Axon is resetting although the voltage is not dropping below 6.4V

[bilals]

Hello,

I have 18 servos and a 6V battery (Capacity:4600mAh) connected to Axon II.

I am trying to run only 2 servos at one time with no load on them (however, all servos are initiated).

There are some act_Setspeed commands with 20ms delay between each act_Setspeed command.

i.e.
 act_setSpeed(&Leg1_servo_1,speed);
      delay_ms(delay);
      act_setSpeed(&Leg6_servo_3,speed);
      delay_ms(delay);
      act_setSpeed(&Leg6_servo_3,-speed);
      delay_ms(delay);
      act_setSpeed(&Leg6_servo_3,speed);
      delay_ms(delay);
      act_setSpeed(&Leg6_servo_3,-speed);
      delay_ms(delay);
      act_setSpeed(&Leg1_servo_1,-speed);
      delay_ms(delay);
      act_setSpeed(&Leg1_servo_1,speed);
      delay_ms(delay);

A delay=20ms did not help.

What can cause the Axon to reset other than the voltage drop , because as i said the voltage is not going below 6.4V.

How can i go about finding where the problem lies.

I tried to used a power supply set between 6 to 7 VDC (Ones used in laboratories at universities), but the axon is resetting with or without the servos being connected to it. The voltage of the power supply is dropping to 1.8V.

Please help. Thanks

[madsci1016]

the axon is resetting with or without the servos being connected to it. The voltage of the power supply is dropping to 1.8V.

That statement seems to show your problem. Are you saying when you connect the Axon to the power supply, the voltage indicator on the power supply goes from 6V to 1.8V? If so, something in your Axon is shorting out.

[Webbot]

Since the servos are all initialised then they are ALL being sent a signal (default would be speed=0 ie stop ie 1.5ms pulse by default) under interrupts.

ie
act_setSpeed(&servo1, speed);
delays_ms(two weeks);

means that the delay will wait for two weeks but during the delay ALL of the servos are still being sent signals based on their last act_setSpeed setting (or 0).

If you want to truly discount the load from a servo then unplug it.
So disconnect all servos - does it reset?
How many do you have to reconnect before it resets?
If the supply is really dropping to 1.8v with all servos unplugged then as madsci has said then you probably have a short somewhere

[Admin]

Tell us more about your power supply.

Do you *only* have servos connected?

In code, how do you define 'delay' and 'speed' . . . for example, an int or?

Measure current draw both with the battery and with the power supply connected and report back.

[bilals]

That statement seems to show your problem. Are you saying when you connect the Axon to the power supply, the voltage indicator on the power supply goes from 6V to 1.8V? If so, something in your Axon is shorting out.

I am not sure because the axon works when using the battery but resets when servos move. While it does not work at all when using the power supply (the led of Axon starts flashing constantly as soon as i switch it on). Could it possibly short out only when using the power supply ?

Since the servos are all initialised then they are ALL being sent a signal (default would be speed=0 ie stop ie 1.5ms pulse by default) under interrupts.

ie
act_setSpeed(&servo1, speed);
delays_ms(two weeks);

means that the delay will wait for two weeks but during the delay ALL of the servos are still being sent signals based on their last act_setSpeed setting (or 0).

If you want to truly discount the load from a servo then unplug it.
So disconnect all servos - does it reset?
How many do you have to reconnect before it resets?
If the supply is really dropping to 1.8v with all servos unplugged then as madsci has said then you probably have a short somewhere

I tried what you said. I disconnected all servos except two of them (I did not remove the initialize code for disconnected servos. does it matter?). The Axon still resets with only 2 servos connected.

Tell us more about your power supply.

Do you *only* have servos connected?

In code, how do you define 'delay' and 'speed' . . . for example, an int or?

Measure current draw both with the battery and with the power supply connected and report back.

1) Check this link (Power supply i am using)
http://www.tti-new.com/press-releases/ex752m-pr.htm

2) Yes, only servos are connected.

3) Here is the code:
SERVO_LIST servo1[] = {&Leg1_servo_1,&Leg1_servo_2,&Leg1_servo_3,&Leg5_servo_1,&Leg5_servo_2};
SERVO_LIST servo2[] = {&Leg2_servo_1,&Leg2_servo_2,&Leg2_servo_3,&Leg5_servo_3};
SERVO_LIST servo3[] = {&Leg3_servo_1,&Leg3_servo_2,&Leg3_servo_3,&Leg6_servo_1,&Leg6_servo_2};
SERVO_LIST servo4[] = {&Leg4_servo_1,&Leg4_servo_2,&Leg4_servo_3,&Leg6_servo_3};

SERVO_DRIVER bank1 = MAKE_SERVO_DRIVER(servo1);
SERVO_DRIVER bank2 = MAKE_SERVO_DRIVER(servo2);
SERVO_DRIVER bank3 = MAKE_SERVO_DRIVER(servo3);
SERVO_DRIVER bank4 = MAKE_SERVO_DRIVER(servo4);

servosInit(&bank1, TIMER1);
servosInit(&bank2, TIMER3);
servosInit(&bank3, TIMER4);
servosInit(&bank4, TIMER5);

int speed=50
int delay=20 //


TICK_COUNT appControl(LOOP_COUNT loopCount, TICK_COUNT loopStart)
   {
   command=USB_GET;

   if (command=='t')
      {
      act_setSpeed(&Leg1_servo_1,speed);
      delay_ms(delay);
      act_setSpeed(&Leg6_servo_3,speed);
      delay_ms(delay);
      act_setSpeed(&Leg6_servo_3,-speed);
      delay_ms(delay);
      act_setSpeed(&Leg6_servo_3,speed);
      delay_ms(delay);
      act_setSpeed(&Leg6_servo_3,-speed);
      delay_ms(delay);
      act_setSpeed(&Leg1_servo_1,-speed);
      delay_ms(delay);
      act_setSpeed(&Leg1_servo_1,speed);
      delay_ms(delay);
      }


   return 2000;
   }

A value of 20-130 for delay does not work (Axon resets). Above that Axon rarely resets.


4) I measured the current draw when using the battery. The maximum I got is about 2.6A (When i move the servos). However, the axon resets at different values of current draw. So, it sometimes resets at 1.5A, 1.8A, 2.3A,2.5A. And at sometimes it draws that much current but it does not reset. i.e. At 1.5A current draw axon sometimes resets and sometimes not.

I did check the current draw when using the power supply (Shall I ?).


Thanks guys for trying to help. I hope we can solve the problem.

[Webbot]

I'm a bit confused by your code - the MAKE_SERVO commands are missing and you appear to be initialising the servos outside of your appInitHardware/Software functions - which are missing.

ie it doesn't compile.

[Admin]

4) I measured the current draw when using the battery. The maximum I got is about 2.6A (When i move the servos). However, the axon resets at different values of current draw. So, it sometimes resets at 1.5A, 1.8A, 2.3A,2.5A. And at sometimes it draws that much current but it does not reset. i.e. At 1.5A current draw axon sometimes resets and sometimes not.

2.6A is a lot for two unloaded servos! With no servos connected, how much current is drawn? What servos are you using?

The link to your psu says "In Mode B it operates as a true single output PSU of 0 to 75V at 0 to 4A", so you must be in Mode B as the other modes max out at 2A.

What battery are you using? (whats the current draw rating)

As for your code, this is wrong:

Code: [Select]

int speed=50
int delay=20 //It should be:

Code: [Select]

DRIVE_SPEED speed=50;
uint32_t delay=2000; //I think you can use TICK_COUNT here, right Webbot? manual didn't say

[bilals]

I'm a bit confused by your code - the MAKE_SERVO commands are missing and you appear to be initialising the servos outside of your appInitHardware/Software functions - which are missing.

ie it doesn't compile.

Sorry, i used some of the code only. Here is the complete code:

In hardware.h:

#define USB_UART UART1
#define USB_BAUD (BAUD_RATE)115200
#define USB_ACTIVATE &uart1SendByte
#define USB_GET uart1GetByte()

SERVO Leg1_servo_1 = MAKE_SERVO(FALSE, G5,1500, 500);
SERVO Leg1_servo_2 = MAKE_SERVO(FALSE, E4,1500, 500);
SERVO Leg1_servo_3 = MAKE_SERVO(FALSE, E5,1500, 500);
SERVO Leg2_servo_1 = MAKE_SERVO(TRUE, E6,1500, 500);
SERVO Leg2_servo_2 = MAKE_SERVO(TRUE, H3,1500, 500);
SERVO Leg2_servo_3 = MAKE_SERVO(TRUE, H4,1500, 500);
SERVO Leg3_servo_1 = MAKE_SERVO(FALSE, H5,1500, 500);
SERVO Leg3_servo_2 = MAKE_SERVO(FALSE, B4,1500, 500);
SERVO Leg3_servo_3 = MAKE_SERVO(FALSE, B5,1500, 500);
SERVO Leg4_servo_1 =MAKE_SERVO(TRUE, B7,1500, 500);
SERVO Leg4_servo_2 =MAKE_SERVO(TRUE, H7,1500, 500);
SERVO Leg4_servo_3 =MAKE_SERVO(TRUE, G3,1500, 500);
SERVO Leg5_servo_1 =MAKE_SERVO(FALSE, L0,1500, 500);
SERVO Leg5_servo_2 =MAKE_SERVO(FALSE, L1,1500, 500);
SERVO Leg5_servo_3 =MAKE_SERVO(FALSE, L3,1500, 500);
SERVO Leg6_servo_1 =MAKE_SERVO(TRUE, L4,1500, 500);
SERVO Leg6_servo_2 =MAKE_SERVO(TRUE, L6,1500, 500);
SERVO Leg6_servo_3 =MAKE_SERVO(TRUE, L7,1500, 500);

SERVO_LIST servo1[] = {&Leg1_servo_1,&Leg1_servo_2,&Leg1_servo_3,&Leg5_servo_1,&Leg5_servo_2};
SERVO_LIST servo2[] = {&Leg2_servo_1,&Leg2_servo_2,&Leg2_servo_3,&Leg5_servo_3};
SERVO_LIST servo3[] = {&Leg3_servo_1,&Leg3_servo_2,&Leg3_servo_3,&Leg6_servo_1,&Leg6_servo_2};
SERVO_LIST servo4[] = {&Leg4_servo_1,&Leg4_servo_2,&Leg4_servo_3,&Leg6_servo_3};

SERVO_DRIVER bank1 = MAKE_SERVO_DRIVER(servo1);
SERVO_DRIVER bank2 = MAKE_SERVO_DRIVER(servo2);
SERVO_DRIVER bank3 = MAKE_SERVO_DRIVER(servo3);
SERVO_DRIVER bank4 = MAKE_SERVO_DRIVER(servo4);


In the Source file (Hexapod.c <---- I named it this way):

#define RPRINTF_FLOAT
#define RPRINTF_COMPLEX

//WebbotLib includes
#include "sys/axon2.h"   //defines your MCU (change to axon.h if you use the original Axon)
#include "servos.h"      //use for software servo PWM
#include "rprintf.h"   //use for UART
#include "actuators.h"
#include "led.h"
#include "hardware.h"

// In my initialising code - pass the list of servos to control
void appInitHardware(void)
{
//initialize UART ports (see hardware.h to change baud/names)
uartInit(USB_UART, USB_BAUD);         //USB

// declare USB for output
rprintfInit(USB_ACTIVATE);

servosInit(&bank1, TIMER1);
delay_ms(500);
servosInit(&bank2, TIMER3);
delay_ms(500);
servosInit(&bank3, TIMER4);
delay_ms(500);
servosInit(&bank4, TIMER5);
delay_ms(500);
}



TICK_COUNT appInitSoftware(TICK_COUNT loopStart)
{
   rprintf("Axon Initiated....Awaiting command\n");
return 2000;
}


int command;
int speed=50;//rprintf("Absolute speed=%d.\n",speed);
int delay=200;//rprintf("Delay set to %dms.\n",delay);



// This is the main loop
TICK_COUNT appControl(LOOP_COUNT loopCount, TICK_COUNT loopStart)
   {
   command=USB_GET;//get a single character

   if (command=='t')
      {
      act_setSpeed(&Leg1_servo_1,speed);
      delay_ms(delay);
      act_setSpeed(&Leg6_servo_3,speed);
      delay_ms(delay);
      act_setSpeed(&Leg6_servo_3,-speed);
      delay_ms(delay);
      act_setSpeed(&Leg6_servo_3,speed);
      delay_ms(delay);
      act_setSpeed(&Leg6_servo_3,-speed);
      delay_ms(delay);
      act_setSpeed(&Leg1_servo_1,-speed);
      delay_ms(delay);
      act_setSpeed(&Leg1_servo_1,speed);
      delay_ms(delay);
      }


   return 2000;
   }

[bilals]

2.6A is a lot for two unloaded servos! With no servos connected, how much current is drawn? What servos are you using?

The link to your psu says "In Mode B it operates as a true single output PSU of 0 to 75V at 0 to 4A", so you must be in Mode B as the other modes max out at 2A.

What battery are you using? (whats the current draw rating)

As for your code, this is wrong:

Code: [Select]

int speed=50
int delay=20 //

It should be:

Code: [Select]

DRIVE_SPEED speed=50;
uint32_t delay=2000; //I think you can use TICK_COUNT here, right Webbot? manual didn't say

1) When no servos are connected, the current drawn is 0.2-0.3 A. The servos are of this type: JR SERVO DS8911. There is no datasheet for them (At least I could'nt find it anywhere online).

2) I am not sure on what mode the power supply was set. I'll set it to mode B.

3) The battery I am using (Actually, two batteries connected in parallel (capacity 2300mAh each) ) are of this type: TENERGY NiMH Rechargable Battery 6v 2300mAh. This is a link for the battery:
http://www.trossenrobotics.com/store/p/5916-NiMH-Rechargable-Battery-6v-2300mAh.aspx

4) As for the code, I dont think it does make a difference, does it? The code compiles with no problems. I have tried not using them, but axon still resets. i.e :
act_setSpeed(&Leg1_servo_1,50);
      delay_ms(20);
..etc


Some questions: The axon is supposed to work normally when connected to a power supply, right ? Could it be a problem with the power supply ? Can the axon be incompatible to some power supplies ?

[Admin]

1) When no servos are connected, the current drawn is 0.2-0.3 A.

Have any LEDs turned on? This seems a bit high, otherwise.

3) The battery I am using (Actually, two batteries connected in parallel (capacity 2300mAh each) ) are of this type: TENERGY NiMH Rechargable Battery 6v 2300mAh. This is a link for the battery:
http://www.trossenrobotics.com/store/p/5916-NiMH-Rechargable-Battery-6v-2300mAh.aspx

The batteries should work fine. Sure they are fully charged?

4) As for the code, I dont think it does make a difference, does it? The code compiles with no problems. I have tried not using them, but axon still resets. i.e :
act_setSpeed(&Leg1_servo_1,50);
      delay_ms(20);

Its very easy to reset a microcontroller with bad/buggy code. Try our fixes.

edit:

The servos are of this type: JR SERVO DS8911.

Forgot to comment on the servos . . . high torque digital servos that require heat sinks . . . yea, 2.6A draw sounds about right for these guys . . .

[bilals]

Have any LEDs turned on? This seems a bit high, otherwise.

The batteries should work fine. Sure they are fully charged?

Its very easy to reset a microcontroller with bad/buggy code. Try our fixes.

1) Just the green dot (The one in the LED Display).

2) Yeah, they are fully charged. How can i make sure that both batteries are actually connected. Is there a way to verify ?

3) I'll try the code. But what i was saying is that before using int speed and int delay, i used to directly put the value in the command itself (I Did not use integers at all).
   act_setSpeed(&Leg1_servo_1,50);
   delay_ms(20);
Axon used to and still resets.

[Webbot]

Can you 'bear with me' - and try putting your servos into say 2 bank of 9. My train of thought is that you are currently using all of the 16 timers for servos. Should be ok but just humour me!

Finally: How do you know the Axon is resetting? Are you getting the "Axon Initiated....Awaiting command\n" msg over and over. - ie is resetting or just hanging?

[Admin]

3) I'll try the code. But what i was saying is that before using int speed and int delay, i used to directly put the value in the command itself (I Did not use integers at all).
   act_setSpeed(&Leg1_servo_1,50);
   delay_ms(20);
Axon used to and still resets.

Thats 'cause, as Webbot pointed out in the very beginning, you are asking the motor to invert its direction every 20ms or so. With a digital servo, that'll cause massive current draw, and as such, a major voltage drop - and therefore a reset.

Just fix the code and don't make me grab you through the computer screen!!!

[bilals]

Can you 'bear with me' - and try putting your servos into say 2 bank of 9. My train of thought is that you are currently using all of the 16 timers for servos. Should be ok but just humour me!

Finally: How do you know the Axon is resetting? Are you getting the "Axon Initiated....Awaiting command\n" msg over and over. - ie is resetting or just hanging?

1) I am sure that the Axon will reset when i use 9 servos in one bank, because with only 4 servos in one bank the Axon currently resets. But I'll try it if you think it will help.

2) Yes, i keep getting the Axon intiated..... plus the green led flahses over and over.

Thats 'cause, as Webbot pointed out in the very beginning, you are asking the motor to invert its direction every 20ms or so. With a digital servo, that'll cause massive current draw, and as such, a major voltage drop - and therefore a reset.

Just fix the code and don't make me grab you through the computer screen!!!

, Ok I'll fix the code, but as i said there is no voltage drop, at least the voltage is not dropping below 6.4V.

[chelmi]

, Ok I'll fix the code, but as i said there is no voltage drop, at least the voltage is not dropping below 6.4V.

How are you measuring this? The drop might be too quick for you to see it on a multimeter.

[bilals]

How are you measuring this? The drop might be too quick for you to see it on a multimeter.

The multimeter i am using has an option that can store the minimum value the voltage drops to.
Today the lowest value it reached was 5.9V, But i think this is because the battery needs charging.

[waltr]

Even a min value storing multimeter can be way to slow to see voltage drops from sudden current surges. Multimeters do average the measured voltage to a response time of many msec.

The proper instrument is an Oscilloscope and is the only way to see a voltage with usec duration.

[bilals]

Even a min value storing multimeter can be way to slow to see voltage drops from sudden current surges. Multimeters do average the measured voltage to a response time of many msec.

The proper instrument is an Oscilloscope and is the only way to see a voltage with usec duration.

Thanks for the information. I'll try to measure it using an oscilloscope, but wont the reading be analog, it doesn't matter, right ?.

[waltr]

Right, the voltage to a processor is analog. That's what your multimeter reads!

[Soeren]

Hi,

Thanks for the information. I'll try to measure it using an oscilloscope, but wont the reading be analog, it doesn't matter, right ?.

Actually, all voltages are analog if you look at them in detail.

Analog voltages is not static however and a dip to 0V in say 1ms (which is a very long time in some electronics) will not even register on a multimeter.

On a 'scope, you can set the trigger to fire at a dip in voltage, so it will be very visible.

If you have access to a digital storage 'scope (DSO) you can set it up to catch single events and just let it listen until a dip happens, then come back and study it later.

[bilals]

Hi,

Thanks for the information. I'll try to measure it using an oscilloscope, but wont the reading be analog, it doesn't matter, right ?.

Actually, all voltages are analog if you look at them in detail.

Analog voltages is not static however and a dip to 0V in say 1ms (which is a very long time in some electronics) will not even register on a multimeter.

On a 'scope, you can set the trigger to fire at a dip in voltage, so it will be very visible.

If you have access to a digital storage 'scope (DSO) you can set it up to catch single events and just let it listen until a dip happens, then come back and study it later.

Thanks for the tips. It is helpful.

[bilals]

Hello guys, i still didnt get the opportunity to measure the voltage drop using the oscilloscope. But do you think that if bought 2 more batteries of the same type (connect them in parallel) the problem will be solved ?

4 x 2300mAh = 9200mAh

[Admin]

Can't say if you didn't fix the code, first

[bilals]

Thanks for the quick reply  . But Admin, can you clarify more regarding the fix.

Thanks for your patience.

[Admin]

Read the above posts by admin and Webbot.