Author Topic: PWM Heating load controller based on frequency (Read 24165 times)

Madscientist267 · « **Reply #33 on:** February 11, 2011, 12:15:30 AM »

Quote

thats one reason I make so much of my own stuff, I tend to overbuild but it always lasts

And you know what's going on with it!

Sometimes, commercial equipment leaves you asking WTF about the designer when you run across one of their idiosyncratic 'features' or something...

Chances are, if you designed, built, and tested it, you've got a really good grasp on what is happening inside of it.

Unmatched by the commercial world.

Steve

wpowokal · « **Reply #34 on:** February 11, 2011, 06:02:57 AM »

Matt I know I am a stupid old F##t, I get told daily but in your schematic "pwm mill controller.pcb" R2 is 12K and on your list of parts R1,R2 R3 =680 ohm
R 4 is listed as 100k and on your list of parts 10K 5 W so and so forth.

Clearly I have missed something, I can accept that, in any case you have given me great food for thought so thank you.

allan

methanolcat · « **Reply #35 on:** February 11, 2011, 04:52:31 PM »

wpowokal,

I understand what you are saying now, took me a bit to figure it out because you was confusing me as well, you said,

"Matt I know I am a stupid old F##t, I get told daily but in your schematic "pwm mill controller.pcb" R2 is 12K and on your list of parts R1,R2 R3 =680 ohm
R 4 is listed as 100k and on your list of parts 10K 5 W so and so forth."

And I was confused by you saying "schematic" and "pwm mill controller.pcb" What confused me was that they are 2 different things, "pwm mill controller.pcb" and "pwm controller.jpg". One is the board layout showing all components and values as they are on the board, and the other is just the schematic. But you are right, The list I made was from the pcb layout and although the schematic has some but not all of the components as the pcb, they are labeled differently. The only reason I posted the schematic was because commanda wanted to see how it functioned and I HAD to give the components numbers in the schematic or the software wouldn't let me finish it. So yes, same components, different part numbers than what is in the schematic, I would change the schematic so that it was the same as the pcb layout numbers if I knew how to swap it out, might try and edit that post and see if I can upload a new changed file with the same part numbers.

I somehow knew that schematic was going to get me in trouble somehow, never drawing them for myself and all. I've made quite a few boards and projects, some pretty involved making this one look simple and have never drawn up a schematic until now, so forgive me for that mistake.

Matt

methanolcat · « **Reply #36 on:** February 11, 2011, 05:31:46 PM »

Ok, seems I cannot edit the old post so here is a new "schematic" with part numbers matching what is on the pcb and parts list.

Note, this is only the controller and not the bar graph display or the main or auxiliary regulators.

Matt

Madscientist267 · « **Reply #37 on:** February 12, 2011, 08:55:58 AM »

Quote

schematic was going to get me in trouble somehow, never drawing them for myself and all

I'm as guilty as you; people don't understand how it can be done. I usually don't bother with one until I'm done either, because the parts can change significantly as development progresses, and I get tired of redrawing them. When I do make them, I always draw by hand rather than on a screen, because I have a tendency to take them with me for study when there is a bug but circumstances don't allow me to work on the circuit itself. It's also easier reference later down the road when I need to look back at one to do a mod or something.

It's apparently a rare breed that does things this way, based on what I'm gathering (mostly out of context from the crowd here)... Dunno what else to say about it.

The valid point made by others however (which I agree with) is that without a schematic, it can be very difficult if not impossible to follow another's footsteps.

Regardless, it looks like you've got a design going here that works well for you. 'Grats man! Maybe it will catch on!

Steve

methanolcat · « **Reply #38 on:** February 12, 2011, 10:01:26 AM »

Madscientist267,

Thanks for the kind words, at least I'm not the only one designing electronic circuits this way, I would have been the lonesome odd ball out by myself, now I know there is at least two of us.

Designing that way is easy for me, and since it is, I assume others can look at what I have done and see what I saw without a schematic and that isn't the case at all apparently.

I'm not sure I would change much on the circuit at this point as I think it works good for me as it is and its adjustable so it can work for others as well. I do not like that the regulator gets as hot as it does but since I am using the heat it doesn't really bother me.

Matt

joestue · « **Reply #39 on:** February 12, 2011, 10:25:23 AM »

couple questions, i'm not seeing the toroidal inductor in the schematic. and where's the snubber for the fets?
(You could use a 330v mov since you have 600 volt fets.)

Madscientist267 · « **Reply #40 on:** February 12, 2011, 03:38:23 PM »

Quote

regulator gets as hot

Got the feeling your regulator is getting so hot because it is dissipating a lot of power even though the current levels are reasonably low.

If you're dropping 105V (120-15) across a transistor (or two, irrelevant) at 1A, you're over 100W of power dissipation! Definitely explains the heat!

Since you're not worried about it in terms of efficiency ('using' the heat, as you said), it might not be a big deal here, but while what you have works on paper, and to a limited extent, on solder, you've kinda pushed the limits as far as you can take it.

Don't forget - an amp out of a linear regulator also reflects an amp in! It's easy to see the 15W at the output... equally easy to forget that the rest is being turned into supercompressed magic smoke just itching to get out of your pass tranny...

I'd use another small transformer if you have one, and regulate the output from that instead, just for reliability purposes. As you know, if your 15V supply goes, the rest becomes useless until it reappears.

Steve

methanolcat · « **Reply #41 on:** February 12, 2011, 03:50:43 PM »

joestue,

Your not seeing it because its not there, in reply #36 it clearly states "Note, this is only the controller and not the bar graph display or the main or auxiliary regulators."

I gave a description of the inductor in the parts list and it doesn't have to be exact, close would work fine.

The snubber in this case (diodes) is inside the heater connected there, probably not the best choice but my lines are short going to the load so I'm not worried about the inductance of them. Theres also protection for each mosfet internally, I scoped the output to the heater and didn't have any harmful spikes showing up but I did put diodes across the load just because its good practice and diodes are cheap anyway.

Matt

methanolcat · « **Reply #42 on:** February 12, 2011, 04:07:17 PM »

Madscientist267,

You make a good point about the smoke wanting out and the rest of the circuit being useless after that, thats why I used such a big heat sink on the pass transistors, for now I am well within the temperature limits of the transistors even though they don't like the punishment. Thats really the only thing I don't like about the circuit, Later on (spring, summer and fall) I am going to be charging batteries with the turbine and the heat would really pose a problem, however I can reconfigure the winding connections for lower voltage (charging 48v) and the regulator would have no problem with it then (second circuit, same as first, different resistor values in regulator).

Its hard to find a regulator that works well over such a wide range of input voltage, a transformer would help if it could be used, however I think the low frequency from the turbine would cause problems for a transformer. I wanted to keep this controller under power from the turbine only and not some auxiliary power source that may go out or cause other problems, unloading the mill.

Matt

Madscientist267 · « **Reply #43 on:** February 12, 2011, 04:30:24 PM »

Quote

auxiliary power source that may go out or cause other problems, unloading the mill

Eek. Yeah, but if you're pushing the limits to the edge of reliability...

Tomato, Red Squishy Fruit...

(nope that's not going to work either... LOL just doesn't sound right)

What kind of temp are you seeing at the heatsinks (more accurately, the tabs of the pass trannies)...?

I hear you on the keep-it-simple front, but personally I don't like heat where I haven't intentionally put it. Too many crash-courses ending questionable reliability with confirmed deadness, if you catch my drift.

There are SMPS designs that can use a very wide input voltage range, and some you can probably stretch even further. Not uncommon to see 100-250V as an input spec on some of these 'switching bricks'. A few mods may be able to get you down in the couple-dozen to buck-fiddy range so that you're closer to your operating window. Those little 120VAC -> 5VDC wall-wart type switchers for cell phone charging are a good place to look. Doesn't take much to tweak them to the voltage you want, and some of them are capable of an amp or more without further modification.

From there you could pan it out by charging a small SLA or similar to run the logic so that the controller stays 'hot' regardless of what the mill is doing.

Just thoughts...

Steve

methanolcat · « **Reply #44 on:** February 12, 2011, 05:51:01 PM »

Madscientist267,

The max temp on the transistors is 302 degrees F or 150 degrees C, I checked the heat sink as close as possible to the tabs at 150 degrees F.

I've not seen any kind of smps that works from 30v to 200v and thats kind of the range I was looking for. I could charge a small sla from the wall and hook the circuit into it, circuit would be drawing from grid as long as grid was up and battery would last long enough to give me time to either fully load the mill (mechanical bypass of controller) or shut it down, I was kinda trying to avoid something like that though.

I'm not real good when it comes to power supply (design) side of electronics. The electronics field is so big that one usually picks an area or two and specializes in those areas, mine being what I would call "controls and dc power management". You need a 25kw or even bigger dc motor control, I can do that, from design to pcb layout, power supplys, smps, etc... ka - put, need help, or tons more research time and hands on.

The big thing for me was CHEAP and LOW PARTS COUNT, I didn't want to resort to buying pre made smps, as that could be troublesome for some obtaining the same smps I used, and didn't want to make one either as that makes the part count go up, not so much part count as much as research on my part, and in a short time before this winter kicked in, so I did what I did, and would/ could make changes to the regulator in the future if I could get some help or research time.

I may fiddle with some cell phone chargers and see if I can come up with something adaptable without too many changes.

Matt

ghurd · « **Reply #45 on:** February 12, 2011, 07:48:40 PM »

Can get 100-240VAC input, 12VDC 500ma lighter outlet output things for cheap.
Example, $1.71 delivered, ebay #220735076282.

I ordered one a couple weeks ago, not here yet.
G-

Madscientist267 · « **Reply #46 on:** February 12, 2011, 08:05:25 PM »

Oh, ok. I remember seeing 150 degrees, and didn't recall F or C, which of course makes a big difference!

You're right, that is still hot as the depths, but within SOL for most transistors.

If you're considering looking into them for the 'summer' version, SMPS in basic forms are fairly simple actually.

A lot of the control is all done on the chip, and so external parts count is minimal.

Really, the trickiest part in working with them is the inductors, and keeping high power traces and paths as short and thick as possible to reduce loss. Success and failure are influenced a lot more by physical part proximities and such than they are on 'normal' circuits. Appropriate part usage is also key.

Small cell phone chargers are an excellent place to begin, because they are usually as bare-bone as it gets, and are relatively cheap, even to buy brand new and destroy for resurrection of something greater.

Don't get me wrong on any of this, you've got an impressive piece of hardware there. Definitely built in the right spirit - anything that is home-made, does what it was designed to do, and can do it reliably is better than ANY commercial product, regardless of availability.

I hear ya in the electronics 'width' problem - Had to concede for myself a long time ago that I was never going to understand all of it. LOL Doesn't stop me from trying sometimes haha with all the futility that comes territorially from being a human... $:-\$

I don't mind playing with SMPS design to try to solve a particular problem... Just had help from Amanda (Commanda) on the design of a buck converter for solar. I don't mind providing this genre of favor to another one of my kind in turn.

Confirm for me: 30-200V input range, 15V @ 1.5A rated output. If I have that right, I'll see what I can conjure up from this.

Nothing to lose, and in fact much to gain... SMPS designs fascinate me, and there are so many basic flavors to choose from, just depends on your needs and what concessions you're willing to make with them in the process. Guarantee just playing with them will be of value to me, even if something truly practical does not come of the experiments.

Let me know, might be able to get you ready for summer by squeezing that extra 100W out as charging current!

Steve

commanda · « **Reply #47 on:** February 12, 2011, 09:49:03 PM »

To stay with the existing regulator, I'd be looking at reducing my current as much as possible.
Are the unused inputs on the LM339 grounded?

Have you adjusted the led current from the LM3914's as low as practical. There's a single resistor which sets the current for all the leds.
Could you run the LM3914's in dot mode rather than bar mode.

To build a switchmode from scratch, to get the input range required, the usual approach would be a PFC boost converter followed by a buck converter.
This is the approach taken by those wall wart (plug packs) with universal input. They are usually specc'ed as 90 - 264 volts ac, and will normally run on a dc input as well. It may be possible to hack one of these to work on a lower input voltage range, although I haven't tried this myself.

Amanda

oztules · « **Reply #48 on:** February 12, 2011, 11:33:54 PM »

Well done Methanolcat.

I too have to own up to not doing schematics first, and have come under the watchful gaze of Commanda some years back for the same sin.

You have done some really dandy stuff, and this is no exception..... although I liked that massive pwm you built for the trike.... truly awesome.

Best wishes with it.... and somehow I feel an affinity with of your" keep it simple approach". It may not always be best practice, but is my first choice...... but I don't wish to tar you with my very basic building skills..... you make it simple ..... and you make it look good too....... I seem only to do the simple part ....

.............oztules

methanolcat · « **Reply #49 on:** February 13, 2011, 01:41:55 AM »

Madscientist267,

   "Confirm for me: 30-200V input range, 15V @ 1.5A rated output. If I have that right, I'll see what I can conjure up from this."

   Yeah, could probably get by with a little less than an amp now, since I changed the bar graph display to dot mode and dumped the full scale flashing part of it (even though it was neat, it did consume too much power for this and could further cause reliability problems).

   I don't think I can live without the 2 fans though and they draw 200ma each, the heater would definitely run too hot without both of the fans since in higher winds I have seen over 6,000 watts for short bursts and a pretty continuous 4,000 watts. I can though drop the voltage going to them though some and cut down the ma draw and that would help as well.

   Just did a quick test, hooked up both fans to my test power supply and at 6v they seemed to move enough air for my needs and only drew 100ma each.

commanda,

   I've been working on reducing current draw, changing the bar graph display was the first thing I did (bar to dot) since it wasn't something absolutely needed. Led current even when I had it hoked up as bar mode was kept to a minimum so not much more I can do there.

   Also reducing voltage to fans is going to help, I can cut 200ma right there and still be fine.

   Yes the unused inputs of the 339 are grounded, I read about doing that in the datasheets quite some time back.

   Also checking into smps of some sort, I have some old cell phone chargers I will take apart and see if theres anything I can do to them to make one of them work.

   Thanks for the tips.

oztules,

   "I too have to own up to not doing schematics first, and have come under the watchful gaze of Commanda some years back for the same sin."

   I hear ya on that, drawing up a schematic wasn't something I wanted to do, it took me around 3 hours to figure out that software and draw that schematic. Commanda means well though, I know.

   I have to thank both you and Commanda since you were both there to help me a few years back when I really got started in electronics and was having some problems, especially Commanda, she stuck in there with me until I got the circuit I was working on to function as it should, even though she didn't agree with the circuit design (which wasn't mine by the way, that time anyhow, lol).

   So a BIG THANKS to you both for helping me and getting me going on my own, much appreciated.

Matt

commanda · « **Reply #50 on:** February 13, 2011, 03:54:48 AM »

What if you put the 2 fans in series instead of parallel. This would halve the current.

Amanda

methanolcat · « **Reply #51 on:** February 13, 2011, 09:30:58 AM »

I thought about the fans in series for that reason but also thought that if one fan was to open circuit for some odd reason then I would lose both fans, so I think halving the voltage to both of them in parallel would have to be my best bet for reliability.

Matt

Madscientist267 · « **Reply #52 on:** February 13, 2011, 08:19:50 PM »

Also wonder about 'runaway' here...

If for some reason only 12V (just saying) gets 'loose' in the fans, they both might not operate.

One could spool up and the other never sees squat, little to minimal airflow could go.

Not a fan myself (no pun intended) of series fans.

Steve

joestue · « **Reply #53 on:** February 13, 2011, 09:15:44 PM »

you could run some 16 volt zenars across the series fans.
i'm not sure if that would save any money however.

zvizdic · « **Reply #54 on:** February 15, 2011, 10:00:03 PM »

Quote from: methanolcat on February 13, 2011, 01:41:55 AM

I don't think I can live without the 2 fans though and they draw 200ma each, the heater would definitely run too hot without both of the fans since in higher winds I have seen over 6,000 watts for short bursts and a pretty continuous 4,000 watts. I can though drop the voltage going to them though some and cut down the ma draw and that would help as well.

These are some impressive numbers for a 15' turbine I am surprised no one noticed.
What was your voltage at 6000 watts or at 4000?
Your airfoil and blade specifics might be of interest for some people.
I would like to use windmill to help with heating to (hate batteries).
Thinking of strait s1223 airfoil what is your take on that? 10 HP 580/220v 3 phase motor to adopt so any help less testing.

Impressive Impressive!!!!

methanolcat · « **Reply #55 on:** February 15, 2011, 11:33:43 PM »

zvizdic,

   "These are some impressive numbers for a 15' turbine I am surprised no one noticed."

   Thats probably because those that know its possible don't get all excited about it like you are. The 4,000 watts I see is in about 34mph wind which we see here quite regularly during the winter, the 6,000 watts I'm guessing cause I don't sit there regularly with meters hooked up watching them, is in like I said bursts of 50mph + winds which is quite possible I'm sure. I have measured 100v across the 2.5 ohm load quite enough times that I know it does happen quite often. I have set up the bar graph display on the controller to show about 4,000 watts when the top bar is lit and the top bar is lit for extended periods, meaning that I am getting at least 4,000 watts, who knows how much higher at times, doesn't really matter to me as long as there is heat coming out of the heater. Note.... my mill is a motor conversion which does help to absorb and dissipate the stator heat.

   I don't use furling of any kind and someday it might bite me for it but until then......This mill is only experimental anyways, this spring I am going to start on the big one, I'm thinking 25 to 30 foot diameter would suit me with a geared up generator 3 or 4 - 1.

   Nothing fancy about the blades, they weigh 15 pounds each, made from laminated 1 x 4's, tsr 7, 3 inches thick and 18 inches wide at the roots and I think about 4.5 inches wide at the tip. They do have twist, I used alton-moore calculator found on the net.

   "Thinking of strait s1223 airfoil what is your take on that? 10 HP 580/220v 3 phase motor to adopt so any help less testing."

   http://worldofkrauss.com/foils/1421 I'm no expert here but that profile looks mighty thin to me, not sure the gains if any would outweigh the strength of the blades, let me know how that works out for you. Although mine ended up thinner than I had hoped for so it may be fine.

Matt

Madscientist267 · « **Reply #56 on:** February 16, 2011, 12:56:24 AM »

Quote

I'm thinking 25 to 30 foot diameter would suit me with a geared up generator 3 or 4 - 1

I'm thinking screw building my own genny; I was seriously considering taking a good hard look at an experimental mill of my own, but after reading this...

Think I'm just gonna figure out my losses and wire sizes and see what it's going to take to tap my system into yours

Whew, dude. If you're already kicking around in the 5K range, and you want to go, what's that work out to, 4x the swept area? 20kW? hahaha

That's enough to power the hungry demands that my whole family takes from the grid at times!

Send some my way once it starts spinning!

Steve

zvizdic · « **Reply #57 on:** February 16, 2011, 06:55:39 PM »

Otherpower 17' start to furl at 3Kw. Now I understand motor conversion does help to absorb and dissipate the stator heat so you can go higher.
http://www.otherpower.com/new17page5.shtml

That's probably because those that know its possible don't get all excited about it like you are.

I know it is possible but how many are able to achieve 40 to 45 % power out of a prop?

methanolcat · « **Reply #58 on:** February 16, 2011, 07:27:55 PM »

zvizdic,

My math shows differently than what you come up with. I'm going by the readings (volts x amps) that I take, maybe the wind is a little higher than I said, but.....According to the alton-moore blade calculator, 15 foot diameter, tsr 7, at the 34mph wind I stated the calculator shows approximately 11,000 watts available at the prop, I said around 4,000 watts, thats around 27.5% power taken from the wind and not 40 to 45% as you stated. I also said I sometimes see peaks up to and over 6,000 watts and I said we have wind gusts of 50mph and higher at times. The alton-moore calculator only shows wind speed up to 44.7mph but at that wind speed there is 24,597 watts of power at the blade, I can easily see how I could be getting 6,000 watts from that, You?

What math did you use to come up with 40 to 45%?

One thing to note here is that this mill does not furl in any way, the faster the wind the faster it goes.

Matt

methanolcat · « **Reply #59 on:** February 16, 2011, 07:32:23 PM »

Madscientist267,

"Think I'm just gonna figure out my losses and wire sizes and see what it's going to take to tap my system into yours"

Where do you live? I suppose I could rectify the output, convert it to ac, use a transformer and boost it up to 10,000 volts and send it to you, that way you can use smaller wire and there might still be some voltage left by the time it gets to you, lol.

Matt

zvizdic · « **Reply #60 on:** February 16, 2011, 09:52:38 PM »

You take compliments the wrong way.

Alton's calculator 4.5 m at 7tsr 15M/S =10000W 4000W = 40%

methanolcat · « **Reply #61 on:** February 16, 2011, 10:20:14 PM »

zvizdic,

I don't try to, to me it sounded more like as, it was too hard to believe. Sorry for taking it the wrong way, thats one thing that happens with text on the internet, everyone takes what they read a different way.

No harm intended, my numbers and yours are partly different most likely because I put in 4.7244 meters as I am actually 15.5 feet because of the big blade hub, actual wood blade is 7.5 feet. Small differences, no big deal, the sheet I printed from the calculator at 15/ms shows closer to 11,000 watts and thats where I figured from. Percentage wise it seems to make what looks like a fairly big difference when in reality I think its not. Anyhow, thanks.

Matt

Madscientist267 · « **Reply #62 on:** February 17, 2011, 02:13:20 AM »

Quote

rectify the output, convert it to ac, use a transformer and boost it up to 10,000 volts and send it to you

I wish you could, my friend. I wish you could...

Although, that's what they do...

If you're feeling froggy on the math, eastern Virginia...

Maybe I can use it for this handy-dandy night light right here...

Steve

zvizdic · « **Reply #63 on:** March 11, 2011, 08:08:33 PM »

Hi Matt!

Any update regarding the controller?

methanolcat · « **Reply #64 on:** April 04, 2011, 06:39:21 PM »

zvizdic,

Sorry for the late reply, been pretty busy. Theres no real update I guess other than its been working great for me, it still gets kinda cool here in the spring, especially when windy, so still using it for now. I may or may not change it for battery charging, since I don't have the batteries just yet, (I got a friend working on that), depends on how soon I get them. The mill is gonna come down here in a month or 2 and I'm gonna build a new tower that can handle a mill around 25 to 30 foot diameter, I don't think the current tower would be able to handle one much bigger than 15 feet diameter.

Soon to start on the hub for the new mill, much like the newer jacobs with centrifugal pitch control. I'm also gonna need to build a bigger generator and a gear box. I was thinking about changing the connections on the current generator to low voltage, then gear it up to get my high voltage back along with more amps, but I have 60 new 1" X 2" X .5" magnets with the hole in them just waiting to make lots of power so I think a new generator is in order.

Matt

zvizdic · « **Reply #65 on:** April 04, 2011, 08:46:28 PM »

Welcome back Matt

Now 25-30' mill , you can supply a village with electricity.
Is it going to be used as help to heat a house or much more?
So about 20KW or more Maaaaan that is a big undertaking . What is a tower idea the same as one you have but bigger ?
Wood blades or fiberglass

? and many moor

??

Please post your progress!!!!!!!!!!

I think your controller would be good for my mill and I am gathering parts needed.
Transformer is a issue and I wonder if you made a mistake to get 1.5A 12V 500-Ohms is to much it should be lover.
I got a 12V 400mA and is only 125 Ohms and more Amps less Ohms.
If you would please think of changes to circuit for 48v system.

Thanks

News:

Author Topic: PWM Heating load controller based on frequency (Read 24165 times)