linuxgod

At the beginning of this thread you mentioned you were reading the temps through the OBD2 port. What did you use to do that? Do you happen to know the PIDs that are used to get the temps?

Mkarl

Initially used this. IMHO these units are the original HDS (from somewhere) not clones. The amount of electronics in these and the complexity of the boards is just so ridiculous and complex that cloning them would be very unlikely. There are however, small clone units out there that also run the same Honda Diagnostic software.

Honda HDS.html

webbch

I know this is a bit late to the game, but just an FYI for attaching thermocouples....You don't have to attach them in an invasive manner and install them in the fluid flow. In live in AZ, and after boiling the trans fluid in my 2003 Honda Oddyssey, have become pro-active in monitoring trans temperatures and adding external coolers based on my results. I have attached TC's to several of my vehicles for monitoring purposes using the following method:
1) Find a metal section of the transmission fluid line (conducts heat very well)
2) Using a bead-style thermocouple, put a little thermal mastic (or CPU grease) on the end of it to ensure heat transfer between the line and the thermocouple even if they're not totally mashed together.
3) Secure the end of the thermocouple to the metal line by wrapping it with aluminum foil tape (the RIGHT kind of tape for wrapping HVAC duct work). The foil tape will hold up to the temperatures encountered on both the inlet and return lines without a problem.

You should easily be within a degree or 2 of accuracy of the actual fluid temperature using this method - well within the objectives of monitoring trans temperatures IMO.

Mkarl

If you want a general reading of ATF temperature this I fine. The metal pipe you use has to VERY short and not directly attatched to any part of the transmission or any other metal object in the engine bay - IE it has to isolated by rubber ATF hose - if this is not done, the readings will be incorrect at highway speeds and cold temperatures.

I have done this for cooler return temperatures and for the most part it is very close.

However what you can not do is monitor the almost instantaneous short duration spikes in TC outlet temperatures - sometimes up to 20 deg.

I have charted the difference, unless you correctly place the tip of a probe into the ATF stream right at the TC outlet, you will be oblivious to the marked difference in ATF temperature tracking, this is especially useful for monitoring for low idle TC flow faults (about 3 seconds duration) in earlier transmissions and for monitoring rapidly changing clutch pack temperature spikes.

This is likely of no interest to most owners, however those of us who are interested, for whatever reason, will still need to measure the fluid directly at the outlet port ATF stream.

webbch

On my 2002 Volvo V70, there is a metal section of tubing (part of the trans line) directly attached to the transmission that worked great for attaching TC's to. I monitored the trans exit and return temps prior to installing an external cooler and found them to hit upwards of 230 deg F on the exit line and 215 on the return line on a reasonably warm day of 105 deg F, which is by no means record breaking.

Quite honestly, the temperature of true concern is the sump temperature - that's what the transmission components we're worried about are subjected to - and is generally somewhere in between the exit and return temps. IMO, a return temperature of greater than 220 deg F for an extended period of time is cause for concern in my book.

Granted, I haven't taken a look at the trans lines on my 2006 TL yet (just got it 2 days ago). If I can't find a way to monitor the trans temps from something like a scangauge II or UltraGauge, I'll likely install some TC's in a similar manner on the Acura and see if I need to add an external cooler on this one as well.

webbch

...The thing I like about the foil tape method of installing TC's is that it gives you the info you need without having to cut into the lines. Running the wires through the firewall is normally the biggest nuisance of the job.

Mkarl

Just so were on the same page, TC in my post = Torque convertor. I've gone over sump / pan temperatures in great detail before, the Honda OEM PAN (bulk ATF) temperature can be at showing 190 deg while you TC outlet it running at 250 deg - you will be slowly killing the TC or clutch packs if you go by sump temps. Do we encounter this in normal driving - no, can we encounter this in steep mountain driving or spirited driving - absolutely. Knowing when to back off or select a lower gear and see an instant change in TC outlet temps (a reflection of either Torque convertor or clutch pack spiking depending on what your doing) and avert the problem is important for the conditions I drive under and the age of the vehicle and its problematic 5AT.

I currently monitor Torque convertor outlet, Trans Cooler outlet and the OEM pan temperature sensor - they are monitored by a micro computer and display on my screen with the push of a button, I can graph/record all 3 and display the results against ATF cooler line pressure / flow rate. The OEM Pan temperature is slow in it's response and will not reflect the current internal (clutch pack and Torque convertor) operating temperatures. As an general long term bulk ATF indicator it is fine but registers too low when cruising because of the transmission case temperature drop.

The Honda I have has very efficient under hood air flow, the case of the transmission is aluminum and it almost never gets above 150 deg when moving, a metal line attatched to the transmission is cooled by the transmission case (measured). Ignoring that, unless you can wrap/insulate the entire metal line, it will be cooled by the great airflow over the motor, transmission and all metal work, and that includes the Transmission cooler if it is attatched to the metal line. I was seeing 10 to 15 deg drops from actual steady state fluid temperatures in the extensive testing I preformed. Of more concern, there is also a delay in ATF heating the metal line that flattens higher temperature peaks.

Again it depends on what you actually need or want to measure and why, as the early transmissions had a lot of problems I needed to catch temperature rises under my harsh driving conditions before they could cause a problem. This obsession with measuring ATF with utmost accuracy was to actually test and find the best ATF cooling system for my Vehicle and dispel the myths, speculation and general quoted transmission cooling setups that do not apply to the Honda 5AT.

Mkarl

One thing I forgot to mention - a difference between other vehicles and Honda transmissions. The Honda ATF cooler circuit runs at relatively low pressure, this is especially true when cruising as pressure at operating temperature can be as low as 3 PSI, obviously the flow rate is also very low. Now the transmission TC is "hopefully" locked as are the clutch packs, so it's basically a manual transmission and very little heat is being generated, the flow rate is low enough to cause the metal lines to cool too transmission Case and engine compartment temperatures, even the OEM transmission temperature sensor readings slowly drop below bulk oil temperature - I can see this any day of the week - with a fully locked Auto operating like a manual transmission, Torque convertor outlet temperature drops to match the OEM pan temperature within a minute or so, however after about 10 minutes of 70 mph cruising, the pan temps drop way below the AFT temperature measured at the Torque convertor outlet port, which is the external cooler circuit outlet. For this same reason - low flow that only peaks at gear changes and higher revs (lower gears) - you run into all sorts incorrect assumptions about cooler Delta or efficiency and reading correct ATF temperature when attatched to metal lines.

I was able to get almost correct readings by having the sensor mounted against a small metal section wrapped and fully insulated from surrounding metal piping, however that still missed 20 second high spikes on hill climbing or consecutive high throttle gear changes, showing as much as 20 deg lower than the true ATF temperature. Again most people don't need this type of response. Like I said, my aim was different, I posted some results to highlight accurate measurements that were not influenced by the variances of mounting methods and external factors that normally skew readings.

Along with the temperature readings I also monitored TC outlet pressure and Cooler line return pressures - IE the pressure drop across the cooler, ATF lines and temperature bypass valve (although reading cooler return pressure in this way is not that accurate) but it does give a repeatable indication of relative flow rate when doing comparative testing.

You are lucky to have a later model transmission and a lot of the early model problems were rectified, but the basic operation of the transmission is still the same.

webbch

I can tell you've dug into this more than myself, and I thought I had been pretty thorough! Like I said, after boiling fluid on my '03 Honda Odyssey and having it rebuilt the 2nd time, I started paying more attention. On that one, I didn't bother to take readings before installing the external cooler as it clearly needed one, although I did install thermocouples after the fact to confirm reading were far more reasonable. The later (Odyssey) by '05 I think had resolved some of the root cause issues pertaining to insufficient flow to torque converter. I suspect the TL shares a good portion of transmission design though I've not researched it to confirm....of course the odyssey is a much heavier vehicle, making it even yet more susceptible to overheating the trans fluid.

Mkarl

You are correct, they solved the idle problem that caused the TC to be bypassed - virtually no flow at HOT idle in gear - but many people are still surprised by the low cooler circuit pressure design in the Honda - Many early rebuilt TC failures were the result of testing the TC at normal transmission pressures, once it was realized that the Honda TC had to be tested at a much lower pressure as well, that part of the rebuild problem was resolved. Interestingly they found the internal radiator cooler drops it's efficiency by a huge margin as it ages - just another cause of overheating ATF in older vehicles like the Odyssey that used that cooler design.