I have checked the circuit using ir 2101,
result is that only the bottom FET working
FET above does not work . I'm having trouble

how did you test it?

what frequency are you using in IC 2101:HIN,LIN?
does the bootstrap cap is big enough?

have you tried to check that all your IC is working?

"only the bottom FET working
FET above does not work"
how do you know that?


*IC- Integrated circuit

Originally Posted by goly

how did you test it?

what frequency are you using in IC 2101:HIN,LIN?
does the bootstrap cap is big enough?

have you tried to check that all your IC is working?

"only the bottom FET working
FET above does not work"
how do you know that?


*IC- Integrated circuit

thank you, I'm really disappointed with the new circuit ir 2101
 I killed many things: 3 ir 2101, 8 fet 3205 and 1 chip 12f 675

I have abandoned this design

you use schematic circuit with ir2184, it is work well

it use two input : PWM and ON/OFF

hi Bruce abbot
blink frequency of LED version 12f 617 near 1s., but sometimes I see LEDs blink very fast, I do not know the error is due design my circuit or for some other reason?

why is capacitor 50p? I replace by 100n capacitor, result is pic does not receive the input signal

Originally Posted by fantasy1988

thank you, I'm really disappointed with the new circuit ir 2101
 I killed many things: 3 ir 2101, 8 fet 3205 and 1 chip 12f 675

I have abandoned this design

you may try to use pnp npn h bridge instead
less efficient but works great, i have one on wildthumper


pnp- of the same type as the npn at the high side of the h bridge
it uses IC-4027 to drive the mosfet and IC-AND gate for PWM speed ctrl

Originally Posted by future88

hi Bruce abbot
blink frequency of LED version 12f 617 near 1s., but sometimes I see LEDs blink very fast, I do not know the error is due design my circuit or for some other reason?

Not an error. The LED toggles on and off for every 8 bad pulses. If there is no signal the pulse measuring routine times out after 33 milliseconds, and the flash rate is quite slow. But if there is interference the pulses can be very short and rapid, so the flash rate will be a lot higher.

why is capacitor 50p? I replace by 100n capacitor, result is pic does not receive the input signal

Its purpose is to bypass high frequency interference (eg. RF output from a 27MHz transmitter). If the value is too large then it will also filter out the throttle signal!

Originally Posted by Bruce Abbott

Not an error. The LED toggles on and off for every 8 bad pulses. If there is no signal the pulse measuring routine times out after 33 milliseconds, and the flash rate is quite slow. But if there is interference the pulses can be very short and rapid, so the flash rate will be a lot higher.

you can show me the command line makes the LED flashes quickly.

I only see the command makes the LED flashes slowly

BadPulse:
ifdef FAILSAFE
movi16 MIDTHROTTLE,ServoCount ; set to neutral
endif
incf BadPulseCount
movf BadPulseCount,w
sublw 7
skpnc ; toggle LED every 8 bad pulses
goto bp_done
clrf BadPulseCount ; reset count
btfss GPIO,LedBit
goto bp_LED_on
bcf GPIO,LedBit ; Toggle Status LED
goto bp_done
bp_LED_on: bsf GPIO,LedBit
bp_done:

Originally Posted by future88

you can show me the command line makes the LED flashes quickly.

I only see the command makes the LED flashes slowly

It's the same code.

The rate at which the LED flashes is dependent on how long the code takes to measure the servo pulse. Normally that will be around 10 to 20mS (depending on the frame rate of your receiver), but the LED doesn't flash when there is a good signal.

If no pulse is detected then the 'Measure' subroutine will time out after ~33mS, and the flash rate will be 8*33mS off, 8*33mS on, ie. about 2 flashes per second. This is the normal 'no signal' response.

However, if you are using an AM or FM radio and the receiver picks up interference, then it may put out a lot of short pulses. The LED will then flash on and off 16 times slower than the average frequency of those pulses (which could be several kHz).

This behavior is good because you can tell the difference between interference (fast flash) and no signal (slow flash).

Originally Posted by goly

you may try to use pnp npn h bridge instead
less efficient but works great, i have one on wildthumper


pnp- of the same type as the npn at the high side of the h bridge
it uses IC-4027 to drive the mosfet and IC-AND gate for PWM speed ctrl

thank you,I had given up all

thank Bruce abbot
why your website have not comment for the viewer?
you should integrate this feature, I am very interested

I have write a small program in version 12f617
create tone "beep" when the loss signal.
create tone "beep beep beep" when the set is complete for neutral
but I'm having difficulties in interrupt of PIC, pic do two jobs at once."Blinking leds and create melody "
I'll post the asm file, hope people will help me

Originally Posted by rc my life

I have write a small program in version 12f617
create tone "beep" when the loss signal.
create tone "beep beep beep" when the set is complete for neutral
but I'm having difficulties in interrupt of PIC, pic do two jobs at once."Blinking leds and create melody "
I'll post the asm file, hope people will help me

I think, 12f 617 program has completed, you should use it, because change program is very difficult

hello bruce abbot
I using the program winpic 800 ( jdm propramer ) to load the hex file for 12f675


during the loading of software appear notice of the alternative addresses for pic: select YES or NO. ?
I do not understand, and it seems to affect the certification of the neutral position of the signal pulse
I loaded the program for two pic 12f675. but a neutral position in two different places.

Originally Posted by fantasy1988

during the loading of software appear notice of the alternative addresses for pic: select YES or NO. ?
I do not understand, and it seems to affect the certification of the neutral position of the signal pulse
I loaded the program for two pic 12f675. but a neutral position in two different places.

Sounds like it is overwriting the osccal subroutine at address 0x3ff, so you must select the other option which doesn't overwrite it. For the PICs you have already burned, you can determine the correct calibration value using the circiuit and code found here. Alternatively you could just try different values until you get close enough to 1500uS.

If you decide to change it manually then the following information may help:-

REGISTER 2-7: OSCCAL — OSCILLATOR CALIBRATION REGISTER (ADDRESS: 90h)
bit 7-2 CAL5:CAL0: 6-bit Signed Oscillator Calibration bits
111111 = Maximum frequency
100000 = Center frequency
000000 = Minimum frequency
bit 1-0 Unimplemented: Read as '0'

example:-
ORG 0x3ff
RETLW 0xfc ; osccal value = maximum frequency (b'11111100')