Davies Craig EWP - preliminary results

Diesel Do said:

I thought so
Measure the resistance of the sensor and you can calculate the resistance at any given temperature! What more do you want

Click to expand...

Are you sure that Lacey hasn't been handing you cheat sheets ?

So, if the prope is a simple positive coeff thermistor, could we unplug it and put in a variable resistor to work out what voltages correspond to what temps are displayed on the controller?

So say for example that the controller show 85 degress when the voltage from the probe is 1v - but when the actual temp is 85 degree the probe sends through 0.8v. Once we know how much the probe is out by we can recude the resistance to a very accurate level. Or am I bonkers?

Bonkers.
Yu don't know that the sensor is resistive. It could be the impedance that varies. Or else it could be a thermocouple - T or K? Just stick to what DC have said.

Yep thought so . I'm sure the DC advice is correct - just couldn't hurt to know why it's correct . I'm eager to get hold of the IR thermo now, it'll be interesting to see where this goes.

But cheers for all the input, as always .

Yu could ask them?

The Mad Hat Man said:

Yu could ask them?

Click to expand...

Might well do, this bloke seemed to know his stuff. It would at least helps us better understanf the system.

Wow fair play to DC, they are sending me a new sensor - now that's service! They also sent me this chart which shows what the sensor should be doing.

That's what I said! Goes down with temp hence using it in parallel! Not sure a new one will be better tho?


DD

It's an NTC thermistor. With a 20K pot across it you will be able to adjust it but only one direction. With the pot in parallel it will make it come in at a lower temp. If the pot was in series it would come in at a higher temp!!

Guys. You're not going to flippin believe this.

I thought to myself "well, maybe I do have a dicky temp sensor which thinks the water is colder than it really is, so I'll set it to the lowest target temp possible and see what happens."

Steady temperature? Sure thing bro.
Constant hot air? You bet.
Post shutdown run on every time? Bitch please.

I don't effing believe this, it's working, the fooking thing is working!!! I still don't know what the real temperature is but the controller is working SO MUCH BETTER with the temp set to 75. The EWP is always in variable voltage mode, unless I stop at light for more than a minute or so. Once the damn thing warms up the heater works perfectly unless the EWP goes into pulse mode and then it drops a bit. SO re-routing the heater return hose has clearly worked in correcting the flow.

So, with a properly calibrated sensor I'm betting this system will be utterly transformed !!! My only caveat is the slightly long warm up time, but this will be improved by accurate temp readings I expect.

Asd for the heater, one of those small merc booster pumps set to come on with the blower should do the trick - it'll keep hot water going through the heater even when the EWP is in pulse mode. Which should also mean that hot air comes through very quickly during warm up.

I knew I was going to get a breakthrough! So the moral of the story? Make sure all your components work properly before rebuilding the system!!!!!

So I'll be able to get a clearer idea of all this when my IRT arrives but I think maybe, just maybe, this could be the solution we were looking for.

Will. Very happy.

Well that was illuminating!

I just took the car for a short spin around the area, couldn't join the main road as it's rush hour but I got the engine up to temperature on the dash gauge. The controller was still set to a target of 75 degrees, as the earlier results were impressive.

When I stopped to take some readings the controller was showing that the EWP was on constant run mode and that the temperature was below 75 degrees. The IR scanner gave me the following readings:

Upper rad hose : ~90 degrees c
Lower rad hose: ~20 degrees c
Heater inlet: ~70 degrees c
Heater outlet: ~40 degrees c

I didn't check the IRD circuit as the runs was probably too short to take a worthwhile reading.

But don't that beat all eh? That sensor is way waaaaaaaay off! So lets say that the replacement sensor is the same, can I take a resistance reading from the sensor and then add a suitable resistor to bring it where it needs to be on the graph supplied by DC?

Will.

Btw MHM, I owe you a beer for suggesting that scanner (on top of the small brewery's worth I already owe DD.)