TeamSwift

I've been struggling w/ poor gas mileage in my '91 1.0 for over a year. Went from around 40 to 30 mpg. Thoroughly cleaned the EGR, it was a mess. I've tested vacuum with a simple carburetor cleaner spray test. A compression test revealed low compression on #3, surprise, surprise. So I had the head rebuilt and rebuilt the bottom end, minus the crank, this fall. My last 200 miles of driving were still 30 mpg. A visual inspection of the O2 sensor reveals that the car is definitely running rich. It also has a high idle and just barely passed smog here in CA in October. So I suspect a bad sensor somewhere or an internal vacuum leak in the throttle body.

I just picked up new IAT and coolant sensors since they were only $35 total and since the car has 200,000 miles it probably won't hurt to replace them. My question is how do I test them to make sure they're working. I was able to test the O2 sensor for resistance. But when I try testing for ohms on the IAT and coolant I get no reading at all, both on the new and old sensors. Any advice would be greatly appreciated. This problem is driving me nuts.

Thanks. - Brian

_________________
'93 Geo Metro

Okay. I'm going to start documenting my efforts to improve mileage. Hopefully it will help someone in the future, and it will let me keep track of what I've checked. Some of this info has been covered on other posts, but not all of it I think.

Last night I went at the sensors again. I guess I was tired the first time, because this time I was able to detect resistance on the coolant and IAT. I then back probed the sensors. Back probing these things is tricky because of the way the wires are sealed on the connector. Here's a pic of how I back probed the coolant sensor:



The green wire provides 5 volts from the ECU; the brown wire with the paper clip probe sends the signal back to the ECU. Using a multimeter, I attached the negative end to the battery ground and first, with the connector off the sensor, probed the green wire to make sure 5 volts present. Then reattached connector to sensor and attached the positive multimeter probe to the paper clip. I then tested first with the ignition off and then the ignition on. Coolant sensor works but I still need to test it with the engine warm.

Other sensors, valves, solenoids, etc:

Intake Air Temp Sensor: back probed, works fine
Throttle Position Sensor: .4 volts w/ throttle closed, 3.7 volts with throttle wide open. Second number should be around 4.5 volts. Need to investigate this.
Idle Speed Control solenoid: 12 volts volts from the ECU, 1.4 volts out when engine cold and no electrical load (lights, radio, etc). Need to recheck voltage with load and see whether air passes through ports.
Evaporative Emissions Control canister: Uh oh. I removed it from the firewall and vacuum checked it on the bench. Here's a pic:



The top tube brings vacuum from the throttle body, the tube on the bottom brings in fresh air, the port w/ a "2" near it brings fumes from the gas tank, and the tube to its right sends those fumes to the intake when vacuum is present. The valve would not hold vacuum. Neither Autozone nor Rockauto appear to have replacement canisters or valves. I will pull the whole thing apart tonight and investigate the valve.

More to come...

_________________
'93 Geo Metro

oxygen sensors can be tricky to test, and resistance is not an indicator of function. At your mileage i would replace the oxygen sensor before doing anything else. It should be cheap $30-50, easy to replace as well. They are wear items, although they hang in there for a long time they do wear over time and use.
I gather it is running well now, smooth with normal power? If power is down, cat could be plugged, especially after burnt valve issue.

_________________
1995 Swift w/16V 4.39s, 3tech cam, Esteem t-body, Header, needs more.
1995 Gt Mustang "Boss Shinoda" package.
1999 F150 4x4 Supercharged
1967 Mustang 428 auto, never ending expensive project
1993 Civic si h22a, fell in my lap, couldn't resist!

before you get taking EVERYTHING apart ( like i did, LOL )

make absolutely certain you have coolant flow in the passage leading to that sensor you're testing.

the lines are on the opposite side of the TB from the sensor. there is one up high on the tower and one down low on the lower half of the TB. if you can blow air thru it you are probably good. if not the block is usually at the back of the lower line, in the Tb where the passage turns up.

if you don't have flow thru this passage the cold start idle valve will hang open and the sensor will never tell the ECM to lean down after warm-up.

i cleared this up on my XFI and gained about 8 mpg.

good luck,
steve

never use a multimeter switched to ohms to check leads going to the ecu. the ecu uses cmos circuits that will absolutely explode when you put voltage to them. your multimeter probably uses a 9 volt battery, the cmos circuits are very low current and a max of 3.3 volts.

never take system sensors to chassis ground as the ecu uses a dedicated ground for those circuits. referencing the sensor circuit to chassis ground won't give you any useful data and you take a very good chance of smoking the ecu's dedicated power supply for those sensors.

the iat and coolant temp sensors are pretty generic thermal resistor (thermister) type. all you need to do is to check the resistance across the 2 pins on the sensor (disconnected) at reference points, i.e. in ice water and in boiling water. you can find the resistance readings for those points in the factory service manual. there's no magic involved, the sensors will give resistance readings within 1% or they're bad.

given the age of your car, i would point you in the direction of the fuel pressure regulator in the throttle body. i'd bet a dollar to a donut that it's full of rusty crap and isn't doing it's job properly. if it hangs up and the fuel pressure remains high, your engine will run pig rich and be hard to start.

there are write-ups on these issues in the stickies at the top of the section. good luck.

_________________
1991 Blue Geo Metro Convertible highly modified 1.0L Turbo3 5 spd. - 1991 Red Geo Metro Convertible customized 1.0L 5 spd. - 1992 White Suzuki Swift GT

My Turbo3 Project

My Cardomain Page -Ol' Blue

My YouTube Channel

My Photo Gallery

SAAB Sonett II

Thanks all for your feedback. This forum has been an invaluable resource for me since I bought the car some four years ago, even though I've rarely posted myself.

codyb76: I replaced the oxygen sensor summer of 2008, assuming it would fix the poor mpg. It didn't, so I did a compression test and rebuilt the engine. The sensor reads 0.93 volts now w/ the engine warm, indicating rich. I'll replace the sensor again once I've gone through everything else. Here's a pic of the sensor. The tip is black as charcoal.



geonewbie: I double checked the tb coolant lines last night. Produced a nice puddle on the garage floor

t3 ragtop: Thanks for info on probing sensors. For future reference, where should one ground the multimeter, if not the battery? I think I've seen all the posts related to the problems I'm having, including your helpful post on the evap canister. My workbench is currently covered w/ printouts from this forum. I'm glad to see you think the problem is in the TB. I was leaning toward a problem with the fuel pressure regulator or the air control valve in the TB. I'll probably buy a gasket kit from Rockauto and rebuild the TB in a couple of weeks. Rockauto lists several pressure regulators for the 1.0, but the pics don't look like what I see from the top of the TB. Here's the link: http://www.rockauto.com/catalog/x,carco ... ttype,6124 Can anyone confirm that these regulators will work?

Here's an update from this weekend:

MAP Sensor: Green wire is 4.98 volt reference, Green/Black is ground, Green/Red is to ECM. The G/R shows 4.83 volts at the connector, when I plug the connector to the sensor and back probe G/R w/ a paperclip it shows 3.68 volts at 0lbs vacuum and 2.28 volts at 10 lbs. I'm not sure why there's a 1 volt difference between the connector and the back probe at 0 lbs vacuum. At this point I assume the MAP is working.

Coolant: Shows under 450 ohms resistance w/ engine hot. This is w/in spec according to Chilton's.

Evap Emissions: I pulled the canister apart Friday night, and in the process broke the valve cover. Here are some pics:

This is the cap for the valve. The tube supplies vacuum. The spring and metal plate push the rubber seal against the purge hole when vacuum isn't present.

This is the rubber seal the spring rests on.

This is the hole for the purge line. The fuel vapor line is below and to the right of the purge. When vacuum is present, the rubber seal lifts up and fuel vapor is sucked into the purge line.

There are two vacuum lines going to the canister, one directly from the TB and one from a bimetal valve located on the back of the TB below the ISC valve. If I understand how this system works, the combined vacuum from these 2 lines overcomes the vacuum from the purge line and opens the rubber seal to allow the purge to draw in fuel vapors. I tested the bimetal valve (BMV) this morning and was unable to blow through the line with the engine warm. Looks like the BMV is shot but I will pull it tonight and test it on the bench with hot water.

Looks like my ISC valve might be shot too. I need to pull it also and test it.

_________________
'93 Geo Metro

I would start with that..............
http://geometroforum.com/topic/2617654/

_________________

• EcoMudder - Geo Metro Mudder
• Project page
• Build Video
• 1991 4 Door Truck Conversion
• Modified Suzuki 1.0L Engine
• High Stall 3200 RPM Converter
• Auto Trans with Welded Differential
• Custom Exhaust and Intake System
• 6" Suspension Lift & Weight Reduction
• 27" Kenda Executioner 6 Ply ATV tires

Thanks JohnnyMullet. That's a great writeup. I have to admit I haven't spent much time on GeoMetroForum. Before I tackle the ISC I'll research that site. Four years ago I replaced the heater core and with it the heater hoses at the firewall. Since I'm able to blow through the TBI heater hoses, the hoses feel warm to the touch when the engine is warm, and the coolant sensor is within spec when warm, I think I've eliminated the sensor and TB coolant as possible causes of my high idle and poor mpg.

_________________
'93 Geo Metro

Removed the BMV last night and checked it in hot water. Actually I have a spare and checked both at the same time. Both are open below 140 F and closed above it. So my BMV works. This blows my theory about the evap emissions though. When the coolant heats up it closes the BMV which causes vacuum from the TB to be applied to the charcoal canister.

EGR: Checked it this morning by pressing on the EGR w/ the engine running. The engine sputters and almost stalls. EGR is still working one year after cleaning it out.

ISC: Disconnected the hoses from the valve and blew through it with a hose. Air passes through valve. Idled engine and air still passes valve. Turned on lights, radio and fan full blast and air still passes through valve. Drove car to warm up engine and w/ light, radio and fan on air still passes through valve. Obviously valve isn't working properly. I disconnected all the ISC hoses and plugged up the TB port w/ a valve cover from an inner tube. Drove into work this morning and it is idling right around 1000 rpm. Some people choose to just cap the ISC but I don't like this solution since it masks the problem not fixes it. I'm also still showing about 0.8 volts at the O2 sensor w/ the engine warm, so it's still rich. I suspect the motor in the ISC valve is all gummed up. I'll pull it tonight and try to disassemble it. Hopefully I don't break it. Since I haven't found much info here or at Geometroforum.com on the ISC, here's some links I've found helpful:

http://geometroforum.com/topic/1051411/2/#new
http://www.aa1car.com/library/idle_spee ... _valve.htm
http://www.misterfixit.com/iacpbm1.htm
http://www.autoshop101.com/forms/h60.pdf

Edit: My mistake. I didn't realize the ISC is the same across many years and models. I had been searching just MK2/3. A search across the entire site reveals much more info.

_________________
'93 Geo Metro

Last night disassembled the valve, cleaned w/ SimpleGreen, let dry over night and reassembled this morning. Spring was not broken and valve wasn't particularly dirty. viewtopic.php?t=25522 shows a broken spring. Here are pics:

ISC valve has blue connector. My improvised inner tube cap is just below the green connector for the coolant sensor.

ISC solenoid valve.

Followed the diagnostic procedures for the ISC on this thread: viewtopic.php?f=22&t=43527&hilit=isc+valve. Solenoid shows 30ohms which is in spec. W/ 12 volts on the solenoid I can hear the motor moving but air passes through the valve w/ or w/out power.

Testing the solenoid.

At this point I'm going to leave the ISC port capped and drive for 200 miles to see whether mpg changes. Then I'll change the O2 sensor since it's cheap and drive another 200 miles.

_________________
'93 Geo Metro

Last tank got about 35mpg, so things are heading in the right direction. This weekend I replaced the O2 sensor and inflated the tires to 40psi. They were at about 22psi. Oops! Next is to run some Seafoam through the TBI and gas tank and double check the ignition timing.

_________________
'93 Geo Metro

While adjusting the ignition timing I discovered that my cam timing was retarded one tooth, because I was able to set the ignition timing to 20degrees with no knocking. Darn it. I thought there was no way I was off when I put the head back on. It really is easy to miss by one tooth, and the car still drives well just lacks a little low end power. I advanced the cam timing one tooth, clockwise, and retarded the ignition until I didn't hear any pinging (by the way the pinging is much more subtle than I expected it to be; I really have to listen closely to hear it while driving). Tip I learned about adjusting the cam timing: before you pull the belt off, use your wrench to feel for any tension on the belt. Yesterday when I pulled the belt off the crank pulley popped back, counterclockwise, and I had to start all over by aligning both pulleys with their marks. After that, I checked for no tension on the belt and then used white out to mark the cam pulley and the head. After pulling the belt I rotated the cam one tooth clockwise past the mark I had made on the head.

MPG yesterday was 30, but I've been messing with the timing so much I'm not surprised. With the timing now adjusted I'm going to leave everything along until the next refuel. I need to work on my truck anyway. Geo is sucking up way too much time.

_________________
'93 Geo Metro

I haven't been driving the Geo as much as I used to, but the last two tank refills have resulted in 35 and 31 mpg. I tested the compression, after letting the engine break in for a few hundred miles, and it was 150/160/160. Not great but I suspect the readings were a bit low because the engine had cooled some before I could do the check.

Checked out the ignition system and found that the points in the cap were well worn and two of the spark plug wires had corrosion and rust at the plug ends. I replaced the cap, rotor, wires and plugs this morning. I'll drive it for a few weeks and see whether the mpg improves.

I also tested the ignition control module (or ignitor) in the distributor. The ohms were a little above the recommended range. A new module costs the same as a distributor at Autozone, about $180. I'd rather not spend that money if I don't have to. I'm going to post the exact numbers here and at Geometroforum.com to see whether folks think I need to replace the distributor.

_________________
'93 Geo Metro

hi

Do you have any updates ?

My mileage has hung around 30mpg for the last few months. Unfortunately, I discovered that my top-end rebuild didn't go as well as I had assumed. Compression was low (about 110) across all cylinders as of last December and I was losing coolant. I pulled the valve cover a couple of weeks ago and discovered that the head bolts were loose. Retorqued them but haven't had a chance to check compression again. I hope I don't have to have the head or block decked (can't remember whether I checked the block when I honed the cylinders).

_________________
'93 Geo Metro