So I rewound collossus today and I got 8 turns of 14awg per tooth. I did a AaA wind then run the wires out on both ends. Then wound the other side AaA and did the same and connected them in parallel on the out side leaving me the option to have 1/2 the KV with 4x the uH! My question is does it matter how I connect for delta? Can A+ go to any other wire other then A- for instance?
Inductance What is it what does it do? Collosus has 8uH!
- Thread starter Arlo1
- Start date
Thud
1 MW
interesting Arlo. Did you measure with all 3 teeth rwond in the same direction?
That is slightly in conflict with the video you linked in the 1st page of this thread. :|
about your AaA & terminating....I think if you keep the phases in correct sequence you willl be fine. Gonna be a little confusing the 1st time LOL (thats a lot of wires to bring out of the can)
That is slightly in conflict with the video you linked in the 1st page of this thread. :|
about your AaA & terminating....I think if you keep the phases in correct sequence you willl be fine. Gonna be a little confusing the 1st time LOL (thats a lot of wires to bring out of the can)
I have them all out and terminated. I Labled them on the out side of the can so I can chage them around if I want later. 
Lebowski
10 MW
efficiency wise a motor should have as much copper as possible... you still got lot of room for more windings !
Not for the rpm/volt I want with the wire I have avalile. I tried to wind it with 10.5 awg. with no luck. I ended up with 14awg with what should be lower resistance than stock and 30% more inductance and the same 75 rpm/volt as stock! I might be able to get it to work with 13 awg but I dont have any so next time when I wind the other one I will order some! You might think there is a lot of room for more copper but with 14awg you cant get 9 turns because of the cross turn from one tooth to the next!Lebowski said:
While were are talking about inductors....
Ok so I have a video I would like some of you to watch and hooked my oscilloscope up to the posistive and negative on the primary side of the coil and watched the charge time while it was running. I see some times there is a negative pulse off the coil and sometimes its not but when I zoom in it seems to always be there so it must just be chaging the lengh of negative pulse. I think but not sure if the negative pulse is short when a spark happens? And long when there is no spark??
SO this is a ICE engine sorry bout that lol. But I need some smart dudes to help me. This is a 2009 Yamaha grizzly 700efi it shows no error codes Yamaha wont do squat for the customer.... So here I am people are starting to bring me the most challanging electrical problems from all over BC!
What is happening is when I run the atv with the pug gapped at normal gap -.020-.035" it stalls at random at idle, bigger gap is worse and smaller is better untill I get to .005" which is where it will run perfact! I have a power comander installed and tried all kinds of settings on it. The power comander is helping with lean idle and lean start up and throttle responce but it still stalls at idle at random unless I gap the plug to .005"
[youtube]tQcPPT1QHv8[/youtube]
Ok so I have a video I would like some of you to watch and hooked my oscilloscope up to the posistive and negative on the primary side of the coil and watched the charge time while it was running. I see some times there is a negative pulse off the coil and sometimes its not but when I zoom in it seems to always be there so it must just be chaging the lengh of negative pulse. I think but not sure if the negative pulse is short when a spark happens? And long when there is no spark??
SO this is a ICE engine sorry bout that lol. But I need some smart dudes to help me. This is a 2009 Yamaha grizzly 700efi it shows no error codes Yamaha wont do squat for the customer.... So here I am people are starting to bring me the most challanging electrical problems from all over BC!
What is happening is when I run the atv with the pug gapped at normal gap -.020-.035" it stalls at random at idle, bigger gap is worse and smaller is better untill I get to .005" which is where it will run perfact! I have a power comander installed and tried all kinds of settings on it. The power comander is helping with lean idle and lean start up and throttle responce but it still stalls at idle at random unless I gap the plug to .005"
[youtube]tQcPPT1QHv8[/youtube]
Lebowski
10 MW
Back in my moped days first thing would be to put in a new capacitor.... does a modern ignition still have a capacitor ?
I have the feeling polarity is reversed on your scope, so the voltage should be high most of the time (90%), pulled to 0
10% of the time to charge the coil with the funny spike being an overshoot to above the 12V... can you check ? If I'm
correct the capacitor would be the first thing I would look at...
I have the feeling polarity is reversed on your scope, so the voltage should be high most of the time (90%), pulled to 0
10% of the time to charge the coil with the funny spike being an overshoot to above the 12V... can you check ? If I'm
correct the capacitor would be the first thing I would look at...
bigmoose
1 MW
Arlo, here is a little write up on ignition systems and their waveforms I did a while back. There was a time I was fiddling with these and designed and built an ignition system for a pulse detonation engine for "some folks." Enjoy...
http://mtg-technologies.com/automotivesubpages/ignition/ignition.htm
http://mtg-technologies.com/automotivesubpages/ignition/ignition.htm
No the polarity is correct. You can even watch the pulse stay hi for a few seconds when the ATV dies then it shuts off the power to the coil (so the coil doesnt get burnt)Lebowski said:
and the trace goes back down on the oscilloscope.
Thanks for the replies and bigmoose I did not get the same looking scope trace as you..... Thanks for the link very interesting!
Here is 3 more pictures. I installed a plug from a 2008 model (this is a 2009) because we don't have a 09 plug in stock ATM. And at the proper gap it is working good...
So it is possible this whole time something may have been wrong inside the plug!
I'm not convinced that Yamaha has enough dwell time.... Or like Lebowski is saying the cap is NFG might have been the wrong cap in the first place. Yes In most ignition systems they still use caps in the CDI or EFI module as far as I have seen. And yes I know the plug for a 2008 is shorter then the plug for a 2009 but it runs...... :|
Here is 3 more pictures. I installed a plug from a 2008 model (this is a 2009) because we don't have a 09 plug in stock ATM. And at the proper gap it is working good...
So it is possible this whole time something may have been wrong inside the plug!
I'm not convinced that Yamaha has enough dwell time.... Or like Lebowski is saying the cap is NFG might have been the wrong cap in the first place. Yes In most ignition systems they still use caps in the CDI or EFI module as far as I have seen. And yes I know the plug for a 2008 is shorter then the plug for a 2009 but it runs...... :|Attachments
John in CR
100 TW
Funky looking plug like that...no wonder it doesn't work right.
Lol yup. I will try to recover all my pics and links and re post them on this thread, Bigmoose would it be possible you you to put the spread sheet back up? I will look and see I might be able to do it as well but if you could edit your own post that would be best. We have John here who wants to try getting his bike to run with external inductors.John in CR said:
John in CR
100 TW
Arlo1 said:
My bike is already running with a choke on each phase, so no need to go to any trouble. Let the computer geeks try to do their stuff. I understand the benefits (at least I think I do), and it's really just a matter of whether mine are providing meaningful inductance.
bigmoose
1 MW
Arlo, I edited this post and put it back in as an attachment. If that isn't the one you were looking for let me know.
http://endless-sphere.com/forums/viewtopic.php?f=30&t=29852&p=433095#p433095
http://endless-sphere.com/forums/viewtopic.php?f=30&t=29852&p=433095#p433095
There you go John. go down load the motor currentsV2 and you will find its pretty cool to play with.
Lebowski
10 MW
It would be interesting to try another method to determine the inductance of this Collosus motor.
Arlo, if you have a transistor lying around you can try to build this simple oscillator and use your
scope to measure its resonance frequency (at collector or emittor of transistor). Try it with
C = 100 nF and 2200 ohm in the emitter for a start...
Arlo, if you have a transistor lying around you can try to build this simple oscillator and use your
scope to measure its resonance frequency (at collector or emittor of transistor). Try it with
C = 100 nF and 2200 ohm in the emitter for a start...
I'm not sure the reason for this. Lets say my meters are out ok.... So one is out of spec the same as the other... Not likely but slightly possible. Then from the air core calculator I wound up 3 15uH inductors and measured them at 15uH with my meters so the meters will read what they are supposed to on the hand wound inductors and not in colossus Very Very unlikely. They also read the inductors sent to me at the numbers I would expect! Then I measured my X5 and a HS35 and they seem to be where most people say they should be for measurement so the probability of my meters not reading colossus properly but everything else about right VERY VERY VERY unlikely. Its not that I won't build this circuit you suggest its that waisting my time on that when I need to finish my pickit lessons finish my brothers bmx build my self another race bike and or mod my bmx because Luke want me to race in SO-cal lol. And all the other projects I got on the go its just not smart for me to build this and mess around any more.Lebowski said:
I have blow over 100 4110 mosfets and destroyed 3 24fet controllers and 1 18fet controller on colossus there is a reason for this. Its not just low resistance like the spread sheet shows the resistance doesn't not effect amp flow nearly as much as inductance time dependently speaking!
John in CR
100 TW
I've done enough reading now to come up with some questions, hopefully good ones, not stupid ones.
1. The 100uH target inductance obtained from Cicero documentation seems arbitrary. Is that a useful target across a broad range of brushless motors and cheap controllers? With the means to measure inductance of DIY coils, varying the inductance by changing the number of turns is the one thing we have an easy means of control, but that's not very useful if the target inductance isn't a good one.
2. When I see different core types being more sensitive to frequency, would that frequency include our controllers' PWM frequency (pulses of DC) or just our motors' electrical operating frequency (the AC frequency) that increases with rpm? If it's the later, then ideally wouldn't we want a non-linear inductor for more torque from retarded timing and better controller protection at low frequency, and then the inductance rapidly rolls off with e-rpm for better high speed timing and controller protection coming from the motor's BEMF?
3. Core saturation- To me this is the biggie as well as the biggest unknown. On the motors I have, I want to get to where I can safely run phase currents of 300-500A. With relatively low value inductors in the range of 20-40uH, is it reasonable to avoid saturation at these kinds of currents without coils so big that 3 become ridiculously large and heavy. What material core would be best and still keep turn counts within reason for say 3ft of large gauge wire?
I'm using some 4.5" diameter ferrite ring magnets as toroidal cores. I want to stick with toroids to keep stray flux down as well as reduce the turn count required for a given inductance. My motor seems happy with the 6 turn coils I'm running, as it is definitely quieter on take-off and runs cooler but at higher performance (though that may also be attributable to the larger phase wires I installed). The controller hasn't blown, so I guess it's happy. Also, the magnets haven't lost their magnetism or gotten the least bit warm, which makes me think I'm not close to saturation. If there's a better material for a DIY toroid, I'm game. Candidates include a doughnut shaped roll of thin steel wire with insulating electrical motor paint as the insulator before forming the rolls. I can easily get sufficient magnetite to mix with epoxy and mold to almost any size, and I'm sure with a magnet I can go to my buddy's machine shop and get plenty of steel powder/filings to do the same. If necessary to increase the saturation limit, I think air gaps can be cut or made into any of these toroids.
John
1. The 100uH target inductance obtained from Cicero documentation seems arbitrary. Is that a useful target across a broad range of brushless motors and cheap controllers? With the means to measure inductance of DIY coils, varying the inductance by changing the number of turns is the one thing we have an easy means of control, but that's not very useful if the target inductance isn't a good one.
2. When I see different core types being more sensitive to frequency, would that frequency include our controllers' PWM frequency (pulses of DC) or just our motors' electrical operating frequency (the AC frequency) that increases with rpm? If it's the later, then ideally wouldn't we want a non-linear inductor for more torque from retarded timing and better controller protection at low frequency, and then the inductance rapidly rolls off with e-rpm for better high speed timing and controller protection coming from the motor's BEMF?
3. Core saturation- To me this is the biggie as well as the biggest unknown. On the motors I have, I want to get to where I can safely run phase currents of 300-500A. With relatively low value inductors in the range of 20-40uH, is it reasonable to avoid saturation at these kinds of currents without coils so big that 3 become ridiculously large and heavy. What material core would be best and still keep turn counts within reason for say 3ft of large gauge wire?
I'm using some 4.5" diameter ferrite ring magnets as toroidal cores. I want to stick with toroids to keep stray flux down as well as reduce the turn count required for a given inductance. My motor seems happy with the 6 turn coils I'm running, as it is definitely quieter on take-off and runs cooler but at higher performance (though that may also be attributable to the larger phase wires I installed). The controller hasn't blown, so I guess it's happy. Also, the magnets haven't lost their magnetism or gotten the least bit warm, which makes me think I'm not close to saturation. If there's a better material for a DIY toroid, I'm game. Candidates include a doughnut shaped roll of thin steel wire with insulating electrical motor paint as the insulator before forming the rolls. I can easily get sufficient magnetite to mix with epoxy and mold to almost any size, and I'm sure with a magnet I can go to my buddy's machine shop and get plenty of steel powder/filings to do the same. If necessary to increase the saturation limit, I think air gaps can be cut or made into any of these toroids.
John
I think 100uH will be a goal untill i can make better tests results to show how much we realy need. I say this based on the rc motor guys having issues with 40-60 uH and my x5 having no isues at 255 uH and my brothers hs35 having no isues just over 100uH
John in CR
100 TW
Arlo1,
Do you know how many turns the HS35 has? My main motors are essentially the same stator with 5mm more length and only 2 turns of right at 8ga equivalent magnet wire, so if his doesn't have issues pushing past 100A, then the 100uH additional should be a good target for me, since that should also cover the lower resistance of mine. Rhitee05 did have me a bit bamboozled about the R/L ratio stuff.
Do you know how many turns the HS35 has? My main motors are essentially the same stator with 5mm more length and only 2 turns of right at 8ga equivalent magnet wire, so if his doesn't have issues pushing past 100A, then the 100uH additional should be a good target for me, since that should also cover the lower resistance of mine. Rhitee05 did have me a bit bamboozled about the R/L ratio stuff.
John as far as I'm concerned I don't need to dig into the L/R ratio stuff
All you need to do is get the lowest resistance possible and the right amount of inductance.
In the end you will have numbers to punch into the spread sheet bigmoose provided and see the l/r ratio and how your motor phases should react to a given voltage over time!
All you need to do is get the lowest resistance possible and the right amount of inductance.
In the end you will have numbers to punch into the spread sheet bigmoose provided and see the l/r ratio and how your motor phases should react to a given voltage over time!
rhitee05
10 kW
John in CR said:
The frequency effects of the core material apply to losses, not to the inductance. Losses are mostly hysteresis losses, which are a property of the material itself, and eddy current losses, which are a function of how the core is assembled (lamination thickness, etc.). The inductance value will not vary with frequency.
John in CR said:
Saturation will be a major issue at those high currents. It will cause the inductance to decrease at high currents, which is exactly when you need the inductance the most. Saturation depends on the material. Most steels have saturation flux density in the range of 1.6-2 T (T = Tesla), depending on the alloy. I did a quick Google search and it looks like most ferrite materials saturate around 0.3-0.4 T, again depending on the exact material. It's important to point out that, due to the B-H curve shapes of these materials, the inductance value will begin to decrease well below saturation flux. You'd probably want to keep a steel core below 1 T, and a ferrite below around 0.2-0.25 T to keep some useful inductance.
Fortunately for designers, the inductance of a coil is directly linked to the flux density. L = N * Psi / I, where N is the number of turns and Psi is the flux [Wb = Weber = T*m^2] (total flux, not flux density), and Psi = B * A, where B is the flux density [T] and A is the core cross-sectional area [m^2]. Using this formula, given a max current and a target max flux density, you can figure out how much core area is required for a given number of turns. For example, let's say N=10 for your 20 uH inductor that you want to carry 500 A. That gives Psi = L * I / N = 0.001 Wb . Allowing 0.2 T for a ferrite core, that will require a cross-sectional area of A = Psi / B = 0.005 m^2 = 50 cm^2, roughly 7 x 7 cm. That's a pretty big toroid. Using steel at 1.0 T, it would be 0.001 m^2 = 10 cm^2, which is still not exactly small.
An iterative design process is probably needed, where you assume a geometry, figure out how many turns are required to get the inductance you want, then check for saturation, modify the geometry, repeat. When specifying how much inductance you want, you should calculate the turns for a higher value because any online calculator will assume no saturation. 50%-100% higher would not be excessive. Obviously there's a trade-off between fewer turns for less resistance, but a larger core.
John in CR
100 TW
Nice post Eric. That's how text books should be written, understandable and complete. Does PWM throw a monkey wrench into looking at it from the current viewpoint? I was looking at the saturation info and example here http://thedatastream.4hv.org/gdt_saturation.htm, and that leads me to think that maybe we should look at the volt-seconds vs the cross sectional area of the core to determine a minimum number of turns to avoid saturation, since the electrical flow is turned on/off at a frequency of I believe 40khz. At a 50% assumed max duty cycle in the region where extra inductance is most needed and my 80V peak voltage off the charger, I came up with a minimum of 2 turns required with my big toroid ferrite cores to avoid saturation. My 6 turns is adequate in terms of frequency.
Does PWM help dodge the bullet in terms of high current causing saturation, or does it force us to look at it from both aspects to make sure the core is adequate?
I'm still unclear about how I need to test for saturation. Can a cheap inductance/capacitance meter measure inductance while the coil is in use? Would saturation show up as a warm or hot core like happens with stator saturation?
John
Does PWM help dodge the bullet in terms of high current causing saturation, or does it force us to look at it from both aspects to make sure the core is adequate?
I'm still unclear about how I need to test for saturation. Can a cheap inductance/capacitance meter measure inductance while the coil is in use? Would saturation show up as a warm or hot core like happens with stator saturation?
John
rhitee05
10 kW
It's a little more complicated than the volt-second analysis. As they note: "If this were an inductor carrying a DC component, the DC current would play a significant part in calculating the saturation." What you get with the PWM is an AC waveform - the current ripple - superimposed on top of a DC current - the average current. Hopefully, the DC current is much larger than the ripple (that is the point here, after all), so the DC current is much more important for saturation. I'm ignoring here the much slower commutation frequency. So, unfortunately, there is no free lunch and PWM does not get us off the hook for saturation since the DC current is more important than the AC waveform.
You should be able to rig up something to test this using a simple LCR meter. All you need to do is use some capacitors to isolate your LCR meter from the inductor. It would depend on what frequency your meter runs at, but the biggest ceramic caps you have around should be good enough, I'm thinking something like 10 uF. You want the impedance of the caps to be much smaller than the impedance of the inductor at the measurement frequency. Then you'd apply a DC bias to the inductor. One way to do this is just use the controller like usual, except lock the motor so the commutation stays fixed. By measuring the inductance at zero and increasing amounts of current, you can come up with an inductance-vs-current curve. The inductance should be constant for low values of current and will start decreasing as the core saturates.
You should be able to rig up something to test this using a simple LCR meter. All you need to do is use some capacitors to isolate your LCR meter from the inductor. It would depend on what frequency your meter runs at, but the biggest ceramic caps you have around should be good enough, I'm thinking something like 10 uF. You want the impedance of the caps to be much smaller than the impedance of the inductor at the measurement frequency. Then you'd apply a DC bias to the inductor. One way to do this is just use the controller like usual, except lock the motor so the commutation stays fixed. By measuring the inductance at zero and increasing amounts of current, you can come up with an inductance-vs-current curve. The inductance should be constant for low values of current and will start decreasing as the core saturates.
Similar threads
