Page 1 of 1
Posted:
Fri Oct 14, 2005 9:54 amby cspeier
Bruce, John, I have a problem. I have a friend who has a Go-Power Dyno system. His water brake is a manual type. Is there anyway to computerize the brake? In general, what can we do to improve the system? This is the place to ask...
Chad
Posted:
Fri Oct 14, 2005 3:14 pmby 86rocco
I've given this problem a fair thought recently, a friend of mine has a Stuska dyno, like your friend's dyno it's also a manual water brake. Full automatic control and data acquistion is out of our price range but mere data aquistion can be done cost effectively. I'm thinking of using an LM-1 wideband O2 sensor with LM-1 wideband O2 sensor with their LMA-3 auxbox, to measure and log EGO, RPM, and a few other engine metrics as well as adding a pressure sensor to the load cell so that the LM-1 could also measure and log engine torque too. That would still require us to manual control the load and rpm but it would give us fairly complete data logging capability. And as an added benefit, away for the dyno, the LM-1 is a useful piece of test equipment in it's right..
Posted:
Fri Oct 14, 2005 7:42 pmby RRBD
I like the LM1 as well, but as I understand the Rpm thing is a problem with them. I have a Stuska that I've been working on myself and am using a self made load cell using a cheap ($100) pressure transducer with a 0-5 vdc output and a Dataq DI-194 rs($25) for dataloggging. As far as controlling the load, some people leave the manual valve set and use a ball valve to load/unload, this will help with repeat tests. I think the best deal would be to use a dc motor/waterpump, to regulate incoming water, with speed control instead of the valves. I havent figured out the Rpm thing yet either, I have been thinking of just using an optical prox switch and feeding it 9-10 volts and counting the "pulses", but dont know how well it would work with the Dataq datalogger. Rocco, or anybody with some ideas on this please chime in. I'm planning on making a couple more load cell cylinders, an if someone can figure out how to get a good rpm signal in the 0-10 volt range I will gladly make up an extra one for payment. I use quad seals so the cylinders are very nice as well as being compact.
Scott
Posted:
Fri Oct 14, 2005 7:59 pmby 84-1074663779
Chad,
I am unfamiliar with a manual water brake, how is it actually controlled ? Is it a lever operating an internal diaphragm plate, or some sort of water control valve ?
Whatever it is, some type of suitable drive motor and position sensor should be able to control it.
Posted:
Fri Oct 14, 2005 8:10 pmby 86rocco
On the dyno I'm most familiar with the load controller is simply a 1/2" npt needle valve which allows you to precisely adjust the amount of water entering the absorber unit.
Posted:
Sat Oct 15, 2005 1:13 amby Rick360
Scott,
Won't the Tach output from the ignition (MSD I assume) work for an RPM input? It's 4x rpm but use scaling to convert?
Why not electronic load cells instead of the pressure type? E-bay has S-type load cells all the time.
Chad,
I assume you want to have it control the load to do controlled accelleration sweep tests. Are you using a hand controlled water valve? A friend of mine had one and it was very inconsistent loading it like that. A computer could certainly be programmed to do the job if you had a fast acting pneumatic or servo water valve. It wouldn't be easy to build. Controlled accelleration tests is the only way to dyno an engine like you're building. The WOT time is kept to an absolute minimum.
Rick
Posted:
Sat Oct 15, 2005 10:05 amby RRBD
Rick, I've thought about using the MSD output on a larger dyno but the one I'm working on is for smaller engines, like a Briggs Animal. I have been thinking about giving the RPM converter from Innovate a try for myself to see if I can get it to work. As far as load cells, is the output voltage in the 0-10vdc range on those s-type? I picked the pressure transducer so I could use it on another project first and it had the right voltage output.(I know enough about this stuff to be dangerous and thats it, but I'm learning fast.
Using a servo or stepper on a water valve is what most Dyno's out there have, but many are going to a DC pump or an inverter driven AC pump to control the load by varying the pump output.
Most dynos run tests at 300 or 600 rpm/sec, I dont have a clue of how to set this up interface wise, but if someone with electronics experience would like to team up on a project like this, I can offer machining (manual and CNC) as well as share in any costs. I have also designed and built absorbers in the past(waterbrake) that are currently in use on large Dyno's but all of those have been set up with Depac or TAT set ups which are too expensive for my " hobby"
Scott
Posted:
Sat Oct 15, 2005 12:54 pmby Rick360
Scott,
Jameco has some inductive proximity sensors that would be suitable for the rpm application. It would need a ferrous metal "flag" to sense each revolution.
$20 and uses 10-30VDC and works to speeds several times what a normal high rpm engine runs. Uses 3 wires +DC -DC and output. This particular sensor is PNP, meaning it will drive the + to the output when sensing metal.
If I had a place for a dyno I'd take you up on the dyno project. I hope to have a place with enough room in a year or two. I have always planned on building a dyno once I have the place. I've built a number of electronic and uP projects as well as programming.
Rick
Posted:
Sat Oct 15, 2005 1:44 pmby RRBD
Thats exactly what I've been playing with, except mine are made by Peperel & fuchs(sp?). I swiped them from some machines that were being scrapped at the place I used to work. I havent had the time to mess with it as much as I'd like, but I did set it all up with a 9v battery and my Dataq seemed to record the pulses, now I need to get the info into a format that is useable. I thought, I would just put the prox flag on the shaft (1") that the Rotor of my absorber is attached to, it being such a small diameter, I think the little bit of imbalance will not even be seen. Thanks for the help, if you need anything built, just PM me.....
Scott
Posted:
Sat Oct 15, 2005 5:03 pmby bruce
I used an inductive pickup for my speed sensor (2 wire, no power required), took a gear and machined the teeth down to get the face length I needed for the sensor. The inductive sensors put out a Mv signal so you would have to build an amp. I feed mine into an industrial digtal rpm tach, this tach has an output if and when I ever go "digital".
I use an "S" beam load cell for my torque, this outputs Mv/V signal that feeds into an Omega Strain gage meter which has an outout for digital signals.
Posted:
Sat Oct 15, 2005 6:45 pmby 84-1074663779
For Rpm pickup there are plenty of low cost do it yourself ways. Direct optical or magnetic pickups could even be home made.
Or how about adapting a standard automotive sensor such as used in an electronic speedometer, cam position sensor, or breakerless ignition distributor ?
Once you have the pulses, some sort frequency to voltage converter is all that is required.
Torque measurement is simple too. Just fit a suitable spring to the dyno torque arm to generate some usable motion. Wrap a wire around a small drum a few times so it cannot slip, and that can then rotate a potentiometer. Output voltage from the potentiometer will then be proportional to torque. It sounds horribly crude, but built with care it can give excellent results at almost zero cost.
The other big advantage is the output will have a decent voltage swing without needing any amplifiers. A digital voltmeter could be calibrated to read directly in torque fairly easily, and it can also be used as a signal source for a data acquisition system.