Announcement

**hamburglar** · 11-22-2014, 06:31 PM

I read through everything and I think I'm a bit confused. It seems like your gauge was malfunctioning, but when you list new parts that were installed you said that the gauge was now working. Then later down again you mention that it might be bad. Can you clarify this?

Either way, this was an amazingly helpful way to ask others for their assistance and ideas. I'll keep checking back in until this gets resolved.

**dragonsung** · 11-22-2014, 06:41 PM

From what I was told, the last engine over heated and died. From what I read once the gauges get pegged hot they don't work right anymore. Every time the car was started it was like the sending unit wire was grounded as it would shot to hot. I tested the wire and it was sending ohms in the right range to the gauge. I concluded the gauge was bad.

I got a new "used" unit and put it in. It seemed to be working right, but as I am having issues still with the cooling I wonder if the new gauge is showing the right temp.

Did this clarify?

**johnl** · 11-22-2014, 09:01 PM

Originally posted by dragonsung View Post

car cooling system when acquired:
No thermostat
fans wired to switch
Bad wiring harness (connectors to the thermo switches)
Bad heat gauge Would peg hot. (not sending unit)
New water pump and timing belt (after market)

I don't know what's going on with your engine (though I have a suspicion), but in my experience the lack of a thermostat is nearly always because someone has deleted it in a futile attempt to prevent the engine overheating.

The only time removing the thermostat is not futile is when the overheating is being caused by a failed thermostat. If the overheating is being caused by a radiator problem, or a blockage, or corrosion scale inside the water jacket, or fans not working, or a failed head gasket, then all removing the thermostat will achieve is to increase the time it takes for the engine to overheat, but probably not by all that much.

More often than not, when I've seen a thermostat having been removed, it turns out the root problem has been a leaking head gasket. This isn't to say that this is definitely your engines problem, but it's on the table.

That the fans have been wired to a switch, and that the water pump has been replaced is also a little suspicious. The previous owner may have been trying a number of things to try and keep the temperature down. Of course it's possible that these are 'innocent' changes, the thermo switch and the pump may have possibly both needed replacement, but...

**dragonsung** · 11-23-2014, 03:33 AM

my dad drove the car over 200 miles in one trip, with no tstat and it did not over heat. He wired up the fans, as the gauge was bad. He is very cheap. His plan was to drive it into the ground.

As i drove it to my house, 70 miles it did not over heat. This was with no tstat.

He changed the timing belt because he did not like the way it looked. Why he was in there he done the water pump.

Does this info help you, help me johnl?

I really like this car, hoping for good gas mileage. I will fix what needs to be fix for sure.

I think the gauge is bad or sending unit.

Test 1:
I opened the bled bolt and let it drip. I then put a thermometer in hole and at the same time measured ohms of sending unit.
data:
temp 185 F, the sending unit was at 63 ohms, and the gauge was about 1/8 away from the red line.

My thoughts:
Im not sure the temp difference from block to tstat but the temp could only go down. The sending unit should be at 49 ohms @ 185 F so if any thing it should be lower than 49 ohms. I'm also not sure why gauge was so high, from what I read normal operating temp is around 185 and the needle should be at about d3 to d4. So from the data I got If anything the gauge should read colder than it does, as the gauge should climb when the ohms drop.

any thoughts?

Test 2
I loosened the bleed bolt to get the coolant temp in Tstat housing, then I temped the left and right side of the Tstat housing (aka the two half's of the housing).
Data:
coolant temp 185 F
left side housing 143 f
right side housing 175 f

My thoughts:
First I know the Tstat starts to open @ 18x F and is fully open around 190F. I tested it on the stove. To me the data says the tstat should be open/opening, but is not. I check the left side for 2 minutes it did not change much. so the question is why does the tstat open on the stove ( I tested many times), but not in the car?

I know I installed it right, this is how I installed it.

I got a cheap heat gauge that i plan to add to the heater core inlet by the block, next to the sending unit. This should give me accurate temps installed there right?

**johnl** · 11-23-2014, 11:30 PM

"I got a cheap heat gauge that i plan to add to the heater core inlet by the block, next to the sending unit. This should give me accurate temps installed there right?"

Not quite sure what you mean, but if you install the sensor in the 'heater circuit' then it may not give an accurate reading unless the heater coolant valve is open and coolant is flowing through the heater core (or rather, flowing through the part of the circuit in which the sensor is located). This would especially be the case if the sender were in that part of the circuit where coolant is being returned from the heater core to the engine, i.e. after the coolant temperature has been lowered passing through the heater core.

If the thermostat is opening in the saucepan then it will be opening in the engine. It responds to temperature and nothing else, and it won't matter whether this is installed in the engine or sitting in the bottom of a pan of water.

You ought to be seeing the thermostat start to open about the time you start to see the first signs of small bubbles forming on the bottom of the pan, then increasingly open as these bubbles get larger and more numerous. By the time the bubbles are moving off the bottom of the pan the thermostat should be fully open (or at least nearly so). This is a very rough guide, but if it's not fully open by the time the water is properly simmering then it's not opening soon enough.

Installing an extra stand alone temperature gauge should tell you if your stock gauge is reading correctly. The sensor should be installed in the engine coolant outlet hose as near the engine as conveniently possible, though anywhere in the upper hose should be good enough.

Note that a sensor installed in the hose will only give a reading reflective of coolant temperature in the engine after the thermostat opens and coolant starts flowing in significant quantity through the hose to the radiator (the gauge will read low as the engine is heating up, then jump up quite quickly as the thermostat opens). Ideally the sensor should be exposed to coolant inside the engine prior to the thermostat, but this may not be easily possible.

Since you have a switch to operate the fan you can just run it all the time. If you still have problems then this will rule out the temperature at which the fan comes on as being the problem. The fan is only really needed when the car is at a standstill or travelling slowly, and shouldn't run at all when the car is travelling at reasonable speed (though the second fan will run all the time that the AC is on, not that it really needs to if car speed is causing adequate flow through the radiator and AC condenser).

Years ago I owned a 'Nota Sportsman' (similar to this; http://www.notasportscars.com/Images...ta_Clubman.jpg), which had an electric cooling fan operated by a toggle switch. Driving in city traffic I just left it on all the time. At speed it made no difference to engine temp, as you would expect.

**dragonsung** · 11-24-2014, 02:04 AM

in the photo above, part # 11 leads to the cutoff for the heater core, then the cut off leads to the heater core. I assume this is the inlet. If i install the gauge basically at #11 and have the cutoff open (aka full heat) this should be a correct reading as the coolant bypasses the Tstat so we get heat faster in the car.

Does this seem right?

(I have not looked this up yet,)
Any one know how I can test if the water pump is working with out taking it out of car?

Thanks

**johnl** · 11-24-2014, 04:22 AM

Originally posted by dragonsung View Post

in the photo above, part # 11 leads to the cutoff for the heater core, then the cut off leads to the heater core. I assume this is the inlet. If i install the gauge basically at #11 and have the cutoff open (aka full heat) this should be a correct reading as the coolant bypasses the Tstat so we get heat faster in the car.

Does this seem right?

By "cut off" I assume you must mean the valve that controls coolant flow through the heater core. I'm not sure whether the valve is on the inlet or the outlet side of the core, in theory it shouldn't matter much, just so long as the flow can be controlled to control heater temperature.

If fitting a sensor into the heater hose, it would concern me that when the valve is closed there will be no flow through the core, meaning no flow through the hoses. So, fitting a temperature sender in the 'heater circuit' (by which I mean a coolant 'circuit' not an electrical one) won't expose the sender to the same coolant temperature seen inside the engine's water jacket (unless the coolant is flowing through the hoses, i.e. the valve is open and the heater on). The further away the sender is from the engine the more likely it is that this would be the case, but even if it's quite close to the engine the coolant at that point may well be significantly cooler than the coolant actually in the head (also keeping in mind that the coolant temperature at the heater hose orifices is going to be less than that at the head outlet orifice).

It would be preferable to fit the sender unit either directly into the upper heater hose (there are fittings that allow this, i.e. fit a sender directly into a hole made in the hose), or into an adaptor tube spliced into the hose. At least then you know that when the thermostat is open that the sender is exposed to coolant that has flowed through the engine.

Note that with these engines, the thermostat blocks flow on the engine inlet side of the circuit, i.e. it doesn't prevent coolant from exiting the head, but prevents coolant from entering the block. Either way, the affect is more or less the same, unless the thermostat is open coolant can't flow through the radiator. With these engines the OE sender is located at the outlet orifice, but very near the coolant that is actually in the head, so normal turbulence (as the pump moves coolant around inside the engine, even if it's not flowing through the radiator) causes the sender to 'see' the 'in head' coolant temperature. If the sender were located further away (say near the radiator inlet), then this would be less the case.

If you were to get disparate temperature readings from the OE gauge and the additional gauge (say the OE gauge reads hot but the second gauge reads significantly cooler), then you might suspect that the thermostat is not opening enough. However, if you managed to mount the second sensor very near the OE sensor then they should both see more or less the same coolant temperature.

Comparing the readings between the two gauges, 'normal' would be to see the OE gauge get hotter first, until the thermostat opens, then for the second gauge to quickly 'catch up' with the OE gauge. 'Abnormal' would be for the OE gauge to read hot when the second gauge reads a normal temperature (after the thermostat has opened), in which case suspicion would point to an incorrect reading from the OE gauge (i.e. some fault).

The ideal placement for a second gauge sender would be as close as possible to the existing OE sender.

Originally posted by dragonsung View Post

Any one know how I can test if the water pump is working with out taking it out of car?

Yes, but not all that easily, nor without wasting coolant. If you took the thermostat out then replaced the cover, then disconnected the upper hose at the radiator, and then cranked the engine you should see coolant pumped out of the upper hose (at least until the fluid level drops below the level of the pump). You can't do this with the thermostat in place because it will block the coolant flow.

Pumps fail in two ways, they wear and leak and / or seize, or the impellor blades rot away (either corroded, or in rare cases eroded by the action of coolant cavitation at the edges of the blades). Damaged blades is unlikely if the coolant has been changed reasonably often, but if they are damaged then the pumping action will be impaired in some degree. I'd be surprised if the pump were rotating freely that it wouldn't be working acceptably, but it can happen.

**dragonsung** · 11-24-2014, 12:16 PM

Well I flushed with Preston flush chemical, got preoccupied and left it sitting longer than it should. The engine was not running but the water I got out was awful looking. Do you think this could cause an new after market pump blade to erode away? If so I would rater lose some coolant than not know I hosed the pump.

I was given an old gauge that I planed to install, but dumb dumb that uninstalled it cut the line, it was mechanical....

I got a new electric one. I'm going to try and install the new gauge where the coolant bleeder nut goes. if not there, it looks as if it will install in the OEM spot.

If all else fails I may not be-able to compare gauges, but I will be able to see the true temp.

What is the the over heating temp for the engine f22?
Thanks

**johnl** · 11-24-2014, 01:54 PM

Originally posted by dragonsung View Post

Well I flushed with Preston flush chemical, got preoccupied and left it sitting longer than it should. The engine was not running but the water I got out was awful looking. Do you think this could cause an new after market pump blade to erode away? If so I would rater lose some coolant than not know I hosed the pump.

I can't give you a definitive answer, but IMO even though you left the cleaning agent in for two weeks, I'd be surprised if it would have caused significant damage to the impellor blades (especially if the engine was cold for that two weeks, since the effectiveness of the agent will be less at low temperature). Most of these cleaning agents are alkaline, which can be corrosive but not as corrosive as acidic formulations (my understanding only, I'm not 100% on this).

Mixed in with a system already full of water the agent will be relatively dilute. When a radiator is professionally cleaned off car they are immersed for hours in a hot concentrated alkaline solution that will be a much stronger than you're likely to get from a bottle of cleaner poured into the system.

Having said this, the only way (I can think of) that you can know for sure that the blades are OK is to see if the pump actually pumps (as I described above), or take it out and have a look. My suspicion would be that aluminium parts (e.g. the head casting) would be more likely to be damaged by prolonged exposure to corrosive agents than steel parts like the impellor blades...

Originally posted by dragonsung View Post

I got a new electric one. I'm going to try and install the new gauge where the coolant bleeder nut goes. if not there, it looks as if it will install in the OEM spot.

If all else fails I may not be-able to compare gauges, but I will be able to see the true temp.

You really want to locate the sender in coolant that is in or exiting the cylinder head. The coolant is not as hot lower down, where the bleeder nipple is, so you won't get a true reading from it if mounted lower in the water jacket.

The cylinder head is the hottest part of the engine. Relatively cold coolant passes from the radiator into the engine block and then rises upward. As it rises it picks up some heat from the cylinder liners, but the liners are relatively cool compared to the metal surrounding the combustion chambers in the head, and around the exhaust ports. It's not until the coolant has flowed past these upper passageways in the head that it reaches it's maximum temperature.

If you take out the OE sender (to use the orifice for the new gauge), then since the sender is also used by the ECU to determine engine temperature, it's most probable that this will affect how the ECU 'trims' the AFR. My guess is that the ECU would assume the engine is cold and enrich the AFR substantially. It might also cause a CEL since the ECU may 'see' an open circuit on the sender (unless you left the sender attached to the wires).

Originally posted by dragonsung View Post

What is the the over heating temp for the engine f22?
Thanks

I think the thermostat opens fully by about 90�C. With a cooling system in good operating order the coolant shouldn't really get any hotter than that, though it probably will if heavily loaded for some time, e.g. climbing a long hill. Note that the thermostat is mounted lower on the engine than the sender is, so by the time the coolant is exiting the head it may be somewhat hotter than at the thermostat.

My old Nota had a thermostat that fully opened (from memory) at about 80�C, and the gauge was calibrated in degrees, it never indicated higher than this on the gauge (unless I forgot to turn the fan on in traffic...).

'Overheating' per se would be at a higher temperature than the thermostat opening temp. From an engine safety perspective anything over boiling point is dangerous for the engine. Since the coolant is pressurised and the coolant itself raises the boiling point, boiling will occur at some temperature somewhat above 100�C, depending on system pressure and coolant concentration and age. It's a moving target.

As a rule of thumb, higher engine temperatures tend to be better for fuel consumption since generally higher temperature helps to atomise fuel better (more of a consideration with carburettors, but still significant with injection). Lower engine temperatures tend to be better for power output, since generally lower temperatures means denser inducted air, so more oxygen with which to burn more fuel.

**dragonsung** · 11-24-2014, 02:34 PM

As far as i'm aware, there are 4 coolant sensors. #1 to switch fans when running, #2 to switch a fan on via timer when car is off but to hot, #3 sending unit for gauge, #4 sensor for ECU.

So I don't think swapping sending units for a new gauge should cause any trouble.

Dont know if I can post a link, but this is what I have been using.
http://techauto.awardspace.com/overheating.html

**dragonsung** · 11-24-2014, 05:54 PM

So I'm currently at a loss. The sending unit for OEM is 1/8 npt and the after market sending unit is 1/8 npt. They look almost identical, but the after market sending unit will not fit in the oem hole. Also the Honda unit fits in the after market adapters, so I don't know. Maybe the end is just a tad bigger and doesn't go all the way in the hole.

I checked temps two ways with sending unit installed at bleed bolt (does not fit all the way, just a few threads), and with sending unit installed in heater line. Both times I had a thermometer in top of radiator to see what temps where there.

Temp data:
top radiator / gauge @ bleed bolt
180-205 / 180-200~ish
top radiator / gauge @ heater hose
180/153

I quickly threw out the heater hose idea. Fans kicked on around 202-205, heat gauge in car was at the top not red but pointing at p when temps where ~200.

My issues seem to be the gauge. I want to install the gauge and test drive.

I need to find a spot to install the gauge sending unit that will not cost me to much more. Any ideas?

**johnl** · 11-24-2014, 07:19 PM

Originally posted by dragonsung View Post

I need to find a spot to install the gauge sending unit that will not cost me to much more. Any ideas?

Yes. You need a short length of steel tube (about three inches long) the same OD as the upper hose ID (or slightly larger, so long as you can force the hose over the tube), and a threaded adaptor to fit the sender thread.

Drill a hole in the tube that allows the sensor probe to pass through, then weld the adaptor to the outside of the tube. If the adaptor is brass, then silver solder the adaptor to the tube (this may require the base of the adaptor to be filed 'half-round' so it fits the tube contour). Lead solder would be OK too, as long as the joint is well formed, but I prefer to use silver solder where possible because it's stronger (though a silver soldered joint also needs a well shaped connection). You should now be able to fit the sensor into the tube so that the probe protrudes into the tube.

Cut the radiator hose so that the steel tube (and sensor) can be fitted as close as is possible to the original hose spigot (using the short section from the cut hose). It's OK for the steel and alloy spigot to be butted up against each other. Attach the rest of the hose to the other end of the steel tube. Retain the tube with 'Jubilee' clips (worm drive hose clamps). Make sure the clips are tight, and recheck the clip tightness periodically.

I've done this with other cars (to fit thermo fan switches) and it works fine. I've also used short sections of steel tube to connect straight and curved sections of rubber hose together, which can allow you to 'replicate' more complex hose shapes using simpler sections from other hoses (useful when the OE hose is not readily available or stupidly expensive). I've done this quite a few times over the years, and it's never caused a problem.

**johnl** · 11-24-2014, 07:32 PM

Originally posted by dragonsung View Post

As far as i'm aware, there are 4 coolant sensors. #1 to switch fans when running, #2 to switch a fan on via timer when car is off but to hot, #3 sending unit for gauge, #4 sensor for ECU.

So I don't think swapping sending units for a new gauge should cause any trouble.

I've only come across the main sender at the engine outlet spigot, and the thermo fan switch at the thermostat housing. I'm not aware of any other switch or sender, which isn't to say they don't exist.

**dragonsung** · 11-25-2014, 02:08 AM

I forgot all about having a tap. Yea I can just get some thick pipe and make a splice in the hose and take it from there. Thanks for the idea.

Announcement

ANY BUYING/SELLING IN THIS FORUM WILL RESULT IN AN INSTANT BAN!

Beginner Forum Rules - EVERYBODY read! (old and new members alike!)

91 accord se cooling issues (long winded)

91 accord se cooling issues (long winded)

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment