DIY Tesla Impulse Turbine

Home Model Engine Machinist Forum

Help Support Home Model Engine Machinist Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Steam flow rate was 0.83 LPM (Liters Per Minute) at just over 100 C and at nearly zero pressure. The coupler melted before I had a chance to measure RPMs and the voltage reading was made after RPMs had already dropped off. This test ended too quickly to have provided adequate data.

I THINK the coupler melted from heat created by friction as the stainless steel TT drive shaft spun inside the aluminum coupler, after the steam heat loosened the connection. I don't believe the steam heat alone melted the coupler's plastic spider. I'll tighten the coupler onto the shaft a bit tighter for the next test,...hopefully, that will be enough.

Both TT bearings are full ceramic which run "dry" (no grease or oil).
Wait a minute, wait a minute. .83 LPM? you must mean .083 LPM? at .83 you would have a veritable explosion. You'd need a huge boiler.
 
Wait a minute, wait a minute. .83 LPM? you must mean .083 LPM? at .83 you would have a veritable explosion. You'd need a huge boiler.

Compared to most boilers found on HMEM, my 10" diameter x 13" long monotube flash boiler, containing 30 feet of 5/8" diameter copper tube, is huge.

The following video was taken just before I redirected steam flow into the Tesla turbine via a 3-way valve. Midway through the video I focus in on the K24 flow rate meter (Blue face plate) which measures feed water into the boiler and is displaying 6.0XX liters total flow (big number) and 00.83 to 00.86 LPM (bottom smaller numbers). So, unless the K24 flow meter is off by a factor of 10, which seems unlikely, the boiler is indeed making 0.83 LPM steam at a 53% power setting.

 
Hi, Can you explain Boiler temp out at "91", and 31psi? - sounds like hot water, not steam. If inlet pressure is 35psi, and outlet 31psi, then the inlet pump is providing the pressure, not the burner "adding enthalpy"(or something...??) if my thermodynamics is right? - SO I am not sure what this is telling me. (I stopped video at 15 seconds run-time).
1730446117417.png

Thanks,
K2
 
Hi, Can you explain Boiler temp out at "91", and 31psi? - sounds like hot water, not steam. If inlet pressure is 35psi, and outlet 31psi, then the inlet pump is providing the pressure, not the burner "adding enthalpy"(or something...??) if my thermodynamics is right? - SO I am not sure what this is telling me. (I stopped video at 15 seconds run-time).

Thanks,
K2
I still haven't been able to set the temp and pressure sensors to display accurate readings; view them as +/- 30.
The video clearly shows steam shooting out of the 16mm copper tube at the end of the video, which shows open flow steam with no pressure and no water sputtering out along with the steam. After the boiler reaches this level of steam output, using a manual 3-way valve, I redirect the steam flow from the copper tube, into the Tesla turbine.

BTW, I didn't have a problem with water building up inside the housing; the turbine spun up quickly with little to no hesitation,...before the coupler failed.

Luck permitting, the steam gods will shine upon me during the next test and allow me to get some good results and good video :)
 
Last edited:
Thanks, I understand!
Do you have a mechanical pressure gauge you can affix? - may be within a few %...
And maybe rest some soft solder on the pipe - when it reaches melt temperature you have a control standard for resetting your fancy digital thing?
K2
 
Thanks, I understand!
Do you have a mechanical pressure gauge you can affix? - may be within a few %...
And maybe rest some soft solder on the pipe - when it reaches melt temperature you have a control standard for resetting your fancy digital thing?
K2

I have the liquid filled dial guage shown in the video in post #305 at 11 sec in, and I have been using that guage as my "gold standard" for pressure values, for adjusting pressure sensor values. But the real problems are electrical noise generated by the 3 motors, the two DC to DC power supplies I use to drop the 36 vdc down to 12 vdc and 5 vdc, and the very small voltages from the sensors which represent pressure and temperature.
The Arduino micro computer I'm using as the ECU (Engine Control Unit) can only use 0 to 5 volts dc to read the full pressure range of 0 to 500 psi. The range on the pressure sensors is 0 to 1000 psi, meaning their 5 volt output indicates a 1000 psi pressure, a 2.5 volt reading represents 500 psi, and 0.5 volt reading represents 100 psi.

The 35 & 31 psi readings seen on the display during the test run meant the two pressure sensors were sending out 0.175 and 0.155 volts respectively. Reading those small values accurately when the electrical noise level is 0.200 to 0.300 volts is challenging at best, and impossible at worst.
 
Probably a working temp around 75°c according to some generalized data sheets.

Well, that's not good,... as I suspect the stainless shaft will always rise to at least 100° C.

The next test run will be interesting,...I replaced the melted spider with quick-set epoxy, essentially making a replacement spider using epoxy. If this too fails, I will use either a high temp epoxy or high temp silicone.
 
I have the liquid filled dial guage shown in the video in post #305 at 11 sec in, and I have been using that guage as my "gold standard" for pressure values, for adjusting pressure sensor values. But the real problems are electrical noise generated by the 3 motors, the two DC to DC power supplies I use to drop the 36 vdc down to 12 vdc and 5 vdc, and the very small voltages from the sensors which represent pressure and temperature.
The Arduino micro computer I'm using as the ECU (Engine Control Unit) can only use 0 to 5 volts dc to read the full pressure range of 0 to 500 psi. The range on the pressure sensors is 0 to 1000 psi, meaning their 5 volt output indicates a 1000 psi pressure, a 2.5 volt reading represents 500 psi, and 0.5 volt reading represents 100 psi.

The 35 & 31 psi readings seen on the display during the test run meant the two pressure sensors were sending out 0.175 and 0.155 volts respectively. Reading those small values accurately when the electrical noise level is 0.200 to 0.300 volts is challenging at best, and impossible at worst.
You can use some RC filters to get rid of the noise.
 
Well done for trying, but looking at the voltage values for your instrumentation it does seem to me to be " impossible at worst." But are these pressure values used in the controls, or just for interest? If control parameters, then you need to invest in some more suitable sensors... but for now I suspect you are just tickling the tiger's tail, not really trying to push for 500psi steam! So maybe when to develop more pressure and reduce noise things will start to work sensibly?
GOOD LUCK!
K2
 
Well done for trying, but looking at the voltage values for your instrumentation it does seem to me to be " impossible at worst." But are these pressure values used in the controls, or just for interest? If control parameters, then you need to invest in some more suitable sensors... but for now I suspect you are just tickling the tiger's tail, not really trying to push for 500psi steam! So maybe when to develop more pressure and reduce noise things will start to work sensibly?
GOOD LUCK!
K2

Displaying temperature and pressure values even when you know those values aren't very accurate is how software development works; I now know the code (aka software) which fetches the temp and pressure readings which are actually values from 0 to 1023, and then changes those values into degrees C and PSI numbers, is all working. The displayed temperature and pressure reading aren't there for interest, they're shown to demonstrate the code is working even if the hardware has problem. I know I still have hardware problems, so I haven't allowed the software to use those values to control the boiler.

However, the software does read an input I call Power Request which reads a voltage level derived from a variable resistor I control, and then sets the PWM speed settings for both Fuel and Blower motors to the appropriate levels for best fuel-Air mixture for best heat output.

You are correct that I'm in no rush to reach 500 psi; I much prefer insuring all is working reliably at lower power settings first.
 
I've already tried hanging 1000uf caps and various resistor values onto the power supply outputs with very little success. I've just ordered some LC filters made for power supply noise suppression,...hopefully these will work :)
Dealing with the noise from the PSU is a good start.

I'd suggest also putting filters between your Arduino pins and their respective sensors. An RC low pass filter is what you want, this will help remove high frequency electrical noise from the signals.

Also with your motors did you put a capacitor in parallel with the motor? It's such a common method to suppress noise from brushes that even cheap toys use it. Motors that only run one way can also have a flyback diode to improve matters further.
 
Dealing with the noise from the PSU is a good start.

I'd suggest also putting filters between your Arduino pins and their respective sensors. An RC low pass filter is what you want, this will help remove high frequency electrical noise from the signals.

Also with your motors did you put a capacitor in parallel with the motor? It's such a common method to suppress noise from brushes that even cheap toys use it. Motors that only run one way can also have a flyback diode to improve matters further.

All good ideas.

I've had limited success using RC filters on the sensor input, but this is an area I need to revisit as the tests I performed were all done when I was still using the bad wires from China. Now that I've hopefully found and replaced all the faulty wires, I need to insert various RC values while watching the sensor signals on my O-scope.

Adding caps and diodes to the motors was the first step I took ,...I even added MOVs to the blower motor.
Air Blower Mods b sml.jpg
 
All good ideas.

I've had limited success using RC filters on the sensor input, but this is an area I need to revisit as the tests I performed were all done when I was still using the bad wires from China. Now that I've hopefully found and replaced all the faulty wires, I need to insert various RC values while watching the sensor signals on my O-scope.

Adding caps and diodes to the motors was the first step I took ,...I even added MOVs to the blower motor.
View attachment 160691
One other thought: what's the situation with your signal wires? As you probably know they can act as antennas for radio frequency noise, which could be coming from the brushes of your motors I suppose.

Makes me glad of brushless motors, no noisy commutator!
 
One other thought: what's the situation with your signal wires? As you probably know they can act as antennas for radio frequency noise, which could be coming from the brushes of your motors I suppose.

Makes me glad of brushless motors, no noisy commutator!

As shown below, both pressure sensors and temperature sensors came with shielded cables. In the photo, the silver wire is for the temperature sensor, (RTD) and the black wire is for pressure. Both shields are connected to chassis ground at the ECU connector.

P&T Sensors sml.jpg
 
Back
Top