mfi thermostat hacks?

[jkeyzer]

I have come to believe that the MFI thermostat is causing some of my troubles in getting my 2.4T MFI system set up. Recently I have been having a situation occur where even after running the car for 30+ mins, it will intermittently idle very rich, but then if I lean the idle out at other times it will be far too lean. I am starting to think that the warmup thermostat is not consistently and reliably engaging to the same spot, giving me some uncertainty in my overal mixture settings.

I get a small amount of hot air blowing through the thermostat, and all the hoses look ok although I do have a torn driver's side heat exchanger to flapper box hose. I have wondered if maybe there isn't enough backpressure on the heat exchanger air to get good airflow all the way to the thermostat.

In any case, removing the inspection cover and pushing on the control stub inside shows that there is still some movement left. The control stub is not pushed all the way to its stop by the thermostat, even when the car is warm. I haven't tried the propane torch/hairdryer test yet.

Has anyone tried modifying their thermostat to give it some kind of adjustability or to force it into a certain position for testing? I remember reading about threading a screw into it in the MFI articles and am curious if anyone has actually tried this. It would be great if I could get it to engage fully so that I can have more confidence that the mixture settings I am working on will "stay put". Beyond this maybe I could get rid of the flakey behavior by pushing it in further for a given temperature.

Any ideas welcome... I have a flapper box hose on order but I doubt that will fix the problem. (it will keep my toes warmer)

[Rot 911]

I'm not sure if I am an expert, but I took the MFI on my 72 E from totally screwed up to mostly in spec. There is not much air that runs thru the hose from the heat exchanger to the thermostat housing. If you can feel warm air on the exhaust side of the thermostat housing then it is getting enough air. I would first try cleaning the disks inside the housing and the housing itself to make sure the disks are not sticking. Complete directions can be found here: http://www.pelicanparts.com/techarticles/911_MFI/TipMFI.htm#MFI%20thermostat%20R%20&%20R

[jkeyzer]

Kurt -

Thanks for the tip but that is the first thing I did in my extensive MFI troubleshooting. I have been through just about every piece of the system short of having the throttle bodies or pump rebuilt!

It is possible I don't have a gasket in the right place as there were no gaskets on my thermostat when I took it apart. I added one between the body of the thermostat and the pump cut out of sheet gasket material. After reading the literature it seemed like there should be one there.

Maybe I've got it wrong?

Aside from this problem it is running fairly well. I've got things synchronized pretty well by the book. I have a little bit of hesitation at part throttle once in warms up but if you really stomp on it there is a lot of power there. I'm also getting intermittent backfiring through the exhaust but I have an exhaust leak somewhere.

[jkeyzer]

I'm curious now - anyone know where there are/are not supposed to be gaskets on the thermostat?

Also are people seeing the thermostat pushing the lever inside the pump ALL THE WAY back to its stop when the car is warm?

I am not seeing this consistently and I believe this is the cause of my woes - a problem that has gotten a LOT worse recently and I am not sure why!

[Early_S_Man]

Guys,

There is supposed to be an 'Insulating Piece,' pn 901.110.925.00, between the thermostat housing and the body of the injection pump. There is a replaceable bushing for the themostat shaft rod, and wear of the bushing, or on the rod could cause inconsistent results of the injection pump mixture.

I suggest conducting some measurements and tests to establish baseline characteristics for the thermostat assembly:

Required tools:

1. Digital Multimeter with temperature capability and Type-K Bead Thermocouple Probe for fast measurement response time (Sears sells a DMM with included bead probe for $59.95 ... another DMM model with Type-K temp. capability and serial port and PC cable, and data logging software, but no probe is $99.95, separate bead probe -- $14.95)

2. Dial caliper with depth-measuring capability or dial indicator and holder

3. Heat Shrink Gun or Hair Dryer ... with adapted hose assembly to fit snugly on thermostat housing ... locally fabricated

Tests:

1. On fully warmed-up engine with oil temp. reading of 180° F, measure the temperature of the bimetallic disks by direct contact of a bead probe into the hot air exhaust of the thermostat assembly.


2. Remove the thermostat assembly from the injection pump, and cover the pump where the thermostat mounts with duct tape.

3. Allow thermostat assembly to cool to room temperature.

4. Draw three tables with columns for temperature and shaft extension ... labeled Test 1, 2, 3.

5. Measure bimetallic disk springs' temperature to establish room temperature baseline within 5° F. Record Room temperature reading at disk 'stack' for Test 1.

With dial caliper depth-measuerement probe, measure shaft extension past base of thermostat housing, and record the extension reading at room temperature row of Test 1.

6. Connect adapter hose from hot air source and turn on to lowest heat setting -- so the heat rise is slow enough to measure at 25° F intervals ... extend hose from heat source if temp rise is too quick!

7. At approximate 25° F intervals, measure temperature and shaft rod extension distance and record on chart for Test 1. Repeat measurements until disk assembly temperature exceeds measured maximum temp on engine after oil was warmed up.

Repeat Tests 2 and 3, then enter data into spreadsheet, and plot X-Y graph with all 3 sets of data. The best results that could be expected would be a perfectly linear curve, and extremely close correlation between all 3 sets of teat data! Do linear regression on the data with temperature as the independent variable, and extension distance as the dependent variable ... then compare the regression data to test data curves.

[Roland Kunz]

Hello

I get a small amount of hot air blowing through the thermostat, and all the hoses look ok although I do have a torn driver's side heat exchanger to flapper box hose. I have wondered if maybe there isn't enough backpressure on the heat exchanger air to get good airflow all the way to the thermostat.

Yes thats a comon problem even with new exchangers and doubled ( isolatet ) pipes runing up.
Only the "Karl Wagner fix" for it seemed to work clever enough. Basically you use the alternatorfanblade as vac and run a pipe to the lower end from the thermostate.

I also tried it with a electric fan in the tube but only found a small one whos blades meltet away. Maybe I will stumble over a other one.

Grüsse

[jkeyzer]

Warren -

If the extension of the rod isn't enough, is there a way to shim the thermostat to push it out further for a given temperature I wonder?

Roland -

So a hose was connected from the exhaust side of the thermostat to somewhere around the fan? How was the fan or fan shroud modified to make this possible?

[jkeyzer]

On a related subject, has anyone on this BBS toyed with the idea of replacing the thermostat with a solenoid or linear actuator? I could definitely handle a switch on the dash for warmup enrichment. One step further would be an engine temp sender and a servo but first things first.

If you have thought about this or actually tried something, or maybe heard of it being done before, I want to hear about it!

Also, how difficult to push back should the lever inside the pump be? When I am poking the lever with a screwdriver (through the access port on top of the pump) to check movement it does take some effort to push it all the way back to the stop. I assume if the lever became very difficult to push in because of a gummed up or worn pump, this would cause the thermostat to have trouble doing its job.

[Roland Kunz]

Hello

Well it is hard to descripe "Karl Wagner" device witout picture but mainly the heatexchangerside keeps stock. You have a exhaustestud on the thermostate showing downwards.
There you slip over a hose and run it to the front from the alternator fan right there is a tube runing around 180 showing inward and also pressed to a flat retangular opening. This end is moved to the other edge from the fan blade and clears that by 0,1 to 0,2 mm ( Closer is better )
The tube as a little base showing to the left and screwed under the stud from the coil. Next Time I´m there i will try to get ahold off one or show pictures. The Wagner fix was also used from the factory ( When he inventet it he was one from the factory mechanics, troubleshooting was there buisness ) on some cars who only had been used on short trips

A other solution is the "Kadach" way. he just removes the complete hose to the exchanger and places a 10 Watt lightbulb from the underside and closes the top with a rubberhat.
The bulb is runing via ignition. This was mostly used on raceengines with headers and with changing the lightbulb or using a resistor it could be finetuned to suit the engine.

Kadach was also a factory mechanic and worked mainly on race engines so he had a other problem and a other solution for that.

My fix was to get a cheap fan and place it inline. My first test was on my 2,7 carrera engine using a small 1" Papst device. But they didn´t lived to long due the heat or due the overreving if theflow works on full rush.
Some might have seen it and sales my idear as E-Ram

The problem is that your heatexcangers will flow nearly nothing at idle and even with higher revs it will not flow enough heat. Heat and flow eill only come with load on the engine.

In the 70´s there wasn´t that much traffic and stop & go like nowerdays so you normaly have much more idletime heating up the engine slow but not the Thermostate witch leads to rich runing ( more wear and thined oil plus lower MPG ) and a very bad throttle response.

From my view the Wagnerfix is the best way as it will not burn up/trough and always runs as long the engine runs. It has a little flaw spot as it needs still good exchangers as it just helps draw that heat trough the thermostatecase. If your exchanger is bad it will not work like it could.

Adusting can be made a bit by changing the distance to the fan but normaly it is very accurate to work on normal MFI engines doesn´t matter if it idles or you rev it trough it always seem to work exactly on the line.

Grüsse

[jkeyzer]

Roland -

Your post is very illuminating - thank you very much for the reply! I think I understand the "Karl Wagner" device and also your idea of putting a fan inline. Perhaps I will try to implement one of these in the near future, I would like to experiment a little bit.

Something like a computer CPU fan in a hose on the output of the thermostat might work but might also melt as you say. It would be easy to try. If it melts a fan with aluminum blades and construction might do the trick.

Replacing the thermostat with a solenoid is still tempting. The light bulb idea is primitive but very cute!

I do frequently drive short trips so I notice the ill behavior of the thermostat often - usually the idle will drop significantly due to over rich condition. For some reason now even on longer trips I am getting the same behavior. Not sure why.

Thanks again.

[jkeyzer]

Bump!

Warren -

I haven't used a thermocouple yet, but I measure 0.583" extension cold and 0.7" after being warmed gently with a torch for a minute or so.

I am not sure what movement is within spec. I will search through the MFI notes but I don't remember seeing a number.