test object 4 inch Garret
|Howard Lewis||14/09/2021 17:37:27|
|5298 forum posts|
Good Work, Werner!
Very pleasing that your work enabled Jurgen's Garret to be so improved. Electronic sensors allow measurements to be taken and shown at higher engine speeds.
Your work is pioneering! And others will, also benefit from it.
I like the "Driver" on your Merryweather!
On high speed (2100 rpm ) diesels, in 1959, we measured cylinder pressures by balancing a floating diaphragm with Nitrogen from a gas cylinder until it balanced cylinder pressure, and there was no movement of the diaphragm.
On the flywheel was a Tufnol disc with magnets in slots (one the opposite way round to show tdc ) to give angular position, and speed, relative to a pick up coil)
Diaphragm movement changed capacitance and the frequency of a VHF amplifier, so that when the frequency became constant, we knew that the diaphragm was balanced and the pressure could be read off the gas cylinder. Slow by modern standards, but this was cutting edge technology at the time. (The amplifier was a valve one, 300V DC on the anodes! The Mullard OC71 transistor was new! )
Progress has made life easier!
|Werner Schleidt||14/09/2021 19:53:00|
133 forum posts
Hello all and many thanks for your interest!
I want to add more information about the sensors.
The pressure sensors are china made ones in a stainless steel housing with diaphragma and strain gage amplifier and ready calibrated. They are made for 5 V excitation and have a range from 0.5 volt to 4.5 Volt. I compared them with a normal Pressure gauge with good results. For my software i marked them with the parameters i have to add in the software for the correct span and zero adjustment.
I made adapter from 1/8 BSP and 1/4 pipe thread to M6*0.75 mm for easy use with my model applications. The sensor is very cheap ( < 20 &euro for the quality and good for field measurements. I found no noticeable drift.
The connectors had been changed to rc model type ones they are simple and cheap. If they were pulled out by accident there is no damage possible. I secure them with tape and add numbers or a discription to know hwere they are installed.
|Werner Schleidt||15/09/2021 17:58:00|
133 forum posts
The displacement sensor or the piston travel sensor is more tricky. I want to have a sensor who is on small and bigger engines possible to install. I had an incremental angle sensor this is very accurate but it have to be attached to the axle or flywheel and that is not so easy without changes at the model or object. I get by chance a very fine LVDT displacement sensors with a carrier amplifier and inbuilt electronic. To meassure 100 mm the sensor is about 300 mm long. This is difficult to mount. So i came to the results that the sensor mechanism with a rope like in the old indicators is very favourable. So i thought about about how i can build a sensor. In my former job i know rope displacement sensors based on a rope a return spring mechanism and a rope drum and a potentiometer. This parts are very expensiv , the range is from 500 to 1000 €
So i construct my own. The base is a multiturn wire potentiometer with 5 Kohm , a drum which is coupled with the potentiometer axis and a return spring i got from a identicard jo jo. This were my parts which play together to operate the rope sensor. My modelplane colleague Herbert sponsored me with thin multiwire control line rope.
It was very tricky to bring all parts together and after the first one i made two others all in a range of max 300 mm. I tested them in comparison to my digital gauge at the milling machine. The measurements were very repeatable and with a good accuracy. I added a small amplifier with a filter for cutting noise and with a level shifter for easy zero shift adjustment. The span was precalibrated at the amplifier and a fine calibration can be made by software. All sensors have a unique calibration file, but the rope sensor are very similar.
To give you an idea from my sensor
The spring is in the frontside of the drum . In the front cover of the housing there is a screw which is the spanner for the spring. The tricky part is adding the rope winding by winding in the right direction and spanning the clock spring and setting the potentiometer at the correct position. This is a four hand job, but with several try's it worked good. The amplifier ist then attached to the potentiometer and then it was added some heatshrink tube. Then the parameter and calibration was measured and marked at the sensor.
In the next picture it is adapted to my loco for measurement.
The pressure sensor is connected to the cylinder cover. The rope sensor is mounted with some improvisation with clamps and wooden bars at the front. The rope is connected to the crosshead by a chain and a adapting screw. The white PA part allows clamping of the sensor and some adjustment.
Edited By Werner Schleidt on 15/09/2021 18:03:05
|Werner Schleidt||16/09/2021 14:08:32|
133 forum posts
In the small box is an arduino nano with the connector bank for the eight analog channels , the bluetooth modul with 115 Kbaud and the power supply based on old laptop cells. The power supply is able to work with eight sensors for two to three hours.
On the laptop screen there is a test measurement of 30 secs to see.
In these many curves is the start , some braking with load and the stop to see. This is a kind of overall info as you can see it during measuring .After the stop the data have to be stored. I can measure several tests and store them with usefull short names. After a stored measurement i can reload it and look in the time frame for interesting parts. In the time frame i can look if there are two to three similar curves and with the start and end point I can plot them then in a xy diagram. Is this in running condition on the track i calculate the average pressure and the rpm of the engine. With the piston size , travel ,rpm and average pressure the power in this cylinder chamber can be calculated. This is possible but it is difficult to run the engine under full load. I use the system as a tool to see the begin of steam come in and the shape of the curve. So i can adjust the valve gear. I do not like try and error methods as i made it befor i have this system.
My experience at the moment is, it work good, gives good results , but the software and the procedure how to measure need some fine tunning for more comfortable use in the field. And one thing is at the moment clear, i see it as i measure Jürgens engine. All visitors are exited to see the result immediatly and this is only possible to see in the overview as it is in the picture above. I have to add in the software during measuring some markers for the constant driving or something like that.
Over the winter I think i can make some fine tunnig based on the experience i had now.
Edited By Werner Schleidt on 16/09/2021 14:11:42
Please login to post a reply.
Want the latest issue of Model Engineer or Model Engineers' Workshop? Use our magazine locator links to find your nearest stockist!
You can contact us by phone, mail or email about the magazines including becoming a contributor, submitting reader's letters or making queries about articles. You can also get in touch about this website, advertising or other general issues.
Click THIS LINK for full contact details.
For subscription issues please see THIS LINK.