues.
Our membership has effectively documented the recent problems with some new and some remanufactured MAFs, as well as the ongoing issue of parts databases showing the 1985 MAF (TPI verson) being the same part numbers as the 1986-up MAFs.
As many know, the typical Hitachi-style (patent 4,369,656) thermal MAF has a thermal sensor in contact with a heated element, and when in operation system power heats the element to a predetermined temperature. The current required to achieve the design operating temperature establishes a "zero" point in many meters. In some designs, the element temperature is also compared to an ambient temperature sensor as verification and precision. As air flows past the heated element heat is extracted, and the system will increase current to maintain its design operating temperature. The level of this heating current is directly proportional to the heat being removed, which is also proportional to the mass (*- not volume) of air moving through the unit.
* - This important distinction means that the sensor accommodates differences in barometric pressure and altitude, humidity, and any other variations. Vane air sensors like those used in earlier systems from Bosch, Nippondenso, and Ford could not inherently compensate to altitude, barometer, moisture, etc.
As the element heating current varies, the internal circuitry samples and maintains the temperature and also conditions that current signal level via a scalar to generate an output voltage, or via an analog-digital converter to produce a digital or frequency-modulated output at a fixed P-P voltage. The digital scheme eliminates potential problems from resistance, poor connections, low system voltage, or other conditions which could affect the output to the powertrain control computer. Thus, regardless of the communication system, most MAFs, rely upon thermal transfer/heat dissipation to measure the mass of air moving through.
Anyone reading this has probably already realized that the cleanliness of the sensing elements is critical to accurate measurement, as any contamination becomes an insulator. This is why a hot-surface sensor is subjected to a burn-off cycle periodically to shed any contamination which might accumulate on the element. There are sealed design units using a thick-film semiconductor (Stoltman-Kabasin patent 4,445,369) which are more immune to contamination, but still not completely. Cleaning is a requirement, eventually.
You may also have realized that despite the incredible speed of electrons flowing in conductors, turning on the sensor to establish a "zero" point can take a few milliseconds, and there are also delays in the powertrain control computer in recognizing this. This is yet one more reason to turn on the ignition and wait briefly before operating the starter.
As this relates to recent problems, it should be easy to understand that it is critical to accurately measure and control the element temperature and heating current, and perhaps even more critical to develop the signal conditioning to recognize the "base" levels and scale the outputs accordingly. A little bit of carelessness in placing components, isolating elements from other heat-leaching surfaces, and selecting a base operating temperature that is adequate to account for seasonal temperature variation at any lattitude would result in an incorrect signal level. I can just imagine that the A-I and "team engineering" (a.k.a., LCD engineering) would perform all the R&D in a design lab at 72°F and 35% R.H., and completely overlook the fact that the design would need to work correctly on top of a 9,800' hill in Saskatchewan in January, as well as 280' below sea level in Death Valley in July. GM managed to do that themselves, and in later working with Bosch and Hitachi.
Whether Malaysia and Taiwan managed to go that far, no on knows, but the evidence is that some of the recent offerings may not have closely controlled that to avoid 800mV signals with no air flowing.
Our membership has effectively documented the recent problems with some new and some remanufactured MAFs, as well as the ongoing issue of parts databases showing the 1985 MAF (TPI verson) being the same part numbers as the 1986-up MAFs.
As many know, the typical Hitachi-style (patent 4,369,656) thermal MAF has a thermal sensor in contact with a heated element, and when in operation system power heats the element to a predetermined temperature. The current required to achieve the design operating temperature establishes a "zero" point in many meters. In some designs, the element temperature is also compared to an ambient temperature sensor as verification and precision. As air flows past the heated element heat is extracted, and the system will increase current to maintain its design operating temperature. The level of this heating current is directly proportional to the heat being removed, which is also proportional to the mass (*- not volume) of air moving through the unit.
* - This important distinction means that the sensor accommodates differences in barometric pressure and altitude, humidity, and any other variations. Vane air sensors like those used in earlier systems from Bosch, Nippondenso, and Ford could not inherently compensate to altitude, barometer, moisture, etc.
As the element heating current varies, the internal circuitry samples and maintains the temperature and also conditions that current signal level via a scalar to generate an output voltage, or via an analog-digital converter to produce a digital or frequency-modulated output at a fixed P-P voltage. The digital scheme eliminates potential problems from resistance, poor connections, low system voltage, or other conditions which could affect the output to the powertrain control computer. Thus, regardless of the communication system, most MAFs, rely upon thermal transfer/heat dissipation to measure the mass of air moving through.
Anyone reading this has probably already realized that the cleanliness of the sensing elements is critical to accurate measurement, as any contamination becomes an insulator. This is why a hot-surface sensor is subjected to a burn-off cycle periodically to shed any contamination which might accumulate on the element. There are sealed design units using a thick-film semiconductor (Stoltman-Kabasin patent 4,445,369) which are more immune to contamination, but still not completely. Cleaning is a requirement, eventually.
You may also have realized that despite the incredible speed of electrons flowing in conductors, turning on the sensor to establish a "zero" point can take a few milliseconds, and there are also delays in the powertrain control computer in recognizing this. This is yet one more reason to turn on the ignition and wait briefly before operating the starter.
As this relates to recent problems, it should be easy to understand that it is critical to accurately measure and control the element temperature and heating current, and perhaps even more critical to develop the signal conditioning to recognize the "base" levels and scale the outputs accordingly. A little bit of carelessness in placing components, isolating elements from other heat-leaching surfaces, and selecting a base operating temperature that is adequate to account for seasonal temperature variation at any lattitude would result in an incorrect signal level. I can just imagine that the A-I and "team engineering" (a.k.a., LCD engineering) would perform all the R&D in a design lab at 72°F and 35% R.H., and completely overlook the fact that the design would need to work correctly on top of a 9,800' hill in Saskatchewan in January, as well as 280' below sea level in Death Valley in July. GM managed to do that themselves, and in later working with Bosch and Hitachi.
Whether Malaysia and Taiwan managed to go that far, no on knows, but the evidence is that some of the recent offerings may not have closely controlled that to avoid 800mV signals with no air flowing.
That is some great insight @Vader . I will have to see how the new BSE works. BSE also has a reman unit. Rock Auto carries that. On Rock Auto both the BSE new and Reman Unit show a picture of the Bosh Unit.
I WinAldl and can look at the engine while running. What should I look for to know if the Mass Air Flow Sensor is working properly ?
A very revealing indication could be the MAF signal with KOEO (zero air flow).
My suspicion is that a factory Bosch core that is properly re-wired with fresh platinum is most likely a safe bet, but only if the reman includes testing upon completion.
I attempted to re-wire an OEM MAF years ago, mainly to see just how difficult it could be. It was only partially successful, and although it would produce an output signal it was consistently low and not as linear as a correct unit would produce. I had struggles sourcing lengths of the proper alloy platinum wire in 41-gauge, and ended up with a couple feet of several various wires that were just not quite right. The ranges of 6.84 - 7.66 Ω per foot in alloys from NiPt, PtCr, Alloy 60, and a few others were just not quite right, but I got to learn what it would take to re-string a sensor and reassemble for testing. It was not difficult, just tedious. I don't imagine that anyone would be doing more than 4-5 of these in any normal work day.
From: Franklin, KY near Beech Bend Raceway, Corvette Plant and Museum.
Car: 1992 Pontiac Firebird
Engine: 5.0L L03 TBI
Transmission: 700R4
Axle/Gears: 2.73
Re: Maf reading
Howell EFI has used OEM MAFs with a money-back guarantee. I thought 1985 MAFs were different from 1986 to 1989 MAFs. I'm not into the pre 1987 cars though.
Used with production TPI systems from 1985-89. Recommended for replacement on stock systems or for any engine with non-standard displacement, or low-idle vacuum camshaft. Used units sold with a money-back guarantee. $150.00
Well the new BSE did not work.
car will start and idle. But the second you touch the throttle, car dies. Can not get it to run. So back to trying to figure out what is going on with the Code 34.
