Water Pump
 

My 98 Boxster is very quickly reaching 60K and I would like to replace the water pump this winter. Is it worth paying the vast difference to get the pump from a dealer, or has anyone had good luck with an EBay seller. Thanks in advance for you thoughts. Ed

what is the benefit of doing hte 160 degree thermostat?

does it make the car run faster by making the car run cooler? i've heard of guys claiming that on other cars.... but h aven't heard any definitive proof that it does that...

if the car runs better/faster/more efficieintly/etc on a 160 thermo why didn't Porsche put one in from the factory?

Besides potentially saving your engine by keeping it cooler, as well as substantially lowering oil temps ( a HUGE plus); yes is makes slightly more power by improving volumetric efficiency to the tune of picking up 5 HP. LN Engineering, who developed the 160 F stat in conjunction with Jake Raby, has data from dyno pulls on their website……………..

And Porsche does use lower temp stats in their "performance" engined cars such as the GT2, GT3, and Turbo cars................

I have the 160 LN stat in my Box S, so I'm biased. This has been debated many times on this and other forums, but the bottom line is the car's coolant temp is reduced by some 20+ degrees, thereby reducing oil temps at the same time.

As for why Porsche didn't design this like the GT3s, turbos, etc?
I think the answer lies in that the typical use of a US Boxster driver is shorter trips as a daily driver vehicle. The higher temps help burn off containiments in the oil, at the cost of higher heat, oil breakdown and engine wear. JFP, Raby and Navarro have dissected these engines and know their faults and weaknesses...I trust their judgement in the "everyday" real world.

I want my engine to last a long time (I'm approaching 16k miles). I change my oil at least annually (3k miles) and use only high quality synthetic and OEM Porsche coolant.

If I were changing my water pump, it wouldn't be with some knock-off.

Hi, I replaced the crappy OEM pump with one bought on ebay, it has been 2 years since and haven't had any problems.

I paid about $125.00 shipped. :)

I have to disagree with those urging oem replacement.

While I usually always go oem, the waterpump is simply inferior to several of the remans or aftermarket ones out there using a metal impeller instead of the oem plastic one.

Not only is coolant flow reduced and debris forced through the system (possibly clogging it) as the impeller degrades, but it also throws the shaft out of balance shortening the life of the bearing, seal, and the pump itself.

The pumps with metal impellers don't cost any more, in fact, they're usually less than oem.

:cheers:

If you call Suncoast, they carry the low temp thermostat as well. One stop shop if you did want to go that route.

I recently purchased an OEM pump from Suncoast for about $200 delivered. A good enough price to make me not think much about the $70-80 "savings" buying one from an iffy seller on eBay.

Sorry, but IMHO, the 'iffy' part is the oem replacement.

:cheers:

------------------------------------------------------------------------------------------
I always thought tolerances in an engine were optimized to operate best at a specific, predetermined temp. Motors are made of lots of different alloys any metals. They all expand at different rates. In this case, the state of the alloys at 180 degrees is where the engineering is done and optimized.

Just in a piston and rod assembly, you have 5,6...10 (?) different alloys all expanding at different rates and amounts as they heat up. The engineers decided 180 was nice and designed the parts, taking into account they're expansion, to fit best when they are all at 180. At 160, your goung to have different tolerances. Maybe the rings have expanded but the cylinder hasn't caught up, more compression, more hp, what ever. But harder on other stuff that's not fitting right. Of course far to be it from me preaching tips for longevity, I've disregarded all common sense pushing mine towards a D-chunk.

I could be totally wrong, after all you don't really have to warm up a car any more. But, justifying this change because a GT runs at 140 is apples and oranges for the reasons above. The engineers liked 140, for whatever reason, and designed everything to fit best at that temp and the accompanying degree of expansion of the various alloys etc.

Aside from that, the GT and Turbos are entirely different engines. Water cooled flat sixes yes, but I believe they were derived from the 959 which was based loosely on the old bullet-proof, air cooled 963(?).

I'm not saying don't do it, with a good explanation, I might even do it myself when the time comes. I'm only saying I doubt that the 180# was an arbitrary # for this motor. Probably very minuscule, but I don't think you do it without some long term consequences.

Coffee's gone, Saturday sermons over.

Regards, PK

Jake and Charles have shown that the M96 engine actually operates at around 210*, not 180*. The factory t-stat doesn't even open until 185* and isn't fully open until around 205*. The LN stat opens at 160* and is fully open by 180*...cooler coolant and cooler oil; makes your engine last longer.

As far as the water pump; if I ever have to replace mine, I will compare the options and purchase the best replacement part regardless of cost.

jmatta

In all sincerity, why would Porsche knowingly put a thermostat in that they must know (if you do) is much hotter and is therefore going shorten the engines life ?( As if the motor wasn't a bit of a disaster as it was.) And if it's running at 210, why not say so instead off consciously papering over the dash gauge to read 180?.

Do you think they play the same hi-jinx with the GT and turbos that, as is stated in this thread , are supposedly running at 140?

Also , cooler oil sounds nice (especially on a hot day (grin)) but what difference dose it make...it breaks down sooner and might be harder on rubber seals but what else that isn't compensated for? These things have a pretty aggressive oil change regiment which should mitigate any oil degradation, and rubber seals arn't your fathers

This isn't rhetorical and I'm not being a smart a__, I'm truly curious.

Regards, PK

Well, 60% of the heat produced in the engine (and some of the most critical heat, think bearings) is carried away by the oil, not the coolant.

In the Boxster, the heat exchanger transfers much of that heat from the oil to the coolant which dispels it from the radiators.

So, if the coolant runs lower temps, it does a better job of aiding the oil in shedding the heat.

:cheers:

pk2, if you are curious, do a search on the topic. The merits of the t-stat have been debated time and again. The temp reading on your gauge is a joke.

Porsche is more concerned with burning off oil containiments (for those extremely long oil change recommendations) than engine longevity, hence their choice in operating temperature.

temp gauge, etc.
 

http://img.photobucket.com/albums/v2...er_cluster.jpg

Some people seem to think that the Boxster cruises at 180 degrees because the temp gauge needle typically sits between the "8" and "0" of the "180" label when at cruise. I don't think that Porsche labeled the gauge so that 180 degrees is indicated when the needle is in that positiion...even if they might have made it deliberately misleading. The temp gauge has a blue-tinged tic in the vicinity of the "100" label, then an unlabeled tic, then a tic above the left side of the "8" in the "180", then another unlabeled tic, and finally a red tic in the vicinity of the "250" label. That being the case, the common temp gauge needle position at cruise (between the 8 and the 0) might be inferred as 200 or maybe more. I suppose that the unlabeled tic to the right of 180 could be inferred as 225...

The Boxster engine actually runs above 200 degrees at cruise, but (in stock form) has a 180 degree thermostat. This is possible because the Boxster's thermostat is on the INLET side of the cooling system. This is made clear by the cooling system diagram on page 19-2 of the Bentley service manual. So if the coolant returning from the radiators has to exceed 180 degrees to begin opening the thermostat, then the coolant leaving the engine (headed for the inlets of the radiators) must be significantly hotter than 180 degrees...and that's why an engine with a 180 degree thermostat actually runs in excess of 200 degrees at the temperature sensor location. This also explains why the 160 degree thermostat can result in cruise temps in the 170s - 180s.

Jimatta, Lil_bastard, blue-Thanks for your interest.

Jimmata, As far as an “ongoing debate” about water temp, I’ll check it out. If there’s anything definitive (Blu_s seem plausible), I’m in. If it’s another XYZ headers vs. stock, K&N vs. stock filter vs. cold induction …bla bla bla…and they all the add hp… I’ll figure it out, not that hard to take an accurate coolant temps.

(Blue_s) Interesting facts. With temps all over the place from heads to block to radiators, could the 180 on the dash just be a calculated average of pertinent areas?

But (lil_bastard), I just cut to the chase and got a read on the oil temp in my car (seems to be the real issue). Unless someone finds fault with my methodology, mine is 188f… under pretty harsh conditions. Don’t know if 188f is good or bad or how closely it’s linked to coolant temp. I’m confident it’s pretty accurate, and from what you’all say, it is the real bottom line.

Anyway, below are picks of my little measurements..
Details of methodology are below pics.

http://img30.imageshack.us/img30/8943/boxtemppost.jpg
By Kroak

I drove moderately hard about 4 miles with a few final full boost red lines, gauge read 180, ambient temp 97f (if not on fire) and dry, 22% humid, asphalt 157f (proximity to oil sump, might matter). Intake temp 175f (forced induction, Eaton M63)*, internally stock original 2,.5L, 32k, an extra 50-80 hp (depending who you talk to (like these threads)) but, a significant extra load worth mentioning.

To measure the temp I used a laboratory digital laser pyrometer and pointed it at 1“ from around the middle of the bottom of the crankcase. Most all the oil, cooled, or back from circulation should be there after a short rest (3mins.). Plus aluminum is a very good conductor of heat; After 3 mins. Oil inside should be about the same temp as the bottom of the case I figure.

So I don’t know, if my methodology is right, is 188f to hot for the oil under somewhat extreme circumstances, what should it be?

Regards, PK

* I only took 1 reading on the blowers exhaust to intake pipe@175f . Like to take 5 or 6 to verify accuracy & repeatability but, trust me, it’s frigging hot..

JFP_PA,

Glad to entertain. I'm all for running cooler.Indeed, I'm on the verge of h20 injection/intercooling. But as blue_s stated, the reason for allegedly running it hot was to burn off oil (theory?). We have the most stringent emissions standards in the land...might not fly.

But, the point made here several times is you want your coolant cooler so the oil is cooler. Well, they don't seem so directly related. Obviously my oil temps of 188f (assuming good methodology) are lower then the 210 every body is talking about & the oil has it's own cooler (no?). So again, begs the question,Forget the water, How hot should your oil be?

The Digital laser pyrometer can indeed be "entertaining" at frat parties especially, drawing a laser bead on your friends crotches and zapping temp readings...laugh riot. But actually, it's a very accurate scientific instrument. I snagged it straight from the family biz/laboratory's where it's used daily while they accelerate, smash and other wise mess with atoms and ions at thousands of degrees and at the speed of light(not exactly frat. stuff). Also, can't think of a better place to take a read, can you?. (if ya do I'll do it).

Anyway you make some good points. OBCII reading sounds interesting, I'll try it. but you realize your taping into a pretty primitive sort of sensor, electro mechanical, through Porsche's apparently less than honest electrical system.

You'all are probably right, you've made some compelling arguments.... And god knows I can use all the cooling I can get. A blower generates lots of top end heat (I know, have 2 EGT sensors). But, so many people swearing they feel the unsubstantiated but advertised 5-10-15 hp with their new ebay headers, greasy air filters, what ever. Ya gotta be more skeptical the more logic is defied. So when someone here said the thermostat is good for another 5 hp...Ya gotta think "oh god, here we go". But as the science trickles in it seems compelling.

Thanks again, PK

BTW; In my BMW3.0 coup days, we used to just drill holes around the perimeter of the thermostat. They were really accessible so trial an error was easy.

180?
 

When you say the dash gauge indicates 180, where is the needle pointing? Is it straight up, between the "8" and the "0"? I contend that the tic mark above the "8" is the 180 degree position, and the common cruise position (straight up, between the 8 and the 0) is intended to indicate HOTTER than 180 degrees. The gauge is certainly non-linear, and was apparently not designed to provide a precise temperature indication. If they had used a more serious-looking font, maybe they could have placed the number markings in a centered position below each respective tic mark. Maybe they should have just left the numbers off of the gauge altogether...

The M96 stat already has a hole in it………..

Typically, we use the middle of the “8” as our reference point, and when the needle is there, the coolant is 210-215F; go past that and it gets even hotter, often reaching 225F by the time it is over the “0”. Because of the non linearity of the gauge, the relationship of position to actual temp is very misleading……

JPA_PA

Yup you do, Your a "CARB state", California has special status, its sets it's own standards, other states can choose between weak Fed standards and tough CA standards. PA and others follow CA. Lucky us lucky you. Unfortunately with the motor totally re-tuned for the blower, I barrrrrely squeaked by ...at idle no less!?!

A water to oil cooler makes them (the fluids) the same temp and that is 220-210, O.K. Why does my oil appear to be 188f. If you read my methodology & you get the logic, I think it's pretty sound. If you got a better method, let me know. Punch holes in mine. I don't care as long as the the numbers add up.

Funny thing though, I suppose if I already had a 160 thermostat (or 140 (?)), oil at 188f might be just about right and we could all go home. (Gotta get a good read on the coolant temp I guess.)


Regards, PK

BTWx2. Porsche has "a' hole in the thermostat, quaint. We just kept drilling till it worked right. 3 or 4 holes I think was the magic number. Did take longer to warm up though.

The problem is your methodology; you are not measuring the actual temperature of the flowing oil, you are measuring the outside metal of the sump using and indirect (non contact) method…………..there is no way your oil is circulating at 188F if you are running the OEM stat……………The M96’s coolant is pulling heat out of the oil once the engine is up to operating temperature, the oil typically is running 20-30F higher than the coolant…………………….

temp display
 

I will have to drive my 2000 "S" this evening and look at how the HVAC control panel diagnostic mode display for coolant temp compares with the needle position on the gauge. I should look at oil temp, too. Right now, ambient temp is in the neighborhood of 100 degrees in the inland areas of Southern California.

Here's the procedure for putting the HVAC panel in diagnostic mode:

Blue_s.

Let me know what you find. I think I have that setup too but quite frankly, I never saw much of interest. Looks like you have though. Be nice to see some alternative numbers.

Only a hundred in Corona eh? It was 97 the day I did my test, about as hot as it's been, I live about 40 mi. south west of you. Figured at least 110 further inland.

Regards, PK

JFP in PA,

Working on a rebuttal. Have dyslexia but good. Takes forever to write something legible.

You make some good points but I think I have you on a few. To one point, you might look-up "digital laser pyrometer" on Wikipedia or wiki-answers (for simple answer).

Regards, PK

Actually, as I already own two of them, a look up is not required, I know how they work, and their level of accuracy..................

Ya I guess not, ... perfect. You just saved me about an hour and a half trying to explain and key in what it's all about.

So, with this one, Ive measured quite a few different materials (including my head, including aluminum) and from different distances in different ambient environments etc. pretty accurately (used a standard scientific thermometer to verify a few). At the Lab I've seen them take pretty precise measurements at at least 1400 degrees. The only thing I can't measure is flame.

Since all it it sees is infared (IR), not the material or you or me, and the IR is all the same no matter where it comes from, and all it does is measure the intensity iradience) of the IR. Why do you think it can't accurately read the temp of this particular aluminum alloy of the sump?

Listen, I really appreciate you humoring me and sticking around. I think I just have one more thing that doesn't quite compute for me, and then I'll leave you alone.

It's sounding like this may be a genuine good thing.

Regards, PK

After recently installing a new water pump, 160F thermostat and new coolant in my 2000 2.7 I've been monitoring my coolant temperatures. With ambient temperatures in the mid-eighties and speeds in the 70 MPH range I'm running between 80-85C (176-185F). When I slow down or stop in traffic the temperature quickly rises to 95-99C (203-210F). I haven't cleaned my radiators in 5 years so I assume they aren't ideal. I haven't yet looked at oil temperatures but will on the next outing.

Curious to see what you are running with your S.

ddb

That's typical. The only way to keep the coolant at the cruise temp is to make the fans come on earlier. There's a thread on the 996 forums at Renntech on how to do the mod:

http://www.renntech.org/forums/index.php?showtopic=28093&st=0&p=148785&#entry1487 85

I plan on doing this mod to our Boxster myself as I see the very high trigger point for the fans as the real limiting factor.

I will have my car back from Raby soon with, amoung other stuff, a 160 Stat, however lets take this thinking to the extreme.

If lower is better, what would happen if the Tstat was removed and water was allowed to flow unhindered?

Not to hijack, just to understand better.

I think the water (and maybe oil) temperatures could be too low for too long. Maybe not a big issue in your climate, but in colder areas you wouldn't have any (or enough) heat. Engine oil and coolant temps. too low probably isn't good for the engine just like too hot isn't ideal.

Removing the T stat makes the engine run HOTTER.. The coolant goes through the radiators too quickly to cool.

-----------------------------------------------------------------
No problem

I like that attitude. I can see your ratcheting it up bit by bit.

I did take the thermostat out of my old 6cyl bmw cs coupe a long time ago. The problem was it never really warmed up. In fact it was down right cold, in the blue at speed somtimes.

With alot of trial and error, I think 3, 1/8th" holes through the thermostat gave me the temp I wanted (140??...maybe?). To be honest though, I didn't really notice a diff. Only that as db said, it took a little longer to warm up. I guess we're talking 5 hp here? I built the whole engine and tweaked quite a bit. 5 hp just might not have registered.

But, I think I agree with db, It's used to be, and I pretty sure it's still true to a degree, only when the engine was warmed up and running within a certain temp window that all the moving parts have expanded and fit together correctly as designed (we hope). So in a way I think driving around a cold engine is just as hard on it as driving it to hot.

Anyway, pop in any time. I'm the hijacker here. Till I arrived this thread was about changing a water pump. PK

DB, your fans aren't turning on very early. Mine come on within a minuet or 2 of hitting traffic. Once It''s warmed up, for what it's worth, the needle doesn't move much.

Don't know what your using for coolant, pretty sure it's not your prob but, your ready for a change. Might consider something like a quality version of GM.s Dex cool. Most of us traitors use Xerex dex-cool (compatible), about 1/5th the price of Porches exclusive juice and chemically identical where it counts (no silica or phosphates).

You probably knew all that already. If not, don't trust me, do some research. That's how I figured it out. It may still be controversial, so is the current issue at hand though.

Regards, PK

I had some time this evening to drive a bit and collect some data from the HVAC control panel's diagnostic mode as explained above. The car was a 2000 Boxster S 6-speed, with a new Porsche water pump and a 1-year-old standard replacement thermostat (presumed to be 80 deg C). The A/C compressor was off throughout the testing.

* During warmup, with the temp gauge needle exactly on the "tic" mark above the "8" in the "180" label, the HVAC panel (in 6c display) read 79 degrees C coolant temp. That converts to 174 degrees F.

* After allowing coolant temp to stabilize by driving approximately 10 miles with cruise set at 71 mph indicated in the digital speedometer display, I saw the following in the HVAC panel diagnostic mode:

Outside Temp Average (5c display) = 30 deg C (converts to 86 deg F)

Coolant Temp (6c display) = 91 deg C (converts to 196 deg F)

Vehicle Speed (20c display) = 111 km/h (converts to 69 mph)

RPM (21c display) = 2700

The coolant temp gauge needle in the instrument cluster was pointing straight up between the "8" and the "0" of the "180" label.

Oil temp was unavailable in the HVAC control panel diagnostic mode, because (I think) the 986 Boxster does not have an oil temp sensor in stock form.

------------------------------------------------------

When I returned home, I idled the engine for a while and gathered some more data:

* at 100 deg C (6c display on the HVAC) the gauge needle was approximately in the middle of the "0" in the "180" label.

* at 102 deg C (6c display on the HVAC) the right side radiator fan came on. This correlates with the specification in the Bentley manual for HIGH speed (stage 2) fan. The coolant temp crept up to 104 deg C (219 deg F) and stayed there for at least 5 more minutes with the right side radiator fan running. The gauge needle was at the right edge of the "0" of the "180" label when the HVAC was displaying 104 deg C coolant temp.

It is apparent that I have a radiator fan problem on my car. According to the wiring diagrams, both left and right side fans should be controlled at the same speed (stage 1 or stage 2) at the same time. It should be pretty simple to begin testing at the relay locations under the dash, since the relays seem to follow the standard ISO terminal definitions. I'll start with my Fluke 88 meter and a fused jumper wire...

So, what's going on here? These coolant temperatures do not match what JFP in PA, cnavarro and Jake Raby have posted on multiple occasions. Is the DME lying to the cluster and/or the HVAC control panel? Is the cluster lying to the HVAC control panel? Is the HVAC control panel lying to me? Or is the HVAC panel diagnostic mode accurately reporting the coolant temp as the DME reads it from the ECT sensor? None of the scan tools I have access to will read DME data, so this might be the push I need to buy the Durametric package...

Blue s,

Great work. So what has been said appears to be true. Have to ponder a bit...Think I'll see if I can get the same in my car... maybe I can even get the oil temp.

Let ya know whhat I come up with.

Regards, PK

I have never used the climate control panel, so I cannot vouch for its level of accuracy, or how it compares to those seen on a scanner. You also need to recognize that these cars do vary somewhat in the temperatures they run at, even with the LN stat, due to differences in the cooling system (debris in the radiators, condition of the pump, non linearity of the dash unit, etc). I do know that the water temperature senders, at least those I have had the chance to evaluate, are pretty accurate. When using an OBD II scanner to monitor coolant temps, you are seeing the actual sensor data being relayed to the DME, so the numbers are also pretty reliable, and have correlated well with aftermarket gauge setups in the same cars.

On the Boxster, you cannot see the M96’s oil temperatures with out adding an oil temperature sensor to the engine, and while it can be done a couple of ways for testing, it is not a practical setup for day to day driving; so we only have used it for test purposes.