If this is your first visit, be sure to
check out the FAQ by clicking the
link above. You may have to register
before you can post: click the register link above to proceed. To start viewing messages,
select the forum that you want to visit from the selection below.
Hi friends,
I'm back now. Still few time left. Still digesting contributions from Ufo's other thread since I was in vacation.
Sorry for neglecting this thread :-) and you all - of course.
First of all I'd like to encourage you all to follow Ufo's hints - especially to start with the less complicated circuit if you have tiny knowledge in electronics only.
I ask you kindly to trust in Ufos's hints. There are very few items I could argue with him if I desired to do that - but I will NOT! Not important enough.
The main task is to get that monster of a coil pulsing at all. Improvements can be added later on.
My circuit suggested earlyer is a sophisticated one (not my invention) and it gives nice signals but it is not essential to use it for first step.
I'd like to add a hint for you. It is highly ercommended to learn to read and draw circuit diagrams. The current tends to flow from top to bottom (like water) and the components can be layed out in order to understand the function readily.
What about using Eagle free version You can draw one sheet only but this will suffice.
Please agree what to use and do it. It will make life for your tinkering much more easy.
Larry! - How are you?
Adilu - nice to see you being active!
Anoop - Do you still have problems with your switch off time? If yes - where can I find your current circuit?
I will look into this thread about once a week.
rgds John
Mr Stone, please give us newbies (with very limited electronic knowledge) more info concerning how to implement your last quoted sentence above.
Thanks for any help.
Rick
Hi,
Larry gave you the circuit. 18K / 82K as voltage devider. It should suffice in order to protect the sound input. But please be careful and measure your circuit without the coil for now.
Can anybody post the max. input voltage in order to get max. display on PC-Scope? If I know the value I can suggest additional protection circuits.
What PC-Scope do you use. It would be advisabele to not use different ones in order to get better help.
1. Please do not expect miracles from PC-scopes. They sample the signal with 44KHz. This gives a good signal on scope up to 5 Khz input signal - sine. Much less for square wave. It will be displayed distorted.
So please understand the scope as minimum viewing tool for those not having ANY funds for a real scope. Just basics for debugging the oscillator.
You shurely will not evaluate the switching speed of the FET
2. Those with some small funds shall look for a small storage scope i.e. this item. You will measure up to 1 MHz sine and this will suffice for most circuits here.
Neverthelless a 20ns switching edge will display like a 500ns.
I do know it is off topic, but earlier Larry mentioned he was looking for a solution to restore his health from cancer.
Came across something that may be of interest to you and anyone else having health trouble and searching for a solution.
This information about Vitamin C has been known by the medical doctors for many years but kept out of the mainstream and few know about its wonders similar to Ufopolitics way of winding motors and few if any of the "doctors" here will prescribe the treatment or get in and do surgery themselves on the motor and share it with the rest of us. Thank goodness for those who have stepped up to support this "alternative medicine" Ufopolitics has brought us and are openly sharing.
This is a 1 hour 25 minute educational video with lots of medical journal references to back it up on how to cure many diseases very affordably with 4 items mixed together that can be obtained at a health food store along with an ultrasonic cleaner to microencapsulate the Vitamin C:
Doctors treating Auckland farmer Alan Smith had decided it was time to turn his life support machine off, until a timely intervention by his family and Vitamin C, saved his life.
Here is a formula I located to make Homemade Liposomal C (Liposomal C is referenced in the above movie):
Anoop - Do you still have problems with your switch off time? If yes - where can I find your current circuit?
Actually it turned out that there was really no problem! I was using a 10K ohm resistor as load, which I had connected across the coil. My scope was hooked up across this load resistor. It seems that at switch off the Mosfet internal capacitance charge was discharging too slowly as it's only route was through the 10K resistor and also through the hi-impedance scope probes. I reduced the resistor to a few hundred ohms and Wow, I got a beautiful sharp drop on the curve at switch off!
In the beginning I just could'nt figure how I was getting good switch on times (which is normally the more difficult part) and yet obtaining very slow switch off. Exhaustive searching on Google findly led me to the answer.
Thanks for remembering!
I've been mainly on the other thread recently, and IndianaBoys' "off-topic" post at that thread brought me here, where I noticed your message.
Hi friends,
recently there were some short discussions regarding PWM and generators in another thread of UFO. I decided to shift the contributions here to in order to separate them from items being dedicated to moters. I feel this to be essentiel in order to give newbies a chance to follow.
Experts spend hours a day in order to question their doing while others stopped thinking feeling they were professionals.
10.01.2012:
Question:
... Most of the video I took, I used 4 ( 2 per driver) fets, each rated at 600v, 70 amps. I would recommend higher rating, higher price fets, instead of half the rating, at half the price (and using 2x as many).
How would/What would you use to drive jumbo, with 20 fets?
.....
Answer:
One essential property of FETs is their ON resistance. Higher voltage FETs tend to have higher R ON (see data sheet RDSon. And they show some additional drawbacks.
600V 70 amps is way enough for a single FET as long we operate our toys.
Perhaps you should decide to add an individual driver for each set of 4 or two FETs.
Thorough scope measurements are essential in order to build a FET driver from scratch as long you use no template like UFOs suggestion.
Experts spend hours a day in order to question their doing while others stopped thinking feeling they were professionals.
10-14-2012
Hi ALL,
some of you have no funds in order to buy an oscilloscope. This fact makes me feel sad. This tool - normally - is essential in order to check the performace of your pulsing circuits.
Additionally I'd like to direct you to my posts in UFO's very first thread where I posted a lot of important knowledge regarding circuits,diodes, FETs and driving FETs.
Today I want to give you some essential know how regarding driving FETs along porposals for you guys with restricted funds. You should be able to build performance drivers as well.
Tutorial:
1. FETs are modern electric valves with some very superiour properties compared to transisors. This makes them a primary choice in order to switch high currents along low loss.
But FETs will perform well only if they are kept within their area of wellness. Unfortunately many of you do not now these conditions and therefore you torture them without any malicious intention.
2. Any valve performs well only if you switch it fastly. Any intermediate state will perform excessive losses. You experienced it before if a switch (valve) in your home does not perform well and the contacts get hot - and possibly ignite your home. So the question is: How do we get FETs hurry up in their switching time.
3. FETs are extremely fast electric valves. They can perform (but not easily) within ps (picoseconds = 10 power -12 seconds) - But they show up some drawbacks we need to take in account.
4. For better understanding let's recall the name of a FET (exactly a N-FET). This is the type we usually use. The leg being connected to electrical minus or GND or ground is called the "source". The leg where you connect your load is called "drain". Where you control the FET is the "gate".
5. The abbreviation FET stands for Field Effect Transistor. This term tells you that you can change the state of ON /Off by controlling an electric field within the structure of the FE-Transistor. see additionally This might give you the notion that a fet will not draw current but the field will be suffitient. This notion is true and false at same time - sorry - given at what time you look at your FET.
6. Gate capacitance: There is no FET (or transistor) without it (app. 1nF = 1 nanaofarad). As you possibly know a capacitance is a bin for electrons where you can put them in and extract them later on (in reality it is not - but let's take it as thinking model). We charge a capacitor and discharge it. In this respect it behaves like a rechargable battery. If you have lots of current you can charge it within short time and if you have a weak power supply you need to wait long time in order to use the charged object.
Now please understand that you can have no natural feeling of what is slow or fast for a FET and what currents will flow. All this matter defies your daily experience and therefore we need to talk about it.
7. Charging a capacitance is no linear job but the more charge you have gathered in the cap the slower it will increase its voltage. So please understand that it will be no good idee to supply your FET driver with 12V while your FET needs 10V for full ON state. Additionylly your driver will eat up some voltage and supply somewaht less than your battery supplies.
Ans beware of long thin wires -they will kill the rest of your switching spe.
For discharge you unfortunately have no negaive voltage inorder to speed up the cycle. Then teh low driver stage needs to be strong enough.
8. Oscillations are another enemy of your FETs. As you learned above every transistion generates losses and you can imagine that some additional oscillations (wires are inductance, capacitance and act along FET capacitance) will add losses and eat up performace of your PWM circuit. These oscillations may go up to MHz! But there is a drug for this - an additional resistor (10....30 Ohm) - look forward to schematic coming soon.
9. Now let's recall some usual nominal propeties of a FET.
Threshold voltage for ON state : higher than ca. 10V
Threshold voltge for OFF state: lower than ca. 4V
This tells you that you need to travel as fast as possible through the lossy zone between 4V and 10V and vice versa and additionally exceed the thresholds by some volts in order to stay in a secure zone.
The bad news is that you do not have a certain amount of loss once only but at every transiton ON/OFF and OFF/ON. The frerquency of 10 KHz tells you i.e. that a FET will experience 20000 times pulses of heat every second because of switching only. Imagine this facts like driving your car without oil in the engine /gear -> friction + heat + damage.
I do not want to derange you with math. If you want to know more see this calculator.
Anyway you can understand that if we have a weak, slow current source as driver and possibly no good conductors it will take longer time to switch a FET ON/OFF.
As you own no oscilloscope it is of no value for you to enter into calculations and figures.
Let's focus on what we can do in order to enhance your FET driving.
10. There are some other facts to be considerd but stay with this knowledge for now.
ToDo:
A: Professional drivers are able to drive up to some amps for very, very short time in order to speed up the switching time. This tells you that you need thick wires from driver to your FET (GND included) -> AWG 18 or 0.8 square mm will be a premium choice . Keep them as short as possible.
B: Any FET can switch a certain amount of current depending on its internal construction. It is a good idea to utilize more than one in parallel in order to get higher currents switched. But please note that you multiply the gate capacitance and additionally make the wires to the Gate / Source of the FET longer. You will shurely can handle it if - yes - if you owned a scope but you unfortunately do not. So take the secure way below.
C: What about dedicating a FET driver i.e. a LM555 / NE555 do any FET you decide to use and in close proxity. You will get short wires, you will have a standard load (1x FET) dedicated.
Build modules and test them separately. Assemble a plurality of them to perform as monster driver.
Thus you can drive with your PWM generator (using 555 if you decided to) multiple 555 inputs and the good news is that the 555 components do not need high current at input. The only requirement is to have wires of equal length between PWM output and 555 input in order to get the FETs switched at the very same time (wires perform as time delay as well).
D: Get hold of defective PC PSUs. There you have a case, a fan, heatsinks, thick felxible wires and possibly FETs being ready for use mounted on heatsink.
------
And please understand that UFO gave you a good schematic based on a PCB (see thread mentioned above) But the schematic itself is no garantee for functioning. We talk here of dynamic switching and therefore your wiring is an ESSENTIAL part of the actual circuit (wires are cap, resistor, coil at same time!). If you follow my hints you will fight successfully against some major drawbacks you experienced before. You will get the optimum out of your funds being restricted.
OK guys! Now you have some stuff to ponder on. Digest the idea and the chances you have. You will get a simple schematic soon along some test procedures.
JohnS
Experts spend hours a day in order to question their doing while others stopped thinking feeling they were professionals.
johnStone:
Tanks for your schematic. What is the schematic of your pulse generator? Have you got a pic of it? There are some promizing measures you can take in order to optimze it without applying a scope. They will apply to your and all other replicas of any pulse generator.
Member:
For the pulse generator I have used this
DIY Homemade Power Pulse Controller - RMCybernetics
I have a few different caps for C1 which I can switch from one to another, and have used the NTE2397 mosfets which Ufo recommended.
10-15-2012
JohnStone:
Yes this circuit is quite good given you do not connect several FETs and given you do not drive above 1 kHZ. Above those limits you might run into issues. A pullup resistor of 2.2K at gate side is not a very impressing driver capability at all.
You can handle all issues if you have meters and scopes in order to detect weak parts.
Those who have no swcope may be lucky and succeed sometimes and sometimes not. It is partly like a blind flight. Therefore I suggest some moderate overkill in electronic parts in order to get a reliable replication for ALL. Take this as emergency line.
And yes an opto might be fine but first of all we need to have a rugged power switcher being tested under ohmic load and performing well. My suggestion for building modular power blocks includes opto as well - if requried and later on of course.
@Netica: You mentioned to have connected several FETs in parallel - right? So you might run into oscillations at gate side because of additional wires. We will cover this issue as well.
NTE2397: You need to transit readily between 4V and 10V. 10V is the minimum you need in order to switch 10A. Unfortunately RDSON is 0.5Ohm -> @10A -> 5Watt loss!!!!! 10A is absolute maximum for 25 cent degeree. For 100 cent degree -> 6.5A
This is a very nice recommendation for a fast FET. But You are not encouraged to turure it outside it's specs.
Please consider that absolut maximum values are NOT valid all at same time!
Oscillator: If you operate it with 12V you might not get your FET in a cool area within short time. You remember the capacitance inside the FET. The charging curve is not linear, it flattens at a voltage higher than 66% considerably. It is strongly recommended to operate your oscillator at 15V!!!! - in oder to get the flattened curve above 10V
Please digest my hints because soon we will need pulsed current at our motors. It is essential to not start rebuilding your circuits then - especially those being red necks in electronics. I want to help you all in order to evade from this uncomfortable status.
JohnS
Experts spend hours a day in order to question their doing while others stopped thinking feeling they were professionals.
10-15-2012
Hi Friends,
today I dismatled a defective PC PSU and built a PWM driver for you by adding a simple NE555. The document is preliminary and there are some smaller issues left. The results are very promizing. see
I hope I could demonstrate some facts I posted another day.
Hints and corrections welcome.
PLEASE BUILD YOUR PWM CIRCUITS VERY COMPACT AND WITH THICK WIRING.
I'd like to continue with my motor measurements and therefore I will continue the document mentioned above on request only.
rgds John
Experts spend hours a day in order to question their doing while others stopped thinking feeling they were professionals.
10-18-2012
JohnStone:
There are many options to make sutch circuits. You are lucky to have succeeded. Do not change it as long it works fine.
Others do not succeed and do not know why. They are stuck. Those unlucky persons are my concern.
My recommendation to those guys:
Build the oscillator cirtcuit mentioned above - or use one you have built.
TEST: Increase the timing capacitor up to the value you can check - visually - the slow oscillation with a LED connected
Build the driver FET cluster I suggested
TEST: it statically with DC and resistive load for switching properly. Use meter and a LED/resitor in parallel to the FET. In case of FET being active switched on the LED shall not glow at all in dark environment (red LED will glow above 1.6V)
Connect both building blocks and check the function with low frequency
Replace big capacitor timing mentioned above with teh origianl one
TEST: Recheck with resistor load for heat
You may check addtitional driver stages in parallel
Now you are ready for kicking your coil or motor
Those two building blocks can be tested and rechecked separately and combining tehm does not generate additionl implications beacause of wire length and generator drive capability. I will add the procedure above to my document sometimes.
The principle behind is to go step by step in order to not be faced with all enemies at same time. Romans had a saying: Divide et impera = divide and rule. This is not true in politics only but in technical matters as well.
Experts spend hours a day in order to question their doing while others stopped thinking feeling they were professionals.
10-19-2012
JohnStone:
....
2.
I do not prefer any caps - I am friend with all of them - but there are different makes dedicated to different performance. Wound electrolitics do not accept higher frequencies. (if pulsed they may exhibit superiour cold / hot conversion (who knows). Ceramic ones absorb frequencies eagerly. Sometimes it is advisable to cluster different makes in order to perform as team. i.E NE555 needs 100nF ceramic and 10µF electrolitic in order to perform correct. Some caps are lossy and others not (i.e. mica). Losses are important at high power flow like motors and high frequencies.
3.
Please understand electronics like our world of colours. If you focus on a green pasture you see green only. But if you take a photograph you will observe a blue shift from the blue sky.
So at any dumb circuit ALL frequencies are potentially present but very often you can ignore them. As soon you have components which do not - you need to deal with that sensitive area.
Remember Tesla's earthquake. He operated the buildings and their ground outside their intended performance. The wet sand below NY started moving......
i.e. a NE555 timer perfoms up to 500kHz. Some single transistors inside the IC might do some strange things up to 5 MHz if stirred by some spurious noise - therefore proper blocking of power input.
Same at FETs - performing inside much faster than the datasheet tells you - but they will peform not like you want them to. Therefore good push / pull driving very near to their gate. You put them on a short dog leash. Don't be surprised your dog jumping with a long leash.
4.
Please accept that all frequency generators discussed here and corresponding electronic components do not apply to ns pulses (GHz range) It is like comparing a slug with a jet plane. Look at cheap older frequency generators. Most of them end their range at 2 MHz sine. Now you know why. They were bild with components we discuss here.
Of course you can buy at eBay DDS generators up to 30 mHZ - but that corresponds to 330ns - and sine not pulse.
This drwback is of no major importance. Fortunately we do not talk of a single effect being elicited in excess but of a bunch of helpful effects, vibrating in concert.
5.
UFOs teachings relate to a viable research being tangible. Of course there are lots of other effects to be researched. But please do not perform like that guy who tried to save money in order to own a million bucks. Ten years later he sadly looked at his savings and decided to start the second million trial .......
6.
Compare OU mattes to the strange effects of giros see . A rotating mass exhibits less inertia (La Palma / US). Rotating masses defy most of Newtonian physics because Newton relates to a very special effect of masses being NOT in rotation.
ALL of you shall watch those giro lectures in order to get a small glance on how narrow minded our standard science is. And you will get a glance of the opportunities we have in OU matters. They need to be researched and not shortcut.
JohnS
Experts spend hours a day in order to question their doing while others stopped thinking feeling they were professionals.
10-20-2012
JohnStone:
1. Caps: Caramic caps are readily available up to 100nF .... 1µF. Electrolitics start usually at .1 µF. Somewhere inbetween we get foil caps. At my generator C1, C3 is of ceramic type, C5 is a foil type (usual technology make is MKT), C6 is an electrolitic one. You can see the types were adjusted to the coresponding frequency they generate along the current source.
2. PNP Q1 (blockA): Any PNP - not being a power type - should do well. Max. current is of no importance here because it will switch some mA only.
3. Please note that I intentionally suggested two different ICs for generator section and pulse section (U1 / U2) (block C/D). It is possible to mix the circuit in one quad generator but that will make ypur life not easyer at all bacause of additional implications.
4. Driver stage NPN / PNP (BlockE) This is just an example how you can do it. Meanwhile I prefer a NE555 there because you get an output with known outout power and speed and debugging is quite easy. If using transistors (everybody uses others) power and speed is not predictable. NE555 has known parameters.
5. You know that we intend to build serious power PWMs. At that time when I proposed the circuit we talked of just pulsing a coil. In this case it was viable to build the generator and the PWM driver as one unit.
But now we prepare our circuit for additional tasks and I strongly recommend to do it separately. Along that I want you all newbies in electronic art to succeed as well!!!!
Imagine the starter motor in your car. It takes more than 100A in order to give your motor some revolutions. Do you imagine that this current flows through your ignition switch? Not at all - it would evaporate. The manufacturers use a serious relay being attached to the starter motor in order to get as much battery power as possible to the armature of the starter.
With this pic in mind you should build your gen / driver blocks.
A: Build any generator you prefer.
B: Add a NE555 at output (even if the generator itself is a NE555) in order to securely decouple all spurious noise from your generator - you do not want it to make uncoordinated jumps with power stage attatched - killing FETs. (This is similar to your ignition switch mentioned above). Now you have a rugged circuit, modular, easy to repair and debug.
C: Build driver stages - modular - conforming my recent suggestion. (This is similar to the heavy duty relay attatched to your starter motor).
The measures suggested above decouples all three functional units. The NE555 attached to FETs performs with known steep edges at output independently from your signal quality at input. These building instructions are the simplest working ones I can imagine. If you have additional questions please ask. And no I have no circuit schematic containing the mix suggested above.
My suggstion is to move for further generator / driver discussions to UFO's first thread in order to keep this one dedicated to motors. Please PM if you start posting there
rgds JohnS
7: Regarding 555: The standard type is NMOS technology (NE555) and can source / sink at output 200mA.
The CMOS types TLC555, LMC555, TS555 are faster but can source 8...10mA only and sink 50 ...100mA depending on manufacturer. Advantage: fast!
So you should own both types in order to test out what model performs best in a certain application.
8: Regarding µs pulse. Interesting! TB seems to have gained additional knowledge since earlier times. But please note that it is no problem to build a generator with µs output. Any digital circuit can do it easily. The vintage standard CMOS series 4000 performs up to 3 MHz. There is a very different story to convert this digital pulse in 10 Amp pulses.
@bobfrench: Correct: a NE555 (NMOS) will not perform pulses of 1µs. TS555 yes (3...6 MHz depending on power voltage) But please stand by and build our current circuits. You will have lots of fun and learning. We possibly will change the technology if we go further.
If a 555 performs at its max frequency it will not be possible to additionally still change the duty cycle to less than 50%. Any edge of a signal is in fact a ramp being more or less steep. At absolute max. frequency the positive ramp will join the negative ramp (high and low as well) so you get a triangle signal only.
I have no easy instruction now but fortunately we do not need it at current experiements. Ufo will tell you if and when we are going to enter this advanced matter.
Experts spend hours a day in order to question their doing while others stopped thinking feeling they were professionals.
Comment