DIY Dyno Tuning!

[RobertISaar]

EDIT by EagleMark for new thread.

i really need to move somewhere so that i can build a load cell and run engines at all times during night and day at various loads. basically, a dyno without any kind of torque measurement, but you'll be able to see the power changes through the change in speed of whatever load you have connected to the powertrain. then when done with that, go to an actual dyno for final numbers.

[EagleMark]

i really need to move somewhere so that i can build a load cell and run engines at all times during night and day at various loads. basically, a dyno without any kind of torque measurement, but you'll be able to see the power changes through the change in speed of whatever load you have connected to the powertrain. then when done with that, go to an actual dyno for final numbers.I found one last year for $300! When I got there it was sold and being loaded. All it was, was a drivability rollers with ramps above floor, I have seen many in floor. It also had an old big box (forget what they were called?) with timing light, vacuum gauge, dwell, tach and a scope. All I wanted was the rollers! Had a big flywheel and five clutches with plates and pressure plate for load, was hand controlled by big lever so you would need someone to operate it while someone drove for load....

[RobertISaar]

was hand controlled by big lever so you would need someone to operate it while someone drove for load....

spouses make excellent lever pullers. :happy:

[gregs78cam]

i really need to move somewhere so that i can build a load cell and run engines at all times during night and day at various loads. basically, a dyno without any kind of torque measurement, but you'll be able to see the power changes through the change in speed of whatever load you have connected to the powertrain. then when done with that, go to an actual dyno for final numbers.

Not to hijack the thread, but......
I have an old 'bread van' chassis, with a rockwell rearend. I have thought about fabbing some rollers to bolt onto the hubs and using the brakes connected to master cylinder and lever.

[93V8S10]

Not to hijack the thread, but......
I have an old 'bread van' chassis, with a rockwell rearend. I have thought about fabbing some rollers to bolt onto the hubs and using the brakes connected to master cylinder and lever.
With garden hose over each hub for cooling... :)

[gregs78cam]

With garden hose over each hub for cooling... :)

:happy:I like the way you think!

[RobertISaar]

DIY dyno planning thread? or continue in here, since it would be quite relevant?

[EagleMark]

Already started new thread so we could get back to spark tuning in one and home made dyno in another...

[1project2many]

I've owned a dyno. I was given a driveability and IM testing dyno a few years back. The rollers were very small diameter and I'm sure any substantial power would have been searching for traction. Software was somehow locked to the state where the dyno came from but chances are that could have been fixed. And max HP according to manufacturer SPX was somewhere around 120. I donated it to a local and asked that if they ever get it running I be allowed from time to time to use it.

If you're in an area where old paper mills exist you might be able to find some big rollers cheap. Up and down the east coast old mills are closed down and being salvaged. A friend of mine got a couple of 24" diameter rollers for $600. He said the money didn't so much buy the rollers as pay the scrap metal guys to load 'em on a truck and deliver 'em to his place.

Upstate NY might have a few Mustang emissions dynos floating around. NY invested millions in an IM240 program that never materialized. Dynos were sold off cheap with hardly any time on them. I know of a Sunbird turbo guy that was trying to get one several years back.

Some realy smart guys on DIY once talked about building a strain gauge out of the driveshaft. Determine the modulus of elasticity for the shaft then measure how much twist can be introduced to work out torque. Calculate from there to obtain power. That conversation digressed into discussing how much twist can be introduced in a 426 Hemi block making 1000 hp (a lot, it's been measured) and nothing was ever built.

ECM can be used to measure time, rpm, delta rpm. Mass of car is known. EGT, coolant temp, oil temp can all be monitored from in cockpit or sent to laptop DAQ (heat = wasted energy). I have actually thought about writing code which can monitor performance based on a few of these values to help with tuning.

[EagleMark]

This isn't the first time I've seen a truck axle used...


http://youtu.be/7g4jPoyT0SQ

[PJG1173]

thats just scarry...

[RobertISaar]

those crazy scandanavians...

anyways, while rolls would work for an interia dyno, what options are out there for a DIY static-state dyno? those would be considerably more useful for tuning spark when not at WOT.

i've seen water and oil pumps mentioned, seems reasonable, seen eddy-current mentioned as well(but need to be able to dissipate HUGE amounts of heat), not sure what else to google for.



my theory(and it could be(probably) wrong), would be that say you're "cruising" with some type of load at 1600RPM and ~50kPa, since those are beleivable numbers, and lets keep the ECM in closed loop for this, since spark changes can necessitate fuel changes as well. now to find the optimal amount of advance for a specific RPM/kPa combo (or RPM/airmass, for MAF applications), when you make a change that increases output, you'll see MAP drop due to the power required to keep the vehicle "rolling" staying the same but output has increased.

i'm sure there is something wrong with that method, but i can't put my finger on it.

i'll give an example of extreme mesaures, since it's generally easier to compare things when they are at their extremes. let's say with 20* of advance at 1600RPM and 50kPa, you're making 10HP(32.8lb-ft at that RPM). now let's add 2* of advance without changing the load. now let's say we're making 11HP at 1600/50 (36.1 lb-ft). now, since the engine is at the same speed and no gearshifts have happened and the load itself hasn't changed and throttle hasn't moved either(or IAC) (and anything else i haven't mentioned), all other things being equal, won't the change be forced to show up as a difference in manifold pressure? or will the increased power from the engine actually speed up whatever load is applied?

[EagleMark]

Couldn't you do that with say a fixed throttle so TPS % stays the same and increase spark while logging and see changes in MAP and Speed? Would require long flat straght road... could output data to spread sheet...

How else could it be done without rollers?

[RobertISaar]

those kinds of road conditions simply don't exist around here.

however, if you were to live in an area with no elevation change, little to no wind, unchanging road surface... then as long as the throttle stays perfectly still...

then you should pick up speed with more power... i think. i mean, you'll be making more power without the throttle being moved, which means you'll achieve a higher speed before wind/tire resistance prevent more acceleration(steady-state).

assuming a manual trans or a locked TCC with no slip in an automatic, RPM will also obviously rise if your vehicle speed rises. and to get back down to the speed you were at before the spark change, you would need to close the throttle slightly, which will increase vacuum (decreased MAP kPa).



maybe i'm just rambling though. :laugh: but it seems like increased power will increase MPH, which will allow you to drive with less throttle, meaning more vacuum(at the same RPM), meaning less fuel required to sustain a given speed, meaning..... better MPGs.

of course, that's only looking at the MPG side of things, in the end, it's making more power using the same amount of fuel, meaning better power AND better fuel economy.


so i guess my above theory was most of the way there, but didn't properly explain WHY you would see certain changes.



in the end: if you don't move the throttle, and don't increase load, yet a change you make causes an increase in vehicle speed(or roller speed, or brake speed or whatever), that change caused more power to be generated.

[1project2many]

i'll give an example of extreme mesaures, since it's generally easier to compare things when they are at their extremes. let's say with 20* of advance at 1600RPM and 50kPa, you're making 10HP(32.8lb-ft at that RPM). now let's add 2* of advance without changing the load. now let's say we're making 11HP at 1600/50 (36.1 lb-ft). now, since the engine is at the same speed and no gearshifts have happened and the load itself hasn't changed and throttle hasn't moved either(or IAC) (and anything else i haven't mentioned), all other things being equal, won't the change be forced to show up as a difference in manifold pressure (http://shop.ebay.com/i.html?_nkw=manifold+pressure)? or will the increased power from the engine actually speed up whatever load is applied?

Your cause and effect are getting tied up. Witha the preconditions above, if you make spark changes and begin to measure 11 HP, then the alter throttle angle to bring power back to 10, you'll see the expected change in MAP. But if the throttle angle doesn't change neither will MAP. You haven't altered anything on the input side of the engine, only made more power from the available fuel and air.

But you're on the right track, though. Lower MAP / throttle angles does indicate more efficient use of fuel. However there is a big consideration which goes unnoticed by most people. When MAP is low, pumping losses are greater as the piston works harder to draw air around the closed throttle. Pumping losses work against engine efficiency by requiring the engine's output to be put back into keeping the engine running. Diesels by nature are inherently more efficient as they regulate power through wide changes in fuel delivery, not by limiting airflow. A diesel intake has no throttle blade. And smaller engines for a given job are often able to be more efficient throughout a wider range as they will operate with the throttle angle open more thereby reducing pumping losses. The interpretation should be this: Don't assume the best way to maximum economy is through maximum spark advance. In some cases an engine will produce better overall efficiency with slightly less advance. This seems most common in cases where the control system is forced to run in closed loop with AFR based on O2 readings. Older systems or lean cruise / open loop cruise strategies still benefit from higher AFR and more advance.

[RobertISaar]

how about my "revised" explanation?

and i would think this takes throttle losses into account already, since there is a load to be tested against? if the engine did lose more power through the throttling loss than it gained from the difference in advance, wouldn't that slow down the load/rollers?

also: EGR counteracts throttling losses from what i've seen, since the engine makes less power with already combusted air/fuel in the chamber, so you have to open the throttle further(reducing the amount of throttling loss) to maintain the same output. IIRC, it will also push the engine a little closed to it's peak BSFC zone as well.

[1project2many]

1project, could your engine problem have been prevented if you ran higher octane fuel?
Maybe? I read statements which were correct but I could not deduce how cohesive the bond was between them. :)


and i would think this takes throttle losses into account already, since there is a load to be tested against? if the engine did lose more power through the throttling loss than it gained from the difference in advance, wouldn't that slow down the load/rollers?

Power measurement will take throttle losses into account. But unless you're measuring the rate of fuel used or waste heat generated you're not going to know if 11hp at 50 kpa and 25% throttle angle uses less fuel than 11 hp at 45 kpa and 22% throttle angle or 11 hp at 55 kpa and 27 % throttle angle. I read a question implied in your post as "If I get the same hp reading with less throttle and less MAP, doesn't that mean I've found more efficient settings?" I'm saying "Not necessarily. The engine may turn out to be more fuel efficient at the larger TPS and MAP values. This is a case where typical dyno tuning doesn't always tell the whole story."


also: EGR counteracts throttling losses from what i've seen, since the engine makes less power with already combusted air/fuel in the chamber, so you have to open the throttle further(reducing the amount of throttling loss) to maintain the same output. IIRC, it will also push the engine a little closed to it's peak BSFC zone as well.

I don't know why EGR would improve torque production. BSFC follows the torque curve and I can't see EGR changing that or improving torque significantly. I'm open to reading about it though.

EGR is able to counter pumping loss some but of course we can only R so much EG before we create a driveablilty issue. IME EGR is best for introducing carbon deposits and scale buildup into an engine over 100k miles of operation. It would be awesome if we had a way to measure NOx and only use EGR when needed. Reducing timing during cruise and choosing a cam with a small amount of overlap are also viable methods for keeping NOx down.

[RobertISaar]

Power measurement will take throttle losses into account.

wouldn't the speed of some load being turned by the drivetrain be considered a power measurement?


But unless you're measuring the rate of fuel used

this, i've done via the datastream, though i don't know how accurate it is compared to how accurate it would need to be for something like this.



I don't know why EGR would improve torque production. BSFC follows the torque curve and I can't see EGR changing that or improving torque significantly. I'm open to reading about it though.

i wasn't trying to imply that it would increase torque on it's own, only that because EGR being introduced, more throttle would be required to maintain the same amount of power, which generally(and i may be wrong here) moves the operating point of the engine closer to it's peak BSFC.



this subject seems to be tough to wrap my head around...

[EagleMark]

this subject seems to be tough to wrap my head around...It's a subject car manufactures try to fine tune every year and look how much money and time they spend on engine dyno time? We are trying to come up with a cheap way that is never going to compare with what they have already done for that years engine.

[1project2many]

My reply disappeared?


wouldn't the speed of some load being turned by the drivetrain be considered a power measurement?

Correct. I was agreeing as in "Yes, power measurement will take throttle losses into account."


i wasn't trying to imply that it would increase torque on it's own, only that because EGR being introduced, more throttle would be required to maintain the same amount of power, which generally(and i may be wrong here) moves the operating point of the engine closer to it's peak BSFC.

I have never heard this statement. There seems to be an overlooked argument here and I'd like some time to consider it fully.


It's a subject car manufactures try to fine tune every year and look how much money and time they spend on engine dyno time? We are trying to come up with a cheap way that is never going to compare with what they have already done for that years engine.

Mark, one advantage we have over OEM's is the ability to spend substantial time with one car learning the ins and outs of it. OEM's need a fast and relatively accurate way to generate thousands of driveable calibrations. We only need to make one.

I have seen some computer modeling related to tuning and I am very impressed with it but for hobbyists such as ourselves this is still an art form. There are some cool things which could be adapted from the OEM world such as ion based knock detection and predictive management for boost control and power output but for the most part we are left to our own senses and interpretation when it comes to getting it right. IME dyno tuning is a very small part of getting everything right. With large power to weight cars it's almost mandatory. Some of these outfits aren't street legal and even if they were it's nearly impossible to get data accurately. But for most street vehicles you can do a bunch of good with careful, effective street tuning.

[EagleMark]

My reply disappeared?

I didn't do it, I was asleep, maybe you did like me on long posts and forget to submit reply?


Mark, one advantage we have over OEM's is the ability to spend substantial time with one car learning the ins and outs of it. OEM's need a fast and relatively accurate way to generate thousands of driveable calibrations. We only need to make one.
I don't know from what time period your talking, or your definition of "fast and relatively accurate"? Maybe that was the way during this OBDI period. There are still huge differences in 1227747 cals. But I can say from talking extensively with a EFI engineer from Detroit during the last 10 years that after the engine has done thousands of hours on dyno and days of time on dyno in wind tunnel he hated spending a week in Colorado driving same car from flatlands to Denver and back down to flatlands fine tuning. Then when done it still had to pass dyno emissions test. All on same engine and model of car.

One advantage we have is we don't have to pass emissions dyno testing in end... I don't know how many MPG or HP could be gained without that restriction.

I can say that when PE is delayed, like we found on some 16197427 PCM for a minute? Is basically turning it off! I turned off PE on my 1227747 the other day to check BLM for larger area of cells and it still came up accurate, I've done this before but never really given it a WOT test from stop and the differnce in power was HUGE!

I also played with MPG history table I made in TunerPro to see where my brick (Suburban) had best highway MPG and did tests from 50 to 75 MPH back and forth on same stretch of interstate and found 55-65 to be about the same. 65-75 was a big drop. So I cruised at 65 and added 2 degrees spark from stock and saw no difference in MAP or MPG reading. I did see a drop in MAP and MPG when I took away 2 degrees. But when I enabled highway lean cruise MPG did go up about 10%, much to my surprise MAP went down (vacuum up) and TPS went up? So how does TPS increase yet vacuum also increases? Just from leaning out AFR ratio? It also adds 3 degrees spark for leaner mixture.

This is not a totally accurate scentific test but it is all I have right now for DIY Dyno...

[1project2many]

I don't know from what time period your talking, or your definition of "fast and relatively accurate"? Maybe that was the way during this OBDI period. There are still huge differences in 1227747 cals. But I can say from talking extensively with a EFI engineer from Detroit during the last 10 years that after the engine has done thousands of hours on dyno and days of time on dyno in wind tunnel he hated spending a week in Colorado driving same car from flatlands to Denver and back down to flatlands fine tuning. Then when done it still had to pass dyno emissions test. All on same engine and model of car.

Getting a calibration online in 18 mos is fast. In the old days it took years to get a cal market ready. There would be many calibration engineers in the field plus lab/shop guys developing and fixing calibrations (why do you think there are so many?) But OEM is getting the cal ready for everyone and able to operate in every condition. They're not spending the time on one specific vehicle working out the ins and out of that particular car/truck. That's the advantage we have. We don't have to get the program moved forward by a certain date, or have to stop playing with a cal when we reach a target mileage/power/emissions level.

I did have to deal with emissions testing for quite a while. The states around me were actually performing dyno sniffer testing so out of stat customer cars had to pass. That's where I got the dyno. :)

[RobertISaar]

Correct. I was agreeing as in "Yes, power measurement will take throttle losses into account."

I have never heard this statement. There seems to be an overlooked argument here and I'd like some time to consider it fully.

ok, i thought you were implying that a roller/load test wouldn't suffice for evaluating changes. good to know.

and i state this due to seeing a few BSFC charts before, it SEEMS to be a common trend, though most of the charts i've seen were of boosted vehicles.


But I can say from talking extensively with a EFI engineer from Detroit during the last 10 years that after the engine has done thousands of hours on dyno and days of time on dyno in wind tunnel he hated spending a week in Colorado driving same car from flatlands to Denver and back down to flatlands fine tuning. Then when done it still had to pass dyno emissions test. All on same engine and model of car.

i've sent quite a few emails back and forth with a former GM powertrain engineer myself, i guess they actually split up the calibration guys into two teams: 1 for engine control, 1 for transmission control. it's gotta be tough to work with multiple people on the same calibration, i would think.

[EagleMark]

There's so many working on differant areas before final driving, testing, adjusting, then the final dyno emmissions... I would think it has become more stream lined then early days since they have so many models to work from?

[daleulan]

And OBD. Around half the code in many ECMs are for the newest OBD requirements. Intrusive tests, passive tests, lots of crazy stuff. Those can take a long time to calibrate.

For low-cost, you might be able to find some used dynos, maybe a waterbrake or a nice eddy current unit that can't do all of the newer tests. I used to tune without a dyno (on-road tuning) but spark timing was almost impossible to map accurately. That part of the curve is very flat and you get minor changes in power output. The barometric pressure and humidity made it even more difficult to see - a three degree change in timing would be swamped by a 20% change in humidity. I used to use a handheld 5-gas analyzer and tune by precatalyst NOx. There's a point around 2000 or 2500ppm where you advance the timing and NOx goes up rapidly but power doesn't. Plus when you get close to knocking the NOx skyrockets and you know when you're getting close, even before you hear it.