Transmission cooler system for all seasons

[Huskymaniac]

Anyone know if this would make a good diverter valve:

http://www.ebay.com/itm/Parker-1-4-3...7#ht_376wt_850

[Huskymaniac]

The more I think of this, the more I like injuhner's approach. It also cools the oil, a little. So long as the cooler has a high enough capacity, it will handle the extra heat exchanged from the engine oil to the ATF.

On KMD's original post, I don't think he even needs to go through the radiator cooler/heater at all. When his ATF is cool, the valve he uses goes into bypass. I think the, quite lossy, tranny will generate plenty of heat to heat up the ATF to the lock up temperature pretty quickly. What we don't want to do is cool the ATF in cold temperatures with a high capacity cooler. In cold temperatures, I'm not sure if the stock cooler is heating or cooling the ATF. But if it is heating it, it is only heating it a little. The stock cooler is located at the bottom of the radiator where the coolant must be pretty cool in cold weather. In warm temperatures, it is probably cooling the ATF, a little.

Has anyone with bypass coolers like the B&Ms or non-bypass coolers plus bypass valves like the one KMD showed here ever had problems with lockup in cold temperatures even when completely eliminating the stock radiator cooler?

[Huskymaniac]

Good deal on a reputable bypass thermostat valve:

http://www.cumminsperformanceparts.com/der13011.html

I ordered one and I am going to install it and completely eliminate the stock radiator "cooler". If I experience delayed lockup next winter, I will either put the stock radiator cooler in series or I will install the oil-to-oil cooler like injuhneer did.

[oab_au]

Originally Posted by Huskymaniac

Good deal on a reputable bypass thermostat valve:

http://www.cumminsperformanceparts.com/der13011.html

I ordered one and I am going to install it and completely eliminate the stock radiator "cooler". If I experience delayed lockup next winter, I will either put the stock radiator cooler in series or I will install the oil-to-oil cooler like injuhneer did.

As long as you remember that the oil that is flowing in the cooler circuit, is to cool and lubricate the gears and bearings in the gear box. The oil pressure that is driving it is only about 10psi, and using too many 3/8 connectors can reduce the flow to lubricate the gears adequately.
Reducing the flow, will reduce the gears survival.

Harvey.

[Huskymaniac]

Originally Posted by oab_au

As long as you remember that the oil that is flowing in the cooler circuit, is to cool and lubricate the gears and bearings in the gear box. The oil pressure that is driving it is only about 10psi, and using too many 3/8 connectors can reduce the flow to lubricate the gears adequately.
Reducing the flow, will reduce the gears survival.

Harvey.

Really? You think a couple of extra 3/8 fittings will kill the system? These thermostat bypass valves have been used by many car companies for years.

And I think Tom concluded ages ago that the stock cooler is the biggest restriction to flow. I seriously doubt this bypass valve will be more restrictive.

Finally, no one has ever proven that restrictive flow in the "cooling circuit" has been a cause of failure aside from the obviously totally blocked screens in the older models. Once that problem was eliminated it seems the main problem was either tranny materials or line pressure (including TCU control of line pressure) or both and you, yourself, have said in the past that flow in the cooling circuit has no impact on these.

These trannys do run slightly hotter than most, when running normally. Some of that may be due to insufficient cooling which may be, in turn, due to insufficient flow in the cooling circuit. But, again, it is likely that the crappy stock cooler with its crappy cooling capacity and high flow resistance would be the problem, not 3/8 fittings. And, even given all of that, the cooling is just marginally insufficient. Worst case, a good tranny will run only 10-20F above the normal range. Since that will occur only briefly, assuming one isn't constantly driving up and down mountains, regular fluid changes will guarantee that the fluid in the system is not excessively oxidized. It seems to me that very high temperatures have always been a symptom of an already failing tranny, not the cause.

[oab_au]

Yes really.

Harvey.

[sketch]

bump - pictures in first post are broken. anyway to get them fixed, OP?

[Tapani]

Let's bump this one since I've done a similar arrangement.....

A installed a Hayden 163 thermostatic bypass, which starts to close at 160F and is fully closed at 180F. Then an external plate and fin type cooler inside the front fender and a fan controlled by a programmable Marshall gauge with the sensor fitted to the pan drain plug. The rad cooler is not connected at all.

The 4EAT cooling flow is an extraction from the converter circuit - it is parallel to it, not in series down stream the converter. See the flow diagram attached. The converter circuit is an extraction from the pump output, i.e. the line feed. The cooler return is used to lube most gears and bearings and finally drained back to the pan to cool the app one gallon of fluid in there.

We know that when not locked the converter is THE major heat source in the tranny. According to the diagram the coverter flow is used to lube the input shaft bearings and then drained to the pan. Pumping losses in the pump are another heat source - this is why the line pressure is modulated as per system requirements - saving quite a bit of fuel.

The TCU temp sensor is attached to the valve body.

According to my so far rather limited trials the pan temperature is app 25-30 F lower than the TCU sensor temp when cruising (app 140F). Naturally this depends a lot on ambient conditions - the pan itself dissipates quite a bit of heat to the air flow. W/o the cooler fan running and letting heat soked engine/tranny idle the temperatures approach each other - 185-190F. The cooling circuit flow rate is probably very low on idle. With the fan running the pan temp drops to app 170 F after a few minutes with the TCU still showing 190F.

Starting with a cold tranny the TCC is engaged when the pan temperature reaches 125F. Very consistent. I believe the TCU threshold is 150F.

If I am correct the overheat mode is entered when the TCU read 214 F and the TEMP light is illuminated at 220F. To me these numbers seem very conservative. Especially concidering the fact that the TCU senses quite a higher temperature compared to the pan. Most other manufacturers report temperatures in the region of 275-285F for overheating while normal pan temperatures are in the 170-200F ball park.

The valve body receives the fluid feed from the pump, line feed modulated with various valves at different locations. The pump inlet is from the pan thru the strainer and thus the inlet temperature is the pan temperature increased by heat flux from the passages. Then the pump increases the fluid temperature depending on the line pressure request at a given moment. I would think that the temperature of the fluid fed into the valve body is quite a bit higher than the pan temp - maybe just what I see on the TCU while cruising. Naturally more heat is fed into the valve body by the fluid drain flows and leakages from above - including the converter return.

All the above written just to clear my own thoughts and to make the measurements logical.

Anyone care to comment?

I still have not been able to drive the car hard enough to see if cooler capacity is enough. Also, today it's only been 40F outside. For all I know the bypass valve might still have been closed (i.e in bypass) when cruising today and the pan temp at 140 F.

Rgds,

Tapani

-edit:
1) Some interesing discussion here: http://oldihc.wordpress.com/2012/11/...ion-cooling-2/ and here http://www.rv.net/forum/index.cfm/fu...print/true.cfm
2) I made four different versions of the drain plug in order to minimize the possibility that the pan itself is directly cooling the sensor more than the fluid. The current version has a minimum wall thickeness both on the M20 threaded surfaces (hollowed with a 13mm drill bit) and also at the end wall (5mm thickness) where the 1/8 NPT threads are. I also shortened the plug a few mm and extended the sensor with a 9mm diam piece of brass as long as I could w/o touching the valve body. I further insulated the plug head and the sensor with some pipe insulation to fight cooling effect of the airflow. All this made very little difference in the readings.

[Tapani]

Digging into the maps discussion between Phil and Mike here: http://www.subaru-svx.net/forum/show...erheat&page=15

I think the 214F over heat mode on and 220F light on I mentioned earlier are not correct.


Post #101

I think C024 is the "on above" temperature, C025 is the "off below" temperature.

On above 244 degrees (0xF4), off below 238 degress (0xEE).

I don't know much about transmissions, but it would make sense to me if the "warning light" limits matched the "overheat mode" limits: 233 degrees and 230 degrees (hardcoded at 0xE200). That way the light tells you when the trans is using the overheat shift mode.

It appears that Mike wanted to lower the values quite dramatically.

So, I think the stock numers are:

Overheat mode on 233F off 230F
Overheat light on 244F off 238F

The pan temperature seems to be quite a bit lower than that.

/T

[Tapani]

I studied the "304" Phase 2 training material and found a mistake in my logics.

The cooler flow changes quite a bit depending on whether the TCC is engaged or not.

I tried to illustrate this in the attached diagrams, which are taken from the SVX manual, fourth gear TCC engaged and released.

When the TCC is engaged there seems to be no "out" flow from the converter as such. Only leakages and a pressure relief to the pan. Also, the cooler flow is restricted - the third opening in the converter control valve is closed and the two others have orifices to restrict the flow.

When released the converter is fed from the opposite side and the "engage" channel is actually used as drain - i.e. the flow direction in this channel is reversed - and the hot oil from the converter is fed into the cooler circuit. The third unrestricted path is opened inside the coverter control valve increasing the total flow.

Any comments?

Tapani

[NiftySVX]

I'm going to have to review all of this when I get my manuals and flow charts back. The cooler circuit can actually flow quite a bit of fluid, at least 1 GPM at idle if I recall.

Of course you are right about the heat created by the pump and the unlocked converter. One of the good things about the SVX is that the engine is so powerful and that power is useable that it is difficult to imagine a situation where the converter would be churning the fluid for any length of time. If you were to floor the car from a standstill the car would be moving at 100 mph at least before you got any sufficient heat built up, and it is hard to imagine driving the car at such heavy throttle that the converter is unlocked at that speed. If I remember correctly, the stock US map will lock the conv at above 90-something MPH unless it is on the floor.

So yes the converter generates a lot of heat, but it is not like the thing is being driven with the converter stalling for more than a few seconds.

The thing that you mention about the cooler circuit outlet to the bearings and etc. is important to note because this is what killed the early transmissions, the cooler circuit clogged and these parts got starved.

Of course, there is going to be a bypass in the circuit for the TC because if it were in series with the cooler the car would not move if there was a problem in the cooler circuit. There is a ball and spring on the cooler outlet from the transmission.

Therein lies what I believe to be a serious design flaw with the unit. I think that many of the transmission problems in these cars has to do with some kind of flow problem somewhere in the cooler circuit, either there is insufficient flow at some point (possibly being directed elsewhere) or by design there is just not enough fluid to lubricate/cool the components in the front part of the case (high clt, rev clt, etc.

This may be all due to a breakdown in main line pressure, which is a known problem with early 4EATS (all L series cars, all XTs, early Legacy, VERY early SVX. They changed to a metal gasket to attempt to fix that issue. '

This is part of the reason why I am always telling people to not use a cooler. Theoretically, increasing volume in the cooler circuit should make no difference. But does it? For sure if someone were to add a cooler to any vehicle equipped with one of these transmissions that restricted the flow even a little bit they would suffer bearing and high clutch failures.

I wish that there was a way to install some kind of flow meter inside the system so that we could substantiate the theories I put here. Sadly, I don't see how that could be done.

The modifications to the 4EAT in the second generation suggest that indeed it is a pump problem as there were no phase 2 transmissions built with a final ratio lower than 4.11 (pretty sure).

Keep in mind when you talk about the fluid body, you are talking about a fluid that is designed to dissipate heat.

The next and most useful modification I would like to see is an oversized fluid pan. If we were able to increase the fluid system capacity even a bit it would help temperature stabilization much more effectively than any cooler.

[Tapani]

Nate, thanks for your comments.

The only situation the tranny is heavily heat loaded is on a race track . Two fast laps and I got into the overheat mode - every time.

So, I needed to increase the heat rejection capacity - and to understand how the cooling system really works. Also, the tranny was over cooled at ambient temperatures below 15F. The TCC was engaging and releasing every three miles.

There is lubrication flow which is extracted directly from the converter regulator valve - same source which feeds the TCC lock up control valve. It is quite difficult to say from the flow chart what exactly this is used to lube in the front part.

The main lube flow is from the cooling circuit, this is used to feed the input shaft internal cavity too.

Just out of curiosity I took some steady state temperature data yesterday. App 70 mile drive one way. I drove long enough so that the temperatures were stabilized.

Remember that I have a thermostatic bypass valve (starts to open at 160F) and an external cooler only - fan kicks in at 172F pan temp for now.

Run one:

32F ambient, 65 mph, 4th locked, pan 130F, tcu 159F
37F ambient, 65 mph, 3rd unlocked, pan 156F, tcu 175F
37F ambient, 50 mph, 4th locked, pan 140F, tcu 164F

Run 2:

48F ambient, 85 mph, 4th locked, pan 155F, tcu 172F
48F ambient, city, drive, pan 171F, tcu 180F

After 1,5 hours stand still, pan 122F, tcu 141F

Can't wait to get on a race track and the weather to warm up :-).

The cooler is inside the LH front fender, it clearly need more airflow. The fender is warm to the touch (120F) after city traffic.

I think the lubrication circuit inlet pressure (cooler outlet) would be a relative indication of the flow too - affected by many factors, surely - but an indication never the less. Not impossible to arrange if desired.

And yes, the phase 2 type pump would be a better choice !

A direct control phase 2 VTD tranny with 4.11 gears and a stand alone controller with graphic user interface would be a nice project. Should we do it?

/T

[Tapani]

Update,

We had a track day yesterday.... mild weather, very gentle for all. Only + 15C and cloudy.

The extra cooler certainly improved the balance, but did not solve the whole issue. Fast laps make the pan temp creep up continously.... After 4-6 fast laps the pan temp was 240 F, the max I saw was app 242F. Not sky high, but high enough for me to pull and chill out.

The high temp mode was never activated and light was never on. I wonder what the thresholds truly are?

I had the original rad bottom cooler and the aux cooler in series. I will relocate the aux cooler to the front of the car - I do not think it gets enough air flow inside the fender although I have large holes in the liner and a sort of a scoop at the bottom.

Any comments?

Other than this the SVX ran well, no issues. The head gaskets seem to hold :-).

/T

[processengr]

Originally Posted by Tapani

Update,

We had a track day yesterday.... mild weather, very gentle for all. Only + 15C and cloudy.

The extra cooler certainly improved the balance, but did not solve the whole issue. Fast laps make the pan temp creep up continously.... After 4-6 fast laps the pan temp was 240 F, the max I saw was app 242F. Not sky high, but high enough for me to pull and chill out.

The high temp mode was never activated and light was never on. I wonder what the thresholds truly are?

I had the original rad bottom cooler and the aux cooler in series. I will relocate the aux cooler to the front of the car - I do not think it gets enough air flow inside the fender although I have large holes in the liner and a sort of a scoop at the bottom.

Any comments?

Other than this the SVX ran well, no issues. The head gaskets seem to hold :-).

/T

I think the stock radiator cooler is series is a problem, restrcting flow, and adding heat when you need cooling. I have mine in tthe fender well also, but with a thermostatically controlled cooling fan. Monitoring temp from a laptop to factory sensor, I hit about 212F going up steep long hill ("Mile Hill Rd. in Newtown, CT) on a summer day.