|
Post by Marathonman on Dec 6, 2020 17:23:35 GMT -6
Another good reason is most of the programs or at least a large part of them are shifting in the MS range which is useless in the Figuera device as that is not near fast enough for 60 let a lone 50 hz. with the program posted it can switch in the low microsecond to high nanosecond range which is plenty fast enough for anyone's purposes even high frequency nonoperm cores. the sky is the limit.
Regards, Marathonman
|
|
|
Post by Marathonman on Dec 18, 2020 20:34:34 GMT -6
Received PANDUIT 22 awg ribbon cable today. i bought it from www.electronicsurplus.com and it is from Germany. i got 100 feet for 20 bucks for some top quality ribbon cable that was probably well over 100 buck brand new originally. i will use this for the connections between timing board and transistor board. depending on the outcome of part G mechanical which will dictate just how many taps i have on electronic part G. thus i can modify the 8 shift register chip board to exactly what i need. it seems i underestimated the inductance so it looks like i will more then double the loop count which is more then what the electronic switching code can handle so it looks like i will have to tap every other loop and with the extra lines on the end it looks like 52 digital lines. double that for calculations is 104 up and back. if you take 16.666 divided by 104 you get 160.25 microseconds on time each. since the program can not do .25 i have to use 160 microseconds. 160 microseconds x 104 = 16.64 milliseconds x 60 x 60 you end up with 59.904 hz. i have to say even the power company can not do that good at .1 hz off. so this is my plan B moving forward with the wonderful program PaulRB of Arduino forum cc. helped me "US" the Figuera members with. working on greenhouse tomorrow then Sunday on part G mechanical. Regards, Marathonman
|
|
|
Post by Marathonman on Dec 21, 2020 14:54:15 GMT -6
Quote from Data sheet; "Data is shifted on the positive-going transitions of the SHCP input. The data in the shift register is transferred to the storage register on a positive-going transition of the STCP input. If both clocks are connected together, the shift register will always be one clock pulse ahead of the storage register."
I do not know if anyone has tried this to see if it would work but according to the data sheet the clock pins can be tied together so this would eliminate the need to bring pin 12 high then low in the code since this will automatically be done by the chip it self. it seems like this should work and would also save chip storage space making the code slightly smaller. it would also negate the need for a ss or cs pin from your Arduino.
the code i have provided can be tested with this scenario by eliminating the last three lines of code that make pin 12 go high then low and tying pin 12 to pin 14 on each shift register.
I do not have the time to test this case scenario out so i will ask one of the members if someone can test this to see if it would work.
Regards, Marathonman
|
|
|
Post by Marathonman on Dec 26, 2020 13:51:22 GMT -6
So correct me if i am wrong but according to an on line calc tool that 160 microseconds is only 6250 Hz. i was lead to believe it was a lot higher but in the case it was wrong. if this is correct i should have no problem using the 100 IGBT's i have for switching part G electronically even in a soft switching scenario with a high side driver. of course i will have to test this out on a demo board. working on part G rewinding today. Edit; No, calc tool was wrong as 1 microsecond is .000001 which is 1/1,000,000 millionth of a second. so on that note maybe the IGBT's are not able to handle the speed after all. it seems i might have to purchase N channel mosfets.
Regards, Marathonman
|
|
|
Post by Marathonman on Feb 21, 2021 9:05:43 GMT -6
Found another very good high side driver that is so small and simple to hook up. the IRS25752LPBF is a 600 plus volt high side driver in an SOT 23-6L package that is very simple to trace out. i have been working on a test board to try it out being so cooped up from the devastating storm that hit us. i have copied my previous transistor board then modified it for this chip and 1.25 board connectors from the Teensy controller board. the good thing is i found it at arrow for .88 cents each and .46 cents per 100. if things work out well on the test i will order them promptly. the program for the Teensy is so amazing as it functions perfect as planned so the addition of a good transistor part G tap board will be very complimenting. below is the transistor board i have remodified (work in progress) to get the best possible circuit i can get so many trials are at hand. as you can see i have complete separation of high and low voltage, high being on the outside and low being on inside. each group of signals has its own ground that is traced back to each individual shift register chip on the Teensy timing board. it also has 15 volt power and ground plans for the high side drivers that go high at 2.2 volts max so it is quite compatible with the Teensy at 3.3 volts. the high side drivers has schmitt- trigger inputs for noise and level translating. i could not find the 3D part for part G board connectors so i added the black ones. yes i know they hang over but they are not the ones that will be put on the board as they will be screw attached and fit in pattern. Regards, Marathonman
|
|
|
Post by Marathonman on Feb 23, 2021 21:02:11 GMT -6
Found the right connectors and 3D cad for the board, seems Phoenix contact came to the rescue again. board is now complete ready to be made then occupied right after two channel testing of the high side circuit. boards are expensive because of the 4 layers but i had to to get er done. ended up using Vishay SQP10250E_GE3 TO-220 Mosfets with the previously stated high side driver. boards are sheer beauty. Regards, Marathonman
|
|
|
Post by Marathonman on Mar 6, 2021 20:13:13 GMT -6
Two channel testing board is being made and will be testing soon. if all goes well i will put 20 channel transistor boards in production as well as the Teensy 4.1 timing board. moving is finished as well as the reorganizing of everything. will return to part G very soon.
Regards, Marathonman
|
|
|
Post by Marathonman on Mar 16, 2021 19:28:17 GMT -6
After some free time finally, i have decided to go with the Nexperia 74VHCT595PW,118 because it is in the TSSOP i need for the Teensy timing board but also has Schmitt trigger inputs that most do not have or not in TSSOP. it tops out at 170 mhz which is good headroom.
I can not move forward on the electronic timing until i finish the mechanical part G first. only then will i know exactly how many channels i need. it looks like i will have to double up on the winding's having two per channel as the 64 bit is dictating this requirement. so what that gives me right now is my part G will be right at 100 winds and the shift timing is at 64 so i will have to limit the timing channels to 50 and double up the winding's per channel. i will still have three channels on the ends to mimic the sine wave inductive roll off as the mechanical rotation is on for three to four times longer. the whole key to electronic timing in my book is to mimic the brush rotation exactly.
PS. try to avoid rectangle wire if you can, trust me when i say winding it is not for the faint at heart.
Regards, Marathonman
|
|
|
Post by Marathonman on Mar 18, 2021 4:57:29 GMT -6
It will obviously be much easier to build an electronic part G being you can use any shape closed core and on that note i think i will use a one kilowatt Variac core on the electronic version.
for one it will be much easier to calculate with the link to the inductance tool i posted using the parameters of the core and winding count. all you have to do is have the proper winding's and your finished. all Creasysee had to do was remove windings on both sides of his core relative to his transistor taps and he would have been in the sweet spot. the whole key is to just reduce the current to just clear the secondary then back to full potential as the other is reduced.
as soon as i finish the mechanical version i will purchase the one kilowatt core then attach the taps which is much easier then building mechanical from scratch. being i have the board designs already finished it should be a breeze.
Regards, Marathonman
|
|
|
Post by Marathonman on Mar 23, 2021 21:18:43 GMT -6
Good news today, i was tooling through my Transistor board design after a long days work to have the two channel test board made and i realized i can make the twenty channel high side board with only two layers. all this time i thought i could only get 15 volts to the high side drivers with two extra layers but hell i am only using the bottom for part G taps on the outside of the board so why not take advantage of the center space being empty.
this is great as now with two layers i can save a lot of money and still get high, low and digital separation.
Regards, Marathonman
|
|
|
Post by Marathonman on Mar 24, 2021 17:46:35 GMT -6
Two channel high side test board is being made and will be on it's way shortly. Epoxy for Mechanical part G rewind will be in nest monday.
Regards, Marathonman
|
|
|
Post by Marathonman on Mar 24, 2021 18:58:23 GMT -6
Another reason i chose the Nexperia 74VHCT595PW,118 is the VHC series logic level input was to high for direct Arduino switching which is above 3.3 volts. the VHCT series is set at 2 volts so the Arduino direct switching is capable. also the added bonus is Schmitt trigger inputs for any such noisy environment like on two layer board possibility. i am preordering certain parts from Mouser or Arrow like the caps, diodes, resistors, and voltage regulator. after the test i will decide if the high side passes muster then move forward. Mechanical part G will dictate the amount of channels on the Teensy electronic switching board.
Regards, Marathonman
|
|
|
Post by Marathonman on Apr 2, 2021 16:34:13 GMT -6
received the two channel high side test boards in today. i will work on those also along with part G. i used JLC this time as to try out the board quality. seems ok for now. i will b e trying it out on variac core for now to see how the switching performs. Regards, Marathonman
|
|
|
Post by Marathonman on Apr 5, 2021 17:01:45 GMT -6
Parts for electronic test board are on their way, will receive in the next few days along with a variac core for the electronic switching. if all goes well this week end i will know if high side drivers are good to go.
Regards, Marathonman
|
|
|
Post by Marathonman on Apr 10, 2021 6:35:25 GMT -6
Two channel transistor switching test board is finished soldered and i will begin testing shortly. high side driver was a real challenge to solder as it is a TO23-6 package but my skills and rework station came through.
also my 1960's variac came in yesterday, it seems the seller on ebay tried to pull a fast one on people by painting brown to hide battle scares but after testing it seems the core is just fine. i took it apart, cleaned and reassembled then after testing it passed with flying colors. I then hooked up to the power supply and light set up then swiped the brush back and forth with a wonderful unison rise and fall of lights.
I am in the process of making a video explaining the process of the active inductor controller with it's similarities to the Variac and also the process in which it operates. the Figuera device violates not one physics law thus it just the obscurity of the device being out of the norm that has people so thrown off.
Testing of the electronic switching will commence shortly. Regards, Marathonman
|
|