Author Topic: PWM Heating load controller based on frequency  (Read 24016 times)

0 Members and 1 Guest are viewing this topic.

methanolcat

  • Full Member
  • ***
  • Posts: 134
  • Country: us
PWM Heating load controller based on frequency
« on: February 06, 2011, 12:24:00 AM »
          Several here as well as myself have been looking for a way to hook up low resistance heating elements, (for heating air or water) and not stall their generator in low winds, since the heat elements look mostly like a dead short.

       Having not found anything, I decided to make such a controller that could allow the turbine to start up with no load, then once it is running, PWM the load onto the turbine using the turbines variable frequency.

       So basically, for those that may not understand, no load at start up, then once running the load will start to be applied, the faster the turbine runs the more load will be put on the turbine until the turbine can handle full load, at which point it gets the full load from set rpm's and higher.

       The rpm the load starts to be applied is adjustable with a potentiometer.

       The controller is run from the turbine itself so no outside power source is needed. A must I think.

       The controller also has an auxiliary voltage regulator, output adjustable up to 37v. (in my case to run 2 small computer fans 12v
           on bottom of heater)

       I also added a 20 LED bar graph on the right side of board which starts lighting from the bottom up with increasing RPM's.
          I designed it in bar mode with flashing full scale but once board was made I decided to change it to dot mode to save
          some power. (changes are easily made)

      The controller can take an input voltage of up to 250vdc (max for regulator) Fet's can handle 600v and 20 amps each, I used 6
          of them giving me a max of 120 amps (plenty overbuilt)

      I used a NTE995 frequency to voltage converter, it sends variable voltage according to frequency (RPM's) of turbine to a PWM built using a LM339 which sends its output to a mosfet driver.
 
      Most all components can be obtained through local electronics store's or ebay.

      The hardest part I think was finding some type of regulator that would work from such a wide range of input voltage, the regulator does get pretty hot so a good sized heat sink is a must on the transistors (max temp on them is 302 degrees F or 150 degrees C) I checked the heat sink at 150 degrees F and not a big deal since I am using the heat anyhow.

      If anyone has a better idea for a wide range voltage regulator feel free to speak up. Something from 30v to 200v at about 1.5A I think would be fine.

     I bench tested this controller then hooked up to turbine, it has been in use for about a week and has held up through 50mph winds we had last Tuesday night. I thought I had a problem with it today but it was a human error problem, lol. ;D

    My turbine starts powering the controller at about 30 RPM's since I am wound for high voltage output. If someone wanted to use this design for lower voltage turbines they could either use an auxiliary power source or use a boost converter, I did what works for me and it may well work for many others but feel free to change it to suit your needs if you like.

    I design pcb's with ExpressPCB a free download, and have the file for anyone that wants it, it shows all the part numbers and part values, ex. resistors, caps etc...

   If I have any problems or design updates I will post them and if anyone else has any ideas feel free to chip in.

    There is no schematic, I design from data sheets and prototype boards, a schematic could be drawn up but I prefer not to, it would be pencil and paper if I did. I think its pretty simple though some of you may think otherwise by looking at the finished board, it does look busy but its not bad. One could even eliminate over half of the board space if ya threw out the bar graph display and the aux regulator.

     Heres some pics, criticism and comments welcome.












 


zap

  • Hero Member
  • *****
  • Posts: 1107
  • There's an app for that
Re: PWM Heating load controller based on frequency
« Reply #1 on: February 06, 2011, 12:32:20 AM »
YOWZA!!! :o

Very nice!

I love the last pic... one can never have enough meters eh?

commanda

  • Hero Member
  • *****
  • Posts: 731
Re: PWM Heating load controller based on frequency
« Reply #2 on: February 06, 2011, 12:48:31 AM »
Would love to see you do a proper schematic. Would give a chance for others to follow your lead, or stand on the shoulders of giants, if you will.
Many people who do pcb design for a living, start with a schematic which nominates all packages, generate a netlist from this schematic, and feed that to an autorouter. Boards done this way are often easy to spot. The physical layout bears no resemblance to the schematic, and it's got about 10 times as many vias as would appear reasonable based on the component count.

What is your relationship between rpm & pwm. It sounds like it may be linear. I would have thought that ideally it needed to follow a cubic power law, or less ideally, could follow some sort of exponential curve.

I have played with the thought that the bar graph led driver could connect to an R-2R type ladder network to generate a step curve approximating the cubic power curve.

Amanda

methanolcat

  • Full Member
  • ***
  • Posts: 134
  • Country: us
Re: PWM Heating load controller based on frequency
« Reply #3 on: February 06, 2011, 03:34:36 AM »
commanda

          Not a proper schematic probably, did figure out how to use expresssch though. This looks way better than what I would have drawn.

         I obviously don't do pcb design for a living, I don't know about all that netlist and router stuff, I know enough to get by. I design in my head, read datasheets and design on prototype boards, there are many that know more than I that is for sure.

        The relationship between rpm and is linear, and I am not getting all I can I'm sure, but I know I am getting a lot more than I was with a stalled turbine, having to wait until wind speeds were high enough to break it loose from being overloaded. I don't claim to have the "end all, beat all" controller here, I'm just sharing with everyone what I do have and if many are stalled as bad as I was I think they would appreciate something like this.

        I didn't draw the regulator in the schematic but I have it separately that I can post, also didn't bother with the bar graph display as it isn't really something you or others havn't seen. LM3914 datasheet has it in there. Also its not what makes this what it is.


       If you would like to add or see something that could be changed for the better please let me know. For now this works ok.

2100-0

methanolcat

  • Full Member
  • ***
  • Posts: 134
  • Country: us
Re: PWM Heating load controller based on frequency
« Reply #4 on: February 06, 2011, 03:39:04 AM »
Forgot the regulator.  Both R31 and R32 are 10k and the zener is 15v instead of 16v.

2101-0

commanda

  • Hero Member
  • *****
  • Posts: 731
Re: PWM Heating load controller based on frequency
« Reply #5 on: February 06, 2011, 04:21:34 AM »
I suspect there are a few errors in your schematic. That top 339 is the strangest looking oscillator. It's not obvious (to me, at least) how it is oscillating.
And the LM2917. Normally pin 5 would go to positive, and the output voltage would come from pin 4.

I'm sure your regulator would work fine with one (high voltage) transistor. Don't understand your spec of needing 1.5 amps from it.
Peak gate current  for the switching fet might be this high, but that's just nanoseconds to charge the gate capacitance.
Average current is very low. Wouldn't expect more than a few tens of milliamps. Peak gate current should be supplied by the low ESR capacitor on the output of the regulator.  I've had good success on many projects using basically your circuit as a pre-regulator to a standard 3 terminal regulator (7812).

Amanda

wpowokal

  • Hero Member
  • *****
  • Posts: 1271
  • Country: au
  • Far North Queensland (FNQ) Australia
Re: PWM Heating load controller based on frequency
« Reply #6 on: February 06, 2011, 04:48:52 AM »
Is the LM 2917 wrongly labled should it not be NTE 995?

I like what you have done, I can see a slightly different application for this circuit if I can cap the frequency.

allan
« Last Edit: February 06, 2011, 04:59:46 AM by wpowokal »
A gentleman is man who can disagree without being disagreeable.

methanolcat

  • Full Member
  • ***
  • Posts: 134
  • Country: us
Re: PWM Heating load controller based on frequency
« Reply #7 on: February 06, 2011, 10:47:45 AM »
commanda,

           No errors when it comes to the 339, strange, maybe, but it works. I used what I had on hand and I am only using half of the 339.  LM339, 3 diodes,3 resistors a pot and 2 caps,  probably common parts you have on hand, try it.

          The LM2917 in the schematic, I did use the 14 pin model which is LM2917N or NTE995 which wpowokal pointed out but the schematic program only had the 8 pin version. Looking at the datasheets for the LM2917 the pin out, other than pin numbers is pretty much the same. Pin 3 and 4 of the LM2917N is the same as pin 3 of the LM2917 when tied together, pin 11 and 12 of the LM2917N are also the same as pin 8 of the LM2917 when tied together, which I have done in both cases and could have used the smaller 8 pin model LM2917 but the LM2917N is what I used.

         About the regulator, I am running the circuit, 2 fans (200ma each = 400ma), the LED bar graph (when used as full bar mode, 20 LED's X 20 ma each comes in at 400ma, the big reason I changed to dot mode) plus the mosfet driver, add those up and getting close to an amp there and since I always overbuild to some extent thats why I said "about 1.5 amp", I would never max out a regulator if I didn't have to, but I'm no electronics expert, just self taught so maybe I am wrong.

         When I used just 1 pass transistor in the regulator it quickly blew for some reason, I had used the 2 pass transistor regulator before in another project, running the circuit, a 9 amp driver and 21 fets and had no problems with it so I decided to use it here as well.



wpowokal,


           Yes the LM2917 in the schematic I drew could be the NTE995, but different pin out. I did use the LM2917N (which is the same as NTE995) in the circuit.  Looking closely at the chip I have NTE printed their part number over the top of what was already etched into the chip LM2917N.

           "cap the frequency" ??  If you mean get 100% pwm at a specific frequency and higher, then yes, this circuit can be adjusted that way.   This circuit for me, is adjusted to start pwm at around 30 RPM's  and hits 100% pwm at around 100 RPM's. Add another pot in there and maybe a couple resistors and I think the low end and high end could be adjustable independently.

           

ghurd

  • Moderator
  • Super Hero Member Plus
  • *****
  • Posts: 8059
Re: PWM Heating load controller based on frequency
« Reply #8 on: February 06, 2011, 11:44:53 AM »
Wow!
G-
www.ghurd.info<<<-----Information on my Controller

BrianSmith

  • Jr. Member
  • **
  • Posts: 94
Re: PWM Heating load controller based on frequency
« Reply #9 on: February 06, 2011, 02:15:31 PM »
Your board looks really good.  Thanks for sharing.

fabricator

  • SuperHero Member
  • ******
  • Posts: 3394
  • Country: us
  • My smoke got out again
Re: PWM Heating load controller based on frequency
« Reply #10 on: February 06, 2011, 02:34:28 PM »
Wow!
G-

Yeah, what Glen said :D  Very nice.
I aint skeerd of nuthin.......Holy Crap! What was that!!!!!
11 Miles east of Lake Michigan, Ottawa County, Robinson township, (home of the defacto residential wind ban) Michigan, USA.

Boss

  • Hero Member
  • *****
  • Posts: 551
  • http://outfitnm.com
    • Outfit Renewable energy site
Re: PWM Heating load controller based on frequency
« Reply #11 on: February 06, 2011, 05:29:08 PM »
Yes, beautiful craftsmanship, and it works as well? the fans help move the heat past the heater?
This is a very encouraging project for us amateurs, "just read the data sheets" I love it.

 
Brian Rodgers
My sustainable lifestyle site http://outfitnm.com no ads, not selling anything either

wpowokal

  • Hero Member
  • *****
  • Posts: 1271
  • Country: au
  • Far North Queensland (FNQ) Australia
Re: PWM Heating load controller based on frequency
« Reply #12 on: February 06, 2011, 06:03:16 PM »
I had read your description where you used a NTE995 and the schematic showed LM2917, confused me.

By "capping the frequency" I can see an application for a turbine with say two windings/stators one wound for low winds and one for higher, to me your idea could be a way to protect the low wind winding by capping it's output current, anyway you have filled in the first box on my doodling, it previously just read frequency to voltage.

Even your regulator is just what I need for an automatic gate opener, I know these circuits are out there but there is nothing like being slapped in the face with a cold mullet to awaken one to different ways of doing things, anyway keep up the good work.

Allan
A gentleman is man who can disagree without being disagreeable.

SparWeb

  • Global Moderator
  • Super Hero Member Plus
  • *****
  • Posts: 5452
  • Country: ca
    • Wind Turbine Project Field Notes
Re: PWM Heating load controller based on frequency
« Reply #13 on: February 06, 2011, 07:53:42 PM »
From datasheet to PCB with no schematic?!?  I am totally lost without one.  With one I am only slightly less lost.

I can't use it today, but I'm definitely saving a copy of your posting.  Someday I will need one, but more importantly someday I may have learned enough to understand something so useful!

Thank you!
No one believes the theory except the one who developed it. Everyone believes the experiment except the one who ran it.
System spec: 135w BP multicrystalline panels, Xantrex C40, DIY 10ft (3m) diameter wind turbine, Tri-Star TS60, 800AH x 24V AGM Battery, Xantrex SW4024
www.sparweb.ca

SparWeb

  • Global Moderator
  • Super Hero Member Plus
  • *****
  • Posts: 5452
  • Country: ca
    • Wind Turbine Project Field Notes
Re: PWM Heating load controller based on frequency
« Reply #14 on: February 06, 2011, 08:02:22 PM »
Did you post a list of parts?  If it's not too much trouble, could you upload the file? 

Thanks again!
No one believes the theory except the one who developed it. Everyone believes the experiment except the one who ran it.
System spec: 135w BP multicrystalline panels, Xantrex C40, DIY 10ft (3m) diameter wind turbine, Tri-Star TS60, 800AH x 24V AGM Battery, Xantrex SW4024
www.sparweb.ca

methanolcat

  • Full Member
  • ***
  • Posts: 134
  • Country: us
Re: PWM Heating load controller based on frequency
« Reply #15 on: February 06, 2011, 10:17:56 PM »
zap,

      thanks, and lol, yeah, one can never have enough meters. Ya know, my favorite one pictured there is yellow and gray one on the right, radio shack model with rs232 port on back for data logging of any measurable input the meter has, kinda handy. The reason I bought the other yellow one in the pic is because it has a capacitance meter and a k type thermocouple as well for measuring temperature, both quite handy. And then theres the far right analog meter, everyone needs to have at least 1 analog type meter. And last the clamp on amp meter, also very handy at times.


Boss,

      thanks, and yes it works quite good. Yes, the fans are under the heater, blowing upwards through the nichrome wire elements. The heater is open on the bottom which sets about 2 inches off the floor with 1/4 inch wire between the legs to keep anything from getting under it.  The heat comes out the top, I made it this way so that if one or both fans were to quit working, hopefully the heat rising normally would help bring in air from the bottom and help to keep the unit from getting too hot, where if it was on its side the metal case would definitely get quite hot. I tested it with the fans off and the case got warm but not too bad.


SparWeb,

      Not quite from datasheet to pcb with no schematic, I went from idea, used some previous knowledge of pwm circuits and freq to voltage converters (which is pretty much the main thing here) and added the two together. I laid out both on a breadboard, connecting them with some adjustability so I could fine tune it to my turbine, or anyone elses that wants to use it. From breadboard to pcb design (which takes hours to layout and route traces to minimize jumpers, the very time consuming part many dont realise or never see).

      I had an idea, I knew others had the same problem as I. I wanted it to be all analog components so that anyone could build it and be able to replace parts if need be without having to program any chips. 

      I will post a list of parts, need to make a list first. I will also upload the file, let me number all the components first so a parts list would make since as to what part is what, R1, R2, R3, C1, etc.......The file I have shows all the component values but it would be correct to give them all a number as well.

      I will also make a modified version of it and upload its file as well. Modified version would be less the display circuit for those that didn't want to use that part of it. The board would be somewhat smaller in size.

      I will also try to post a larger image if I can figure out how, something more legible, with outlines of the different sections of the circuit,........ Input filter, Regulator, Auxiliary regulator (for fans or whatever), Display circuit, and the frequency to pwm circuit.

     Give me a couple days or so to get it all going.

SparWeb

  • Global Moderator
  • Super Hero Member Plus
  • *****
  • Posts: 5452
  • Country: ca
    • Wind Turbine Project Field Notes
Re: PWM Heating load controller based on frequency
« Reply #16 on: February 06, 2011, 11:37:53 PM »
Take your time.  This is worth doing well.
The work you've put into it already is greatly appreciated, too.
No one believes the theory except the one who developed it. Everyone believes the experiment except the one who ran it.
System spec: 135w BP multicrystalline panels, Xantrex C40, DIY 10ft (3m) diameter wind turbine, Tri-Star TS60, 800AH x 24V AGM Battery, Xantrex SW4024
www.sparweb.ca

madlabs

  • Hero Member
  • *****
  • Posts: 686
  • Country: us
Re: PWM Heating load controller based on frequency
« Reply #17 on: February 07, 2011, 11:21:48 AM »
To echo the others, wow, looks great. Any chance of getting the gerber  file? My buddy has a CNC router and the layout looks easily within it's capabilities. I'm approaching the need for a mill controller myself.

Jonathan

zvizdic

  • Full Member
  • ***
  • Posts: 180
Re: PWM Heating load controller based on frequency
« Reply #18 on: February 07, 2011, 09:13:42 PM »
This is what i wish to have but no skills to design .

I have  a 48V 10' mill with 9 6V cells to solve stalling .

What woos the voltage you start with ?

WOW is the word !!!!!

methanolcat

  • Full Member
  • ***
  • Posts: 134
  • Country: us
Re: PWM Heating load controller based on frequency
« Reply #19 on: February 07, 2011, 11:38:44 PM »
madlabs,

         Sorry, there is no gerber file, the software I use, ExpressPCB uses their own type of file that is not compatible with anyone elses . What they want is for you to use their software, design your board, and have them to make your finished pcb's. They don't want you to use their free software and have someone else make the pcb, which I can understand.

        I didn't know this when I started using their software, and rather than learn a new software which spits out gerber files I have just kept using their software. Its not all bad though, for the prototype kind of person only needing a board or 2, the software allows the design to be printed out. I either print out the design using the toner method of making boards or I print onto transparencies and use photo sensitive boards, the toner method is much cheaper and I have had good results, if it screws up you just clean the board and start over, however I have trouble making narrow traces with toner method. If I have a pretty compact design then I use photo sensitive boards.

methanolcat

  • Full Member
  • ***
  • Posts: 134
  • Country: us
Re: PWM Heating load controller based on frequency
« Reply #20 on: February 08, 2011, 12:14:50 AM »
zvizdic,

     I'm not sure if this circuit would do what you want as it is, some changes might be able to made to make it work though, I would have to do some serious thinking about it.

    The thing is that with pwm, anything under 100% pwm will be a lower average voltage. For example, if the voltage going into the controller was 48v and the controller was at 50% pwm then the average voltage out would be 24v, in a case such as mine, having a heat element as a load the heat element will take and use any voltage level it see's, whether it is 24v or 200v, doesn't matter, unlike a battery bank which clamps the voltage and nothing happens with lower voltage until it is above the bank charging voltage.

     I'm not saying it won't work, but I'm not saying at this point that it will work.  Its something I will work on though, I'm kinda covered up right now though with a few other projects that I am working on, one being a pure sine wave inverter using all analog parts in the 5000 watt range for a 48v bank, more on that at another time.

    There are many others having the same problem as you and opening the air gap on axial flux designs or adding resistance to the line, both methods create some loss of power. I know, there has to be a better way, and I'm sure there is. I just need to set up a test bed and go to work on it.

     Someone correct me if I am wrong.

methanolcat

  • Full Member
  • ***
  • Posts: 134
  • Country: us
Re: PWM Heating load controller based on frequency
« Reply #21 on: February 08, 2011, 12:33:03 AM »
zvizdic,

     After thinking about this for a minute, is your mill putting out too high of a voltage at low rpm's?   If so then this controller may be just what you want, bleed off some of the voltage through pwm at lower rpm's to keep out of stall, then as rpm's rise so does the pwm, putting you where you want to be.

    If this is the case, I would lower the resistance of 2 resistors in the regulator circuit to allow for the lower voltage your using, or change the regulator to use just one pass transistor, anyhow re-design the regulator, and use mosfets more suited to your voltage and current levels and leave everything else as it is. Both simple changes.

    I'm pretty sure this is what you were asking since a higher voltage mill would stall on a lower voltage battery bank.

zvizdic

  • Full Member
  • ***
  • Posts: 180
Re: PWM Heating load controller based on frequency
« Reply #22 on: February 08, 2011, 07:45:56 PM »
Yes, it is putting out too much voltage when the batteries are at low voltage which slows the turbine down.

If I understand correctly at low rpm controller loads a mill whit slower pulsing and at high rpm pulses are closer together until is on steady.

Also I could use heating  in winter months.

Building your controller would bi a learning experience.

 
 
 
« Last Edit: February 08, 2011, 08:54:36 PM by zvizdic »

methanolcat

  • Full Member
  • ***
  • Posts: 134
  • Country: us
Re: PWM Heating load controller based on frequency
« Reply #23 on: February 08, 2011, 09:08:55 PM »
zvizdic,

         Your thinking is pretty close, however the switching speed always remains the same, this controller switches at 18Khz (which is 18,000 times a second).   At lower turbine rpm's the pulse width (the amount of time that the switch is on) is very short, allowing the turbine to start up, as the rpm's go up the pulse width is longer, giving the batteries more power, also starting to load the turbine harder, if it trys to load the turbine too hard, the turbine would slow some and the pulse width would go down as well, unloading the turbine some and allowing it to speed back up. So the controller and the turbine would be working together, helping each other out so to speak.

         The switching pulse width goes from 0 to 100% and it gets to 100% at fairly quickly as the turbine speeds up, for example, pwm may be 20% at 50 rpm's and 100% from 100 rpm's and higher, just throwing numbers out here as the pwm start speed is adjustable, could be lower or even much higher.

        The point at which the controller starts to work for you is adjustable through a potentiometer, get it set and you shouldn't have to change it again, unless you change magnets or stator configuration on the turbine.

zvizdic

  • Full Member
  • ***
  • Posts: 180
Re: PWM Heating load controller based on frequency
« Reply #24 on: February 08, 2011, 09:49:18 PM »
If you would sell a kit I would be first to buy.

Maybe you could sell a board and help me with the changes for my voltage, I would appreciate it.

I'm also thinking of improving my blades and adding a little dip into the front face to imitate GOE222 airfoil to improve the torque of the blades.


methanolcat

  • Full Member
  • ***
  • Posts: 134
  • Country: us
Re: PWM Heating load controller based on frequency
« Reply #25 on: February 08, 2011, 10:24:52 PM »
                                                            PWM Heating Load Controller Parts List


       Here are the files and a parts list I said I would post, I even looked up component part prices through mouser and digikey and have come up with, if you didn't have any components on hand at all (resistors, capacitors, diodes, etc....) it would cost around $60 to buy what is needed. Some of the parts are even available at radio shack, (resistor assortment 1/4 watt $10 to $12, diodes = cheap, bargraph displays (2 needed, I got mine from radio shack, like $1.50 each) potentiometers, capacitors, regulators, etc......

       A small transformer is going to be needed, I scavenged mine from an old microwave control board, NOT the big heavy one inside, but the small one on the control board. Input voltage is 120v and output voltage is 12v. Try to find one with a primary winding resistance of 500 ohm or higher, mine was a bit over 700 ohms, I found that the lower resistance transformers acted like they were trying to load the turbine a bit and would overheat the transformer I am sure. The transformer hooks up to 2 of the 3 ac lines from the turbine and is needed to drop turbine ac voltage down to a level where the circuit can sense the frequency (rpm's) that the turbine is running.

      Also an inductor is going to be needed for an input filter choke, I just took an old inductor core, iron I believe 36mm O.D. (just under 1.5 inches). I'm not really up to snuff on toroid cores and choke design, I made a few up with different number of number of turns and tested them to see how turns count affected them and ended up using 40 turns of 17ga. magnet wire on the choke. I think its a hit and miss kind of thing with turbines as the frequency is usually quite low and is also variable. The choke is just to help filter out the ripple from the rectified D.C. from the turbine as most integrated circuits don't like too much ripple voltage.

      The mosfets I used I bought from mouser, SPW20N60S5, 6 of them at $3.12 each and they currently have hundreds more, digikey doesn't have them and is saying they are obsolete and won't be getting any more. There are many, many different mosfets that could be substituted but I searched for ones that could do the job and were cheap as well. Good luck finding a cheaper substitute with similar specifications. Many I looked at were in the $5 to $10 range.





These 2 files can only be opened with ExpressPCB program from their website, free download.
 
Open the files to view them, use mouse wheel to zoom in and out and if your mouse wheel clicks when you press down on it, you can drag the pcb around to view while zoomed in, that really helps.

To print the layout, go to upper left corner and click "File" and go down the list and click "Print", window opens, select box next to "Bottom copper layer", Do Not click "enlarge to fit page" and then click "ok".

Print and check that it was printed full scale, get that print copied with a toner copier, dark toner on like thin magazine paper and iron onto a clean bare copper board, then etch and drill and populate.

* pwm mill controller.pcb (52.12 kB - downloaded 312 times.)

The second file here is the controller without the display for those wanting a smaller, and a little bit cheaper controller.

* pwm mill controller no display.pcb (34.41 kB - downloaded 243 times.)

wpowokal

  • Hero Member
  • *****
  • Posts: 1271
  • Country: au
  • Far North Queensland (FNQ) Australia
Re: PWM Heating load controller based on frequency
« Reply #26 on: February 10, 2011, 04:20:31 AM »
OK I waited but seems I am the only one, thank you for the component list but it does not agree with your published schematic, what have I maissed again?

allan
A gentleman is man who can disagree without being disagreeable.

methanolcat

  • Full Member
  • ***
  • Posts: 134
  • Country: us
Re: PWM Heating load controller based on frequency
« Reply #27 on: February 10, 2011, 08:18:46 PM »
wpowokal,


         Not entirely sure what you mean unless your talking about the schematic in reply #3 and if you are the only difference is that the schematic drawing program I used didn't have the LM2917N or NTE995, (both the same chip) but they did have the LM2917 which is the 8 pin version of the exact same thing but with 2 sets of internally grouped pins which I explained in reply #7.

        This taken from post #3, reply to commanda,    "I didn't draw the regulator in the schematic but I have it separately that I can post, also didn't bother with the bar graph display as it isn't really something you or others havn't seen. LM3914 datasheet has it in there. Also its not what makes this what it is."

        The regulator is not in the schematic but is posted in reply #4. The bar graph display is also not in the schematic, but it is in the datasheet for the LM3914N, here.   http://www.datasheetcatalog.org/datasheets/400/50332_DS.pdf
The top of page 12 shows the schematic for bar display with alarm flasher.

        The parts list agrees with the circuit board I built and that is the file "pwm mill controller.pcb"  posted in reply #25, if you open the circuit file you will also see that every part has its parts list number or value either on it or next to it .

                Hope this clears it up for you, also here is a more legible copy of the "parts list" can also be opened with the same program from ExpressPCB.

   * parts list.pcb (4.25 kB - downloaded 191 times.)


Also in the pcb file it shows exactually how to wire it up to your windmill rectifier and load, the top right side of board there is a place to put a switch so that the display can be turned off, or you can just put a jumper wire there and leave it on all the time.

Connections on the left side of the board are as follows,

       top 2 connections go to any 2 a.c. phases from your windmill

       the next 2 down labeled 5k pot, need to go to a 5k potentiometer if you want to hook external fans as I have, the 5k pot varies the output voltage to the next 2 lower down connections labeled fan+ and fan-      this voltage can only go up to 15v, which would not harm the fans but make them run faster than they would at 12v, this voltage can be turned down as well to slow the fans, all adjustable by the 5k pot.

       the next 2 down fan+ and fan-   see above.

       the next 2 down both go to pwr in + from your 3 phase rectifier, either one or both can be used.

       the 2 lowest connections, which you need to make some simple power rails from copper or aluminum, I used aluminum, the top one is - or GND from 3 phase rectifier and the bottom one is pwm output to heater or load, DO NOT MIX THESE 2 UP, MAKE SURE THE TOP RAIL GOES TO GND AND THE BOTTOM GOES TO HEATER OR LOAD.

    I will post a picture of the back of mine so you can see the power rails bolted on to the board.

    An alternative to the power rails could be bare copper wire soldered directly to the full length of the trace, I think making some rails would be better in cases where you are using high current but copper wire soldered on could be substituted.


      Matt

methanolcat

  • Full Member
  • ***
  • Posts: 134
  • Country: us
Re: PWM Heating load controller based on frequency
« Reply #28 on: February 10, 2011, 08:40:03 PM »
zvizdic,

      I don't have the money needed to start up a kit, I would have to spend about $60 of my own money per controller, maybe less if I was to buy in bulk, if I was to buy in bulk I would think no less than 10 kits worth of parts and were looking at probably around $500 and I sure don't have that to spare.

      I possibly could make and sell you a board and make the changes for your lower voltage but I have no experience sending anything out of the country, not sure where you are located. Changes would be pretty easy to make but I would need some time for testing before I made a board, I wouldn't want to send you something that didn't work or failed on you. Thats one thing that really gets me and I hate it "things you spend hard earned money on that are junk or quit working because of poor research and testing", thats one reason I make so much of my own stuff, I tend to overbuild but it always lasts.


Matt

zvizdic

  • Full Member
  • ***
  • Posts: 180
Re: PWM Heating load controller based on frequency
« Reply #29 on: February 10, 2011, 10:39:32 PM »
I'm from Toronto Canada and I understand it well be to much and to early .

What would be your battery charging voltage?

Would you please make changes for us with 48 V I well test.

To me this is a hobby as for lots of other people here so I do not mind doing-et .

Thanks for your efforts to help

Ungrounded Lightning Rod

  • SuperHero Member
  • ******
  • Posts: 2865
Why not just use series capacitors?
« Reply #30 on: February 10, 2011, 10:56:53 PM »
Just out of curiosity, since the stated target was avoiding stall at low winds:

Why all the active components?  Why not just get three motor run capacitors and put them in series with the lines from your PMA to your resistive heating load?  (Or one cap if it's a single-phase PMA?)

No current at startup, current ramps approximately with square of RPM at first (with power at fourth power of RPM), then knees over towards first power of RPM (power rising with square of RPM) around the time-constant frequency, with the rising resistance of the heaters folding it over farther.

It wouldn't be perfect MPPT.  But it would give you your unloaded startup and be a reasonable engineering approximation at most useful wind speeds.

SparWeb

  • Global Moderator
  • Super Hero Member Plus
  • *****
  • Posts: 5452
  • Country: ca
    • Wind Turbine Project Field Notes
Re: PWM Heating load controller based on frequency
« Reply #31 on: February 10, 2011, 11:12:36 PM »
Cat,
Thanks I knew you'd come through.  I had a bunch of questions when I first looked at the list but it looks like Allan's already asked and you've answered.


ULR,
I've heard of that but don't the caps get hot?
No one believes the theory except the one who developed it. Everyone believes the experiment except the one who ran it.
System spec: 135w BP multicrystalline panels, Xantrex C40, DIY 10ft (3m) diameter wind turbine, Tri-Star TS60, 800AH x 24V AGM Battery, Xantrex SW4024
www.sparweb.ca

methanolcat

  • Full Member
  • ***
  • Posts: 134
  • Country: us
Re: PWM Heating load controller based on frequency
« Reply #32 on: February 10, 2011, 11:44:54 PM »
Ungrounded Lightning Rod,


        A couple reasons I guess, first one being I think I would need 3 banks of capacitors, not just 3 capacitors (which could get expensive) at my power levels, this is for a 15 foot diameter mill with a generator rated at 4kw in 25 mph winds, we had 50 mph winds last week, so how much power was that? hard to say, much more than 4kw. (I have no furling system, let her fly and see what she can handle, it took the 50 mph winds fine, still flying today), capacitor banks would have probably gone off like huge firecrackers.

        I have no idea where to even start when it comes to what mf values to even begin with and how expensive would it get finding out what worked and what didn't.

        Reason 2, I had most everything I needed to build the circuit on hand already, my electronics bins make radio shack look like child's play. I don't have much money for this hobby right now and didn't have to spend much to get this working.

Matt