Alimentator PWM analogic

mpursu

Well-Known Member
Trenulist
25 Aug 2007
3.371
51
Oradea
railcrisana.wordpress.com
WEBSITE
https://railcrisana.wordpress.com/
LOCATION
Oradea
YOUTUBE
mpursu
Thank you for the clarification 8)
As I was saying previously, for now we forget about the transistors. Since R2 is disconnected from pin11, the transistors cannot get any signal from CD4093, so now the voltage across OUT+ and OUT- cannot be influenced by the potentiometer, thus it is not important.
So the LED light gets brighter or dimmer according to the rotation of the potentiometer, which is good. Does it behave like the LED's of the locomotive that I showed in the video? If YES, then we can be sure that CD4093 works properly and we can concentrate on the transistors.
 

PrinzEugen

Active Member
Trenulist
21 Ian 2012
64
0
41
Belgrade
LOCATION
Belgrade
yea...led gets diimer or brighter between 11 and gnd.....thats clarified....now transistor is issue

-------------------------------------------------------------------------------------
i have one question...i know im boring but this is mistery for me....

how to stop flickering back leds...forward leds lets say goes yellow but i have also yellow at back, all led lights up depends of direction ...but they flicker a bit when trains move...this is from motor induction...ems i think...

do i have to use a pair of flyback diode for both direction ?
 

mpursu

Well-Known Member
Trenulist
25 Aug 2007
3.371
51
Oradea
railcrisana.wordpress.com
WEBSITE
https://railcrisana.wordpress.com/
LOCATION
Oradea
YOUTUBE
mpursu
The flickering of the back LEDs cannot be avoided, because of the reverse voltage spikes issued by the motor when the locomotive loses contact with the rails for very short moments. These spikes cannot light filament bulbs because of their thermal inertia, but they can light LEDs because they react instantaneously to voltage pulses. This happens only in analog locomotives fitted with LEDs, because the LED groups are connected directly to the wheel contacts, in parallel with the motor itself, so they receive any voltage more or less spike-free within the locomotive. This phenomenon does not occur in digital locomotives, because the digital decoder separates completely the lights from the motor.
 

mpursu

Well-Known Member
Trenulist
25 Aug 2007
3.371
51
Oradea
railcrisana.wordpress.com
WEBSITE
https://railcrisana.wordpress.com/
LOCATION
Oradea
YOUTUBE
mpursu
Now let's check the power section. First of all, remove completely R3 and BC547, reconnect R2 to pin11 (with or without the previously mentioned LED with 1k in series) and connect the BD651 emitter to GND. For now, the PWM device is no longer protected against overload! Rotate the potentiometer to "min", connect the rails to OUT+ and OUT- and place a locomotive with LED lights on the rails. Carefully, connect the 17V power supply.
If the locomotive runs away at full speed with lights brigthly lit no matter how you rotate the potentiometer, then BD651 is internally shortcircuited and must be replaced.
If the locomotive stands still with the lights off no matter how you rotate the potentiometer, then BD651 is internally interrupted and also must be replaced.
If the locomotive stands still with the lights on, then everything is OK, we are on the right track 8), which means the amplified narrow PWM signal reaches the locomotive. Now turn the potentiometer slowly away from "min" position. You will see the lights getting stronger and from a certain point the locomotive will start to move away with increasing speed as you continue to turn the potentiometer towards "max". This means that now both signal section and power section work properly together :aplauze:
 

PrinzEugen

Active Member
Trenulist
21 Ian 2012
64
0
41
Belgrade
LOCATION
Belgrade
im planning to buy new bd651...in case that old one dont work...i cant find on stock BD651..can i use TIP120?

i read that tip120 have protection diode inself and in examination of spec. i found out that tip120 have less Ib base current 0.1A than Bd651 0.3A..


or just use BD649 NPN just slight lower voltage of VCBO and VCEO
 

PrinzEugen

Active Member
Trenulist
21 Ian 2012
64
0
41
Belgrade
LOCATION
Belgrade
ok...here are the results....

ive first changed BD652 with new BD649 with original scheme (no led diode), nothing. ...engine dead...on output 14,4 at max...about 7volts at min.


then ive run mpursu test...

removed r3 and bc457 and return back old BD651 and connected his emmiter to gnd...short circuit...sparks on input...ive checked its about 30ohms of resistance on iput as emmiter connected to gnd without overload....??

same results as BD659 on mpursu test...
 

mpursu

Well-Known Member
Trenulist
25 Aug 2007
3.371
51
Oradea
railcrisana.wordpress.com
WEBSITE
https://railcrisana.wordpress.com/
LOCATION
Oradea
YOUTUBE
mpursu
Shortcircuit?!? Sparks?!? What on Earth is happening there?!? :confused: :confused: :confused: :confused: :confused: :confused:
Come on, it's just an output section with 1(one) Darlington power transistor, with the base connected via R2 to pin 11 (output of PWM oscillator), the emitter connected to GND and the collector connected via one motor to VDD! What can there be wrong?
Of course, the overload test is NOT to be carried out without R3 and BC547, which I most certainly hope that was obvious and you did not even try.
 

PrinzEugen

Active Member
Trenulist
21 Ian 2012
64
0
41
Belgrade
LOCATION
Belgrade
man i wish i know.....i have small hammer behind me. ill save pot and rest into pieces...no use...
 

mpursu

Well-Known Member
Trenulist
25 Aug 2007
3.371
51
Oradea
railcrisana.wordpress.com
WEBSITE
https://railcrisana.wordpress.com/
LOCATION
Oradea
YOUTUBE
mpursu
Take off from the board BD651, BC547, R3, D3 and all the wires between them. Clean thoroughly the excess solder from all the copper pads involved by these components and check for any shortcircuits between the cleaned pads, because they may be caused by small solder droplets that remain in the gaps and touch two adjacent pads.
Check all of the above components of the power section.
Then... put back on the board only BD651, with the base connected to R2 (which goes to pin11 of CD4093), the emitter connected to GND (pins 5-6-7-8-9) and the collector soldered on a pad without any other connections. Connect a 1kohm resistor to the collector of BD651 and a LED with the cathode to the resistor and the anode to VDD.
Now, if you connect the 17V power supply at the inputs on the upper left side of the board (by the way, sign "+12VDC" alias VDD is wrong, please change it to "+17VDC"), there are three alternatives:
1) the LED does not shine at all, no matter of pot position, which means that BD651 is internally interrupted;
2) the LED shines brightly, no matter of pot position, which means that BD651 is internaly shortcircuited;
3) the LED shines brighter or dimmer according to the position of the potentiometer, as you saw in my video.
Cases 1 and 2 mean that BD651 is defective and must be replaced, while case 3 means that BD651 is good and works properly.

This time you simply must get it right. You just can't miss it!
 

oilen

Active Member
Trenulist
7 Dec 2011
74
0
Bucuresti
LOCATION
Bucuresti
I think you can safely use Tip 120 or derivatives as long as you remain in the safe zone of Vceo and Vceo. Personally i tried with TIP 122 and the device performs Ok.
 

PrinzEugen

Active Member
Trenulist
21 Ian 2012
64
0
41
Belgrade
LOCATION
Belgrade
here are first results....before putting bd651 to modification scheme...ive removed r3,d3, b547..ive checked all connections on board...cross conecctions, solder joints....everything is ok....but still short circuit on input on continuous test and is about 30ohms....then ive took out l7809cv...after that of course ive lost short circuit on input....


ive cross checked pins on voltage regulator l7809cv they are all short cutted about 30ohms beetween pins in -out pins, gnd -in pins, and 4ohms gnd-out???



maybe this component is one to blame for faulty work?
 

mpursu

Well-Known Member
Trenulist
25 Aug 2007
3.371
51
Oradea
railcrisana.wordpress.com
WEBSITE
https://railcrisana.wordpress.com/
LOCATION
Oradea
YOUTUBE
mpursu
7809 is certainly damaged. Remove it of the board, but let's NOT replace it with a new one yet, because now it's time to check once and for all the functionality of BD651. Without 7809, CD4093 has lost its 9V stabilized power supply, so it won't work even if the 17V power supply is connected.
Let's leave in place the 1kohm resistor and the LED, connected between the collector of BD651 and VDD. Take a crocodile clips cable, disconnect (once again) R2 from pin 11 and clip its free end with one crocodile clip of the cable. Connect the 17V unstabilized power supply. With the free crocodile clip touch repeatedly the +17V wire and observe the LED. If the LED shines only when the +17V wire is touched, then BD651 is alright. Any other behaviour means that BD651 is defective and must be replaced.
After we make sure that BD651 (or its new counterpart) works fine, disconnect the 17V power supply, solder carefully the new 7809 on the board, solder R2 to pin 11 and reconnect the 17V power supply. Now I am positively sure that the LED will shine brighter or dimmer according to the potentiometer.
 

PrinzEugen

Active Member
Trenulist
21 Ian 2012
64
0
41
Belgrade
LOCATION
Belgrade
ok...figure out..we bypassed ic4093...just emmiter as out and vdd as input ...all this through r2 connected to base bd651....


led SHYYYYYY

when thouch free end of r2 with vdd 14v