Engine Dyno Project

This has been asked a few times and I have searched around. I want to make sure I understand everything and clarify before I proceed.

The controller in our engine dyno at work has died. Local electronics repair shops haven't been able to fix it yet. The dyno is 40+ years old. Manufacture doesn't support it anymore. I thought maybe an Arduino would be capable of doing this work. The engines are 3000hp diesel engines and I think the speeds are up to 2000RPM.

Load cell: Vmin - 0.5, O/P 3.0023mV/V @ 5000 lbs.
speed sensor is a mag pickup that counts teeth on a flywheel.

Questions that I have:
Will the hx711 work with this load cell? is more information needed?
Do I need an op amp for the mag pickup.
Will an uno be able to take these inputs simultaneously and calculate horsepower?

Thank you

I can't really answer your questions...

You'll probably have to do experiments to test one thing at a time.

[quote="wyedelta, post:1, topic:1062768"]
Will an uno be able to take these inputs simultaneously and calculate horsepower?
[/quote]The speed sensor seems like the only thing that's time-critical so you MIGHT have to build something separate for that.

The processor can't do anything "simultaneously" but often it can do things fast-enough that they appear simultaneous. The actual RPM and the load-cell readings don't change that fast so the processor will be plenty fast-enough.

And with load cells and strain gauges the high forces you're working with should make it easier to get a measurement.

I assume that besides the inputs, there is an output to control the load? Do you know if that can be done with a low-power control-signal?

[quote]Do I need an op amp for the mag pickup.[/quote]If it's just the output of a coil you'll need SOME circuitry to get a "clean" 5V logic signal into the Arduino. A coil & magnet will put-out AC and the negative voltage can damage the Arduino and the voltage may also be too high or too low for the Arduino.

I think calibrating the thing might be tricky... The RPM isn't too tricky but you need to know that you have an accurate torque reading. (If the torque can be measured at zero-RPM and then remain accurate with rotation, that's probably not too difficult.)

What sort of brake does the dyno use?

Traditionally, we used a comparator to condition the output of a variable reluctance pickup. Nowadays, you could replace the mag pickup with a Hall effect device.

The processor can't do anything "simultaneously" but often it can do things fast-enough that they appear simultaneous. The actual RPM and the load-cell readings don't change that fast so the processor will be plenty fast-enough.

Sorry I should have worded this differently. Are there specific GPIO pins that these need to be connected to or is it all in the programming? Will an uno have enough pins to connect everything at the same thing? I think that the speed sensor should be fine with any GPIO because it is either going to be high or low. The ADC will convert the analog load cell signal to a digital signal so would I be fine to connect that to any GPIO as well since it will be a digital input?

I assume that besides the inputs, there is an output to control the load? Do you know if that can be done with a low-power control-signal?

The engine load is still controlled through the existing dyno hardware. There is no need to output anything other than a display that shows the engine speed, torque, and horsepower.

It uses water.

If I use a comparator, would this work?LM393DR2G

Connect the output of the mag pickup to the input of the comparator and the output of the comparator to the GPIO on the arduino?

Yes

That's correct. So you'd count the gear teeth for some specific length of time and calculate RPM.

Do you still have the calibration procedure for the old controller? It can be useful to calibrate a new controller.

So does the load cell measure the force of the brake rotating against it like a scale or is it measuring pressure from compressing hydraulic fluid or what?

yes I was planning to use the old procedure to calibrate the new equipment.

it looks like it measures the force like a scale

Ah, okay, I was just wondering as there are several ways to do that. Things have come a long way since the Prony brake from college. :grinning:

40 years ago the name Lebow was the go to for rotary torque sensors. The sensitivity you mention was pretty common as to 3.00 mV/V give or take which is why we calibrated them. What that means is with for example a 10.0 volt excitation at rated max load the output was 30.0 mV. Depends on what you have and what the excitation voltage actually is. There should be a part number and with luck you will find a data sheet. Also, those older early rotary torque sensors used brushes which after years of use needed replaced or cleaned including the slip rings they rode on.

As to the speed pickup (mag pickup) it depends on what the sensor actually is. Airpax was popular back 40 years ago and here is an old example. These sensors were 3 pin, Vcc, out and ground. The output was pulses and no need for any amplification. One pulse per tooth and the rest was in the math. That was done in a controller or tach system. So it really depends on what you have.

The links are merely a few examples of how things may have been done 40 years ago on your dynamometer.

Ron

Is that lb-ft (torque) or load in pounds from which you calculate torque by knowing the lever arm length.

It is the load in lbs. The lever arm is 21"

1 Like

the speed sensor I have here is only 2 wire. I was reading that to get the best result the signal should be a square wave which is why I think I need the comparator. I don't believe it is supplied with any voltage. It just generates a voltage as each tooth passes by it.

The load cell is a ARTECH brand S shape load cell. We have quite a bit of documentation on it so I believe we should be fine with that.

Thanks. Sorry for all the questions but I was involved with mechanical horsepower measurement for all kinds of things, all kinds of ways, for 50 years, so I was just curious. :nerd_face:

No the questions are good! I didn't know much about this when I started researching but all your questions are things that I found answers to before I posted here and so it makes me think I'm on the right track haha.

1 Like

I may be preaching to the choir but I want to make sure you're aware that, depending on the air gap and the pickup, you can get a couple hundred volts peak to peak so make sure you limit the input to the comparator to <= 3V common mode.

Alrighty, two wire speed sensor. Would be nice if you had a scope to actually look at the output. That way you would know actual amplitude and what signal conditioning you will need. As mentioned you are limited as to what you can feed a micro-controller. Yes, a comparator will give you wave shaping and yes, you want a square or rectangular wave shape. The problem now is knowing what you have.

Ron

the speed sensor I have here is only 2 wire. I was reading that to get the best result the signal should be a square wave which is why I think I need the comparator.

It has been suggested before that you can change out the variable reluctance gear tooth sensor with a gear tooth sensor using a digital output hall sensor. Have you considered this?

https://fullfunctioneng.com/info/Hall%20vs%20VR.pdf

https://www.digikey.com/htmldatasheets/production/57408/0/0/1/gs1005-gs1007-series.html