TPI MAF Sensor Degradation

I have a 1227165 MAF TPI on a 72 chevelle. It runs nice when the maf works. They last about 10 hrs and then give max gm/s at idle and make it run super rich. At idle it shows 23 gm/s instead of 5.5 – 6.2. It gives a code 33, max MAF value exceeded. I’ve gone through about 5 Cardone remanufactured MAFs. They run great , then fail.

At startup it either works right, or fails. If it starts as right, it stays right. If starts as fail, it stays fail. It does not start right then degrade to failure when driving. It never starts a failure and then becomes good. If failure, I have to start by using 1/4 throttle and slowly let up. The MAF when new works great consistently. Then over time, every so often it starts as a failure. Then the occurrence of a rich start becomes more often. After a while, it fails all the time. My wb O2 shows the mixture getting richer below 10 AFR as it slows to idle. It is so rich it barely runs. At higher throttle it goes so lean it doesn't have any power. It says 23gm/s till a higher rpm then jumps to low 40’s gm/s and varies a little. It never reaches 50. As rpm goes up, it gets leaner till past 20 AFR. Sometimes this was intermittent where it would work normal showing 5.5 - 6.2gm/s at idle and run great. Then start it again and it acts up. Now it is acting up all the time.

If I unplug a failing MAF, the car starts and runs. It’s sluggish transitioning. You can see the calculated maf values while data logging. The BLMs take a while to adjust. Sometimes there is much less power till the blm catches up. I can get home this way.

Conditions:

I have a Painless harness. I soldered the harness ground wires to a 10 gauge battery cable accessory line so it is direct to battery. The 12V terminal is soldered to the harness and attached to the 12V above the power brake booster. I have a 3-wire electronically regulated alternator and tunerpro displays 13.9V nearly all the time. At idle it will drop to 13.6 sometimes. The battery is about a year old and starts strong.

I tried 3 1227165 ECMs. The latest was a Cardone remanufactured. All work the same.

I replaced the 3 relays w automotive 30A epoxy sealed relays. I swapped out the relay sockets by soldering them in and heat shrinking the joints. Dissecting the factory oval and rectangular relays showed a sloppily built relay where sometimes the contact barely had pressure or a small gap since the relay metal frame was poorly shaped. The factory relays have a resistor in parallel w the coil. The new relays are coil only.

Using a volt meter I verified the fuel pump, maf, and burn-off relays are working properly. Also, MAF burn off works because after turning off the engine, I quickly ran over to see the wire glow red hot. I can hear the relays click each time I turn on and off the key. You can hear the maf burn off click then a second or two later click again. Failing MAFs glow for burn-off as well.

Research:

An author claims the 1227165 was superseded by 16198259, then by 99888194. He says changing to 88999194 fixed this issue. He says the 165 stresses the maf each burn-off cycle and the 194 doesn’t. (He incorrectly typed 99888194.) http://www.corvetteforum.com/forums/...ay-issues.html

Has anybody heard of this? Has anybody used a 88999194?

Reflection:

So why do these MAFs keep failing? Is my system stressing it? Is there something I can check?

If my system is fine, what can I do to substitute a more reliable maf?

I found a ford analog maf from the mid 90s. I doesn’t have the burn-off and looks like it would work w 3 wires. On the web I found a doc that shows a similar response curve as the bosch. I was thinking of this as a replacement. Downside is this requires making a custom housing to fit in the 3” air tubes.

I was also wondering if the LT1 or LS1 mafs can be tapped before the analog to frequency converter and and analog output made? These would be modern, cheap, and allow more flow. I assume much more durable.

I’m sure others have hit the same wall as me. It would be nice if someone has this figured out.

Thanks for your help,

TPI Chevelle

.............

Quote:

---

Originally Posted by billygraves

Sent you a PM.

---

$12P it is an Australian MAP/Speed density code that runs in their 808 which for all practical purposes is a 165.

I converted to an analog slot-style MAF about 8 years ago and haven't looked back. All the work has already been done for you.

No more burn-off, plus all the benefits of a modern maf. 165 ecm with supporting software and simple wiring changes required....870 ecm application still in the works. No maf translator required, no need for a frequency to voltage converter, etc.

Increased measurement range, better signal quality, increased packaging flexibility, low replacement cost, strong aftermarket support, integrated IAT, larger housings available for increased flow/horsepower gains, sensor design is robust to reversion...to name a few advantages.

Everything you need to convert is available from Blowerworks.net: sensors, housings, connectors, compatible tunes and definition files for various power ranges and applications. Feel free to contact me if desired for more info.

Thank you all for the replies. I thought I'd update since working through this. It took considerable time since there were discoveries over many months.

Discoveries:

1. A spark plug on the non O2 sensor bank was bad. The porcelain leaked and lost compression. It rocked in the spark plug base. Charge squirted onto the header. Replacing this helped.

2. The new relays were corroded in the socket. They were about a year old. I think this was the biggest contributor. These were 12V mini relays epoxy sealed. The socket was not gasketed and over a year, the terminals corroded and relaxed. The terminals look like brass but are a metal without springiness. Over time, the terminals lost grip and didn't grip the replay spades tight. Some wobbled loosely. The relays are like this: https://www.google.com/imgres?imgurl...act=mrc&uact=8


3. Vacuum leaks at the throttle shaft/housing let in false air. I have a Professional Products 58mm. The shaft is 9.8mm diameter supported by a 10mm needle bearing. The two orings in the needle bearings are slightly bigger than the shaft and leaked. I replaced these with rubber orings. The orings are a little too tight and cause the throttle return to be damped. I anticipate this will loosen up over time.

Also, the crankcase breather hose nipple leaked where it was pressed in the throttle body. I put rtv on both nipples.

Another leak was the iac. The IAC housing gasket surface doesn't line up with the Prefessional Products gasket support. The gasket can allow air to leak from the water passage side. I used to get water sucked in and bypassed the water. Since the water pressure was absent I assumed air didn't get in on it's own. It does. Also, the IAC distorts over time becomes loose. The srew portion slowly pulls out and the flat gasket surface becomes sloped like a funnel. The IAC becomes loose and leaks.

Remedies:

1. Changed to automobile relays from a 98 Cadillac. These are water tight with a gasket on the wire side of the socket and a gasket between the relay and socket. They are like this: https://www.google.com/imgres?imgurl...act=mrc&uact=8

Also, my HEI coil relay was like the others with loose terminals. Changed this over as well.


2. Changed the TPI IAC housing to an LT1/4 and bypassed the water. The LT IAC has an oring gasket and srew tabs holding it in. It is much better.


Results:

Addressing all of these deficiencies made a huge difference. The mafs weren't bad after all. Their power and burn-off voltage was intermittent. The car starts, idles, and performs like never before. I recommend these relay and IAC upgrades for all TPI applications. Both the oval and rectangular stock relays are not reliable and the IAC is leak prone.

I explored other throttle bodies by purchasing and disassembling:

Holley uses a bronze bushing to support the throttle shaft. There are no rubber seals. Although this will have low to no leak at first, it looks like it will wear like a factory unit over time creating a gap and leak.

BBK uses sealed roller bearings and a shaft oring to seal the shaft vs. inner bearing race. These leak air through the bearing between the bearing seal and inside race. I'm experimenting with a nylon bushing and oring to create a seal. The nylon bushing slides onto the throttle shaft with a tight fit sealing the shaft. The bushing rests inside the hole created for the outside bearing race. A thin oring is inserted between the bushing and housing. A washer is between the bushing and throttle linkage and keeps the oring from working out. The other side is the same, but there is no room for a washer so I need to machine the bushing thinner. I plan to replace the professional products unit with this modified BBK in a few week. The nice thing about this is the sealed bearings are much beefier than the Professional Products needle bearings and there is no wear on the throttle shaft. This will last much longer.

Another update:

One night coming home I noticed it was pitch black dark so I left the garage door closed and used the opportunity to check my spark plug wires for arcs. My wires were generic, spiral wound, 5 years old, and looked like new. With the engine running I lifted the hood under the moonless night and no lights outside made it easy to see arcs. I saw corona on the spark plug porcelain between the boot and spark plug base. I'd never seen this before. There were also a few arcs from wires to the engine. The wire arcs definitely indicated electrically worn wires. So I bought a new set.

A few weeks later replaced my wires and did the same test. No more arcs from wires and the engine ran a little better. However, I still had corona on all plugs.

I pulled the plugs and couldn't find anything wrong. I figures maybe the .045 gap was stressing the wires by requiring higher voltage to initiate the jump. I lowered the gap to .040. The car idled a little smoother and there was less intense corona. I lowered the gap to .035 the same as a 72 chevelle spec, corona gone. The car idles smooth and has more power at full throttle.

I did some research and found:

1. Racers gap at .035 or less. As compression goes up, it's harder to jump the gap so the voltage has to go higher or the gap smaller. Higher voltage stresses the ignition components so gap has to decrease. At 10:1, gap has to be no more than .035, at 11:1 .030, and 12:1 .025. Racers experiment with this and often are less than .035. I have 9.7:1.

2. Racers file their ground straps and plug base to eliminate corners. Corners have less heat dissipation and are first to glow red hot under load causing detonation. I filed the base corner, the ground strap corners and made the ground strap end spherical. I noticed a difference but I can't say it made lots of power, hard to describe.

3. I looked at several old sets of plugs and found: my Autolite irridiums have sharp cornered ground straps, I filed these above. Champions have sharp corners. AC has a bulged side on the ground strap reducing the ground strap corner.

4. A few weeks later I read an article claiming Platinum is a catalyst for the combustion process and will help the burn initiate.

So, I bought a set of AC double platinum plugs, rounded all the corners, and gaped at .035. The car starts, idles, and pulls the best with this plug setup.


Let me know if anyone else has similar results.


Happy sparking

Quote:

---

Originally Posted by tpichevelle

Another update:

One night coming home I noticed it was pitch black dark so I left the garage door closed and used the opportunity to check my spark plug wires for arcs. My wires were generic, spiral wound, 5 years old, and looked like new. With the engine running I lifted the hood under the moonless night and no lights outside made it easy to see arcs. I saw corona on the spark plug porcelain between the boot and spark plug base. I'd never seen this before. There were also a few arcs from wires to the engine. The wire arcs definitely indicated electrically worn wires. So I bought a new set.

A few weeks later replaced my wires and did the same test. No more arcs from wires and the engine ran a little better. However, I still had corona on all plugs.

I pulled the plugs and couldn't find anything wrong. I figures maybe the .045 gap was stressing the wires by requiring higher voltage to initiate the jump. I lowered the gap to .040. The car idled a little smoother and there was less intense corona. I lowered the gap to .035 the same as a 72 chevelle spec, corona gone. The car idles smooth and has more power at full throttle.

I did some research and found:

1. Racers gap at .035 or less. As compression goes up, it's harder to jump the gap so the voltage has to go higher or the gap smaller. Higher voltage stresses the ignition components so gap has to decrease. At 10:1, gap has to be no more than .035, at 11:1 .030, and 12:1 .025. Racers experiment with this and often are less than .035. I have 9.7:1.

2. Racers file their ground straps and plug base to eliminate corners. Corners have less heat dissipation and are first to glow red hot under load causing detonation. I filed the base corner, the ground strap corners and made the ground strap end spherical. I noticed a difference but I can't say it made lots of power, hard to describe.

3. I looked at several old sets of plugs and found: my Autolite irridiums have sharp cornered ground straps, I filed these above. Champions have sharp corners. AC has a bulged side on the ground strap reducing the ground strap corner.

4. A few weeks later I read an article claiming Platinum is a catalyst for the combustion process and will help the burn initiate.

So, I bought a set of AC double platinum plugs, rounded all the corners, and gaped at .035. The car starts, idles, and pulls the best with this plug setup.


Let me know if anyone else has similar results.


Happy sparking

---

I run copper ac delcos at 0.050 on my 11:1 383. I have the 24x coil per cylinder ignition setup on it though. On my 350 swapping from the 0.035" gap and stock single coil & distributor to the 0.050" gap and coil per cylinder setup gave a very noticeable power increase.