Society of Robots - Robot Forum

Electronics => Electronics => Topic started by: raltracer on November 15, 2011, 07:28:40 AM

Title: Newby looking for guidance for making a robot controled engine Dyno
Post by: raltracer on November 15, 2011, 07:28:40 AM
Hi Guys,

Newby here looking for info on a few things. I race single seater racing cars for which we also build the engines. We have a manual dyno for engine testing and I would like to make the dyno run a basic program as follows.

Open throttle slowly over a few seconds whilst applying load (load is by a water valve that is gradually opened) at full throttle ther load should be open enough to hold the engine at 4000 rpm then once this is achieved the load valve should be closed off slowly untill the engine hits 6,800 rpm. Once this is achieved throttle can be closed as can the load valve.

Whilst all this is going on I also need a datalogger for the torque load cell and cross referable with the rpm at points of 500 rpm increments this should then be transferable to a spread sheet.

This above would need to be adjustable for load and throttle position

So does anybody know where I would start??

Many thanks in advance

Benn 
Title: Re: Newby looking for guidance for making a robot controled engine Dyno
Post by: Soeren on November 15, 2011, 10:12:05 AM
Hi,

Newby here looking for info on a few things. I race single seater racing cars for which we also build the engines. We have a manual dyno for engine testing and I would like to make the dyno run a basic program as follows.
Is the dyno for the engine only, or for the entire car?

The water valve for the brake will need to be interfaced to a gear head motor somehow and the position of the valve will need to be decoded.

Likewise with the throttle, a Bowden cable with another gearhead motor should do nicely.
This could be done without an encoder, as you'll get feedback in the form of RPM count, but you'll be able to get a better response from the total system with an encoder.

RPM count is quite simple to do - is the ignition controlled by a Hall element, a konventional Kettering system or...?


Open throttle slowly over a few seconds whilst applying load (load is by a water valve that is gradually opened) at full throttle ther load should be open enough to hold the engine at 4000 rpm then once this is achieved the load valve should be closed off slowly untill the engine hits 6,800 rpm. Once this is achieved throttle can be closed as can the load valve.
When the hardware described above is in place (and tested), the actual algorithms are just that (algorithms).


Whilst all this is going on I also need a datalogger for the torque load cell and cross referable with the rpm at points of 500 rpm increments this should then be transferable to a spread sheet.

This above would need to be adjustable for load and throttle position
A USB enabled microcontroller is one way to go, but personally, I'd prefer a controller with an SD-card, as this can both be used to log whatever amount of data you need AND hold a file with RPM counts etc for controlling the test (assuming a fairly similar test pattern with just the numbers changing) and then you can keep the PC in a less hostile environment.

Triggering a reading and logging of the torque cell at certain RPMs is small potatoes.

The controller can write the data to the SD-card as a comma (or semicolon) separated file and will then be very easy to import it into whatever spread sheet program you have.
Title: Re: Newby looking for guidance for making a robot controled engine Dyno
Post by: raltracer on November 15, 2011, 12:00:27 PM
Hi Soeren,

You make it sound easy  :)

The Dyno is just for the engine, we race an Historic formula ford 1600 car and the engine is a ford Kent pushrod engine. We also need to be able to make it adaptable as we also plan to build Toyota engines for our F3 car that produce much more power.

I have been thinking of fitting a needle valve which I could easily mount a gearhead motor to for the control. We are currently using a gate valve but the adjustment is very tricky. I have also been toying with the idea of using a bleed off to regulate the pump pressure so this could be the load adjustment for different power engines. 

The ign is a simple old fashioned coil and distributor but on the end of the Dyno rotor there is a trigger wheel so I could use a hall sensor and I guess that would be more accurate?



Thanks for the help I'm begining to think this is a possiblity even with my lack of knowledge. :)

Benn



 
Title: Re: Newby looking for guidance for making a robot controled engine Dyno
Post by: Soeren on November 15, 2011, 02:29:27 PM
Hi,

You make it sound easy  :)
"Easy" may be an exaggeration, as there will be some mechanical work to electrify the now manual valve and speeder, bu working on engines, I imagine you won't feel too challenged doing the mech. bits.


I have been thinking of fitting a needle valve which I could easily mount a gearhead motor to for the control. We are currently using a gate valve but the adjustment is very tricky. I have also been toying with the idea of using a bleed off to regulate the pump pressure so this could be the load adjustment for different power engines. 
A pressure sensor could be used as input for the controller, which in turn can control the bleeder.


The ign is a simple old fashioned coil and distributor but on the end of the Dyno rotor there is a trigger wheel so I could use a hall sensor and I guess that would be more accurate?
The accuracy would be the same, but the Hall will be a little easier to interface (not much though).


Thanks for the help I'm begining to think this is a possiblity even with my lack of knowledge. :)
Sure it's possible ;D
Title: Re: Newby looking for guidance for making a robot controled engine Dyno
Post by: raltracer on November 16, 2011, 01:40:55 AM
I am a engineer by trade and making bits to connect the valve to the gear head motor is no problem. Would it be possible to rig a large scale servo (as I have a spare) for the throttle?

I think the main problem will be learning the programming and getting the thing wired up correctly. Do you have any recommendations for the microcontroler I've looked at the Axion on here would that control the motors and cope with the datalogger? I'm guessing I can rig a push button to activate the program or could I interface with a pc and run it from there? I'm in the uk so hopefully they can ship here.
Title: Re: Newby looking for guidance for making a robot controled engine Dyno
Post by: Soeren on November 16, 2011, 10:57:13 AM
Hi,

Would it be possible to rig a large scale servo (as I have a spare) for the throttle?
If the servo has got the power needed, then sure.
Main questions here is: What is the servos torque and what torque is needed to move the throttle.
Whether you'll find it sufficiently precise is another thing to consider.


I think the main problem will be learning the programming and getting the thing wired up correctly.
Yes, but these days program libraries exists for most stuff (a library is an assembly of routines that does the grunt work of something when you call it from your program).

Wiring it up as easy as possible probably means that you should get a board which has got as much of what you need, to minimize what you have to make yourself.


Do you have any recommendations for the microcontroler I've looked at the Axion on here would that control the motors and cope with the datalogger?
I'm thinking more in terms of something with as much as possible already on the board - like eg. an (ยต)SD card slot, an USB server or whatever will let you transfer the measurements to the PC with less soldering on your side.
Something like this board (http://www.microchip.com/stellent/idcplg?IdcService=SS_GET_PAGE&nodeId=1406&dDocName=en543105) (US$60), a board I have myself, or, if more speed and storage capacity is needed, this board (http://digilentinc.com/Products/Detail.cfm?NavPath=2,892,894&Prod=CHIPKIT-MAX32)  (US$50) or even its smaller sibling (http://digilentinc.com/Products/Detail.cfm?NavPath=2,892,893&Prod=CHIPKIT-UNO32) (US$27).
The last two board doesn't have an SD slot (although the larger of them has got USB OTG),but they're blindingly fast and especially the larger of them is crammed full of goodies and you won't be running out of RAM space.

However, selecting the board really cannot be done until you have decided on how you want to transfer the data to the PC, as no two boards are the same.


I'm guessing I can rig a push button to activate the program
Most boards have a reset button and that's all that's needed (or just let it rip when you apply power, which generates a reset event anyway).


or could I interface with a pc and run it from there?
Sure, but do you really want a PC next to a dyno (dirt and oil "fog" in the air will not be the best environment for a PC).

If you really want a PC in there, a better solution might be to use a USB interface/measurement card on the PC.


I'm in the uk so hopefully they can ship here.
I have never had any problems getting stuff sent to Denmark, so I cannot imagine there should be any problems at all shipping to UK. All the hardware (boards, programmers etc.) from Microchip shipped to inside EU, is ab UK anyway ;D


If I had to make this, I'd start with the interfacing hardware (motors, their linkages an the sensors), then add the electronics controlling them and testing that each behave. Then (and not a second before), I'd select a board, get it, program it and put it in service.
This way you know exactly what you need and the time tinkering with the hardware will let your thoughts on how it should all play together will mature.

BTW. wouldn't you need a sensor for measuring relative humidity as well?
(Changes in R.H. from one day to the next can seriously render your measured data untrustworthy, as power depends quite a bit on R.H.).