Lean is less dangerous than rich?

[raz]

Gas temperatures are notoriously difficult to measure accurately, especially under dynamic conditions. The temperature of your header is determined by the balance between heat received by convection and radiation from the flowing gases inside, and losses from convection, radiation and conduction to the ambient air and environment and cold cylinder respectively. This balance will alter with speed, gas composition, throttle opening etc etc. To measure EGT accurately you need a thing called a suction pyrometer.

The sum of knowledge owned by the members of this forum never cease to astonish me

 

OK, the keywords here are "good enough" and we're interested in relative temperatures, not calibrated absolute ones. A standard EGT sensor mounted with a bung is probably good enough. From what you say I guess an outside one is not. One problem with internal ones is that they will make an obstacle in the pipe, taking as much as 10%, or more, of the cross-section area. If nothing else, this probably means we want to put one in each downpipe. Any conclusions gathered this way could be said to not necessarily be valid without that obstacle in the downpipe, I guess.

[luhbo]

...I'm planning to write a simple datalogger for my laptop that logs timestamped ECU output, GPS speed and LC-1 output. If that Beltronics thingy can output to a computer too it would be a cool complement...

 

While you are still planning: why not code this for a PPC?

 

Hubert

[raz]

While you are still planning: why not code this for a PPC?

 

Hubert

I don't even know how to develop stuff for my Loox and it drains its batteries before my Guzzi warms up. Also, it doesn't have a serial port nor USB (as a host). Sure, there are ways around that but I don't want to focus on that. My laptop is an incredibly small Thinkpad X41 and the current plan is doing it Perl. On the other hand, maybe there is a Perl interpreter for PPC? I would be surprised if there isn't. I sure will share my code for free whenever it reaches any state of usability.

 

I'm into Glühwine now. Merry christmas

[dlaing]

I think one of these should be accurate enough to determine maximum EGT.

http://www.zeitronix.com/order/order.htm#EGT

There might be a little lag time, so results may vary between steady rpm/load tests and acceleration tests....or not

Maybe Innovate offers something, too??? These are supposed to be compatible with Innovate products:

http://www.exhaustgas.com/ProductDetail.as...tID=&RepID=

I plan on logging the stock engine temperature sensor output, A:F, rpm, tps, and maybe air temp and air pressure....someday.

[Ryland3210]

The sum of knowledge owned by the members of this forum never cease to astonish me

 

OK, the keywords here are "good enough" and we're interested in relative temperatures, not calibrated absolute ones. A standard EGT sensor mounted with a bung is probably good enough. From what you say I guess an outside one is not. One problem with internal ones is that they will make an obstacle in the pipe, taking as much as 10%, or more, of the cross-section area. If nothing else, this probably means we want to put one in each downpipe. Any conclusions gathered this way could be said to not necessarily be valid without that obstacle in the downpipe, I guess.

 

Industrial thermocouples are available in a wide variety of sizes and mounting systems. For the temperatures expected in the exhaust gases, these generate a high enough voltage to drive meters directly.

 

A thermocouple is simply two different metal alloys which generate a voltage proportional to the difference between the hot junction of the two wires and the cold junction at the meter, in this case. Various alloys are available to handle various temperature ranges. Accuracies of + - 2% are ordinary, + - 1% is relatively easy to obtain.

 

A thin gauge wire type rated for 2,000-3,000 degrees F would impose very little obstruction.

 

Those interested can search Omega Instruments for explanations, and a wide variety of sensors, instruments and related products.

[raz]

Industrial thermocouples are available in a wide variety of sizes and mounting systems. For the temperatures expected in the exhaust gases, these generate a high enough voltage to drive meters directly.

 

A thermocouple is simply two different metal alloys which generate a voltage proportional to the difference between the hot junction of the two wires and the cold junction at the meter, in this case. Various alloys are available to handle various temperature ranges. Accuracies of + - 2% are ordinary, + - 1% is relatively easy to obtain.

 

A thin gauge wire type rated for 2,000-3,000 degrees F would impose very little obstruction.

 

Those interested can search Omega Instruments for explanations, and a wide variety of sensors, instruments and related products.

Thanks. The ones I refered to from my googling are thermocouples and I knew the actual wire junction is thin but it seems it's commonly put it a wider tube. But I misinterpreted some pictures, they are thinner than I thought. A fairly common bung thread size seem to be 1/8" and I didn't parse that before but now when thinking about it, thats just a couple of mm and the actual obstacle is maybe less than a third of that, like 1 mm. Maybe I was also mislead by someone else refering to how much of the cross-section area they take because that was in the context of a tiny kart downpipe.

 

So, a 1 mm obstacle half-way through in a 42 mm ID downpipe, that's roughly 1.5% of the area. Maybe that can be disregarded. Checking out Omega though, when searching for pipe plug probes I only find 1/4" NPT with probe size of 6.4 mm. That would be just the 10% I was guesstimating, if my dilettant skills count it right. OTOH, 6.4 mm is roughly (and even larger than) 1/4" so I wonder if that's really the size of the protruding probe (obstacle). Then there is unsheathed ones and they are really thin but I'm not sure how I would apply one. Probably the 1/8" kart type is good, cheap and easy enough. Stay tuned. Or even better: chime in, take the lead and shut me up!

[Guest Nogbad]

Mineral insulated sheathed thermocouples are available down to 0.8mm in diameter.

[Ryland3210]

Nanmac is a source of an even wider variety of thermocouples than Omega. They also specialize in open junction thermocouples for fast response measurements.

 

A common metal sheathed size is 1/8 inch diameter, in virtually any desired length. These can be mounted using a compression fitting similar to those used for tubing, but made of stainless steel. Available in either 1/4 or 1/8 NPT. One way to do this would be to weld a 1/8 NPT bung onto the exhaust pipe for the compression fitting. The thermocouple probe can be inserted to the desired length, and the compression fitting tightened to hold it there.

 

The metal sheathed variety is more rugged than the open type, but will read considerably lower than the open type because of the conduction path to the bung and farther through the thermocouple sheath. They also will respond much slower. The engine would have to be held at steady state for some time for the thermocouple to reach equilibrium temperature. I doubt this would be of any practical use, because of these issues. An open type, although more delicate, would be a better solution.

[raz]

A common metal sheathed size is 1/8 inch diameter, in virtually any desired length. These can be mounted using a compression fitting similar to those used for tubing, but made of stainless steel. Available in either 1/4 or 1/8 NPT. One way to do this would be to weld a 1/8 NPT bung onto the exhaust pipe for the compression fitting. The thermocouple probe can be inserted to the desired length, and the compression fitting tightened to hold it there.

If the sensor is 1/8", how could it fit an 1/8" NPT bung? No wait, let me guess. Plumbers, right? 1/8" does not refer to the actual thread size but to the inner diameter of a standard pipe with that outer thread?

[Ryland3210]

If the sensor is 1/8", how could it fit an 1/8" NPT bung? No wait, let me guess. Plumbers, right? 1/8" does not refer to the actual thread size but to the inner diameter of a standard pipe with that outer thread?

 

You're on the right track, except the 1/8 pipe size is nominal. The actual dimensions vary with the pressure rating of the pipe. In the National Pipe Thread system, schedule 40 is used for typical household plumbing and schedule 80 and 160 for higher pressures. The outside diameter of 1/8 NPT tapered threads is around .40 inches and 1/4 NPT around .52.

 

There is also a 1/16 NPT and sizes above 1/4 to quite large pipe sizes. Tubing, on the other hand, is dimensioned by the outside diameter.

 

In Europe, these tapered threads are sometimes refered to as "conical" on engineering drawings.

[Guest Nogbad]

You're on the right track, except the 1/8 pipe size is nominal. The actual dimensions vary with the pressure rating of the pipe. In the National Pipe Thread system, schedule 40 is used for typical household plumbing and schedule 80 and 160 for higher pressures. The outside diameter of 1/8 NPT tapered threads is around .40 inches and 1/4 NPT around .52.

 

There is also a 1/16 NPT and sizes above 1/4 to quite large pipe sizes. Tubing, on the other hand, is dimensioned by the outside diameter.

 

In Europe, these tapered threads are sometimes refered to as "conical" on engineering drawings.

 

Also, remember that US NPT and UK and European BSP/RP threads whilst as near as dammit the same in diameter have different thread pitches and are not compatible! (with the exception of a couple of sizes that will mate up!)

 

Incidentally, a standard brass compression coupling will stand exhaust temperatures, and has the advantage that it will not weld itself to your stainless boss in the way stainless to stainless threads tend to do.

[raz]

I'm still thinking about how to measure EGT in a Quick'n'Dirty™ but still Good Enough™ way. For some reason there are M10x1 bungs in the downpipes from factory. Unfortunately they sit in the wrong end of them, like 50 cm from the exhaust. If I put temporary sensors there, how inaccurate or slow would they read? Of course, it would be much more influenced by varying cooling and stuff but for short-term comparisons or on a dyno perhaps it might suffice. One spin-off effect is I will know the temp near my lambda sensor, which might be a good thing to know.

 

If this won't be good enough, I might still be able to use the bungs for some thin sensor that goes 45 cm up the pipe on the inside...

[Skeeve]

...For some reason there are M10x1 bungs in the downpipes from factory. Unfortunately they sit in the wrong end of them, like 50 cm from the exhaust. ...

 

Those are for measuring CO, I think. At least, that's what they were for on the ones with which Yamaha used to equip all their streetbike exhausts.

 

[Ryland3210]

I'm still thinking about how to measure EGT in a Quick'n'Dirty™ but still Good Enough™ way. For some reason there are M10x1 bungs in the downpipes from factory. Unfortunately they sit in the wrong end of them, like 50 cm from the exhaust. If I put temporary sensors there, how inaccurate or slow would they read? Of course, it would be much more influenced by varying cooling and stuff but for short-term comparisons or on a dyno perhaps it might suffice. One spin-off effect is I will know the temp near my lambda sensor, which might be a good thing to know.

 

If this won't be good enough, I might still be able to use the bungs for some thin sensor that goes 45 cm up the pipe on the inside...

 

Since you are looking for a relative but fast measurement, the location is fine where it is, in my opinion. I would be looking for an open thermocouple bead positioned in the middle of the exhaust stream.

 

For what it's worth, the 1/8 NPT tapered thread is 27 to the inch, not too far from 1 mm pitch, and it's maximum diameter is about 0.400", so its tapered leading end might fit into the 10 mm bung. A brass compression fitting would probably make a tight enough seal for temporary use. Just don't get carried away and jam it in too far.

[Ryland3210]

I'm still thinking about how to measure EGT in a Quick'n'Dirty™ but still Good Enough™ way. For some reason there are M10x1 bungs in the downpipes from factory. Unfortunately they sit in the wrong end of them, like 50 cm from the exhaust. If I put temporary sensors there, how inaccurate or slow would they read? Of course, it would be much more influenced by varying cooling and stuff but for short-term comparisons or on a dyno perhaps it might suffice. One spin-off effect is I will know the temp near my lambda sensor, which might be a good thing to know.

 

If this won't be good enough, I might still be able to use the bungs for some thin sensor that goes 45 cm up the pipe on the inside...

 

If you go to the Nanmac website, they offer a thermocouple designed for gas temperature measurement with microsecond response time including 1/8 NPT fitting and electrical connector. Look up:

 

PENCIL PROBE "ERODING" THERMOCOUPLE - E12 Series