Building an engine dyno.

[Boxersix]

Originally Posted by Dessertrunner

Adam do you find engine dyno's handy, I read one guy say "The engine dyno always comes before the car dyno" what do you think?
Tony

It's personal preference for tuning really, as both can provide proper load for engine tuning. The benefit I find with an engine dyno is the ability to make changes and adjustments to the engine hardware far easier than if it were in a car strapped to a chassis dyno. For example, camshaft choice on a purpose built v8 can be finiky and requires multiple cams to test and find the right choice. On an engine dyno it's clearly easier to make those cam swaps without much down time.

The way we have ours setup is in a modular fashion. the engines are mounted in a mobile cart that has adjustable supports and fixtures to fit multiple motor makes/models. We then have an array of adapters that bolt directly to the flywheel/flexplate on the motor and connect to the water brake via a short(adjustable ~12-14") "driveshaft". The cart rolls up to the dyno and is pinned/locked in place to the actual dyno frame.

This all avoids the multiple, and complicated array of bellhousing style adapters. Starter motor engages a ring gear affixed to the brake itself, so there's never a worry about having to find a way to get the engines actual starter motor to work with the dyno.

[Dessertrunner]

Originally Posted by Boxersix

This all avoids the multiple, and complicated array of bellhousing style adapters. Starter motor engages a ring gear affixed to the brake itself, so there's never a worry about having to find a way to get the engines actual starter motor to work with the dyno.

The starter motor idea is a great one, I will do it the same way, thanks for the suggestion.
Tony

[Boxersix]

If you would like more pictures of our setup just let me know. I'll be at the shop putting a subaru motor together and will take photographs of the entire setup if needed.

[Dessertrunner]

Adam would love a couple of photos to help me.
Also I need to get my head around how the dyno load works as regards load managment.
As I understand it I adjust the load generated by the retarder as the engine revs up, what I don't get is do I increase the load on the retarder or decrease it? Or to add confusion is it both ways?
Tony

[dynomatt]

In an ideal world, you would want the dyno to allow controlled power runs and also tuning at static revs.

A lot of dynos do power runs...this is a controlled run from low revs to max revs. You'd want the dyno to lag behind the engine torque so that the engine can still accelerate in a timely and controlled fashion. If you could build a timer so that the retarder would let the engine increment revs in 10 rpm per second...that might make a nice power run. Ie, from Idle (700rpm) to max revs (7000rpm) would take 63 seconds...as I read that, maybe you want it to increment 20rpm per second. The retarder could then vary its load in such a way that if it exceeded 10rpm per second, it would apply more load, and release load if it was taking longer than 10rpm per second.

probably am not making any sense.

For tuning, you want to be able to dial in load to let you hold various throttle positions at various revs. Again, you could build a circuit that allowed the retarder to vary load to keep it at a certain RPM.

Full load on the retarder will stall the engine. Finding the balance will take time.

I would think

[Dessertrunner]

I think you are right Matt, its going to be a bit of suck it and see. As I see it the graph will be over rev's not time so it less critial to increase the revs at a constant rate.

The frame is built and shafts for the retarder are being made, belt drive is ordered. PLC is here so its all coming together.

A question for all you old hands out there. What will exhaust gase temp on each cyclinder tell you & how can you tell which cyclinder is developing the most power,,,, is it the hottest exhaust?
Tony

[TomsSVX]

Originally Posted by Dessertrunner

I think you are right Matt, its going to be a bit of suck it and see. As I see it the graph will be over rev's not time so it less critial to increase the revs at a constant rate.

The frame is built and shafts for the retarder are being made, belt drive is ordered. PLC is here so its all coming together.

A question for all you old hands out there. What will exhaust gase temp on each cyclinder tell you & how can you tell which cyclinder is developing the most power,,,, is it the hottest exhaust?
Tony

EGT's don't necc tell you which cylinder is making the most power... EGT's can change based on ignition timing, cam timing, and fuel quantity.

Tom

[Trevor]

Tony,

I will be accused of boasting but what the hell. I can advise you on the basis of having knowledge of the subject. Many years ago my Company, Relays NZ Pty Ltd was commissioned by Repco Ltd, to design and supply the control gear for a wheel dynamometer they were intending to produce locally.

From your comments I can see that you know exactly what you are about due to your reference to a strain gauge , torque and RPM. You are not building a dyno on a commercial basis and do not requite any form of direct read out, as is required to impress paying customers, and as was required by Repco.

You only have to register engine speed in RPM, and torque in Kilograms Meter, or Pounds Foot as is held/resisted by the retarder/brake. You can draw up a simple graph from which you can then directly read off horse power. This is in point of fact the correct way to achieve properly reliable results.

Absolute simplicity will provide maximum accuracy, and in this regard a lever operating a simple mechanical spring balance is preferable to a strain gauge. You can directly measure the lever and easily check the accuracy of the spring balance, only the calibration of the tachometer will be beyond your control. Therefore you will be able to be very confident with your results.

Direct manual control of the load applied by the retarder (brake), will enable you to record engine performance at any throttle opening. In respect of acceleration, if required this can be established against set values of torque, using a stop watch.

[Dessertrunner]

Hi Trevor,
The strain guage and data feed costs me all of $200 so its worth geting the data feed to a PL to PC. I have also figuring out how to read each exhast gase temp, just trying to work out how to read 6 Narrow band O2 sensors back to the PLC (Convert output to 0to 5 or 0 to 10v).

You are correct in saying I only want a base line number to work from that is why the base will be from a 100,000k engine with standard everything (No AC).

On you project did the load have to be increased as the engine increased rev right up to red line?

Tom I kind of agree that all the things you suggest change exhaust gase temp but feel that equally is a reflection of change in power. To some degree increased air & increased fuel means higher temp & a sort of increased power.

keep the info coming so the interest on the job stays.
Tony

[TomsSVX]

Originally Posted by Dessertrunner

Hi Trevor,
The strain guage and data feed costs me all of $200 so its worth geting the data feed to a PL to PC. I have also figuring out how to read each exhast gase temp, just trying to work out how to read 6 Narrow band O2 sensors back to the PLC (Convert output to 0to 5 or 0 to 10v).

You are correct in saying I only want a base line number to work from that is why the base will be from a 100,000k engine with standard everything (No AC).

On you project did the load have to be increased as the engine increased rev right up to red line?

Tom I kind of agree that all the things you suggest change exhaust gase temp but feel that equally is a reflection of change in power. To some degree increased air & increased fuel means higher temp & a sort of increased power.

keep the info coming so the interest on the job stays.
Tony

Kinda of confused... The heat lost through the exhaust is energy not transferred to the flywheel. In all theory, the higher the EGT's the less power you *could* be making. I un derstand what you are getting at but you can run low on power with insanely high EGT's...

Its a line you have to walk... retarded timing can be a caused for higher EGT's because combustion comes too late and the charge is leaving without a full burn.

Tom

[Dessertrunner]

Tom,
Don't quite agree, heat is a byproduct of creating energy that goes to the flywheel, no heat no energy.

[TomsSVX]

Originally Posted by Dessertrunner

Tom,
Don't quite agree, heat is a byproduct of creating energy that goes to the flywheel, no heat no energy.

I understand completely... If you were measuring the heat produced by engine in a whole, taking into account the cooling system and radiance through the block along with exhaust gas temps the power measured could theoretically be seen through temp readings...

Exhaust gas temps change with the three variables afformentioned... I have adjusted timing on my engine with the Hydra based soley on EGT's in the past. When I drag raced on the ECUTune chip I would see EGT's top out above 1500*F w/ only 8* ignition advance. We all know that no engine is making good power at 8* advance at WOT but it sure as hell made a lot of heat. W/ the new ECU and a decent timing map my temps dropped around 1200-1300 and made twice the power.

Highway cruising I was seeing around 900*F EGT's with my basic maps... Advanced timing a little more and watched them drop to 700*F. A late spark will not allow a complete burn and thus a good deal of unburned fuel and unfinished combustion is released into the exhaust creating a high EGT but will not make any more power. You will need to rely on another source of measurement for how you want to gauge the power output of the engine

Tom

[SVXRide]

it's always a good idea have a good wide band AF gauge installed....the EG33 tends to like to run a little leaner AF than you would expect to produce max power, yet you tempt the engine melt down gods as you start leaning things out...

-Bill

[Dessertrunner]

I accept your point Tom.
Tony

[oab_au]

The EGT will only tell you how much energy is wasted out the exhaust. To tell how much torque each cylinder delivers to the crank, you would need to read the pulsations that the crank has on a scope triggered off No 1, cylinder, but I don't think it is worth the trouble.

I would use the load cell with a digital readout, and a rev counter. Connect a data logger to record the load over the rpm. You can then load this to an Excel sheet.

To run the engine for a torque reading, you run the engine at about 3000 rpm to get it to operating temp, apply a small amount of load to get the retarder to temp. Then raise the rpm to 4000, start applying load as you open the throttle to keep it at 4000. Just like letting the clutch out.. When the engine is on full throttle at 4000 wind on more load till the rpm drops to 2000.
You can then start the torque readings by reducing the load, to allow the rpm to rise up through the rev range to say 7000, to complete the readings. Then start closing the throttle and reducing the load till it is back to 2000, with the load removed and engine running for a cool down time.

To do spark timing or fuel adjustments, you just run the engine to the rpm that you want to do, hold it there with the load, on full throttle, while moving the timing, watch the load reading for the highest reading.

Harvey.