You should need to adjust the second 2 stages of a double pulse test. They are just there to study what the switch looks like under load as long as they are long enough to get a complete transition and they are not to long you are fine. If the second is to long I mean really long the current will drop. If the 3rd switch is to long you could potentially damage something. Desat should protect on the 3rd pulse if you make it to long but not everyone uses desat although they should.SjwNz said:Arlo you doing this double pulse testing is good, so when I get around to doing my tests in a couple of months I shouldn't have to bug anyone about how to do itI am looking forward to making a little Double Pulse tester, maybe a small LCD with 3 sets of Up/Dwn buttons to adjust each 3 stages On ,Off On. Fire button and maybe a Desat input just in case.
Arlo's power stage Leaf controller runs and drives page 103
- Thread starter Arlo1
- Start date
newbiebiker
100 mW
@Arlo1
you used copper for heat dissipating from MOSFET, will silver usage will make any difference ..??
whats the volt and ampere that 6 FET controller handle (Peak) , also 6 FET controller means, it contain 6 MOSFET ? each mosfet need seperate IGBT in IGBT driver circuit?
Offtopic:
i am reading your posts in my freetime, have to understood all things first, have to start output stage soon
you used copper for heat dissipating from MOSFET, will silver usage will make any difference ..??
whats the volt and ampere that 6 FET controller handle (Peak) , also 6 FET controller means, it contain 6 MOSFET ? each mosfet need seperate IGBT in IGBT driver circuit?
Offtopic:
i am reading your posts in my freetime, have to understood all things first, have to start output stage soon
You want to use the best heat dissipating material you can afford with the best medium (paste) to help make the thermal path the lowest thermal resistance possible. But remember the backs of the fets are LIVE they are connected to the drain I have used that to my advantage.newbiebiker said:@Arlo1
you used copper for heat dissipating from MOSFET, will silver usage will make any difference ..??
whats the volt and ampere that 6 FET controller handle (Peak) , also 6 FET controller means, it contain 6 MOSFET ? each mosfet need seperate IGBT in IGBT driver circuit?
Offtopic:
i am reading your posts in my freetime, have to understood all things first, have to start output stage soon
You need a multiple of 3 but starting with 6 power switches IE 6 mosfets 9 mosfets 12 mosfets for a total controller. In this controller design it is best to stick with multiples of 6 because good controllers PWM both the high and low side fets. Some of the cheep controllers would run un even number of fets because they only did PWM on 1 side. It is best to start with a 6 fet controller and build it as best as you can.
The current limit depends on the power switch design (mosfet or IGBT) and how good you are at designing a driver and controlling it.
newbiebiker
100 mW
Arlo1 said:You need a multiple of 3 but starting with 6 power switches IE 6 mosfets 9 mosfets 12 mosfets for a total controller. In this controller design it is best to stick with multiples of 6 because good controllers PWM both the high and low side fets. Some of the cheep controllers would run un even number of fets because they only did PWM on 1 side. It is best to start with a 6 fet controller and build it as best as you can.
The current limit depends on the power switch design (mosfet or IGBT) and how good you are at designing a driver and controlling it.
i am just started noob (started means, daily reading and understanding), still i am not completely understood the basics and components.
what i finalised is, 48V 2000W motor, and 1 x 16Ah 22.2V 25c (will give ~ 355.2 Wh or less) Lipo battery - will upgrade battery once it start running.
as per your advice, i am starting with 6 MOSFET controller.
Microcontroller supplied signal is very low for MOSFET, so we use IGBT drivers to amplify then give it to the MOSFET (this is what i understod),
but 6 MOSFET is connected to 3 phase wires,
but what the components in IGBT driver...??
that is we need some 6 components (in IGBT driver side) for 6 MOSFET ?
Also share your latest working design of OUTPUT, i have to studey yhe components one by one.
Offtopic:
i fixed a deadline, if i am not able to finish controller within August 2016 (next year), then i have to buy one , Ride and FUN...then i will continue my learning/understanding , whatever we call..lolz . but i have confidence
You will need an oscilloscope.newbiebiker said:Arlo1 said:You need a multiple of 3 but starting with 6 power switches IE 6 mosfets 9 mosfets 12 mosfets for a total controller. In this controller design it is best to stick with multiples of 6 because good controllers PWM both the high and low side fets. Some of the cheep controllers would run un even number of fets because they only did PWM on 1 side. It is best to start with a 6 fet controller and build it as best as you can.
The current limit depends on the power switch design (mosfet or IGBT) and how good you are at designing a driver and controlling it.
i am just started noob (started means, daily reading and understanding), still i am not completely understood the basics and components.
what i finalised is, 48V 2000W motor, and 1 x 16Ah 22.2V 25c (will give ~ 355.2 Wh or less) Lipo battery - will upgrade battery once it start running.
as per your advice, i am starting with 6 MOSFET controller.
Microcontroller supplied signal is very low for MOSFET, so we use IGBT drivers to amplify then give it to the MOSFET (this is what i understod),
but 6 MOSFET is connected to 3 phase wires,
but what the components in IGBT driver...??
that is we need some 6 components (in IGBT driver side) for 6 MOSFET ?
Also share your latest working design of OUTPUT, i have to studey yhe components one by one.
Offtopic:
i fixed a deadline, if i am not able to finish controller within August 2016 (next year), then i have to buy one , Ride and FUN...then i will continue my learning/understanding , whatever we call..lolz . but i have confidence
I am building a 6 mosfet controller and I have all the boards here. I can send you what ever you need so you can build your own Or I can build it then send you a built controller... Or I can let you use the files to build what ever you want. This was a huge learning project for me but I am getting there. I have most of the knowledge I need now I just need to work out the last couple bugs and I can fly at it. A 6 fet controller will be a walk in the park for me. If you work hard and read and try to understand as much as you can I would love to help. I have to finish paying the nice lady back for the solar equipment for my house and after that I might shop for a new oscilloscope when I do I can donate mine to you but that will be a few months at the earliest. Let me know, as it is a great journey and you seem to have the passion and for that reason you will succeed Also some info here that I'm trying to keep a little more organized. https://endless-sphere.com/forums/viewtopic.php?f=30&t=63982
newbiebiker
100 mW
i think this is the first electronic product i am going to buy for "Creating other electronic" will buy one within this month or max within next month.Arlo1 said:You will need an oscilloscope.
Also some info here that I'm trying to keep a little more organized.
also thanks for your sweet words, i am inn
basically in india (my country) DIY thing is not popular, because of our culture and things we're following for more than 2000 years, like we have to ask everything with our parents (even we going to use our money) and only if they satisfied means, they allow (but mostly they wont allow, they think these things are waste of money/time),
also our families tell us to study exactly the syllabus (only given subjects) because our education is "Mark" based (spoon feeding like, we have to write the answers without missing a word in it in the book they given), not quality based, so we just memorize and spill all in exam hall without even knowing what the formula it is,
thing like read/listen, understand, then do is something new for us (me)
buddy, for organizing, create one blog (in blogger), i just update/edit the post there about my progress, adding pictures, it help other too, and we can easily understood what we done in this last one month or 6 month.
http://ebikecontroller.blogspot.com/
For PCB boards, here we have CNC machines in private shops/colleges, just we have to supply the diagram, it automatically drill and create a conduction line/components place , so we just need to place components exactly in position and solder it (no etching thing and all).
any theory/books related to IGBT driver circuits, how you understood this things ?
Here is the code we are using so far for the double pulse tester. Its super easy with an arduino uno and i even got a couple knock off arduino uno's from dealextreme for like $8 each
Code:
/*
Button
Turns on and off a light emitting diode(LED) connected to digital
pin 13, when pressing a pushbutton attached to pin 2.
The circuit:
* LED attached from pin 13 to ground
* pushbutton attached to pin 2 from +5V
* 10K resistor attached to pin 2 from ground
* Note: on most Arduinos there is already an LED on the board
attached to pin 13.
/*
int sensorPin = A0; // select the input pin for the potentiometer
int ledPin = 13; // select the pin for the LED
int sensorValue = 0; // variable to store the value coming from the sensor
void loop() {
// read the value from the sensor:
sensorValue = analogRead(sensorPin);
// turn the ledPin on
* /
created 2005
by DojoDave <http://www.0j0.org>
modified 30 Aug 2011
by Tom Igoe
This example code is in the public domain.
http://www.arduino.cc/en/Tutorial/Button
*/
#include <util/delay.h>
// constants won't change. They're used here to
// set pin numbers:
const int buttonPin = 2; // the number of the pushbutton pin
const int led = 13; // the number of the LED pin
int POT = A0;
// variables will change:
int buttonState = 0; // variable for reading the pushbutton status
void setup() {
// initialize the LED pin as an output and pin 8
pinMode(led, OUTPUT);
pinMode(8, OUTPUT);
// initialize the pushbutton pin as an input:
pinMode(buttonPin, INPUT);
int POT = A0;
Serial.begin(9600);
}
void loop(){
int POT = analogRead(A0);
// read the state of the pushbutton value:
buttonState = digitalRead(buttonPin);
if (buttonState == LOW){
digitalWrite(led, LOW);
digitalWrite(8, LOW);
// check if the pushbutton is pressed.
// if it is, the buttonState is low:
}
if (buttonState == HIGH){
// turn LED on:
// the loop routine runs over and over again forever:
int pulselength= (POT+5);
//_delay_us(pulselength); //(pulselength); // wait for the amount of on time we need to build current
delayMicroseconds(50);
Serial.write("\n");
Serial.write("POT value: ");
Serial.write(POT);
Serial.write("\n");
Serial.write("pulselength: ");
Serial.write(pulselength);
digitalWrite(led, HIGH); // turn the LED on (HIGH is the voltage level)
digitalWrite(8, HIGH);
delayMicroseconds(pulselength);
digitalWrite(led, LOW); // turn the LED off by making the voltage LOW
digitalWrite(8, LOW);
delayMicroseconds(25); // wait for 25us
digitalWrite(led, HIGH); // turn the led and pin on
digitalWrite(8, HIGH);
delayMicroseconds(25); //wait for 35us
digitalWrite(led, LOW); // Shut off Led
digitalWrite(8, LOW);
delay(1000); // let things stay off for a long time to cool
}
}Ok I made some great progress today. I started with testing and saw some big spikes on the gate during shut off. The on switch looks ok. But the off is bad.
I still have a few tricks up my sleeve But I'm also all ears as to what I might want to do to make this better.
I Started with a spike that was really big then I tested by moving the probe closer and closer to the gate and skipping the flying leads.
Then I removed the big thick 0awh cable I run from the negative IGBT to the cap which spans across the top of the driver board and I put the thin piece of aluminum that worked somewhat as a laminated shield to go next to the Phase aluminum buss that also crosses above the driver. That helped a lot so I tried a couple shield ideas for the driver that didn't work then I decided to run a positive buss back across just to make a set of 3 laminated buss bars even though the DC + is not connecting to anything and this made a huge difference. They are just quick fixes but it proved its needed..... Even back at 400amps its looking good. Not perfect but much better so now I need to figure out if there is a certain ORDER that the buss needs to be in? IE TP - then Phase then TP +??
Right now I have TP- Then TP+ then Phase....
Anyways here is some pics and screen shots sorry some of the scope shots might confuse you
I still have a few tricks up my sleeve
But I'm also all ears as to what I might want to do to make this better.I Started with a spike that was really big then I tested by moving the probe closer and closer to the gate and skipping the flying leads.
Then I removed the big thick 0awh cable I run from the negative IGBT to the cap which spans across the top of the driver board and I put the thin piece of aluminum that worked somewhat as a laminated shield to go next to the Phase aluminum buss that also crosses above the driver. That helped a lot so I tried a couple shield ideas for the driver that didn't work then I decided to run a positive buss back across just to make a set of 3 laminated buss bars even though the DC + is not connecting to anything and this made a huge difference. They are just quick fixes but it proved its needed..... Even back at 400amps its looking good. Not perfect but much better so now I need to figure out if there is a certain ORDER that the buss needs to be in? IE TP - then Phase then TP +??
Right now I have TP- Then TP+ then Phase....
Anyways here is some pics and screen shots sorry some of the scope shots might confuse you
Attachments
Lebowski
10 MW
I mean to me it just looks like the typical ringing from having too much inductance everywhere... (see my low inductance output stage thread)
Facepalm how many times do we need to go in circles about this.???Lebowski said:
If you read my post I show how I almost made it go away with some quick work to laminate the buss bars.
I am not making 400 0r 500 small controllers to make a 300HP motor run. This is coming together before your eyes. Yes some inductance might be a problem but its what I'm working with and again the laminated setup made a HUGE DIFFERENCE its almost fixed.
You need to quit posting about your low inductance output stage. I have read it many times and what I'm working on is a different animal Nissan didn't need to make 400 small power stages to make it work for them....
PS it only rings turning off... But its almost fixed with the laminations and maybe I can make them better or bigger or put them in a different order...
liveforphysics
100 TW
Arlo1 said:
When you drop back from 400A, is there a point it stops ringing, and then some defined current threshold that you see ringing on turn-off begin?
Lebowski
10 MW
Arlo1 said:
How much power is Nissan pushing ?
And I'm not telling you to build 400 small stages, you can build an output also with big components but at the same time still use the low inductance guidelines / topology.
Yes and I think that is around 50amps.liveforphysics said:
and once I get to ~100 amps it stops getting worse it looks just like this from ~100-400 amps!
I am happy I made big progress but it needs to be better and I will get it the double pulse tester has saved me a lot of time be letting me see this issue right away.
Thanks for the Tip. What do you think about a TVS or schottky diode? I Think it should be fun to sort out.HighHopes said:
But just to point out this is a comparison of where I started to how close I got it in just 1 day.
It was spiking almost 20v at the gate (5v over the fully charged gate) and now its almost gone. And this is just 1 after noon of working on it
Attachments
I went in this morning before work. And found some weird stuff the ground side of the cap that's parallel to the gate shows ~160v AC at a perfect 60hz and its a clean sine wave... This is without pushing the double pulse trigger button. So I tested it with a multimeter and it said its connected to ground (common which is phase) but it might be a bad connection or I made an antenna... I will test more asap. But this makes me think the cap to help smooth things out by being parallel between the gate and emmiter is not functioning correctly and most likely my cause.
HighHopes
10 kW
- Joined
- Mar 28, 2013
- Messages
- 930
yellow is your gate curve?
depends where u thinking to put it. for high power drive like yours, putting a bi-directional TVS gate/emitter is not a bad idea. it would be around 17.5V and 1.5kJ or so. if you think TVS with the ringing, no it should not.
YesHighHopes said:
And I made it better even today. I'm almost there.
I added a jumper wire from negative of the cap that's in parallel to the gate and run it to the emitter. Then I made a few small changes including calibrating the scope I have to remember and here we are the first screen shot is at the same settings as the last tests. Then the second screen shot is from a test with the POT cranked and that gives 400 amps
I'm so close but so tired I need to sleep.Attachments
HighHopes
10 kW
- Joined
- Mar 28, 2013
- Messages
- 930
U still have that inductor chip in series with gate?
This looks like a resonance issue but if so I'm not sure why it appears as related to drain source current threshold.
Experiment with changing gate cap and if u still gave that inductor. Adding resistance can also dampen.
I see same thing if u use other probe same spot?
Yellow trace using diff probe?
Take both probe leads and attach to same Igbt gate. U should read zero volts with some very small 1v or less transient as ground moves but no ringing. Do u?
This looks like a resonance issue but if so I'm not sure why it appears as related to drain source current threshold.
Experiment with changing gate cap and if u still gave that inductor. Adding resistance can also dampen.
I see same thing if u use other probe same spot?
Yellow trace using diff probe?
Take both probe leads and attach to same Igbt gate. U should read zero volts with some very small 1v or less transient as ground moves but no ringing. Do u?
No inductor on the gate... I can try one again if I need but It picked up noise last time I tried it. I have tried 5 probes and 1 of them was the differential probe they all read the same. It seems once the phase starts to transition it starts to ring. I will try hooking 2 probes up to the gate and double check that. I have gone from 3 ohms off to 6 ohms off and it didn't seem to help. Once sorted I will try 3 ohms again. I will look at the cap and maybe try adding a little more cap to the gate.HighHopes said:
HighHopes
10 kW
- Joined
- Mar 28, 2013
- Messages
- 930
just so i'm clear, what i'm talking about is hooking up ONE probe, but both leads physically connected to same point, the gate pin. since they both connected to same point they will read 0 volts differentially.. but .. common mode noise should register. its not all that accurate but it can give u a sense of how much common mode noise is evident. what you're looking for is a signature, does it match that yellow ringing trace?
resonance is sort of anticipated when you have a "pie" shapped circuit. like capacitor to ground, passive component, capacitor to ground. the second capacitor is inside the IGBT, its the gate cap. so you need to find another capacitor outside that IGBT creating the pie network which is an architecture prone to resonance. where is this first cap... it could be the external gate/emitter cap that zombiess put to slow down rise time (i don't recall if you have it or not, in my own circuit i describe it as a "do not populate" part cause i want the pads just incase i need an external cap but i prefer not to install .. due to fear of creating a resonant circuit). if you do not have this external gate/emitter cap.. then the next closest cap is your boost transistor caps. but i doubt this boost transistor cap is creating the "pie" network because between this cap and IGBT internal cap are your ON/OFF resistors which should dampen any resonance.
Thanks. I did another 3 hours of testing tonight and tried removing the gate-emitter cap and it got worse. Then I added 100nf at a time until I got to 600nf. I think 200-300nf is optimal. Yes I know it Taxes the gate driver circuit. I have a couple more things to test tomorrow then up the voltage to see if its different. The IGBT I'm testing is a temporary replacement its 600 amp 1400 volt rated. So running it at 205v is not what it was made for. I have found once you up the voltage the diode in the opposite side of the phase will work better getting rid of some of the ring on turn off and that might be all I need to get it just right.HighHopes said:
Lebowski
10 MW
be carefult with the high gate capacitance, 600nF at 15V, 5 kHz will lead to almost 1W of dissipation in the gate resistors, so make sure they can handle the heat.
Similar threads
