80-100hp/no EGT issues/4-6mpg> miracle inj installed Friday

[Vaughn MacKenzie]

Blake is also running BFG A/T 315s which generally hurt mileage a bit. Vaughn

 

[DLeno]

Ok here's the story. Note that at the time this data was collected, the BD X monitor was not available (thats why it is not included). And since the data was from early 2002, some mfgs may or may not have made changes.



The first part of the test was to insert each probe into a 1700 degree blue torch flame (controlled lab experiment). The time it took for each thermocouple to register 1700 degrees (in milivolt output) was recorded. Here's the data -- the number of seconds it took for each probe to register the correct milivolt output.



SPA technique: 18

Westach: 35

autometer 37-39 (three models tested, 5244, 5249, 5243)

IssPro: 47-48 (three models tested, R630, R607T, R607G)

Pricol C210-001: 78

Hewitt: never reached 1700 degrees after five minutes



What's important is the comparative data, as the experiment itself doesn't correlate very well to real exhaust manifold installations. Its just that there is a clear speed winner followed by a group of acceptables, followed by, well, err, uh, ... . just be aware of how slow your probe may be. DONT interpret this data to mean that it takes 78 seconds for a pricol probe to register the correct EGT in your truck. thats NOT what this data says. it just says that the pricol is about 60% slower to respond than most, and that the SPA is the speed demon.



Now in addition to the probe speed, the gauge face itself has to respond to the signal provided b the probe. What follows is the number of seconds it took for the gauge face to acheive the right reading when presented with a fixed milivolt input:



SPA: less than 1

IssPro digital: less than 1

Westach: less than 1

Autometer 4365: 4 seconds

Isspro R607G: 4 seconds

Isspro R607T: 4 seconds

Pricol C205-250: 5 seconds



Now this data you can interpret directly. you need to know that if you gauge reads 1500 degrees, your EGT has probably been that way already for a few seconds...

 

[GWoody]

Thanks DLeno, appreciate the info, sorry guys, didn't mean to sidetrack the thread.

 

[Badunit]

Now this data you can interpret directly. you need to know that if you gauge reads 1500 degrees, your EGT has probably been that way already for a few seconds...

Interesting data. On this last point, that is true if the EGT's climb very fast from a low EGT to 1500 but maybe not the case if they were in a slow steady climb or were close to 1500 to begin with before they spiked. In those cases, the gauge might be tracking pretty close to the truth.



I'm a little curious about why the probes were so different. Any idea why that was?

 

[DLeno]

... On this last point, that is true if the EGT's climb very fast from a low EGT to 1500 but maybe not the case if they were in a slow steady climb or were close to 1500 to begin with before they spiked. In those cases, the gauge might be tracking pretty close to the truth.



I'm a little curious about why the probes were so different. Any idea why that was?

exactly right. such as during an acceleration when you have added fuel and you are driving by the gauge because you still have the stock huffer.



The reason why there is so much difference in probes is because of the way they enclose the junction itself. In general, thin probes will be faster than thick ones. I think there is a type of ceramic material in the tip itself, transfering heat to the junction itself but insulating it from the block. So if you have a thin layer of ceramic covering the actual junction, the probe will be fast. too much insulating ceramic means the probe will be too slow or even not accurate -- as in the Hewett

 

[WPeschel]

What about Auto meter good or bad?

 

[geusterman]

I would like to drag this back to Chad's comment about Wyo Tech's take on post turbo. The flame tip of a torch can be hotter than the torch body... (my thought here)... with the flame giving a specific temp measurement that may not be telling the whole picture of piston risk. They apparently think that one inch past turbo tells a more realistic story of what the piston temp really is without interference of flame spikes. Now, having had 4 hotrod Cummins (12V 5. 9, 12V 8. 3 and a 24V and all post-turbo) I have learned to just not spend a lot of time in the higher temps because other things start to indicate danger... like water temp and engine compartment temps (melting wires etc) and transmission temps. Never hurt one that I know of.



I think the issue regarding these injectors is that there is a definite increase of fire and temps and whether or not it is an "issue" may depend on DPU's definition of what constitutes danger rather than whether or not there is an increase in temps. I would like to steer the thread to what temps actually caused meltdown from actual experience. Thanks, George

 

[DLeno]

george. Indeed, you have nailed the crux of the whole issue -- correlating measurements with engine failures. unfortunately, not many are willing to do that in a statistically meaningful way because that requires actual engine failures. And those engines that have failed cannot be meaningfully correlated with an EGT measurement because of measurement variability, operating conditions, etc.



of course the tip of a blue torch flame is the hottest part, and doesn't correlate well to piston temp. that wasn't the purpose or result of the study. The purpose was to show how slow the various thermcouples responded when in the presence of a high temperature. thats it. The actual numbers (seconds to reach temp) should not be interpreted literally. The point is that some are slower than others, and that has to be part of your correlation work when you are trying to infer something about piston temp when you measure EGT under dynamic conditions.



Like when you are in an acceleration run and you want to limit the amount of time you spend above a certain temp. If you have a slow probe and a slow gauge, you are spending more time than you think at the temp of interest.



my only point is that when we push the limits of EGT, our measurements are more meaningful when we minimize the amount of correlation work that has to be done. The more direct the measurement, the better. If you go post turbo, you have to do a lot of work to characterize the temperature differential across the turbo and try to correlate that with preserving the mechanical integrity of piston. There are no completely objective, precisely accurate models that tell us exactly what temp to measure with what instrument in what location and how that correlates to actual piston temp or the statistically valid probability of failure. Nor is there anything telling us how hot the piston face can get without putting the body in danger of changing shape.



All we can do is measure the temperature of the air somewhere after it exits the cylinder and try to infer what that means. the more removed the measurement from the parameter of interest, the more risk you take on that you are in error. And the more you understand about the accuracy of the instrument you use, the better equipped you are to do that correlation work. At the end of the day, all EGT measurements consist of correlation work (or trust in someone else's) combined with caution relaxed only by your own desire to take on risk.



Correlation work that can be trusted is that of Cummins, who performed a great deal of pre-post study so that they could declare, with science behind them, what a permissible post-turbo temp was. They chose post-turbo for a variety of reasons, but they correlated that with pre-turbo numbers which they found to be acceptable: 1250 degrees for the 2nd gen 24v engines. 300 degree differential temperature on the stock turbocharger, which means 950 degrees post turbo. For stock production engines operated as intended, there's no reason you cannot use post-turbo. yes, the 3rd gens have been beefed up a bit in the allowable temperature department, as already pointed out.



WPeschel: Autometers are among the good group. Frankly (the way I interpret this data) the only scary one is the Hewett probe. Most of the others with the exception of the pricol (slower) and the SPA (fastest) are all about the same. They all work, and they are all useful for the indended purpose. They're just not all the same in their behavior, thats the point.

 

[DanDee]

Geusterman...



Been a month... . those injectors get any better? or did they come back out??



Inquiring minds want to know!



Thanks, DanDee

 

[BIG BOB]

Geusterman...



Been a month... . those injectors get any better? or did they come back out??



Inquiring minds want to know!



Thanks, DanDee

George email me today. He is in DC until October 21st. He said that his last post is still his views at this point.

 

[DanDee]

Thanks... Guess we will have to wait a while longer to see how it all washes out!



Thanks again, Dan

 

[BIG BOB]

Here you go George, I think you were looking for this thread!



Give us the results!

 

[geusterman]

Thanks BB. Well boys, finally got the flatbed tanks set up right after being gone 7 weeks of the last 2 or so months. Got the tail lights going this afternoon and went right to the diesel pump in Burlington WA and filled away. A hundred dollars later and the fuel licking the threads of my custom aluminum fill neck I eased on over to Starbucks to spend another $5 and then down the freeway to Lynnwood with speedo set to 70. Tires were overinflated (left from having camper on) to 65 and 80 rear. No wind, per se, but raining half the way. Puter was between 16 and 18 mpg. Turned around and came back with some 60 to 65 mph for about 1/3 the way. Pulled back to the identical spot with a slight lean to the inlet and refilled 5. 193 gallons against 86 miles equaling 16. 5 which is two mpg more than factory. Originally, with the claim of 4 to 6 mpg improvement and most trucks running 20 mpg and more the $2000 injectors made sense over 50K miles with extra power thrown in. For some that may be just fine but not for me. More smoke and EGT increases on top of it tells me that I am going to put the stock injectors back in and if I want power I will buy an EZ. One the DPU injectors is ticking madly and the white smoke on startup has become major. I left the rig at Burlington Kar Mart for the TSB on white smoke (as well as losing the auto door lock and a TSB for combining the brake/turn signal in one bulb for the flatbed lights). I would like another highway run when I pick it up tomorrow. I thank Josh at Dishman for his patience (kind enough to not bill me until I fixed and tested the truck) and all you waiting for results. If I get a road test with the TSB as well I will post that soon. George

 

[Greg Boardman-RIP]

I have had 3 cummins.

One was an 01 450 hp

Two is my truck, has had injectors, boxes, now a turbo, ... .

Three is my wifes truck.



all three of them get or got 16 mpg +or- 1mpg since new. The injectors didn't help. The boxes didn't help. The turbos didn't help. The only thing that helps is going slow. Under 2000 rpm's is it. IF you can cruise at 2000 r's then you will see upwards of 18 or 19 mpg.



JMHO



Greg

 

[Rotty]

(Long)I agree with greg boardman on the milage issue. I have been regulary getting 17-17. 5 (empty) hand calc running at 65 on cruise mostly interstate. I tried a little experiment last weekend and set the cruise to 60 for a round trip to seattle.

Hand calc was 19. 2 @ ~450 miles round trip. Now I feel lucky with these numbers as I have heard fuel mpg all over the board.

I definatly believe the if you want more fuel mileage then have better right foot control (impossible with a 600 ). Computer milage was 19. 9 to seattle and 20. ? on the way back.

I did reset the computer when I was traveling down the interstate and it went up to 26+ ( I wish it was actually getting this) at first then settled down after 50-75 miles.



This little experiment only added 15 min each way to my trip, but I definatly had some people passing me at 70-80 mph. I was mostly in the extreme right hand lane, and it felt a little dangerous at times.



Now I have a 04. 5 4wd, quad long, 6 sp, 3. 73, 4300 on the clock.

I will be doing my first tow Oo. thanksgiving weekend and will check my mileage. With my old 99 dodge 5. 9v8 quad 4wd 4. 11 gasser I would only get 8-10mpg towing and could barely get to 60mph.

 

[fbaurley]

One thing if I remember correctly about Type K thermocouples is the response time from around 0 degrees celsius to there max of 1500 Degrees celsius is around 7 to 8 seconds if not longer, so if you convert those numbers to degrees F, and and figure out what a normal operating temp is , and what a heavily loaded exhaust gas temp is you can figure out how long it will take the thermocouple to detect this increase in exhaust gas temp. When the thermocouple is mounted post turbo, by the time you take this response time into effect it could have already cooked your turbo.

 

[KYLE4]

I'm gonna check these out I'm a vendor and I have a banner on this site www.extremedieselperformance.net

 

[grfraz]

as far as the thermocouples refer to: Technical Topics (Issue 34)

Service/Parts Update: Exhaust Gas Temperature Gauge Evaluation

(Issue 34, pages 20 - 29)



this can also be found in the sample article section of this site.

 

[DLeno]

Issue 34 is a good one, and you will see me refer to that in issue 47 during a discussion about measuring EGTs for very short dyno runs. Be careful not to interpret the rise times in issue 34 literally; yes the fastest was on the order of 7 seconds, if memory serves, but that was to 1600 degrees I think. Anyway, the important thing is that some are faster than others. For my work, the importance was that the SPA thermocouple responded in half the time as most of the others. It was like the SPA was head and shoulders above the rest in speed, then the bulk of the market followed at about double the rise time, then at the very end was a couple of devices, one especially, that were very very slow. One didn't even acheive the target temperature. you'd be surprised who the slow ones were.



keep in mind that this is pretty old data, so the themocouples tested in the article may have been improved, I don't know.



There are four major sources of error in a type K pyrometer:



1. gauge speed. the gauge itself has to respond to the signal generated by the thermocouple. Analog gauges can be fast, below or about 1 second, but watch out; some respond slower than others. I think the Westachs are among the faster gauge faces out there.



2. Thermocouple speed. Since the primary K junction has to be protected (usually encased with some ceramic type material), the protection itself has a thermal rise time, mass, etc. Very Heavy protection generally means very slow; you can see this evidence by inspecting the tip itself. As issue 34 points out, some are slow and some are fast.



3. gauge calibration. All type K pyrometers measure the difference in temperature between two locations. you are not measuring the absolute temperature of the exhaust manifold, for example, you are measuring the difference between this point and some other point. how a gauge deals with this second point, and the error it introduces, is important. there are three ways I know of (to deal with this error) that are used typically in gauges available to us:



a) the secondary junction is placed under the hood. The Cheaper thermocouples do this, and is evident by the use of ordinary copper wire under the hood, typically emerging several inches from the main probe. No matter where the second K junction (boundary between "K" wire and copper wire) is, the gauge has to make an assumption about the temperature of this junction. This is because the gauge responds to the difference in temperature between the two locations. whatever assumption this is will be wrong by an amount equal to the temperature variation under the hood. for example, if the gauge assumes 100 degrees under-hood temp, and its a cold day, your EGT might be 50 degrees cooler than the gauge says. or, if its a hot day and you are towing hard, under hood temps might get very hot, and your actual EGTs might be 75 degrees hotter than the gauge reads.





b). the 2nd junction is placed in the cab. This is done by the use of very long type K wire (the unweildy and stiff wire) that extends into the cab. In this case, the temperature of the 2nd junction is much more controlled, and the gauge itself can be calibrated with the assumption that if you are driving the truck, the in-cab temperature is probably 75 degrees or so. So all the gauge has to do is assume that the 2nd junction is at 75 degrees, for example. The gauge reading will be wrong, but only by an amount that the cab temperature differs from 75 degrees. I think Westach uses this method; others probably do as well.



c) electronic temperature compensation. this is overkill for all practical purposes, but it is the ultimate in accuracy. The SPA gauge does this -- inserts circuit that actually measures the temperature of the 2nd junction, much like a Fluke meter would, for example, and calibrates itself against that. This produces extraordinary accuracy within 1 degree or less.



4. install location of the primary junction. what ever location you pick, there are errors. if you go post turbo, then distance from the turbo will matter and there will always be some correlation as to what a safe maximum is. if you go pre-turbo, then the location on the manifold matters, the manifold itself matters (stock, ATS, whatever), etc. there are lots of recommendations out there and many experiments have been done, and still not all agree on the best location for mounting the thermocouple.



well that was a bit long. sorry. the point is to understand the souces of error and know your gauge's accuracy. In general, most of us try to eliminate as much error as possible within our means, and get as close as possible to the physical evidence that represents the vulnerability we are out to mitigate. For most of us, that means, a fast gauge, minimally an in-cab location of the remote K junction, and a fast thermocouple mounted in the hottest part of the exhaust manifold.