My Motors got me to Tap into Radiant Energy

[Midaztouch]

Great!

G'day Team

of the 2 monster driver I made the first worked OK the second ? I need to diagnose it
As I said I would post some pics of my progress so here they are

Kogs UFO Projects Photos by Kogs1 | Photobucket

I had trouble getting my test driver mobile as she had not ridden a bike for quite some time I'm sure she will feel safer when she test Drives EVTrike1


The motor is a standard MY1018 36v 250w motor
I now have to keep using Kogs EVbike1 to see what problems I may encounter then I will exchange the motor with a Asymmetrically wound motor ( I have to buy 2 more motors to do this).

In the meantime I will be Wiring my 1000w Chinese motor to fit in my EVTrike1 of which I will use firstly the same set up then use an Arduino to control the motor to see what the difference is.

Kindest regards to you all



Kogs still progressing trying to catch up, running out of time

Kogs,

Awesome! I loved the smile on your wife's face. That's what we call an EV Grin.


When you use your Asymmetric motor, there shouldn't be any overheating with your boards. Also use a chain from the motor to the gears. If you want to reduce the size of the box on the back by 1/2, buy lipo batteries. Anything over a 20c rating will not be dangerous for your 1000w motor. Your max amp draw will be less that 25% of the batts max amp discharge.
Hobbyking has a lot of choices.
ZIPPY Flightmax 8000mAh 6S1P 30C


Keep it Clean and Green
Midaz

[JohnStone]

RPM sensor

Hi guys out there
still pondering on rpm sensor and how to intergrate it.
What about tachogenerator out from washing machine motors (source = junk yard). They have a magnet ring around the shaft rotating and a pickup coil around it not moving.
Diametrical magnetisation (2) of magnet required! (not like speaker magnets (1)!)

• 1) axial magnetisation
• 2) diametrical magnetisation
• 3) radial magnetisation

Example for source and here .
Transistor and resistor required to adapt to Arduino interrupt. The magnet ring can be glued or screwed to end of shaft if diameter does not fit.
I have no such motor available and do not know how many pulses they emit per turn. Can somebody check how those tacho sensors perform?



Similar procedure if you add a drive motor from CD/DVD device.


JS

[Zardox]

Thanks Kogs for the nice clear photos of your driver. I finally have some of the parts ordered to get started and that will be a big help.

[iankoglin]

Thanks Kogs for the nice clear photos of your driver. I finally have some of the parts ordered to get started and that will be a big help.

G'day Zardox
Thanks for the commendation I used my IPad it takes great photos
The Monsters are the exact ones that John Stone designed with the exception I have widened the gap between C7 and the Mosfets this allows me to put the R7 on top and makes it easier to solder the Mosfets also the lands are slightly wider than John Stone had made to allow more solder as he suggested.
I used "Fab in a box to make the PC Boards.They come out all right

Kindest Regards


Kogs still here

[machinealive]

Good-day

It took a while but I managed to get boxes finished, and sensors setup. It's funny you bring up rpm Johnstone, because for some reason, that was the only a sensor that didn't record right. Reading was quite a bit higher then measured, and slowly increasing. Current sensor seemed good, at 12 v, the current measured b/t 10 and 11 amps, as measured initially, with DMM. So, I'll be looking at the rpm program today.

Machine

[RAMSET]

Gentlemen, a formal invitation

A chance to teach and be taught ,a chance to showcase and experiment.

But most of all some good clean fun.....

Here

Welcome To The 2013 Pulse Motor Build Off!! Ready Set GO!!! Its ON!! - YouTube

Direct link to the forum
International Alternative Energy Center - 2013 Pulse motor build off - Powered by ForumCo.com - The Forum Company

Thx
Chet
PS
Girls are allowed too.......................

[Cornboy 555]

Bugger.

Hello John Stone, Et al, have finished and tested nand gate section, of Monster, i am sorry for taking so long, for some reason i wasn't expecting the very next post to be the next instructable, and i missed it.

So i was waiting for John to post it, thinking he's busy, then going back, i found it, and was very embarresed.

Thanks John, ready when you are!,

Warmest regards, Cornboy.

[GChilders]

RPM Sensor

@John Stone
Hi John,

When I finished the third replication I cut a four inch diameter piece of lexan and two 2 inch diameter pieces to reinforce this piece. I then drilled 4 mm holes in the center and epoxied a 1 inch diameter 3/8 inch thick piece of aluminum, that had been drilled and tapped with a set screw , with this assembly. This is what I am going to use for two purposes. First this will act as the timing mechanism for the motor, and secondly as a tachometer. How will I accomplish this dual purpose. First my choice as I had mentioned before in the forum before I wrote the tiny piece of code that was used to demonstrate the quad fade method. I have chosen to use an infrared transmitter coupled with an infrared receiver which is coupled to the interrupt signal of the Arduino. These are dirt cheap, they are everywhere and ultra dependable. They are used here in the US for safety of garage door closers. The one that I have on my garage door has been faithfully operating without a hitch for over twenty years. The lexan wheel can be manipulated in two ways, in order to adjust the coarse and fine tuning of the motor. First I set the coarse tuning with the set screw, I can change this at any time when the motor is stopped, advance or retard the timing easily. Second, by using any opaque tape I can incrementally change the timing until a sweet spot is found. By moving the motor control code into the interrupt sub process this can be used to time the motor in a very satisfactory way as the code will be executed every rotation of the motor and all adjustments to the duty cycle and such will be calculated upon every revolution of the motor since the interrupt in this instance will only occur once per revolution. Any way that is my solution which may not work for some applications. Hope that it helps.

Cheers,

Garry

[JohnStone]

Instructabel E: FET driver

Hi Cornboy and sthose builder of moster drivers in obscure!
Congratulationss! Now you have successfully built the environment where the true FET driver will dwell happily. It lives like queen bee in a hive.

Now lets look at those inhabitants: FET driver, FETs and protection circuit. In this session we will deal with FET driver only.
But first of all I want to explain why we need a dedicated FET driver while others can drive some FETs from Arduino directly or by a simple NE555 timer circuit or other small bore components. All is a matter of performance.

If we want to drive some few amps is is not essential if we have some few milliwatts of additional loss. It is similar like you drilling some screws in the wall and hang a box on those screws. You will not need a dedicated stress analyst for safety reasons. Very different if you intend to build a bridge over your local river. Plese regard this imagination if we deal with our monster driver.
At several hundred volts and about 100A any tiny overstress might be desastrous. Therefore I designed some safety "ingredients".
The first of these measures is to introduce a strong FET driver.

It's pretentious task is to get the FET readily on and off. Conversely to normal notion this is not easy. Most of you might think that a FET is a component we can switch almost powerless because it opens and closes by exitance or absence of an electric field applied to gate only. In theory this is correct but if we look closer -> the gate owns two hidden and parasitic capacitances - one related to sourse pin and one related to drain pin. For ease of unsderstanding we peculate the third one (source / drain)

Take it for your memory that a FET is a circuit for itself!
Now let's look to capacitance Cgs. It behaves like a bucket. We need to fill it with electrons in order to rise fluid level (voltage). And only the voltage is able to get the FET conductive. You easily can imagine if you open the tap above the bucket to spend droplets only it will take much more time compared to press the button at your flush toilet. Your driver here (IC7) is just an electric flush mechanism. It can pump up to 12A into the bucket in order to fill it in a sudden.
Now do not believe this driver can do it constantly - no for some µs only but it suffices. For comparison a NE555 can pump not more than 300mA!!!!!
Imagine the impact of a 1/2 pound hammer. Constantly it can press with 1/2 pund of course. But if you use it in dedicated mode as a hammer it can push with up to 8 tons - for short itme of course.......... And you know the effects you can achieve!
The driver takes these 12A out of the allocated capacitors. Therefore it is essential to have SHORT copper leads and a cascade of capacitors of different make and value. This team is made to apply brute force - tenderly to the FET.
The advantage is that our FETs will stay very short time in intermediate state of half open / half close - an extreme lossy state. Apart that we foster radiant if we get switches go as fast as possible.
All facts above relate to switch off state as well.
Just as short hint: The other capacitance Cgd is a penetrant antagonist to our FET driver. I do not want to go in detail but please understand that if we want to make the FET conductive it will oppone. If we want to get current flow shut - it will oppone...... So understand the driver is loaded with this fight as well and it will get more seriously if we deal with increased voltage in our setup.
(BTW: A FET is theoretically infinitely fast (some picoseconds). Only those capacitances are responisble for slowness. Silicon Carbide (SiC) technology allowes much less capacitance and hence those components can switch much faster ... and they purge your purse much much faster than you can watch - currently)
Now you possibly understood the right to exist for this tiny FET driver. And you hopefully understand that this team is well tuned. Any modification neds to regard all seide effects mentioned above and some more I will not dicuss here.

Resistor R8 stays there in order to get input securely low at power on time. We do not want any unsolicited switching action while power on state.
The driver requests a steep edge at input signal in order to prevent unsolicited spuirous spikes at output. Therefore we conditioned the signal with the logic IC from last section.
And last but not least we have a LED at output in order to have a visual feedback of operation. (R7, D11)

Now, assemble components mentioned above and and play with your circuit. Operate input manually or by a generator and observe behaviour of LED. It shall be on or off but not glowing with half intensity.
For first PCB it is advisable to use an IC socket for IC7 but for the other boards please omit it. We do not want to add additional drawbacks for our dream team.
If you have a scope please measure the signal at IC7 pin 1 and pin 3. You should not have any spurious oscillations or lazy edges in your signal measured. For reference please note the time the signal goes from low to high (rise time) and vice versa (fall time). You will recheck this measurement after adding your FETs in next section F.
BTW: rise and fall time of signals regards the time from 10% to 90% of high signal amplitude.


BTW: Your seemingly fast scope might be too slow in order to see any slope. The time expected will be far below 50ns!!!!!

Check operation with sine or trinagle signal from your generator. The slopes shall be of same quality like at square wave. You possibly will observe change in duty cycle. Do not mind it.

Apply a signal with low duty cycle and observe the output of driver. Is it same or inverse? Memorize it! Later on you might want to know at what input signal FETs are on or off.

You might feel that a seemingly delicate circuit is very easy to build. You are right - thanks to those scientist who invented integrated circuits and sophisticated this technology constantly. Thanks to them we have today cheap Arduinos in private use and monster drivers with performance not being imaginable some years ago. So please regard this success of global community as well. On the other hand we intend to add with these tools another realm of new knowledge to energy processing.

JS

[JohnStone]

@John Stone
Hi John,

When I finished the third replication I cut a four inch diameter piece of lexan and two 2 inch diameter pieces to reinforce this piece. I then drilled 4 mm holes in the center and epoxied a 1 inch diameter 3/8 inch thick piece of aluminum, that had been drilled and tapped with a set screw , with this assembly. This is what I am going to use for two purposes. First this will act as the timing mechanism for the motor, and secondly as a tachometer. How will I accomplish this dual purpose. First my choice as I had mentioned before in the forum before I wrote the tiny piece of code that was used to demonstrate the quad fade method. I have chosen to use an infrared transmitter coupled with an infrared receiver which is coupled to the interrupt signal of the Arduino. These are dirt cheap, they are everywhere and ultra dependable. They are used here in the US for safety of garage door closers. The one that I have on my garage door has been faithfully operating without a hitch for over twenty years. The lexan wheel can be manipulated in two ways, in order to adjust the coarse and fine tuning of the motor. First I set the coarse tuning with the set screw, I can change this at any time when the motor is stopped, advance or retard the timing easily. Second, by using any opaque tape I can incrementally change the timing until a sweet spot is found. By moving the motor control code into the interrupt sub process this can be used to time the motor in a very satisfactory way as the code will be executed every rotation of the motor and all adjustments to the duty cycle and such will be calculated upon every revolution of the motor since the interrupt in this instance will only occur once per revolution. Any way that is my solution which may not work for some applications. Hope that it helps.

Cheers,

Garry

Hi Garry,
I pondered on this item for long time in order to find a general solution easy to use and I decided that it depends essentially on individual setup. Now our community has the choice from very different approaches and everybody will decide for his preference.
I understand you have a vane wheel and a photo interruptor. Your vane has as many slots as your armature poles. That is exactly what Cornboy plans to do.
I am a visually oriented person and my imagination might be wrong. I still ponder on how you know which driver to operate at which interrupt. I can imagine to either have 4 (6 for Cornboy) sensors or a trick to synchronise full revolution. Would you mind to post a pic of your suggestion?

But I have an additional question:
If we have an interrupt for every pole passing - and given this interrupt occures in advance of pulse start intended. Will it be possible to retard the pulse by program in steps of i.e. degree. This feature would ease massively finding sweetspots without any mechanic provisions.
We know timed distance of vane interrupts and we know how many degrees they represent. This will suffice to get electronic ignition timing run. Just an idea!
John

[Cornboy 555]

Pics

@John Stone
Hi John,

When I finished the third replication I cut a four inch diameter piece of lexan and two 2 inch diameter pieces to reinforce this piece. I then drilled 4 mm holes in the center and epoxied a 1 inch diameter 3/8 inch thick piece of aluminum, that had been drilled and tapped with a set screw , with this assembly. This is what I am going to use for two purposes. First this will act as the timing mechanism for the motor, and secondly as a tachometer. How will I accomplish this dual purpose. First my choice as I had mentioned before in the forum before I wrote the tiny piece of code that was used to demonstrate the quad fade method. I have chosen to use an infrared transmitter coupled with an infrared receiver which is coupled to the interrupt signal of the Arduino. These are dirt cheap, they are everywhere and ultra dependable. They are used here in the US for safety of garage door closers. The one that I have on my garage door has been faithfully operating without a hitch for over twenty years. The lexan wheel can be manipulated in two ways, in order to adjust the coarse and fine tuning of the motor. First I set the coarse tuning with the set screw, I can change this at any time when the motor is stopped, advance or retard the timing easily. Second, by using any opaque tape I can incrementally change the timing until a sweet spot is found. By moving the motor control code into the interrupt sub process this can be used to time the motor in a very satisfactory way as the code will be executed every rotation of the motor and all adjustments to the duty cycle and such will be calculated upon every revolution of the motor since the interrupt in this instance will only occur once per revolution. Any way that is my solution which may not work for some applications. Hope that it helps.

Cheers,

Garry

Hello Garry, thanks once again for joining in like you have, it would be a great help to all here if you could perhaps, include some pics of what you are doing, to give us a clearer understanding.

Thanks once again,

Regards Cornboy.

[Hitby13kw]

temperature sensors

@ Team & John Stone, Been looking at other temperature sensors that are available, came across the DHT11 and his big brother the DHT22 this sensor can also report the humidity and dew point data, and you only need one digital pin for each sensor so multiple data points (both bearings, each of the eight brushes, the motor case, etc.) would be a piece of cake, main downside is this unit is slow ~ can only pull data from it once every 2 seconds.

Another unit that might be worth considering is the DS18B20 it is offered in two versions, just the sensor and in a nice protective case with 3 foot of wire either way, this unit communicates via the Dallas 1-Wire protocol so an unlimited number of sensors (a unique 64 bit ID is burned into each chip) could all be tied to just one digital I/O pin, one of the main advantages of this sensor is the wide usable temperature range it covers: -55 to 125°C (-67°F to+257°F)

Hitby13kw

still working the plan

[Cornboy 555]

Driver section.

Hello JS, everyone, have finished the driver section of the Monster, man it's good to see that led flash to your pulse input.

You are a very good teacher JS, i feel confident enough, to even trouble shoot a problem now, Thanks so much.

I have a scope that goes to 2.5ns/div but am still trying to work out how to use it, but all looks good with the monster so far.

[IMG][/IMG]

Warm Regards, Cornboy.

[JohnStone]

@ Team & John Stone, Been looking at other temperature sensors that are available, came across the DHT11 and his big brother the DHT22 this sensor can also report the humidity and dew point data, and you only need one digital pin for each sensor so multiple data points (both bearings, each of the eight brushes, the motor case, etc.) would be a piece of cake, main downside is this unit is slow ~ can only pull data from it once every 2 seconds.

Another unit that might be worth considering is the DS18B20 it is offered in two versions, just the sensor and in a nice protective case with 3 foot of wire either way, this unit communicates via the Dallas 1-Wire protocol so an unlimited number of sensors (a unique 64 bit ID is burned into each chip) could all be tied to just one digital I/O pin, one of the main advantages of this sensor is the wide usable temperature range it covers: -55 to 125°C (-67°F to+257°F)

Hitby13kw

still working the plan

Hi Hitby,
there are plenty of different sensors but they might not fit to application.
DHT are air sensors they can not be attached to a solid component.

DS... Sensors use one wire protocol. Please check if it can be used at same time with I2C. They might interfere with interrupt or dedicated pins or program space. The canned type is for fluids or air.
At one wire devices it is not possible to set a known address like at I2C. You need to add scan cycles to program in order to read IDs then you need to store found IDs and find out wich ID is which sensor. The only way to discern is to heat one, read all sensors and check temperature. Then you need to configure sensor name to ID. Overall a boring procedure along many pits to fall in. And it needs to be repeated by any user and any replacement, addition of sensors.

Found a I2C level shifter at sparkfun:
You need to add it once to the temp sensor branch. Then you can use multiple TMP102 -> being of 3.3V type. You need to add 3.3V poswer supply (might come from Arduino):

ELSE you get the breakout board PMB102 and replace the IC with TMP101. Then you are 5V ready. But you have still teh challange to centact it thermally and dirt proof to your setup.

I do not want to act as a defeatist. I might be wrong. I still am eager to find viable alternatives. As I do not know one just now, I ordered yesterday TC74A for €2.08 each at Farnell Element14 alias Newark. They are I2C, need a cheap insulation kit but are dedicated to be attachted to solid objects. (housing TO220 - like voltage regulators 7805 or 7812) (see my post #2337
Unfortunately eBay is no source for this component but all distributors sell it: Mouser, Newark, Future,, Alliedelec, Franell, Arrow, RS components, Avnet, Digikey, eBay, Microchip direct, dontscrapit
JS

[JohnStone]

....

I have a scope that goes to 2.5ns/div but am still trying to work out how to use it, but all looks good with the monster so far.
.....

Congratulations!
What type of scope?

Some low end explanations for all:
Imagine a scope to be some apparatus like a seismograph. The paper shifts from left to right and the "electrical" vibrations get the hands with pencil excited.
Different from seismograph you can change the shift speed for paper and the sensitivity of writing hands.
Another difference is that you have not a long paper banner pathing from source paper coil to destination coil but your scope writes one sheets of paper one after another. Imagine a flicker book. But at your scope-flicker book the contence of pics change and the paper frequency needs to be synchronized.
Fortunately modern scopes own an automatic - so called - trigger mode. This special circuit looks thoroughly to edges and slopes of your input signal and tries to get it synchroized with the paper change (beam sweep over screen). How? The beam waits at left hand side being switched off for a trigger. If the trigger unit detects a slope or edge it fires the trigger in order to get the sweep started -> traveling with predefined speed over yoru screen. But there is a timeout available as well firing the trigger even if there was no signal or slope detected.

Hands on exercise:
1. You have a dial for GND position of beam (vertical, x-axis). Set it with no input signal connected to be in the center of screen. You will have 5 devisions of grid above and 5 devisions below.
2. Adjust your generator to 1 KHz frequency and 5V peak to peak output level - that is -> it will swing 2.5V above GND and 2.5V below GND.
3. Your task is to adjust for Y-axis of screen the time dial to about a fittng time base. If you adjust your generator i.e. to 1 KHz it will be suitable to adjust your scope to about 1ms per div.
4. Then adjust your input amplfier (X-axis of screen) to 1V per division. 5. Now regard your probe type. If it is 10:1(will be printed on it) type you shall account for it by setting the X-dial to .1 V per devision.
6. Now connect your generator and play with those trigger dials, buttons in order to explore their function.
7. Change frequency and / or amplitude at your generator and find out what dials need to be adjusted and how in order to get a nice pic on your screen.

OK, now play with your toys for self education.

BTW: Some scopes own an "autoset" button. You need to have the signal connected and then press (grandma-) button. You're done! But learning effect = 0!!!!! Sit back in your grandma chair and doze another hour.
Other tutorial: here

JS