Sweetenough
Veteran Member
Joined: April 2016
Posts: 121
|
Post by Sweetenough on May 6, 2020 12:09:53 GMT -5
Hi Scott, Would love to get the details on your high side driver circuits. I have planned to use highside switches (BTS442) but they are only good to 21A that might be enough but it is never wrong with some margin Have you played around with the PWM frequency? The standard 973hz (I think it was) seems to generate some noice in the electrical motors. Thinking about increasing it to next step that was about 8khz but have not got that far yet. Made a "test board" yesterday so now hopefully I can start taking baby steps i the arduino world. Read your build thread briefly yesterday, had no idea that arduino could drive a beautiful display like that and all the other functions you put into it Will start over and read your build thread from the beginning. I planing to use PI controllers, do you have any advise what to use and what not? Maybe even a code example? All my "normal" turbine control is made in a industrial control system (ABB AC70) but I am out of I/O channels and have no room for more cards. Very very bulky compared to Arduino but have the great advantage of programing with function blocks and then just make grafical connections. Arduino is so small, cheap and usable that I can not ignore it anymore so I plan to "carve" out the lube oil controls and put it into a arduino system. Kind Regards Thomas
|
|
slittlewing
Senior Member
Joined: November 2017
Posts: 458
|
Post by slittlewing on May 10, 2020 4:06:28 GMT -5
Hi Thomas, I apologise for not coming back sooner with circuit details, the weather has been so good that I have been outside the last few days and didnt get the chance to power up my "main PC" where the circuit stuff is stored. Below you can find the circuit I have devised: It is quite simple, and operates as a high side drive (+12v is switched to pump instead of "ground" switching). I know some people had issues before with the cheap chinese PWM's that switch ground, where they accidentally grounded their fuel pump and they went to max speed so I put the little extra effort into going "high side". The Terminal J14 has two connections. Pin 1 is +12v feed which you can fuse separately to the arduino. I am using 30A fuse for my fuel pumps and 40A for oil pump. Pin 2 is the +12v PWM output to go to the pump. Additionally you should add some capacitors on the arduino power feed to reduce noise, I have not altered the standard arduino PWM freq, its more than fast enough even though it does give the pumps an audible buzz kind of noise. Thats soon drowned out when you start engine! In terms of components, you will see there is a P CHANNEL MOSFET, a normal transistor and a DIODE (labelled scottdiode but I didnt have the proper symbol for it). I found that commonly used MOSFETs are not up to the job after blowing up quite a few. Take for example the popular FQP47P06 here: uk.farnell.com/on-semiconductor/fqp47p06/mosfet-p-channel-60v-47a-to-220/dp/2575367It says 60V and 47A but that is not the continual rating, its just for a fraction of time. After looking at the datasheet, at 12v drain to source, the continual DC rating is just 7amps. So, I found the following parts which are working well for me with plenty of headroom: SPP80P06PHXKSA1 mosfet, able to do about 25 or 30amps continous STPS40SM120CT Schottky diode which is rated to 40amps. this is the "freewheeling" diode that stops the motor EMF coming back to the PCB between pulses. Both should be heatsinked. For the oil pump I ran 2x mosfets and 2x diodes so that I got twice the rating, as the oil pump is much bigger than the fuel pump and I found it could draw huge current with cold oil. I had not seen the BTS442 you mention but it looks good. You may need a couple in parallel as the max power is 167Watts which is more like 13Amps continuous @ 12v. Definately worth a go on some breadboard, I would be interested to know how this goes. If you check out my thread you will see that I characterised each system in turn (fuel, oil) of PWM % vs Pressure, using a simple ramp. I then took this as the "base" mapping which the arduino uses. You could just use this open loop, particularly for fuel. However I added PI control on top, using the standard PID library available for Arduino. Hope this helps, BR, Scott
|
|
Sweetenough
Veteran Member
Joined: April 2016
Posts: 121
|
Post by Sweetenough on May 10, 2020 14:01:36 GMT -5
Hi Scott, Thanks alot for taking the time to explain and share your design. My electronic skills is very limited when beyond ohm's law, so prepare for stupid questions and if you do not have the patience to answer I understand Anyway here comes some questions:..... - The "normal" transistor is there because the arduino board is not strong enough to drive the mosfet? - Been staring at the circuit diagram for a while now and I just dont get how the mosfet gets it's power supply and how it is working, where do the mosfet take its +12VDC? Finally down at J14 FUELPWM Pin 1 is + and Pin 2 - ?
Edit, striked that question.... When I read your answer it is clearly stated how it works, sorry Thanks again for pointing out the 167 watt limitation for the BTS442, totally missed that and only saw the Amp limit. Got the advise to use the BTS442 from friends that have used the BTS442 (driven directly from the arduino pwm output) in combination with a bosch 044 fuel pump with success. It I should use 2 BTS442 in pararell, could both be from the same arduino pwm output or do I need some kind of transistor inbetween? Will read your thread thoroughly from start to finish so I don't miss anything. Managed to get all inputs to my little arduino project configured and displayed on my small 4x20 LCD screen today. Nothing advanced but a big step for me Next up is to download the PID library and get a feel for how it is used. In my bulky old industrial control system I have alot of easy to use logical blocks so I am a little bit spoiled with controls. Used to work as commissioning engineer for 25-50MW gasturbines so I am somewhat damaged and would like my small engine to have the fancy things like the bigger turbines During start I just use a fixed fuel value and when the engine is above 25krpm the speed controller takes over and accelerates up to idle speed 40krpm. Then it follows the rpm set point up to max rpm (90krpm) But several controllers (exhaust temp, turbine inlet temp and compressor discharge pressure) are active and if any of them is higher than the set value it becomes the limiting controler. Kind Regards Thomas
|
|
slittlewing
Senior Member
Joined: November 2017
Posts: 458
|
Post by slittlewing on May 11, 2020 5:49:45 GMT -5
Hi Thomas, My electronics skills are similar to yours!! haha. The circuit I used was based on the site www.gammon.com.au/motors, the part where he says "switching higher voltages". But i changed the component types and values. I think you should try two BTS442 in parallel from same arduino pin to begin with, if that doesnt work you can use a small transistor. Excellent work on the LCD display!! It sounds like you have alot of knowledge on gas turbine control strategy. I am not running any kind of speed control, purely throttling with fuelling and I have an oil pressure kill. I have no speed sensing, but I do sense P2 pressure and use that to define the engine being started. I had some code for detecting start on TOT temperature but since the sensors keep breaking I had to remove it for the meantime! What speed sensor are you using? Good luck and keep us updated on your progress. Cheers Scott
|
|
Sweetenough
Veteran Member
Joined: April 2016
Posts: 121
|
Post by Sweetenough on May 11, 2020 10:43:08 GMT -5
Hi Scott, I think your electronic skills is level ninja compared to mine Did not get your circuit description fully until I read it again (striked one of my questions). Thought that the J14 terminal was the connection to the motor. Will try the BTS442 x 2 maybe with Schottky diode and also try to change the frequency to see what it does. Agree that the electronic noise will be covered in the engine noise but it bugs me. My cheap chinese PWM modules does not have the disturning noise (low switching, but they have caused me problems when accidentaly grounding the pump...) Tried to meassure the PWM frequency on the chinese ones and think they are 12,5khz and they operate very quitely so it is worth a try at least. Using a www.full-race.com/store/borg-warner-efr/turbos-efr-series/borgwarner-efr-turbo-speed-sensor-kit-2/Had problems with it to start with but now it is working really good. First when I installed it I wanted to have close to the casing edge to minimize the distance to the compressor wheel. But it turned out to be a tad to close. Showed to high rpm (compared to actual) when the engine was cold and than gradually more correct as the compressor casing got warmer. My hillbilly conclusion is that when the compressor casing expands from the heat the probe does not expand to the same extent. Meaning that the probe is slightly retracted in the housing. But it is really small margins, I think I shimmed it 0,1 or 0,2mm furter out and then it worked fine. Have not tried to move it out more to see where the limit is. Also attching a picture on the programming of my beloved ABB system, really like the grafical programing! Could not somebody make a converter from ABB code to arduino code?... Still finding it very difficult to write the arduino code but it will hopefully improve with more training. Have also read through your thread now, impressive work!! Good luck with the future tests! Kind Regards Thomas
|
|
Sweetenough
Veteran Member
Joined: April 2016
Posts: 121
|
Post by Sweetenough on Jul 12, 2020 12:36:52 GMT -5
Hi Scott (and everybody else...) Have made some progress with the arduino project. Got help for a friend at work with program code for a PI controller so now it is going in the right direction But since my electronic skills is close to zero, would you mind Scott to share the circuit diagram over your lube oli pump design with 2x mosfet and 2xdiodes? Have bough the same components as you have now and have tested with the BTS442 with the scotty diode, works quite well for even the lube oil pump but a lot of heat is generated so I will most likely uppgrade to your components spec. Got some problems with the MAX6675 thermocouple board, it could only handle non grounded thermocouples and the tc I used for oil tank temperature was of course of the grounded type.... When checking my chinese tc stock it seem random if they are grounded or not For this specific project I will solve it by changing to a PT100 element and to a MAX31865 board but it would be good to now if there is a solution? Anybody who nows if there is a TC board for arduino that can handle grounded thermocoouples?
|
|
slittlewing
Senior Member
Joined: November 2017
Posts: 458
|
Post by slittlewing on Jul 23, 2020 16:04:07 GMT -5
Hi Thomas! Firstly I am sorry for the late reply, I have not been on here for a couple of weeks because of House DIY work every single night and weekend! I’m glad to hear you were able to get the BTS chip working and you are making good progress with the PI control For my “dual”/70A lube oil pump circuit, I simply paralleled two power mosfets and two Schottky diodes (still used just one output pin from arduino and one NPN transistor to feed the PWM signal). If you need circuit diagram I can try and go back on my desktop computer though. I am planning to make a mk2 ECU over winter that can handle more voltage and with more functions or spare I/O. In terms of max6675, that is an interesting finding!! I have had problems with them too but I think it is from EMC rather than grounding since mine has worked perfectly until I start igniting spark plugs. For your oil setup a PT100 sounds like a great solution. I have no other solution for TOT measurement other than to use K type and MAX6675 though! edit: been googling and it appears you can buy a chip called ad8495 which provides analog output from k type instead of SPI. Cheers Scott
|
|
Sweetenough
Veteran Member
Joined: April 2016
Posts: 121
|
Post by Sweetenough on Jul 27, 2020 3:33:45 GMT -5
Hi Scott, Thanks for the reply answered all my questions regarding your "dual" design. And thanks for tip of the ad8495, will investigate this further but for the lube oil control project I will just go ahead and use a MAX31865 for PT100. Have not been much progress lately, did a test connection with the same mosfet/scotty diode that you use and tried different PWM frequenecy. When running "standard" PWM frequency 973 hz there is somekind of noise that disturbes me, probably is no problem and will be covered in the overall engine noise but still.... Running at 7,8 khz the sound disappears, but the signal looks much worse on the oscilloscope. Top yellow line is motor voltage, bottom blue is mosfet control voltage. 973 hz 7,8khz I guess some more testing is required. Kind Regards Thomas
|
|
slittlewing
Senior Member
Joined: November 2017
Posts: 458
|
Post by slittlewing on Jul 27, 2020 15:05:24 GMT -5
Hi Thomas, No problem at all In terms of the standard PWM frequency (I think is 490hz for mega), it creates an audible buzzing noise on all of my pumps. However, this has not created any mechanical problems for me so far! I look forward to your future updates. Also, nice oscilloscope! I wish I had one sometimes!! Cheers Scott
|
|
Sweetenough
Veteran Member
Joined: April 2016
Posts: 121
|
Post by Sweetenough on Jul 28, 2020 1:12:22 GMT -5
Hi Scott, I think that it only me creating problems for myself with the PWM frequency :-) Totally agree with you that the pumps will work fine even with the "disturbing" buzzing noise. It a quite nice oscilloscope for my DYI purpose at least, straight from China but fairly user friendly and not that expensive. You should threat your self with one :-) Looking forward to your future bike updates as well
Kind Regards Thomas
|
|
Sweetenough
Veteran Member
Joined: April 2016
Posts: 121
|
Post by Sweetenough on Aug 19, 2020 2:54:08 GMT -5
Hi Guys, Have done some more testing and is pretty happy with the PWM performance so it is time to go away from the ugly OSB board design and make something more suitable to have onboard. Are now using Scotts design for the PWM end with mosfets, scotty diodes etc. Thanks again Scott for sharing your design! Bought a way to large heat sink Really heavy but almost everything else is heavy as well so I will go ahead and use it anyway. A friend made a 3D printed box that uses the heat sink as a lid and next step is to see if all the components will fit before the final box is printed. Also added a relay to control the oil tank heaters, probably not necessary but what aspect of a DIY jet build is? Kind Regards Thomas
|
|
Sweetenough
Veteran Member
Joined: April 2016
Posts: 121
|
Post by Sweetenough on Oct 6, 2020 1:30:37 GMT -5
Hi Guys, The lube oil control box is slowly coming together. It is now bench tested and after correcting a design error it seems to work as planed. Have some modification work left to do on the vehicle with electrical connections but then it time to make some noise and do some testing and tuning of the system. Kind regards Thomas
|
|
Sweetenough
Veteran Member
Joined: April 2016
Posts: 121
|
Post by Sweetenough on Oct 20, 2020 7:56:29 GMT -5
Hi Guys, Made a test run last Sunday of the new Lube oil control unit and also a thrust test with the new scale with logger function Pressure controller works really good as long as the oil temperature is low enough, but.... seems like my cooler/fan arrangement is too small to give sufficient cooling. Oil temp is rising up to 110°C and at that point the pressure pump runs at maximum speed to keep the oilpressure at 5 bar. Thought that I bought a big enough cooler when I purchased a 25 row cooler but apperantly it is not enough, so I ordered a new 50 row cooler and will use dual electric fans. Thrust turned out to be just short of 50kg. Had hoped for more , since I made some ball bearning "feet" to put under the skates to minimise friction and also preloaded the vehicle with a spring. Tested ended when the afterburner spitted out the spark plug. Kind Regards Thomas
|
|
|
Post by racket on Oct 20, 2020 18:32:48 GMT -5
Hi Thomas
50 kgs of thrust is a bit modest for such a large turbo with A/B running :-(
Could you please give me some wheel sizes for your turbo, inducer/exducer of comp and turb , do you have a comp map ??
You may need to install a jetpipe pitot tube to measure the total pressure exiting the turbo , this is invaluable in calculating potential thrust as well as troubleshooting a lack of thrust
Cheers John
|
|
Sweetenough
Veteran Member
Joined: April 2016
Posts: 121
|
Post by Sweetenough on Oct 21, 2020 3:30:36 GMT -5
Hi John, Unfortnuattly i have not been able to source the correct compressor map, found some that I hope is simular Basically it is the same S500 turbo as Anders & Olov used on their dual jetkick. Compressor inlet 82mm, wheel od 116,5mm Turbine outlet 91mm, unfortnuattly I can't find the od among my notes ( I now I meassured it at some point in time...) This map is from an B&W S400-80 See other log data from screen print Running @90000rpm Compressor outlet pressure 3,057bar Compressor outlet temperature 218°C (Probably not true, will get some combustion radiation heat) Combustion chamber pressure 2,835bar Turbine inlet temp 849°C Turbine outlet temp 719°C The pitot tube meassurment, should be done close to the turbo exit or could it be done in the AB exit? Kind Regards Thomas
|
|