Author Topic: Squeezing a bit more power out (Read 16264 times)

leftcliq · « **Reply #33 on:** January 05, 2016, 05:15:38 PM »

Hi,
Thanks for your reply. I was referring to: www.thebackshed.com/Windmill/articles/GordonsCapMod.asp . I cannot create html links on the forum, it just freezes up. I have used a voltage tripler successfully but a voltage quadrupler would be better. I have read the forum page on the back shed about this issue but as this information is now old, a lot of the images are gone.

Regards,
Dale.

SparWeb · « **Reply #34 on:** January 06, 2016, 01:14:43 AM »

Hi there
Welcome to Fieldlines.

I'll try to dig up the old graphics that I posted and refresh them.

SparWeb · « **Reply #35 on:** January 06, 2016, 01:28:16 AM »

Here is the text of my original post, with the graphics restored:

The point of these circuits (for those who haven't seen one before) is to create a circuit that cuts-in sooner than the regular battery-charge circuit. You still have the main rectifiers making DC for the battery, this circuit works in parallel to give you something like 2x voltage output. When tied to their own rectifiers, this circuit can start the battery charge process at a much lower cut-in speed.

Well, that's the promise, so does it work? My results are mixed, but I am leaning toward "yes". I tried working it out in 2 ways. First, I used computer models (Spice = free download), then I tried it for real with my own turbine. The tests on my turbine showed a noticeable improvement, so I thought I'd post them. When I attempted to use circuit a simulator program to see how the effect worked, however, the results were not encouraging. In fact, this is one case where experimental evidence exceeds expectations, so I'm surprised.

Here's what it looks like on my power curve. I was able to collect many hours' worth of data for a day where the wind was relatively consistent the whole time. This meant there was enough time to have the capacitors switched "on" for half the day and "off" for the other half, and get data that I could compare fairly.

The curve is quite a bit higher when it's on and turning slowly. The normal cut-in for my turbine is about 200 RPM, and that happens at about 10-15 kph. Again, what people say about these capacitors is borne out in my measurements - they don't have much effect above the cut-in speed, but below normal cut-in, they get a trickle going where normally there would be nothing.

Trying to see this in a Spice simulation was another thing altogether. I was rather frustrated with this for weeks because all I could see were increasing noise, RF, and spikes. Eventually I tweaked out the flaws in my circuit diagrams, and I can finally see the "positive" results.

Here is the "doubler" circuit, as drawn in Spice. I have also tried a "tripler" circuit, which also works a little.

Here I compare what the turbine would do with and without the capacitor doubler circuit. The first graph is basically flat-lined, the second is getting an undulating 400-500 milliAmps. For a turbine turning at 100 RPM, this means it is collecting up to 50 watts at half its normal cut-in speed.

leftcliq · « **Reply #36 on:** January 06, 2016, 02:18:13 AM »

Hi Steven

Thank you for reposting the images. With the 3 phase bridge in parallel with the voltage doubler, does this turn the voltage doubler into a voltage quadrupler? When you did the tests on your turbine did you use the same 6500uf caps that you have in your Spice diagram?

Thanks again,
Dale.

SparWeb · « **Reply #37 on:** January 06, 2016, 11:24:11 PM »

I just went and checked. The bundles of caps are still sitting on the back of my shelf! It seems I was using 2200 uF and 4700 uF capacitors in parallel in the actual tests. The Spice simulations showed it didn't really matter if the capacitance was exact. I just used capacitors I removed from other things for the tests. I have a different turbine now so I'm not using cap triplers any more.

leftcliq · « **Reply #38 on:** January 07, 2016, 07:44:26 PM »

Hi Steven,
Is your circuit diagram above a voltage doubler or voltage quadrupler circuit?

Regards,
Dale.

leftcliq · « **Reply #39 on:** January 08, 2016, 07:04:14 PM »

Hi Steven,
Would having 2 voltage doubler circuits in parallel create a voltage quadrupler circuit?

Regards,
Dale.

SparWeb · « **Reply #40 on:** January 08, 2016, 11:34:19 PM »

Hi Dale,

I don't think it's that easy.
Bear in mind that if the circuit can store charge at one moment and discharge it from two phases in a way that doubles the voltage, then the impedance of that circuit is at least twice the impedance of the primary circuit. That limits the current that can flow. If you propose to combine them in a way that allows tripling or quadrupling, then there will probably be a greater impedance, and then less current. Ohm's law and the law of diminishing returns.

leftcliq · « **Reply #41 on:** January 09, 2016, 12:20:34 AM »

Hi Steven,
Thanks for that tip about the impedance. I have a tripler that I use that works very well. I am happy with it, I just need a quadrupler circuit to replace the tripler circuit. Having a real hard time locating this circuit. Do you have a quadrupler circuit? Is it a matter of money? Happy to pay for it.

Regards,
Dale.

Mary B · « **Reply #42 on:** January 09, 2016, 02:23:31 AM »

Pretty simple circuit

DamonHD · « **Reply #43 on:** January 09, 2016, 04:06:15 AM »

The generic circuit is a Cockroft-Wilson multiplier as here for example:

http://www.earth.org.uk/wind-power-pilot-autumn-2007.html

It has to be the case that any increase in voltage is matched by a reduction in current / increase in impedance. That that's ignoring losses in the caps and diodes.

Rgds

Damon

SparWeb · « **Reply #44 on:** January 09, 2016, 05:08:08 PM »

Damon, Mary,

You're cutting into my profits.

Dale,
All kidding aside, if you've reaching for tripler, quadrupler... quintupler... then there is less and less to be gained.
Of course, you could be exploring it just for the sake of interest. We would be interested in seeing results when you've tried it.

OperaHouse · « **Reply #45 on:** January 09, 2016, 05:54:47 PM »

Like everything else a voltage multiplier has to he designed for an application.
I prefer to see active electronics do this multiplying, but simple capacitor methods
can be effective. Usually going beyond a simple doubler isn't effective except for
very low currents. All these diodes will have stackup voltage losses. Capacitors
are the electronic component that lest matches the theoretical. Identical values
in different size cases will not have the same current carrying capacity. Like diodes
and wire, these will heat up. They can be destroyed by current and have limited life.
Multiple parallel capacitors should be considered for current over a couple amps.

Could you give us a scale of your application. What voltage battery are you tryigg
to charge? What currents have you seen with present multiplier setup? What wattage
mill is this being used on? There may be better alternatives.

Mary B · « **Reply #46 on:** January 09, 2016, 06:53:31 PM »

Best bet might be a high efficiency switch mode supply that accepts a wide range of input voltages and frequencies... I have seen some that work for ~60-280 volts and ~40-200hz. Would have to dig on the power supply companies websites for the right models.

leftcliq · « **Reply #47 on:** January 10, 2016, 06:31:37 AM »

Thank you all for your kind input.

I have a 3 phase fisher and paykel motor and I require a voltage quadrupler. Basically I need 4 x the output voltage hence the voltage quadrupler. I know Gordon (Gwatpe) has a cap mod page. Is this circuit at the bottom of his page a voltage quadrupler? I have ordered a large number of caps and have the diodes and bridge rectifiers. I just need a circuit to follow and then to experiment with the caps to gain the most voltage out of the motor.

I have tried rewiring the stator but have had the best results with the voltage tripler. Really I just want to try this quadrupler and to let everyone know of the results. The tripler really does triple the voltage and reduces the load on the motor. My rpm is quite low.

In the link you provided DamonHd, the author says that the circuit is more of a tripler than a quadrupler. Have you tried this before?

Thanks again Steven, DamonHD, Operahouse and Mary B for your time and assistance.
Dale.

DamonHD · « **Reply #48 on:** January 10, 2016, 06:57:37 AM »

Hi,

That's my page, and although the circuit was nominally 4x, because of diode losses for a start, it was more like 3x in practice.

If your input voltage is higher than mine was, and you were to minimise diode losses (eg with good Schottkys), then maybe a theoretical 4x could nearly be one in practice.

Rgds

Damon

OperaHouse · « **Reply #49 on:** January 10, 2016, 10:53:34 AM »

Sorry, but you never answered any of my questions.

leftcliq · « **Reply #50 on:** January 11, 2016, 06:32:10 AM »

Hi Operahouse,

I am using a 300w generator. I'm just trying to get the electronics side of it right and experiment with what I can achieve. The scale is really just for domestic personal use. I have thought about using a transformer but I would much prefer to use the cap - diode approach.

I was planning to use 10Amp diodes. I realise there is a 0.7 voltage drop across each standard diode. The shottkys have a very low forward voltage so I don't think they are usable in the instance unless my understanding is incorrect. Please correct me if I am wrong here as my understanding of shottkys is limited.

Regards,
Dale.

leftcliq · « **Reply #51 on:** January 11, 2016, 06:34:23 AM »

Quote from: DamonHD on January 10, 2016, 06:57:37 AM

Hi,

That's my page, and although the circuit was nominally 4x, because of diode losses for a start, it was more like 3x in practice.

If your input voltage is higher than mine was, and you were to minimise diode losses (eg with good Schottkys), then maybe a theoretical 4x could nearly be one in practice.

Rgds

Damon

Hi Damon,

In your circuit, you have the positive AC phase 2 as ground. Is the correct or am I missing something?

Regards,
Dale.

leftcliq · « **Reply #52 on:** January 11, 2016, 07:05:47 AM »

Guys,

What are your thoughts on the circuit at the bottom of this page: http://www.fieldlines.com/index.php?topic=139196.0

I can't insert a link or an image when I reply.

Regards,
Dale.

DamonHD · « **Reply #53 on:** January 11, 2016, 08:12:48 AM »

Quote from: leftcliq on January 11, 2016, 06:32:10 AM

The shottkys have a very low forward voltage so I don't think they are usable in the instance unless my understanding is incorrect.

1) Lower is better when it comes to diode forward voltage, but you also have to make sure that that diodes you use are not too leaky in the reverse direction, which could possibly be a problem with Schottkys.

2) There is no 'ground' on that turbine; there are three phases and I can tie any *one* of them to ground, or one other point in the circuit to ground, with no ill effects, providing I'm very careful what else gets grounded (eg not shorting out some of the diodes by grounding two incompatible parts of the circuit).

Rgds

Damon

OperaHouse · « **Reply #54 on:** January 11, 2016, 09:08:58 AM »

Obviously, a language barrier prevents you from answering my question.

OperaHouse · « **Reply #55 on:** January 12, 2016, 09:30:22 AM »

The datasheet is your friend. I get a little perturbed when I hear Schottky diodes
are the cure all for everything. Fact is in many cases there is no real advantage.
As an example for this F&P circuit, a 1N5820 is rated at 3A 40V. The forward voltage
ie rated at .370 @ 1A (I hear cheers) and that goes up to .625W @ 3A. Drop the
voltage to 30V and the numbers get better. 40V is pretty wimpy and we all know bigger
is better. Jump up to 50V and a Schottky diode radically changes. Look at the data
sheet for a 1N5400, at 3A the forward voltage doubles to 1.2V and even at 1A it is about
.75V. The big advantage has disappeared and many of the applications here will need
over 40V. Before buying a diode check the data sheet.

DamonHD · « **Reply #56 on:** January 12, 2016, 11:11:30 AM »

Yes, agreed, check the data sheet.

And it does entirely depend on the application, expected turbine voltage, etc.

Rgds

Damon

JW · « **Reply #57 on:** January 12, 2016, 06:02:52 PM »

Quote from: leftcliq

I can't insert a link or an image when I reply.

http://www.fieldlines.com/index.php/topic,146470.0.html

News:

Author Topic: Squeezing a bit more power out (Read 16264 times)