A green and clean environment is a human right!
Projects by members => Projects by members => Donaldwfc => Topic started by: Donaldwfc on August 22, 2010, 03:12:35 am
-
I need help with a pulsing circuit, that takes pulses from an Opto-coupler, and pulses a Power Transistor or MOSFET.
I want to switch on/off the 12 volt power supply, up to 5 amps.
I don't have much experience with this stuff, and I can't quite figure it out, I graciously ask for help from anyone who has more experience!
I have tried 6 circuits already, and the one I am using now is my latest, but it still isn't working properly, it only puts about 2.5 volts across my primary, and the mosfet/heatsink gets hot while running around 1 amp.
I have lots of scavenged components from computer power supplies and stuff like that to work with.
Please help :)
-
I need help with a pulsing circuit, that takes pulses from an Opto-coupler, and pulses a Power Transistor or MOSFET.
I want to switch on/off the 12 volt power supply, up to 5 amps.
I don't have much experience with this stuff, and I can't quite figure it out, I graciously ask for help from anyone who has more experience!
I have tried 6 circuits already, and the one I am using now is my latest, but it still isn't working properly, it only puts about 2.5 volts across my primary, and the mosfet/heatsink gets hot while running around 1 amp.
I have lots of scavenged components from computer power supplies and stuff like that to work with.
Please help :)
Do you have the technical specs of that optocoupler?
-
(http://jnaudin.free.fr/steorn/images/orbodrivers41.gif)
This circuit has worked fine for me. I don't know what you want to pulse (inductive loads ? resistive .. ?) but you can replace the 4 inductors with your load and use another mosfet if the current you need is higher than rated.
The optocoupler i use i the CNY 70, the resistor values were changed i think, depending on the opto coupler of course.
I have tried 6 circuits already, and the one I am using now is my latest, but it still isn't working properly, it only puts about 2.5 volts across my primary, and the mosfet/heatsink gets hot while running around 1 amp.
What's the pulsing frequency? If you have a transformer and the frequency is too low most of the voltage will be at the mosfet (depending on how high the RDSon of the mosfet is), therefore heating up. Can you give more data on mosfet and transformer?
If there is current going through the transformer the switching part seems to work just fine.
-
I have a PC817 Opto, and about 6 others like it that came from similar places, but I cant read the number or find data sheets for them. Data Sheet is attached.
The Mosfet I am currently using is a 2SK1400A.
And MUR460 Super Fast Diodes.
It's to pulse a VIC, so inductive load. Pulsing Frequency 1 Hz to 20,000 Hz
I am going to take a look for some other transistors to use, I tried using Power Transistors but I couldn't get them to work properly... figuring out the proper resistors and current to them to get enough current through the load...
-
I just pulled a 40 amp MOSFET off of an old computer power supply, I think this would be good to use since it is over-rated, here is the Data Sheet, comments?
I also found a 12 Amp MOSFET
-
In which range is the frequency?
Also how do you measure the voltage?
Note that my circuit is inverting if used in that way.
I'd prefer the 12A mosfet, as the Uds is not as high (6V is enough).
-
I'd like to be able to use it in the range of a few hundred hz to 20 khz
i have 12 volts from the power supply
2.4 - 2.8 volts going into the primary
5-6 volts at the cell
also measured at the output of the power supply it was drawing between 2 and 3 amps
the heat sink and transformer started getting hot
also i should mention, i have a diode before the primary and a diode after the primary, so this should take up 2 volts, but i am thinking of removing these (i still have another diode backwards across the primary as required)
This is the step up I am currently using, but i will be making another one sooner or later
(http://img.photobucket.com/albums/v81/bigbuba/Picture33-1.png)
-
What do you think of this? ... With the 12 amp MOSFET
Do I need the capacitor in there?
(http://img.photobucket.com/albums/v81/bigbuba/Circuit1.jpg)
-
when the pulse comes the coil is disconnected, is that what you want? then its ok
-
when a pulse-on comes from the microcontroller, i want this to pulse-on the power supply
would this circuit work backwards?
-
use another small npn transistor to invert the optocoupler output. then it should be fine.
at south end of the 10k and gate of the mosfet the collector, the base at the collector of the opto-transistor and the emitter to ground.
you'll need another resistor from 12V to the opto-transistors collector, the value depending on your npn-transistors base-current.
-
The way i imagine the circuit working, an on pulse from the opto gives and on pulse to the primary, can you describe how it's working so i can see what i'm missing?
i am thinking when the opto-led turns on, it completes the connection to the ground on the other side, allowing voltage to go to the mosfet and turn it on, allowing current to flow through the primary.
i'm confused about how it is inverted
-
The way i imagine the circuit working, an on pulse from the opto gives and on pulse to the primary, can you describe how it's working so i can see what i'm missing?
i am thinking when the opto-led turns on, it completes the connection to the ground on the other side, allowing voltage to go to the mosfet and turn it on, allowing current to flow through the primary.
i'm confused about how it is inverted
The opto turns on and the connection of the mosfet-gate is at ground, so voltage = 0V. if the led is off the gate is at high voltage (12V), because the 10k resistor is very small compared to the open circuit the opto-transistor has.
so if the pulse is on the gate voltage is 0V and the current on the transformer is 0A.
you need a second transistor like i described.
-
Ok thanks, I see it now
This is what you described,
(http://img.photobucket.com/albums/v81/bigbuba/Circuit2.png)
I'd rather not add in another transistor if I can get away with it, what do you think of either of the next two, will they work? (10k still make sense?)
(http://img.photobucket.com/albums/v81/bigbuba/Circuit4.png)
this one is similar to what Stan has in one of his patents
(http://img.photobucket.com/albums/v81/bigbuba/Circuit3.png)
-
My advice to you is that you use a mosfet , like the irfp250 . Transistors IMO are inferior for switching purposes .Less rise and fall time , less amps , less durable .
To drive mosfets I like the totem pole method , this is the method used in mosfet drivers .
resistors make slow rise time , remember that you wanna charge that 2000 pf as fast as you can , so resistors is bad , that circuit fom Naudin is not so good , slow rise and fast fall if a big mosfet were to be used . Remember that you need to discharge also , some of those circuits are just non-functionnal and wont discharge . Also you need short distance between driver and mosfet because high rise + inductance + capacitance = ringing . Minimize the inductance as best you can , big wire 1/2 inch long to gate .
My advice is that you use a mosfet driver or study totem pole if you wanna make with available components . That way you can transfer the high impedance source of the opto to a high imopedance input of a mosfet driver or totem pole pair driver . That second pic right there that you claim is from stan is non-functionnal , n channel fets need 10 volts Vgs to be saturated , you would need to drive it with 22+ volts
Thats the method I always use , lazy and small . I use mosfet drivers personally because it uses less board space , max627 for example .
-
The circuits shown are good you just need to add a capacitor between the drain and source, about 200nf works fine for irfp250 transistor (this allow the transistor to completely stop conducting the current during the of pulse and thus make it work cooler... use a mosfet driver is the best...
There are many ways you can do it... Can you tel us what you need this driver for?
In my projects page you find a good schematic of a pll circuit you can use it...
Best regards
-
Also, Take out those diodes and add only a capacitor in parallel with the primary too about 100nf is well ,,,
-
sebosfato, can you draw the circuit you are describing, i can't quite picture it
this driver it to pulse the primary of a toroid set up transformer, for a VIC
-
The circuits shown are good you just need to add a capacitor between the drain and source, about 200nf works fine for irfp250 transistor (this allow the transistor to completely stop conducting the current during the of pulse and thus make it work cooler... use a mosfet driver is the best...
There are many ways you can do it... Can you tel us what you need this driver for?
In my projects page you find a good schematic of a pll circuit you can use it...
Best regards
Nah , neither those circuits are good , nor is your answer . Sorry sebos /:
He needs the driver for driving a mosfet , the relevance of your question ?
Donald , when using a mosfet on the high side ( positive side ) , you need 10 volt Vgs , that means you need 22 volts pulses if you wanna saturate a N mosfet @ 12 volts at drain . So that circuit you claim to be in Stan's does not work ... Neither have I seen in in Stan's coloring book .
Thats why we use P channel on the high-side , these can work with totem pole and dont need much higher voltage source , but they are inverted compare to N type , On @ gate = no current draw from mosfet .
But P channels are more expensive , have less current , has more capacitance , just inferior compared to N channel .
I think this is enough for you to research and have success , it really isnt rocket science to use a mosfet .
-
dankie can you draw a circuit of something that will work good? please
-
I am truely surrounded by pathetic individuals .
And you guys wanna be engineers and inventors ? Hmmph !
-
Ok thanks, I see it now
This is what you described,
(http://img.photobucket.com/albums/v81/bigbuba/Circuit2.png)
Yes, but the base of the middle transistor should be at the collector of the opto transistor and the emitter of the opto-transistor at ground.
the second could work, but you'll have a large voltage at the transistor (Uce + voltage over resistor).
Or just use Dankies's version. I don't know if it works but with the way he presented it it should better work.
Ironically he was laughing at people using bipolar transistors and presents one himself. Good job.
-
turns out the 40 amp and 12 amp mosfet I had don't work, but if i get the circuit switching properly then that should fix the heating problem and the voltage problem, the 7 amp mosfet i have is a K1400A, i think the tech sheet is on the first page.
I've got my circuit torn apart, and a better heat sink for the mosfet, now I just need to confirm a worth while circuit. I don't have any mosfet drivers that I know of, but maybe i can scavenge some of those too.
I think i'll start with the first circuit even though it has an inverting signal, then modify the code from the micro controller to deal with it, if that doesn't suit me i'll track down an npn transistor to add to the circuit.
-
You can use a driver chip for a FET, but its not needed.
FET's are switched on by voltage and not by current.
So, anything that switches on like 5 volts on a FET will work.
Dankie is right when he says that by using a driver chip is a perfect way of doing it, but again, its not needed.
Steve
-
I think i'll start with the first circuit even though it has an inverting signal, then modify the code from the micro controller to deal with it, if that doesn't suit me i'll track down an npn transistor to add to the circuit.
Oh why didn't you say that earlier? ;D
You just have to change the HIGH and LOW values on input or output (if the optocoupler gets his signal from the mikrocontroller you can just switch the output to low when there is a pulse and to high when there is none)
-
haithar if i said that right away i wouldn't have learned as much :)
is the cap in the JLN circuit needed?
10k resistor still good?
-
yes you can use a 10k or something in that range. if you have 9k or 13k or something like this it should work.
the cap is not absolutely needed, but always good for stabilizing everything.
-
I will draw for you a good circuit... this evening...
The capacitor conected to drain and source is a must because this allow the transistor to switch of... otherwise you will aways see it warming a lot...
-
I do not know about the capacitor between d and s , perhaps somebody could find a link to more information and circuits showing this .
My circuit works , just make sure you put a 1k-800 ohm type resistor to protect from over on base of both transistors . I use cmos for everything using square logic , I simply love cmos cd4000 , unfotunately these cannot be salvaged since its too old but digikey has all of them for 25 cents each , most electronics stores also have them for 1$ each . Imo only capacitors should be salvaged , I cannot imagine myself being in a position where I look for small little components on old stuff and then unsoldering , only in a post-apocalyptic world where all silicon ship makers were destroyed .
If someody wants an easy schematic he will have to suffer humiliation , that is the rule . My comment was adressed only to Donald , even if it was in the plural form .
Donald , you should stop being distracted by these trivial activities and return to your studies , mister engineer student .
-
Awesome thread donald! i am lost in this area too and this topic is shining some light.... i would throw some insight in too but most likely it would be off subject lol.
-
The thing about ordering components is I don't know exactly what I'm looking for, and then there is shipping time and the cost of the chips... if I can get something going without spending much, then I can order new components and make a fresh circuit once I know exactly what I need. I have no training in this type of thing at all, so sometimes I miss the obvious solutions when searching for what I 'think' might be a reasonable approach...
Thanks for your help everyone, I should have something together tonight hopefully.
-
wow I can't follow that, try neater
-
is this what you meant seb? i have question marks in areas of question
(http://i82.photobucket.com/albums/j243/outlawstc/101_1016.jpg)
-
Dam that schematic is wak as hell .
What are those useless diode . What is that wasted 13mA half of the time . What is that 20 ohm resistor , it only slows down the charge and leads to more heating . And you dont even let me know what pin 7-6 do there unless I look for myself .
Also ,a heatsink does not have to be used unless you are doing a bad switch .
-
Hello
Sorry i got very little time today to do the drawing but i tried to make the best i could... Tomorow i make it more clear to you...
Yes outlawstc
The left question mark is connected to the positive side (is an electrolytic capacitor)
and yes the unmarked pins were 6 and 7
The diode on the left is good to avoid problems with ground loops thus will protect the driver and should be about 5 amp diode if possible...
Dankie, The 20 ohm resistor is there to limit the current, and is needed as a mean of avoiding undesired resonance or oscillation, false triggering, on the mosfet witch can lead to destruction of the driver and mosfet... (it reduces the Q factor there) (20 ohms is the min recommended on the data sheet)
Outalwstc on thing you miss is the protection diode on the on off switch witch directs the spike from the coil to the battery when you turn off the circuit avoiding damage to the transistor... (the parallel capacitor with the primary is also for protection and is also a mean for make the power to be used more efficiently as the coil instead of discharging back into the battery witch will not readily recharge (the case of having an anti parallel diode in paralell with the coil) Thus the coil will discharge on the capacitor, so the energy will be there available for the next pulse ... there are other strategies thought like adding also a resistor in parallel but keep it simple... This would be called snubber is meant for protection...
(There are other thing called the earth capacitor with is just a big capacitor connected between a small coil and the battery with a diode in series before going to the primary witch serves as this way back to the unused energy... However the parallel capacitor do the job-... )
The diode on the switch is very important because it creates a path for the current to go when the circuit get open (if this diode were not there the spikes would go to the transistor and time after time will destroy it! (the represented switch is a three way switch and you just need to connect it just like i draw it! easy like this
The capacitor in parallel with the drain and source is there for the same reason, when the current change direction on the coil it reverse polarity and thus allow the transistor to completely turn of... It also Reduce the spike peak protecting the transistor during the switching operation for the
same reason as the diode on the switch...
people use to draw a diode in antiparallel, i tried practically but i found that the capacitor is way better and that it makes the transistor to work cooler...
Is very easy circuit
For better understanding i recommend that you google the component names and there will be datasheet with some more information on the devices and there will be some schematics hopefully with the explanation for what every pin stand for....
Is never late to learn something new
Hope it help
see you tomorrow
Regards
-
OHHHH
Sorry I did something wrong
the 1000ohm resistor should be connected from the gate to the ground... sorry for that one i have just noticed
-
Look
i decided to show you another way...
The blue is the direction of the current during the on pulse
the red is ... during the off pulse...
The big diode works like a switch that only opens when the voltage of the battery is greater than the voltage on the capacitor...
The diode in parallel with the mosfet allow for the current or (non used energy) go back recharging the capacitor...
The "switchs" diode protects the mosfet as explained in the other post...
The pink diodes protects the circuit in the case the polarity of the coil reverses, aways recharging the capacitor instead of blowing out the energy...
When the voltage in the capacitor is bigger than the battery voltage the big" diode stop conducting making of the capacitor the source of the energy at least until its voltage is grater than the battery voltage...
This capacitor should be big if possible and rated for maybe 50v, and should be an efficient one (not electrolytic would be better) (it will work as the earth ground witch i mentioned....
If you connect a diode in parallel with the coil, (like on stans papers) simply when the pulse is off all the energy you inputed on the coil will be shorted on the coil because the polarity will reverse and the diode will than start conducting... (if you want to try this way, use high power diodes cause will get really hot)
Best regards
-
this better?
(http://i82.photobucket.com/albums/j243/outlawstc/101_1017.jpg)
-
This is the circuit i named the components so we can more easily talk about them
You could also use irf 540 witch is also a good mosfet with low rds on (the irfp250 is better however)
C1 should be about 15uf if possible a paper capacitor (not electrolytic )
C2 could initially be around 200nf but you can tune the circuit for a certain frequency adding other capacitors in parallel, play with some values and look at the oscilloscope, it will depend on your load, inductance...
C3 from 100nf to 200nf would be just fine
For the diodes you can buy some 1n4007 and use them in parallel so you can make any diode you need... If they get hot add more diodes in parallel...
the 20 ohm resistance should be 2 watts
the 1000 ohms 0,5 watts is fine
-
Circuit from haithar is up and running, more details in a sec
first test, i have a comment about amp restriction and inductive reactance
~1000 Hz , 1.3 amps
~2000 Hz, 1.1 amps
~3000 Hz, 1.0 amps
~4000 Hz. 0.88 amps
~5000 Hz, 0.5 amps
~6000 Hz, 0.17 amps
~7000 Hz, 0.11 amps
~8000 Hz, 0.08 amps
~9000 Hz, 0.04 amps
~10,000 Hz, 0.02 amps
~11,000 Hz, 0.01 amps
drawn from power supply, the chokes are restricting amps!
inductive reactance = 2 pi f L
-
Circuit from haithar is up and running, more details in a sec
first test, i have a comment about amp restriction and inductive reactance
~1000 Hz , 1.3 amps
~2000 Hz, 1.1 amps
~3000 Hz, 1.0 amps
~4000 Hz. 0.88 amps
~5000 Hz, 0.5 amps
~6000 Hz, 0.17 amps
~7000 Hz, 0.11 amps
~8000 Hz, 0.08 amps
~9000 Hz, 0.04 amps
~10,000 Hz, 0.02 amps
~11,000 Hz, 0.01 amps
drawn from power supply, the chokes are restricting amps!
inductive reactance = 2 pi f L
Congratulations!
Did you measure the volts on the cell at 11khz?
Steve