Day/Night Temp Fluctuation: Tuner Trivia

[Wheelman-111]

Greetings:

See Scooter Builds Flash III Chronicles for the full story.

Good morning, kids! In our last episode, a smug Wheelman was ever so certain he'd found the 109 to be the Magic Jet for the stock carb with the 78cc setup. Comfortably cool during testing last night, even after a good spell at WOT. So we're done until break-in stage is over, right? Maybe Not!

Now this morning, with ambient air temps easily 15 degrees cooler, it's running warmer again. Pretty much stuck at 370-400 at steady-state cruising and without throttle abuse. What makes an engine run 325 at night and 375 the next morning? Humidity? Bad eyesight??

Moving the clip down helped a little, but still nudges 400 more often than I'd like.

I need to tune for average conditions, so I'll break out the drills again tonight. Sheesh!

[bradthreee]

Put the drill away and wait a couple of days for the real thing to show up in your mailbox

[Wheelman-111]

Greetings:

Wise B3 suggests:

Put the drill away and wait a couple of days for the real thing to show up in your mailbox

Quite true that the bore size of the jet is only one factor in how much fuel it flows. I'm still at a loss to explain the differing temps observed lately.

[mousewheels]

Haven't found any mfg data for the VDO gauge as to whether or not it is temperature compensated. If it is not, it will read much differently between cold and garaged in the morning vs out in the heat and sun of the afternoon.

Info below is from the Westach CHT meter:
See pic below. Westach describes the ambient temperature impact to readings.

Another source of error is shown in the Westach schematic. They show the thermocouple being wired to a 'patch cable' extension. If the splice is not thermocouple wire and is made where it is hot (like in the engine compartment), there will be increased error. Many would say run an *unbroken* length of thermocouple wire up to the meter.

The wiring splice example important because even if a meter has temperature compensation, it's cannot adjust for an unknown splice temperature.

Info from Westach 2C1 CHT gauge instructions


Edit: 11/25/2013 Pic above was lost due to a photo hosting service change. Pic below is substituted.

Westach_theromocouple_installation.JPG (88.95 KiB) Viewed 7191 times

[Wheelman-111]

Greetings:

Mr. Mouse, you dunnit again!

First of all, the Westach uses a type "J", which is encouraging if my VDO uses the same system. Pretty safe bet?

Second, the "Reads Hot When Cold" idea perfectly matches my observations. I do have a splice pretty close to the sender, inside the rear panels. Evidently there's no thermal compensation on mine.

Some of these issues are about to be resolved, however. I took a chance and ordered this Type "J" from Grainger:



One drawback was the $29 price tag. Another was that the choices were between 4' and 12' of lead wire. At least I can coil up the 7' of excess under the front cover. The good news is that all of it is metal-jacketed, fiberglass-insulated top-quality material. The ring lug is a 1/4", so it'll tuck safely away from marauding Wheelhands attacking the spark plug every other day. It's going to a headbolt instead. I'll try your suggestion and test with a cup of boiling water to get a baseline error estimate.

Thanks Mouse!

[Wheelman-111]

Greetings:

Tried the dip test with the Grainger probe. Transferred boiling water to a travel mug and got it to the hooked-up probe within a minute. Gauge responded with a reading of 150. Close enough?

Nothing to lose so I trimmed the 12' metal-jacketed wire to 6' with similar results. Water sure seemed hotter than 150...air temps around 85, no splice in the wire.

[mousewheels]

Transferred boiling water to a travel mug and got it to the hooked-up probe within a minute. Gauge responded with a reading of 150. Close enough? Nothing to lose so I trimmed the 12' metal-jacketed wire to 6' with similar results. Water sure seemed hotter than 150...air temps around 85, no splice in the wire.

Definitely the VDO does not use a 'K type' thermocouple, it would read even lower.
Edit (addition) The 'J' type may be correct, and the meter reads low. Below, is a 'what if'...

There is a higher output voltage thermocouple Type 'E'. The 'E' Type would read approx (150 * 68/55) = 185 degrees + 10 ambient compensation ==> 195.

How did it look on this mornings ride?

[Wheelman-111]

Greetings:

That E-type sounds like the correct calibration doesn't it. I'm pretty sure that cup o'water was still hotter than 180. Lip-burning hot.

Now who carries an "E" thermocouple?? I have my original VDO that I'd written off because it began working intermittently. Can I measure it somehow? (Have Multi-Meter.) Crimp on a 1/4" ring and slip it under a head bolt??

[mousewheels]

I have my original VDO that I'd written off because it began working intermittently. Can I measure it somehow? (Have Multi-Meter.) Crimp on a 1/4" ring and slip it under a head bolt??

Sure, DMM test is good, use the 200mV DC scale. Be easier to run the test in your kitchen with boiling water. Tested on the scooter, we'd want to know accurate head temperature .

Update
Added tables of input temperature (left col) vs expected voltage output (far right col) for 3 common thermocouple types. In the middle is the process to 'compensate' for 70F temperature at the meter side of the thermocouple. With a 200mv scale DMM limit on a common 3 1/2 digit DMM, the signal is in the last digit of the meter. An oven will help to increase measured voltage.

Also, PM if interested in sending a couple inches of thermocouple wire. Will ID the type with an industrial thermocouple meter. It has switch settings for E,J,K.

[maddog]

All i can say about you kidz is both of you have no clue on how to build an engine

[mousewheels]

All i can say about you kidz is both of you have no clue on how to build an engine

Hi maddog! This should save some $$$ for the the good stuff8) Why be paying $30 or more for a factory thermocouple?
Armed with this table is anyone here can test and identify what their CHT or EGT uses for a thermocouple, then get wire or probes from a lower cost source.

A test of the process was run at home, with an 'unknown probe' in a junk box.
Kitchen was 81F, probe in boiling water. Adjusted the compensation voltage to match. The closest table fitting is the 'K type' the unknown probe was within 1% of expected output voltage for a 'K' probe. It's clearly a 'K' nothing else is close!


The test is this simple
To ID an unknown thermocouple, test with the DMM close to 70F:

1)Put probe in boiling water
2)Measure the themocouple output voltage. 3.25 mV = 'K' type, 4.19 mV = J, 5.06mV is 'E' type.

If your meter cannot read into the decimal places, test at a higher temperature
1) Place in 400F oven
2) Measured themocouple voltage of 7 to 8 mv it 'K', 10mv is "J", 12 mV is "E"
----- Actual values would be (7.47 is 'K' , 9.95mV is "J' , 12.31 is 'E') if your meter could display the extra digits.

[Wheelman-111]

Greetings:

Good info, Mouse. If I could boil water, I'd try it out but lack the cooking skills.

I'll pull out the old original VDO sender and see if it works at all.

If only someone could condescend me an engine-building clue...

Some people dress in tweed. I've sewn your dreams.

[Wheelman-111]

Greetings:

Both my VDO and my cut-down TRW 6' solid wire read 3.5 mV in boiling water.

At least now we know which type the VDO requires. Is it possible my third probe from Aircraft Spruce is a Type E? That would explain alarming readings with good engine performance. Too late to pull it off tonight.

[mousewheels]

Is it possible my third probe from Aircraft Spruce is a Type E? That would explain alarming readings with good engine performance.

Argh sure would, so many different CHT probes!

[Wheelman-111]

Greetings

I went poking around for my CHT probe receipt without success. What do you think my peak temps of 450 would be on a J probe if in fact I've got a Type E? Gauge scale is 0 at 100F.