Junkyard motors parade

AC, DC, amps, volts and kilowatt. It's all discussed in here
User avatar
weber
Site Admin
Posts: 2936
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Junkyard motors parade

Post by weber »

If your motor is capable of more power than your controller can provide (by up to a factor of sqrt(3)), and you have a single drive ratio (no multi-ratio gearbox), then you should wind the motor so you get its maximum torque at your controller's maximum current when the motor is in star or wye configuration. Then if you switch it to delta you will have sqrt(3) lower torque, but that torque will be constant out to sqrt(3) times higher rpm, and will still get the full controller power into the motor.

In this controller-limited case, a star delta switch is equivalent to a gear-change with a ratio of sqrt(3) ~= 1.73.

If changing between star and delta on-the-fly you need to somehow allow for the discontinuity in voltage and phase. Ross Pink has done this with an industrial VFD.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).
T2
Groupie
Posts: 121
Joined: Sat, 08 May 2010, 04:51
Location: ON,CANADA

Junkyard motors parade

Post by T2 »

It's interesting to read about your efforts, but why would you be winding a motor as 2-pole when it is to be used with a VFD?

Unless I am misguided Weber, it appears to be a four pole using the "consequent pole" technique.

I would like to elaborate on your statement equivalent to a gear-change in your most recent post however.

A mechanical gear change will eventually be needed for any motor with a high value of V/Hz which when driven fast enough will require more and more voltage until the inverter ouput is forced to clip the rails as it runs out of available voltage.

In the application here, which Weber is alluding to, the star-delta switch is a method of artificially altering (by lowering) the V/Hz and thus keeping the controller away from "voltage saturation" for a little while longer, whilst more and more rpms are continuing to pile onto the motor.

The way I see it, the more recent availability of 500/650A controllers enables us to avoid this particular "trick" by starting with a motor that is designed with an extra low V/Hz from the beginning, whenever that is possible. And since this happens to be a "motor rewind" thread...

Before I begin to sound like a one trick pony myself, I must take the opportunity to emphasise, something that Weber is quite aware of I am sure BTW, and something that needs to be said, namely that star/delta switching does not improve the overall torque characteristic of the motor itself.

Mostly it is something you might want to consider when something outside the motor is sufficiently wrong. That could be insufficient controller current, insufficient battery bus voltage or the shortcomings of a non optimal gear ratio.

Star delta switches are likely to be bulky, complicated and expensive. In fact high voltage/ high current switching is something to be avoided. From personal experience I can say it attracts problems.


T2
User avatar
weber
Site Admin
Posts: 2936
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Junkyard motors parade

Post by weber »

T2 wrote:Unless I am misguided Weber, it appears to be a four pole using the "consequent pole" technique.
Ah yes. Silly me. Thanks T2.
I would like to elaborate on your statement equivalent to a gear-change in your most recent post however.

A mechanical gear change will eventually be needed for any motor with a high value of V/Hz which when driven fast enough will require more and more voltage until the inverter ouput is forced to clip the rails as it runs out of available voltage.

In the application here, which Weber is alluding to, the star-delta switch is a method of artificially altering (by lowering) the V/Hz and thus keeping the controller away from "voltage saturation" for a little while longer, whilst more and more rpms are continuing to pile onto the motor.

The way I see it, the more recent availability of 500/650A controllers enables us to avoid this particular "trick" by starting with a motor that is designed with an extra low V/Hz from the beginning, whenever that is possible. And since this happens to be a "motor rewind" thread...

Before I begin to sound like a one trick pony myself, I must take the opportunity to emphasise, something that Weber is quite aware of I am sure BTW, and something that needs to be said, namely that star/delta switching does not improve the overall torque characteristic of the motor itself.

Mostly it is something you might want to consider when something outside the motor is sufficiently wrong. That could be insufficient controller current, insufficient battery bus voltage or the shortcomings of a non optimal gear ratio.

Star delta switches are likely to be bulky, complicated and expensive. In fact high voltage/ high current switching is something to be avoided. From personal experience I can say it attracts problems.
I totally agree.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).
User avatar
peskanov
Groupie
Posts: 153
Joined: Sat, 15 Oct 2011, 00:55

Junkyard motors parade

Post by peskanov »

Weber,
as T2 said, this wire layout is a 4 poles one. In a fact it is the layout the motor had from origin, I wanted to rebuilt it with lower voltage.
As I found it too difficult, I moved to lap winding.

About optimizing torque: motor can be cheap (at least if you rewind a motor successfully) but controllers are not. This means it's easy to have a motor which can be pushed in revs or torque a lot more than the controller can provide.
This is the case with this motor:
-If I tune for torque, revs will be unused (torque will disappear sooner, much lower than 3000 rpm). Also, torque beyond nominal is not efficient.
-If I tune for revs, the torque potential of the motor remains unused. In the other hand, as you use higher ratios, you get the same torque at a higher efficiency. But having a high ratio is difficult and balancing your rotor for >5000 revs it's difficult too.

So I am trying to play conservative here, available torque should be close to 2.5x, and revs should be ok having peak power at 2x also.

I heard about the star/delta switch concept and I think it a great way to enhance acceleration, but I am going to keep this design simple as I have little money for it.
Last edited by peskanov on Sun, 25 Aug 2013, 16:30, edited 1 time in total.
User avatar
peskanov
Groupie
Posts: 153
Joined: Sat, 15 Oct 2011, 00:55

Junkyard motors parade

Post by peskanov »

T2,
Well, perhaps the motor is trying to tell you something. That this new winding has made the motor too powerful for the controller.

If you go back a few posts you will find Richo and myself, initially recommending that you start with something smaller, like perhaps 5.5Kw ?

You, however, insisted on going ahead with a 9Kw. Of course, having a slightly heavier motor is not a serious problem to have. You always have the option to switch to a 650A controller later on if things turn out to be successful.
Well, having a 7.5kw or a 9kw should not be a problem for going high revs. They will run cooler and accept more torque from the controller if your controller can afford it! A 7.5kw can be wired to low V/Hz as easily as a 5.5KW, the only advantange of the 5.5 being it's slightly lighter and cheaper to rewind.
In fact, bigger motors have higher efficiency.

The problem with rewinding for high revs is that it is extra difficult. You can check Ivan Bennett threads and see how the 1 turn layout is the most difficult one, forcing you to work with very thick cables in a pretty small space.
The other problem I see is the transmission. Having cruising speed on 9000 rpm seems difficult to built well. Can a common gearbox work that way? I guess that would mean using 2nd gear all the time, is that a sound idea?
Also, I am not so sure about balancing the rotor. The guy who balance rotors, which I mentioned previously, told me he can not balance rotor as big as a frame 132 one. Finding a shop with the necessary equipment could be a problem.

About the torque curves, I reckon I don't understand the matter well. In my Kawasaki conversion, torque is very good up to ~1700 rpm, then disappears at 2200 rpm. This is to be expected, as voltage should provide maximum power at the 1600 rpm mark.
We tried to flatten the curve using field weakening using the method explained in Ivan Garage forum, but we didn't get any progress.
Do low V/Hz motor have a better torque curve?

User avatar
BigMouse
Senior Member
Posts: 600
Joined: Thu, 28 Oct 2010, 02:39
Real Name: Vincent Tannahill
Location: Silicon Valley
Contact:

Junkyard motors parade

Post by BigMouse »

peskanov wrote:About the torque curves, I reckon I don't understand the matter well. In my Kawasaki conversion, torque is very good up to ~1700 rpm, then disappears at 2200 rpm. This is to be expected, as voltage should provide maximum power at the 1600 rpm mark.
We tried to flatten the curve using field weakening using the method explained in Ivan Garage forum, but we didn't get any progress.
Do low V/Hz motor have a better torque curve?
As I understand it, torque is directly proportional to flux density, which is the product of amps and turns. The torque a motor can produce is limited by the magnetic properties of the motor components, not necessarily the windings. A winding with 10 turns and 10 amps will produce the same torque as a winding with 1 turn and 100 amps, and the same torque as a winding with 100 turns and 1 amp. The current you can deliver at a given frequency is dependent on the current and voltage capabilities of your controller.

Field weakening isn't something that you program a controller to do, it's a condition that the system enters as the speed increases beyond point where the supply voltage can provide a constant V/Hz ratio. If you're running your motor at 2200rpm, and your controller isn't clipping, then field weakening is happening. Field weakening is happening whenever you operate in the constant-power region of motor operation.

That's how I understand it at least.

Of course a high-end controller can be tweaked to provide different behavior in different regions of operation, such as field weakening (under-fluxing) at lower speeds to reduce heating, or close to synchronous speed to improve efficiency, but these are different to what you're referring to I think.
T2
Groupie
Posts: 121
Joined: Sat, 08 May 2010, 04:51
Location: ON,CANADA

Junkyard motors parade

Post by T2 »

Thank you for jumping in with that post, BigMouse, a useful contribution if I may say.

Hi Peskanov,
I was going to put up another post as well but I am starting to see a new problem develop here - maybe YOU are getting TOO much advice.

It's keeping you away from the task in hand, which was not my nor anyone else's intent here.

You have to make your own decisions of course, whichever way you decide. However bear in mind that :
The Performance Directive dictates that the design should obtain as many turns of the motor per unit distance as efficacy permits.
Efficacy for the motor (with a 300Hz controller limit) demands a non relenting power band over the range of 9000rpms.

Roughly 1st gear yields 70km/hr and 2nd gear yields 140km/hr.
A choice of either lousy top speed or lousy acceleration. Both sub optimal. Realistically I would recommend for 100km/hr gearing.
T2
User avatar
peskanov
Groupie
Posts: 153
Joined: Sat, 15 Oct 2011, 00:55

Junkyard motors parade

Post by peskanov »

Bigmouse,
I was referring to this:
http://ivanbennett.com/forum/index.php?topic=57.0
I always thought this was similar to the tech applied to DC motors, where armature current is purposefully reduced in order to get more speed. In an AC motor you should be able to reduce armature current reducing the slip, but I have not paid too much attention to the issue...

T2,
as you say, standard gearing does not fit well a >6000 rpm motor. And it has been mentioned (but I am not sure about it) that 1st gear does wear out quite soon, as it's usually designed for non continuous operation.
For a high speed motor, I think the best option would be removing all clutch/gearing, add an industrial reduction and operate gearless.
User avatar
peskanov
Groupie
Posts: 153
Joined: Sat, 15 Oct 2011, 00:55

Junkyard motors parade

Post by peskanov »

Rewinding update,
yesterday we finished winding the coils. Basically, we followed Ivan's method (with a different layout) as all our experiments proved unproductive.

I have a few tips to share for anybody trying to do the same:

- Good lighting. The stator interior is shadowed easily, and you will be performing precision work there.
- A helping hand. Getting help makes the work only a bit faster, but makes it less tedious and tiring.
- Reduce the number of wires. If the original number of wires in one slot is X, I would recommend going for 4/5 or 5/6 of X. It's too hard for a beginner to pack so much copper as professionals do.
- Get a good material for the wedges. We used the same insulating paper used as base in the slots. The shop assistant told us we could use it for wedges also, in order to save money. Well, it's true. It can be used, but it's just too much work in an already difficult task! And it does not look too strong.

And a recommendation: in retrospective, I should have rewound an small motor first, just to get proficient at it (an waste less copper with my experiments). A motor that was in working order, and big enough to fit the hand, maybe a 2KW one. Working they way I did, I was getting the "art" of it at the end of the winding, and that means the motor looks unprofessional…and if it does not work I will not be sure if it was my fault or the rotor is really damaged…

Ok, first the layout we did. We abandoned the concentric one (the motor original one) for another easier to wind:
Image
I think (but I am not sure) Ivan is using double layer designs only, while this is a simple, single layer one.
Originally, the motor had 56 wires in a slot. 14 turns, 4 wires in hand. We rewound at 2 turns, 23 wires in hand (I reduced from 28 to 23 to make the process easier).

First step, calculate the length of the wire for one phase (ex. X-U) and cut 23 wires. I made a roll with it and let it behind me.
Image
You just keep taking cable from the roll, make a lasso with 1-2 feet and pass it through the stator (check Ivan videos). Then try to fit the wires in the slots.
In order to fit them easily, I flatten the wires as much as possible. Any twist of the wire must be removed; combing the wires a bit can help to fit the center, as they tend to comb out of the slot:
Image
Some kind of "ruler" is necessary to push the wires inside the slot, specially when the slot is packed. I used a piece of painted metal I had, but plastic or wood is probably better to avoid damaging the wire. It's important to use one with the correct thickness, not too thin (or wires will exit the slot).
Image
User avatar
peskanov
Groupie
Posts: 153
Joined: Sat, 15 Oct 2011, 00:55

Junkyard motors parade

Post by peskanov »

Here we have the fist phase built:
Image
The insulated parts are the "bridges" that can be seen on the diagram. All the wire is in one piece, not cuts/connections.
I think each phase takes abut 6-8 hours (not for a pro, of course Image).
Making a good coil head is an art, and I am still getting good at it. I would recommend checking as many videos as possible before shaping one.
A view of the second phase / start of the third:
Image
At this point space starts to look scarce. Wiring is more difficult.
Finished:
Image
Image
It looks so packed, the rotor could not enter the stator! Image
Hopefully, when we apply the last step everything will be ok. Working with so many parallel wire is nasty, as you need too much space to jump from to coil to coil or exit the stator.
New week, tweak the coil heads, solder terminals, check for shorts and varnish!
User avatar
Johny
Senior Member
Posts: 3749
Joined: Mon, 23 Jun 2008, 16:26
Real Name: John Wright
Location: Melbourne
Contact:

Junkyard motors parade

Post by Johny »

Fantastic! Make sure to show us the pictures when you smash it with the rubber mallet to make room for the rotor (like Ivan did)!!!
This is really interesting.

Edit Ivan not Ian
Last edited by Johny on Wed, 28 Aug 2013, 04:59, edited 1 time in total.
User avatar
weber
Site Admin
Posts: 2936
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Junkyard motors parade

Post by weber »

Heroic! Image
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).
User avatar
BigMouse
Senior Member
Posts: 600
Joined: Thu, 28 Oct 2010, 02:39
Real Name: Vincent Tannahill
Location: Silicon Valley
Contact:

Junkyard motors parade

Post by BigMouse »

peskanov wrote:I always thought this was similar to the tech applied to DC motors, where armature current is purposefully reduced in order to get more speed. In an AC motor you should be able to reduce armature current reducing the slip, but I have not paid too much attention to the issue.
In a DC motor you reduce the field current, not the armature current. Reducing the strength of the magnetic field in which the armature is spinning reduces the torque, but also the back-EMF, allowing for higher speed from the same input voltage.

Great progress on the motor wind by the way. I'm really looking forward to doing this myself some day.
InductEV
Noobie
Posts: 1
Joined: Tue, 06 Aug 2013, 03:24
Real Name: Stefan Smolenaers
Location: Melbourne
Contact:

Junkyard motors parade

Post by InductEV »

The flux issue is why im a fan of the voltage 'overclock' method.

You keep the same amount of turns, but put each pole winding in parallel reducing the resistance thereby increasing the reachable current. As long as the motor is not near saturation, you can massively increase the (A*T) flux and torque. You also get a higher top speed as full voltage is available across each pole.

Obviously with your winding, because you have gone from 14 turns to 2, you need to increase the current by 7 just to get the same A*T.
Even with a custom winding like yours, i'd suggest looking into placing poles in parallel.
Obviously with more turns though you have less room for 'in hand' wires, meaning the although it may handle the same amount of current, it would heat up quicker and therefore have reduced peak torque time duty. On the plus side though, you would most likely get higher peak power as field weakening would happen at a higher RPM. A higher flux allows you to run lower currents for the same torque anyhow.

Its kind of like running in star vs delta - and similarly you'd have to be weary of uneven/circulating currents from mismatched pole winding resistance - maybe accentuated by DIY primitive wiring methods, but I dont actually think it would pose too much of a big deal - only testing will tell (I plan to 1/4 voltage [not rewind] my 160L WEG motor after successful results with the 2kW).


User avatar
peskanov
Groupie
Posts: 153
Joined: Sat, 15 Oct 2011, 00:55

Junkyard motors parade

Post by peskanov »


Johny & Weber,
thanks for the encouragement, it really helps. Johny, is there any video or thread where I can find about that (I mean Ian's motor)?

GlanDeas,
you are an elegant spambot and your creator is an stylish coder, but some moderator should delete you from this board….

BigMouse,
yep, you are right. It seems I had all the FW thing messed up.

InductEV,
I am not sure I agree with your view, but it would be very interesting to see your idea tested. In my view, paralleling coils is equivalent to reducing turns + augmenting section; but (as you pointed) there is an increased risk of unbalanced currents with parallel coils.
Anyway, it would be really cool to see you or anybody try it and post their results. I also toyed with the idea previously (in order to avoid rewinding).
User avatar
Johny
Senior Member
Posts: 3749
Joined: Mon, 23 Jun 2008, 16:26
Real Name: John Wright
Location: Melbourne
Contact:

Junkyard motors parade

Post by Johny »

peskanov wrote:Johny, is there any video or thread where I can find about that (I mean Ian's motor)?
Sorry I meant Ivan.
It's at the 1 minute 10 second mark on this video.
User avatar
peskanov
Groupie
Posts: 153
Joined: Sat, 15 Oct 2011, 00:55

Junkyard motors parade

Post by peskanov »

Update time,
we are reaching the finish line, and at this point I don't even care if it works or not! I just want to see the damned thing done! Image

Some pics of the coild heads packing. Here we are using the nylon thread.
Image
Here you can see (using some effort) the small wire we use to guide the thread trough the coils
Image
The head coils get much thinner when you pack them well, drawing strongly from the thread. But take care or you will get some blisters (or worse)
Image
When we ended one side of the stator, we got a nasty surprise. This Leroy-Somer is short and light (compared to other motor of this size) but the price to pay is less space for the coil heads.
Look at the end plates:
Image
Nearly flat! And we did not noticed it while molding the coil heads. Fortunately, some rubber mallet fixed the problem, but the heads are now too close to the plates and that makes me unhappy. Some nasty induction could be produced there.
Image
Packing result:
Image



User avatar
peskanov
Groupie
Posts: 153
Joined: Sat, 15 Oct 2011, 00:55

Junkyard motors parade

Post by peskanov »

After packing the coils, we tested the circuits for shorts, bad insulation, etc...Everything was OK but you cannot test if each phase has the same resistance, as winding for low V/Hz makes resistance so small a normal multimeter will not be able to read it correctly.

Now, the varnishing. That's easier, but also messy.
We wanted to use the method of closing the motor and filling it with varnish. However we just bought 2 litres and that was not enough, so we just varnished it in the classical way.
Fist we filled the slots, as these are the parts that really need the varnish (to get rigid).
Image
That took time, a lot of varnish can be pumped in these slots.
Next we painted/filled the coil heads using a brush.
Image
Image
...and it's still drying, as we did use the cheap varnish that dries at ambient temperature.

Next, the motor terminals. We used a gas torch and sulphuric acid ("Mr Rambo" Image ) to clean the enamel. As we are still working on it, that tale is reserved for the next chapter.
Image
User avatar
Adverse Effects
Senior Member
Posts: 1172
Joined: Sat, 01 Jan 2011, 03:30
Real Name: Adverse Effects
Location: Brisbane

Junkyard motors parade

Post by Adverse Effects »

that is some nice looking work / art there mate
User avatar
peskanov
Groupie
Posts: 153
Joined: Sat, 15 Oct 2011, 00:55

Junkyard motors parade

Post by peskanov »

Hi Adverse effects,
in the pictures looks much nicer than it is, we are learning an it shows when you a get close view. Thanks anyway!

Well, I am happy to say the motor is alive! Today we solved the terminals problem and tested it without load.
About the terminals, this task is a bit harder than it looks. We used a torch to burn the enamel an then submerged the wires into the acid.
Sulphuric acid is a nasty stuff, it perforated a plastic bottle, made a nice hole in Vicent's t-shirt, and burned my hand a bit through the rubber gloves.
Imho, the acid is unnecessary. If you torch the wires well, the carbonised remains of the enamel can be removed easily using steel wool.
Once you clean the wires (the best you can or tin will not work) you need to heat them enough to solder. That proved difficult, we tried the gas torch, a soldering iron and an hot air soldering system.
All them worked badly, but we found that a hot air gun we have around was ok for this task.
Image

We fitted the rotor into the stator to check the air gap was still ok.
Image

An then carried the assembled motor to our "testing mule". As the battery is only 60V, 15Ah I programmed to the Curtis to 15% current. I forgot to limit the current of the brake though, and we had an unexpected spike of 100A (end of the video) when touching the brake.
We tested little more...checking the phases with the ammeter we found it one was a little unbalanced, but the difference seemed to reduce with higher amps. We will see when the real testing beings.
http://www.youtube.com/watch?v=zXWnxlgcmME

Next step...the dyno! But it will take some time, as I need access to some serious battery first.
User avatar
Johny
Senior Member
Posts: 3749
Joined: Mon, 23 Jun 2008, 16:26
Real Name: John Wright
Location: Melbourne
Contact:

Junkyard motors parade

Post by Johny »

Fantastic work peskanov. Great documentation too! Thanks for the rubber mallet shot.
User avatar
weber
Site Admin
Posts: 2936
Joined: Fri, 23 Jan 2009, 17:27
Real Name: Dave Keenan
Location: Brisbane
Contact:

Junkyard motors parade

Post by weber »

I recommend a concentrated zinc chloride solution, also called Baker's soldering fluid, tinner's fluid or killed spirits. It will remove the copper oxides produced in burning off the enamel, without removing your skin and clothes. You wouldn't have to work so hard with the steel wool -- just remove the burnt enamel, then dip it in this stuff before soldering. Then wash it in water (after it has cooled) to remove this flux.
One of the fathers of MeXy the electric MX-5, along with Coulomb and Newton (Jeff Owen).
User avatar
BigMouse
Senior Member
Posts: 600
Joined: Thu, 28 Oct 2010, 02:39
Real Name: Vincent Tannahill
Location: Silicon Valley
Contact:

Junkyard motors parade

Post by BigMouse »

Well done! I hope it works out for you once you get a chance to properly test it. If the rotor is good, then you'll have yourself a nice motor!
User avatar
peskanov
Groupie
Posts: 153
Joined: Sat, 15 Oct 2011, 00:55

Junkyard motors parade

Post by peskanov »

Hey, the rubber mallet part of the process rules! It's satisfying to hit those damned copper wires, after suffering so much inserting them Image. Btw, I bought mine in a bazaar, price 2 euros.

Thanks for the tip Weber, we will try it for sure.

BigMouse, just checked the AC-x line of motors from HPEVS and surprisingly this Leroy-Somer corresponds to the top of the line AC-75! (if we get it to work accordingly to specs, of course).
I guess it make no sense to built them bigger, as there no affordable controllers for them.
However, I don't want to get to attached to this motor because now I know it was not well chosen, and was poorly rewound. If it does not work as intended...next one!

A note about the previous steps: the varnish can take eons to dry in ambient temperature, I would recommend heating the stator any possible way. We are going to open it an let it under the sun for some days, but an oven would be what doctor ordered...
User avatar
BigMouse
Senior Member
Posts: 600
Joined: Thu, 28 Oct 2010, 02:39
Real Name: Vincent Tannahill
Location: Silicon Valley
Contact:

Junkyard motors parade

Post by BigMouse »

peskanov wrote:A note about the previous steps: the varnish can take eons to dry in ambient temperature, I would recommend heating the stator any possible way. We are going to open it an let it under the sun for some days, but an oven would be what doctor ordered...
I'm not sure how this would work, but in some industrial applications the motor windings themselves can be used as heating elements for preventing condensation in open-frame motors. Perhaps you could experiment with feeding it a current-limited DC to make it hot to help the varnish dry.
Post Reply