Fisher & Paykel start up torque

[dinges]

What it could be used for? You serious about that question?

How about using motorconversions straight, without rewinding the stator!

Say, they put out 100VAC to 400VAC (no clue here whether this is the correct range, perhaps they go even higher in voltage?). Rectify to DC 140VDC-560VDC. Downconvert to 12-24-48VDC, whatever you need. Perhaps even make it switcheable. In that case, even the axial-flux guys would quickly start building high-voltage stators, I'd guess. Switching over from 12/24/48V would never again require a new stator.

Man, if you ever have a working design, I'd kiss you. Nobody would ever need to rewind stators of motorconversions again, simply add the "Oztules PowerWizzard (TM)" to your motorconversion.

Should I hold my breath, or start ordering wire for that stator-rewind?

Peter.

[commanda]

Peter obviously has a good handle on where I'm coming from.

.....simply add the "Oztules PowerWizzard (TM)" with Commanda's MPPT (TM) to your motorconversion.

Only minor technical hitch is the input voltage range. 2:1 is fine. 3:1 is doable at a pinch. Beyond that you need to cascade converters.

Amanda

[oztules]

Willib,

Agreed, an order of magnitude should be enough to convince you that the calcs were a little wayward.

I still think it should do a little better than that. Check your pulse width on the scope.

Wind 130 turns on a 5/8 inch former. place a ferrite inside of it, and hopefully your current will come up to a few more amps... the resistive losses will be higher for sure, but your frequency needs lots more inductance. That should give you a test to fiddle with. From there you can use more /less (I think more) and see what happens with the power.

Remember that with calculators.... BS In = BS Out.... the real world tells the real story.

..........oztules

[commanda]

Dominic,

Sorry to hear you've not been well. Hope all is well now.

Picaxe is a nice device. Used them in several projects (not related to RE). What you gonna do with it?

EEPROM's are a real dirty word with me right now. Our recent production (where I work) we've got a faulty batch. About a third wouldn't work at all. Of the rest, although they worked initially, are now failing in the field. <insert suitable expletive here>.

1. All the photos I've ever seen of modified F&P's used insulated multi-strand cable. A single conductor will eventually vibrate itself to cracking.
   
2. Allowing for future star-delta switching is very wise.
   
3. Can't give a quantified answer, haven't done enough research. Short answer is yes. If nothing else, you'll have greater headroom before failure, and bridge rectifiers are cheap anyway. Educated guess is it will output slightly more power.
   
   

Amanda

[oztules]

Hi Peter,

For some reason I hadn't even thought of that application.

It currently works as follows.

@300vdc... 12-45v dial up via trim pot.. can go higher. This sets constant voltage. It can get it's output voltage from Amanda's mppt (when it's finished)

The current can be dialed in the same way (trim pot, or shorten the 4" shunt wire.

Current is from 0-20A.... can squeeze more out, but the tranny needs sizing bigger for any more really. so with no real changes from it's present state. it can do

12-50v @20Amax.

If you get protel autotrax from the net (free from protel... dos version) and open http://www.otherpower.com/images/scimages/5171/smps3l.pcb, it is ready to roll. (The transformer info is on mech layer1)

It can be built from the parts in a normal atx or at power supply. You just need to wind the power tranny. The driver can be used from the existing psu.

My camera hasn't recovered from the trip up the mountain (and being left on the mountain top for a few weeks in the rain). I got it going but it failed after a few days.... gets stuck in the boot sequence. So can't take any pics of it yet.

If you appeal to Amanda, she may convert it to pdf for you perhaps. It is how she does hers from the same prog... but I haven't got the hang of it.

Before you get too excited though, Amanda is doing the same thing, only I expect that hers will be better designed. She is also including the power factor correction booster to supply constant voltage for the step down stage which is integrated with the mppt as well.

This one is an offline design, and so is a bit more parts intensive, and uses more inductors to achieve the same thing. It can be plugged into the mains and be used for lab supply 0-50v @ 0-20A.... need cooling above 400w, or larger cooling surfaces than I am using. The fans from a psu should do it. The tranny will require a bigger ferrite to support continuous use over 500-600w (although it hasn't gone over about 90C in 20 minuite tests of 18A @ 41.3v

I guess it would cover a few different stators, but then a boost converter may be simpler to build ie Flux and Matching the load.

I'd wait for Amanda's offering on this one.

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

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

[dinges]

Understood about the range problems.

In commercial PC PSUs, they use direct rectification for 220VAC, but most PSUs can also deal with 110VAC systems. Usually just a simple switchover. The rectifier + elco become a voltage doubler in that case.

Perhaps a switch-over like this (by a relay, driven by an input voltage sensing circuit) could double the range? From 2:1 to 4:1 (without problem) or perhaps 3:1 to 6:1 (in a pinch)?

Now, if anyone would design it in such a way that old PC PSU parts could be used (esp. the transformer and ferrites) I'd be a happy camper.

Perhaps even two completely separate voltage converters (one for the high input voltages, another for the low ones) could be a solution, switched over by relays, again depending on input voltage?

Just brainstorming here.

Trouble with PC PSU parts is that they're designed for a maximum of, say, 250W. The average windmill puts out a lot more, in high winds (say, 1kW).

[dinges]

Hi Oz,

Saw your response after I posted mine to Amanda.

All in all, it looks very promising. Higher power (500-600W) than I had in mind (200W max).

I knew both of you were busy on the issue. But what to think of your advice to wait on Amanda's offering for this one...

Like I said before, I'd be really interested in a (more-or-less) fully finished design. It's not hard for me to build an existing design, or even modify a little, but major electronics design work is not my strongest ability. I won't even begin to start thinking about designing switch mode power PSUs.

High voltage/high power electronics like this is not the kind of thing that I would experiment with myself too much (though I'm crazy enough to try the odd PC PSU conversion, as you already know), but if a good, proven design comes up, I think I'd be one of the first to build it.

BTW, the link doesn't seem to exist.

W.r.t. Amanda's dumpload controller: work has started on the main PSU. Expect it will be finished by the end of this weekend. That means most major parts are finished, except connecting the boards together and building the actual FET dumpload modules. Slowly but surely, we're building a system here.

I can't stress this enough, but I think one of the greatest barriers at this moment for advancing the field any further is no longer the mechanics (axial flux gennies or motorconversions) or propellors, but the electronics: dumpload controllers, power conversion (wild/raw HV AC to low voltage DC) and MPPT.

[commanda]

Peter,

In the long run, I really think you're better off spending the few dollars and buying the ferrites. No two PC psu's are ever the same. Mass-produced budget items like that, they use the absolutley smallest bit of ferrite they can get away with. And with buying them, you at least get a published datasheet.

Relays and Murphy's Law don't mix.

All of this is doable right now. Flux's boost converter, modify the feedback to give constant voltage output; Oztules converter; and my mppt circuit which I've passed on to several people now.

Just wish I had all day every day to devote to this.

Amanda

[domwild]

Amanda,

Thanks for reply. Will use the 28X for RPM measurement and Glenn Littleford has the code + schematic for a more sophisticated load shedder (dump controller) on his site. To keep it simpler I am not going to output to a display. If all goes well, the 28X will also become the star/delta switcher.

You must have seen Michael Lawley's torque measurements in ReNew magazine. I could have saved you some work (and mulla) by quoting his torque figures measured with a spring balance.

The EEPROM is a 24LC256 with 64k; I will keep my fingers crossed then!

dom

[commanda]

I won't get into the long term reasons for not using a micro-controller. I've practically written a thesis on the subject here several times.

Without a display, how will you know if it's working correctly or whether lightning has killed the micro?

Don't get renew magazine. Nothing like doing it yourself to really get a feel for it. Different to just reading some figures published somewhere.

You're EEPROM is most likely fine. We just got a bad batch.

Amanda

[oztules]

Peter, sorry about the link, looking at it, it appears to have a full stop after the .pcb at the end. try this:

http://www.otherpower.com/images/scimages/5171/smps3l.pcb

if this don't work, look in my files for smps3l.pcb.

It is a protel pcb file . Will require protel autotrax, or protel 99se  or protel for windows etc to open it. As I said, the transformer data is on mech layer 1.

It will give you an idea of what it looks like.  The output inductor (filter) is not shown. It is currently 30 turns of heavy  wire on a torroid former.  This requires further work as I think this simple coil is saturating at high power levels... it gets too hot. The shunt is also not shown. This is 4inches of ...hmmmmm lemmesee... the yellow wire on a psu...about that

The board is a work in progress when I designed it, and so it is poorly labelled as it was only for me to do the board easier.

The only thing to mess up (other than the power tranny) is the orientation of the leads of the driver tranny.

All in all, I'm happy with it. Will tidy up the coil data for the filter at a later date. It's dirt cheap to build.. because of the psu parts.

It will need a bigger transformer if higher power levels are sustained.  So it is not really ready for general use.

Like I said, Amanda is doing a comprehensive system, if I were you I would be inclined to follow her lead.

I'm silly enough to enjoy the challenge, so will stray a bit into the unknown. I will tidy this unit up so that you can build it if you want, but it will not be tested in unison with the rest of your system.... hers will be.

If you just want a lab supply with excellent stability, then this may be for you. The efficiency has not been tested yet either.

For the service you mentioned, I would redesign the board without the prefilter, and use this space for a larger transformer perhaps..... more power... more power

.........oztules

[dinges]

Amanda,

It's not the expense of buying ferrite cores that I fear. It's the usually lousy availability of them. As long as the ferrites are available in small quantities over here & elsewhere (Australia, USA, Europe) I can live with the fact that I'll have to pay for them

Should post a picture someday of my parts bins. 12 racks of 48 drawers, filled to the brim with components. All 'pre-used' parts That's what you get if you start tearing up TVs at the age of 8, I suppose. The buckets with ferrites are somewhere in the attic. Goes to show how often I use them, I suppose.

Peter.

[commanda]

Peter,

Check out Farnell or Radio Spares. Both based in Britain. I buy a lot of stuff through their Australian offices (mostly Farnell). On-line ordering, accept credit cards, small quantities no problem. The cores I'm using are E42 or E55. They also carry the ETDxx cores oztules uses.

Try farnell.com, and select your country from the drop-down list.

Amanda

(Born to shop :-)

[dinges]

Yep, Farnell is the one that I use when stuff seems to be hard to get via my regular electronics suppliers. They're quite expensive (2-3x 'normal' prices), but still, better that than 'not available' as a reply. Good to hear that you & Oztules use their ferrites.

[oztules]

Hi peter,

actually I have a few (lots) of etd39 cores from  http://www.neosid.com.au/ who claim to make them over here. I have a few bigger ones ee65 and ee55 to play with this (i managed to get it to pdf for you) http://www.otherpower.com/images/scimages/5171/xx.pdf  or xx.pdf in my files if that doesn't work.

The use of the larger ee's may take us up to the 1kw mark fairly safely. Being able to control the current and voltage outputs, no change is required to increase the power. The only thing is the diodes on the output may require better cooling, and the switching trannies. (some bigger caps on the input filter would be a step in the right direction as well.. maybe 800uf would be better.)

Being that the windmill has mostly transient power at these levels, may get away with just holding the current back to reasonable levels, and ignoring the higher winds.

Like I say, Amanda's entry is what I'm looking forward to, so I wouldn't move until that is on board if I were you...... me ... I got plenty of learning to do with these things still, and more fets to torture.

There's a million and one ways to skin a cat, or build a pwm... i've got a few more to go I think. Meanwhile I will continue to avoid any Netherland folks with amorous tendencies.

.........oztules

[commanda]

Oz,

Quick question on how you've wound that main transformer. What gauge wire and how many strands in hand?

Amanda

[oztules]

Amanda

 The wire through out the transformer, primary and secondary is 1.7mm. The aux winding is 0.4mm.

These sizes with the packing tape fill the winding area. It is one in hand winding.

To keep leakage down to a minimum, use the primary on the inner and outer as per factory winds.

The output filter inductor ( shown on the pdf) is 28 turns of 1.7mm wire on a yellow torrid approx 33mm outer, 22mm inner... This will require some more work, as it gets hotter then I would like. I think it saturates at high power.

To use more of the high end power, two or three in hand on a 65 would be nice, and perhaps 3mm wire on a etd39 core (with an airgap)for the output filter.

Am involved with some work with my big caterpillar at the moment, and will get back to this project in earnest a little later.

Will do it as a diary perhaps.

..........oztules

[mrpackethead]

I hear that the Type 4 F&P has been made so it does't suffer from cogging nearly as much as the type 1/2/3's..  does anyone know?

Regards