A new project that i am looking forward to, I have many projects started, finished and left by the way side but this is one i have always thought of without having the time or resources. All this changed after contacting a chap with more projects than me, and that can sometimes be the problem. So Iliya (in UK) now has one less ToDo job and i have a mammoth task to work on. "Megga !"
Were do we start, well a good clean always make the job look as if its achievable
Ok so im not great at environmental protection but that birds nest just had to go.
A quick look over the engine found a few of the connectors either broken or missing, so off to eBay to look for parts. Now i did mention the engine was from the USSR era ?. ok so eBay does not have what im looking for and that what i can find indicates there are a hundred or so different types :(
So i put a call out to a few potential companies that either break or may have spares, i also remember seeing an old souse capsule at the Leicester space centre so that was also on the list. Out of all the contacts only one came back with a maybe and that was the Space Centre. wow! didn't expect that, i was put in contact with a chap in America that had some background links to Russia and can actually read the lingo. Did he come through, I have managed to replace all the duff connectors and he has helped with some of the translation. Big thanks to Ilya (US) !!
Checkout those pesky Acrylic characters. They actually mean something, (And look cool)
Re-assembling the main coupling was a job for steady hands as the older cables are covered in muck and have become oxidised from corrosion
We need to look at the fuel system. This was a little unusual, i expected an electrical pump etc but found that the bottom primary pump was a centrifuge with a upper vain pump for the starting.
As the engine has sat around outside for about 6 years and the fuel line had been severed without being sealed it will have to be stripped down. Surprisingly hardly any dirt and it still smelt of aviation fuel. All the seals and O-Rings have been replaced with exact or as close as possible sizes
Top Starting Fuel Pump
Its about time we spun the engine over to see if we get the dreaded rumbling noise. To do this i need 24v with about 180amps all running through a little connector with pins that are only 3mm wide. Again Ilya came to the rescue with the correct fitting, I added some 16mmSq cable and we are ready to go.
Now the engine is supposed to have a soft start of sorts!, the wiring diagram shows the power being shunted through a large resister for about a second before full power to the starter (well i did say sort of) so im using just 1 battery, this gives me about 90amps at 12v, that will do for a test.
Not all of this project will be be without pain, connecting the cables gives a medium sized spark as expected and the engine starts to spin up. Good its sounds good and then an abrupt stop!, disconnect the cables, turn the engine over by hand seems ok. Re-Connect, larger spark but nothing.
Stripping the started down identifies the problem, water has leaked in between the input shaft and the starter, the bearing shield failed and water contamination of the bearing has rusted the balls. which are now seized. New bearings (£7.00) and we are back in business.
This is the intake with the motor removed, there's a little leak on the seal but that will have to wait.
Its looking a little cleaner now, although no guarantees yet, time to have a look around for any other potential points of failure
Now im no aerospace engineer but the rear combustion turbine looks to have some damage on its tips. The turbine blades on this engine are a solid one piece design, this makes them easier to produce without blade keys etc, but the entire blade disk has to be replaced if damaged. Lucky we have another engine! Nice, spare looks good so that will need replacing, but i think a test run first just in case we have a problem and damage the spare.
I have seen a few other engines on eBay etc and most have parts missing, a common part that goes that is either damaged or missing is the the ignition coil. Luckily I have 2, both look in a dreadful state but after a strip down and clean both seem to work
KP-9V Air bypass valve after clean up, this is the main working part of the engine. it has no electronics or controls and could be overlooked as irrelevant to the workings of the engine. But this valve provides the means to start the main larger turbine engines of an aircraft. It syphons off air from the compressor turbine which is either expressed to the atmosphere or channelled to its outlet, a diaphragm inside detects if the outlet is blocked which then opens a side valve to bleed off the excess air, and we are talking about allot of air !
Electronics
The AI-9 electronics are rudimentary, designed to be resilient and trustworthy. How they did this back in the 1970s was to use a mechanical motor turning small cams connected to relays. This allowed for timing of various relays to control the engine start. sounds like Meccano if anyone remembers that. 
I found quite a few schematics along with timing charts that were almost but not quite what was described in the manual. Various planes and helicopters use the same engine, all with variances in the start sequence.
I came across a Russian site "In Russian" that has a fair amount of technical data and some schematics, this one seems to match the manual and starts to make more sense
Putting all this together i have laid out my version of the schematic, It has changed a few times, specifically the diodes on the supply to the Buck converters, this is to allow the USB to power up the Arduinos without back feeding to the other controllers and causing overload of the USB ports.
Trying to get one of the original controllers, and their are quite a few revisions will be near on impossible. Modern solution needed !
What is the criteria ? Cheap, Can manage multiple inputs, has enough outputs, easy to replace and is cheap, enter the Arduino micro controller (£ 6.00) now that's cheap. Now don't knock it, this little controller has loads of potential and as I have loads of them knocking about and i need to use my old stuff before buying new. Challenge accepted!
To put this to work i need to build two housings, a wiring centre to bring the bulky items together and a control module to house the Arduino, buttons and relays. The first will be a wiring centre, a box with 2 solenoids, a heating coil (to dump the load while soft starting) an isolator and some connectors.
I have kept the engine as original as possible (no extras) so connection is via the existing USSR 14pin plug, from there its into the wiring centre. After putting it all together and bolt it to the frame. Nice !
That was the easy bit, now for the tech. I needed quite a few bits but what else am i going to do while covid is stopping me from doing all the house work ? Planning is the key to this and probably took the longest to do (about 3 weeks) drawing schematics, creating timing charts and programming. The basics are done and here is my test loom.
While adding the buttons i suddenly ran out of inputs, this happens when you keep thinking of new stuff to add. But I have a cunning plan, use a series of resistor and sample the input when a button is pressed. Magic i can now use plenty of buttons from one input.
ok testing done and I now have all the input values, so I can hard wire them into the test loom, this will end up on a small board later.
One of the issues in building something the first time is you begin thinking of new things that will make it even better. But I had already bough the control box for all this to go into ! Ok time for some real estate planning, if those planners can fit 10 houses on the back of a postage stamp ill be damned if I cant get it all in.
Phew, its a tight fit, a little re-design of one of the circuit boards and more drill hols got it in. You may notice a little slip-up on the top plate but my welder will put that right.
I also need to debug the code so lets add a port hole for the USB, Ill turn up a plug with an O-ring later.
Keep in mind i still need wire for an LCD display
First dry run fitting produced a real hum dinger as its the first time i have tried to close the lid.... The bloody switch button is to high and the lid will not close.
Some milling and glue will be needed .
Ok over to my CNC router to make it look pretty. Now that looks nice even if i say so myself. I have added a TFT screen which im hoping to connect to a slaved Arduino uno, this will off load any graphics work making sure the master does not slow down and miss something.
To debug I need to set relays in positions and short out wires etc. this is going to be a nightmare as I don't want to try out the launch control on my only (potentially) working engine. So an additional side project is to build a Diagnostic Box that can emulate all the relays and sensors of the main engine.
Studying the Russian manual gives all the timings that the Launch system needs, reverse this and I can replicate it in my diagnostics box.
So another £6.00 Arduino. 
Prototype boards always look ugly, but with a little planning you can cram on loads of stuff, just get the board the right way round (Vertical or Horizontal that is the question!)
Opps missed that one, One of the relay outputs was not getting picked up, bad track
Dry test of the Opto isolators and relays, i get to use a few more of the 50 I bought earlier.
Im going to use a mains transformer for the 24v, I need enough juice to power the diags box and the controller box and all the lamps etc without the ups and downs of any power being drawn from its supply.
Time to cram it all into a box i had spare, this is becoming a habit. But the smaller boxes do make it look better, almost as if I had planed it that way.
Im trying out a cheap 4 line display £6.50, Ideal for feedback to the user. Selection Button and a reset and off we go. Hang on, lets add a bit more.. An LED to show its working and one to let you know it has all gone to the dogs.
If you want more info on the coding follow my link dbeacroft - Arduino Project Hub
And with an updated cover makes the box look the part
Now to update the TFT control panel, I will be running another Arduino to run the display, mainly due to updating the screen takes to much time (relatively slow) and the main processor has other things to do. So i have programmed an Arduino nano to manage all the screen updates and linked it to the primary Arduino via I2C line
Here is the first go without colour, i will add that later.
.
Build a small board to hold it and fix it to the rear of the TFT.
First Cold Scroll.
I have all the parts together now and am ready for the Cold Scroll test. Cold Scroll is the process of exercising all the controls and start procedure but for the fuel and ignition. The core shaft is spun up for 20 seconds and then shut down all from the control
No bangs or rattles, so im happy with the test, Next will be the False Start, Similar to the previous tests with fuel but no ignition.
No bangs or rattles, so im happy with the test, Next will be the False Start, Similar to the previous tests with fuel but no ignition.
And a bonus!
The electric motor brings the core up to about 17500‐20500 rpm with new oil (the real stuff AiroShell 750) the oil pump has pushed lubricant around the bearings and triggered the oil pressure switch (yellow lamp in video) This is a good sign as it comes on to show good pressure unlike a car which would be definitely bad.
Fuel
It is possible to run gas turbines on diesel but I have a small stash of A1 Jet fuel keeping as close to original conditions as possible while testing. The AI-9 engine requires fuel supplied between 0.6 to 1.7 bar. I found a 24v transfer pump on eBay with a plate pressure of 1 bar which should be ideal as the delivery fuel pump. Add a filter and pressure gauge and we should be good to go.
Fuel connectors can leek and cause fuel spills just waiting for ignition. making connections to the engine fuel pump had an unusual metric thread which i could not get hold of. To make things fit I needed a BSP thread which will open up a variety of available connections. Using the original nut I turned up a new union with a new mail thread, now I can use quick release connectors. This should stop air getting in and fuel from leaking when connecting and removing.
False Start
All the kit is now moved outside for this test, although no ignition there is still the possibility of a fire so the addition of extinguishers will be mandatory.
Remember this blog is a work in progress so things happen in-between, well the in-between this time was an issue with relays causing Back EMF, this is the effect of a small power spike being created when a relay coil is switched off, the larger the coil the larger the spike. Most relays will have a diode placed across the terminals of the relay, this arrests the spike as its created.
The engine fuel relays are 24v and consume about 3Amps, but no diodes. So my first test did not go to plan. as one of my relays had some timer electronics. Not any more, after blowing that the Arduino carried on and attempted to spin up at full amps (about 180A). With no relay the low amp start coil melted so no more tests!
A re-design brings in additional diode protection and a swap around of relays and controls, no more timer. The next test should go smoother with fuel but no ignition. I am looking for wetting of the combustion chamber and turbine fins. this should dislodge any fine grit or soot before attempting to run it.
Attempt 2 was a little more successful, fuel runs through injectors fine which should also clear out any fine particles still leftover in the main engine.
Inspection
A review of the test through up a few minor issues with O-ring seals, which were quickly replaced. No oil leeks and the processor did not drop out during peek load on the batteries (buck converters doing their job)
Micro Start
Im now ready to start the engine but am weary that things could go wrong. So im intending to part start the engine. I have rigged it to cut the main fuel early to stop it hitting peek revs, Should no issues be detected then Full Start will be on the books.
The engine still needs temperature control, if the combustion turbine exceeds 850degrees then internal damage could occur. The existing thermocouples were damaged some time ago when the engine was removed. I have a pair of K type thermocouples rated to 1250 degrees which need a little adjustment to allow fitting to the exhaust nozzle.
Until I have the temperature fix there will be no long running tests to prevent any damage.
As the existing thermocouples had been cut and connectors damaged. I have managed to source some K type thermocouples with a 0 to 1250 degree range to work with a thermocouple Amplifier (MAX31850K) for the Arduino. The housings have been welded and tapped to take a NTP thread. Hopefully they will react quickly enough to allow the engine system to detect issues and automatically shut down the engine if necessary (we will see)
I am still having issues with the display controller running low on RAM and then crashing the display controller. As it happens the Arduino range has just announced an Arduino Nano Every, this little chap has a slightly faster processor and the all important increase in RAM. Reprogramed, Fitted and now it seems to respond fine with RAM to spare.
Some modifications to the old thermocouple housings (welding and tapping) has produced a nice connection to the exhaust cowling. A pair is now fitted and cabled back into the wiring centre, a basic test with a blow torch to see if the control centre responds to the temperature changes.
As the weather is warming up i managed to find a gap in the weather. Out with all the kit for a quick blast. The run went well with a good start, the afterburn reduced back into the combustion chamber and the central core ran upto speed. I then got a loud pop and shutdown.
The electric started has a centrifuge that did not trip my control software, so an auto shutdown and the AN-9 typical pop.
Fix of the 20 second shutdown.
I have attached the starter to the workbench with a clamp and spun the motor up with an LED attached to the centrifuge switch. Sure enough the switch opens an around 2700 RPM which is good so the issue is in the controller.
I have added some additional lines of code to the diagnostics and we seem to be back in business.
Successful Full Run
Pictures to follow but this weekend was a big day, for one the engine runs great still lagging on the temperature but continuous running until commanded to shut down.
Second item of the day, while building the engine above my donor engine has been under review and i have re-built all the missing parts (which were many) still some teething problems with that one but had that running too.
But they do sound great !
Its been a while but loads of work behind the scenes
I have a second engine that looked like it had been thrown out of the helicopter at height! This was my donor engine for parts i needed, between the two some parts good, some parts bad and some parts missing. I have had the main engine running nicely but the main combustion fan does have some damage. Taking the main drive shaft out is like removing a piston ring from a fully working engine (not so simple). No manual or help, so strip the spare and rebuild helps to familiarise yourself with the beast. Turned out the bearings in the spare are pretty good so i started to re build and make the missing parts with good results, used all the control gear created in this blog and it runs (sort of !) more on that another time.
So its time to upgrade the appearance, i have been practicing TIG welding so decided to build a more robust frame for transport and protection. This will lead to a more modular approach to move it around and sit on a fixed bench for running and display.
Ahhhhh!
I striped both engines to switch the main turbo shaft and found the main bearing in my good engine had failed at some point and spun in its housing. This is not a good thing!. I have used the bearing from the donor engine but that could have been the end of all play time so lucky to have found it. The lugs of the carrier have broken off allowing it to rotate and some how did not find there way into the race way (Lucky). The bearing housing is marked but not badly and the bearings still looks good so what went on here i don't know ?
For those who like to see how they are taken apart, here is a video (25 mins) with a little commentary of what is going on. I just love the size of that turbo (Mega)
Fast rebuild for the weekend
Im in a hurry now, I have a display event to show this thing in action so the bench needs to be sorted fast. Turns out the missus wants new kitchen worktops. Hmm opportunity at hand, and this is what i came up with. "Yes luv lets have some new work tops"
Great day and lots of noise from bikes and the Engine :)
ToDo...
Temperature Calibration & speed of UIRecording Engine run data from the Arduinos.
Second Engine Rebuild
New Control System ??
Additional Stuff ...
Arduino Site for more info on the computer build The Arduino Project
More images at YouTube https://www.youtube.com/@derekbeacroft
Need to use a workshop in the midlands ? The-machine-shop.co.uk
And if you just want to follow my updates find me on face-book "Derek Beacroft"
Comments