Servo control woes

I'm trying to get down the basics of PIC servo control, and not having much luck. I'm using a 16F690 and trying to get some kind of meaningful movement out of my servo. However, the only thing my servo will do is move to one extreme of its range and sit there, perhaps scornfully.

The offending code is:

@ DEVICE MCLR_OFF,WDT_OFF,PROTECT_OFF
DEFINE OSC 8
servo var PORTC.5
i var byte

loop:

for i = 200 to 400
low servo
pulsout servo,i
pause 18
next i

goto loop


I thought that would get a wiggle out of it, but no. I'm using the internal oscillator there, so I think my timing is correct - the 16F690 is running at 8MHz.

I've tried many different timings, and I've RTFM and RTFDataSheet and RTF-half-the-contents-of-google. I've tried so many tweaks and variations that I'm worried I'll hit the PIC's write limit soon. I've also tried using different servos (from other manufacturers) and a different 16F690. It just moves in one direction to its extreme and sits, every time.

What is it I am missing here? Any help would be greatly appreciated.

[QUOTE=Eriswerks]I'm trying to get down the basics of PIC servo control, and not having much luck. I'm using a 16F690 and trying to get some kind of meaningful movement out of my servo. However, the only thing my servo will do is move to one extreme of its range and sit there, perhaps scornfully.

The offending code is:

@ DEVICE MCLR_OFF,WDT_OFF,PROTECT_OFF
DEFINE OSC 8
servo var PORTC.5
i var byte

loop:

for i = 200 to 400
low servo
pulsout servo,i
pause 18
next i

goto loop


hi,Eris

@ 8 Mhz, pulsout resolution is 5µS ( 10µS *4 / F.Mhz)

so, correct for that.

for i = 200 to 400 means i should not be a byte ( i max = 255 !!! )

so, what's happening ??? i value goes from 200 to 255 then ... back to 0 , out of the for - next loop !!! program ends.

just change i for a WORD !!!

Alain

Thanks for the reply. I've changed the variable's type from BYTE to WORD, but there must be something else wrong as well - my servo still goes to one extreme and sits. Even if I do away with the loop entirely and just send it one position to center it, I'm still getting this behavior. That suggests to me that it must be timing-related, but I've tried timing some LED pulses without doing a DEFINE OSC and it confirms the 8MHz I see in the datasheet.

Maybe if I buy the PIC some chocolates?

Looking through the 16F690 datasheet for anything I might have missed, I found this little nugget:

The output of the 8 MHz HFINTOSC and 31 kHz
LFINTOSC connects to a postscaler and multiplexer
(see Figure 3-1). The Internal Oscillator Frequency
select bits, IRCF<2:0> (OSCCON<6:4>), select the
frequency output of the internal oscillators. One of eight
frequencies can be selected via software.


So, it seems like in order to make sure that the PIC is running at 8MHz, I should be telling it to use HFINTOSC and not LFINTOSC, and according to the diagram mentioned I should be setting the frequency select bits to 111.

How do I go about doing that in PicBasic? Is it actually necessary to do so, or does the "intrc_osc_noclkout" that's in the include file for this chip have me covered already? I've also found mentions of putting "OSCCON = $70" in the code to set 8MHz, but I can't seem to find any real documentation on it. Is this controlling the frequency select bits? If so, do I still need to specify HFINTOSC somewhere?

Sorry for the deluge of questions...

1. Why not take the whole 8MHz issue out of the equation and run it at 4MHz? Once you get it working at 4MHz...THEN change the oscillator speed.

2. I have some servo code at home. For a 12F629 I think. I'll dig it out and post it later.

3. From what I remember, sending a single pulse of 1.5mS won't center the servo. You have to KEEP sending pulses to get it to move. So if it's at the extreme end, then you have to keep sending 1.5mS pulses until the servo reaches center. AND you need to keep sending those pulses if you expect it to STAY in the center. Some servos are quite a bit slower than others (check the spec sheet).

regards

Eriswerks, milestag is right, the servo requires updates every 50 milliseconds or so othwise most servos will shutdown and produce no torque.

Dave Purola,
N8NTA

Thanks for your suggestions. I've whittled down the code to the bare essentials:

@ DEVICE MCLR_OFF,WDT_OFF,PROTECT_OFF

servo var PORTC.5

loop:

pulsout servo,150
pause 18

goto loop


That's the entirety of it, no configuration or anything left out. That ought to center it and hold it there so long as the PIC is powered, but once again it moves all the way to one extreme and stays. I tried other pulsout values from 1 to 2000 just for the fun of it, but I keep getting the same result.

The only thing I can think of is that the internal timer is in its LFINTOSC mode, which is 31kHz and would probably cause any pulse I care to send to look the same to the servo. There must be some language in PicBasic to tell the PIC to which frequency to set the internal oscillator, but I haven't been able to find it.

What am I (still) doing wrong here?

Hi,Eris

Seems you have to read some closer the µChip Datasheet section 3 ...

Lots ( ! ) of registers to set " by hand " that PbP do not handle !!!

to confirm that, just program a dusty 16F84 ... will work properly with your example !!!

Alain

The internal oscillator doesn't automatically run at 8MHz. You have to
write to OSCCON to switch to 8MHz.

The default value of OSCCON frequency select bits is for 4MHz. See the
datasheet for the default value in this register at power up.

Also PULSOUT toggles the pin twice to produce your pulse, so the initial
state of the pin determins the polarity of the pulse.

You want a high-going pulse for your servo, so clear the pin before the
PULSOUT command.

See if this works.

Code:

---

@ DEVICE INTRC_OSC_NOCLKOUT, MCLR_OFF,WDT_OFF,PROTECT_OFF
DEFINE OSC 8
servo var PORTC.5
i var word

OSCCON = %01110000 ' Now it's going to run at 8MHz

PORTC.5 = 0 ' <-- make pin low before PULSOUT (toggles high then low)
TRISC.5 = 0  ' make pin an output

loop:

for i = 200 to 400
low servo
pulsout servo,i
pause 18
next i

goto loop

---

Hi Eriswerks,

Try using another servo or at least make sure that your servo is ok.

Also, as you know, for most of the servos, the pulse range is from 100mS to 200mS.

For some recent ones, range is from 60mS to 240mS (wider movement).

Also, below is a code to consider while playing with servos.

servo var portc.5
freq var byte
pulse var byte


loop:

pot portc.3,255,pulse
pot portc.4,255,freq

if pulse>240 then pulse=240 'Max pulse protection.
If pulse<60 then pulse=60 'Min pulse protection
If freq>30 then freq=30 'Frequency protection
IF freq<5 then freq=5 'Frequency protection

pulsout servo,pulse
pause freq


goto loop
-----------------------
If you also use an LCD, you can see at what value(s) your servo does what it does and how it does.



------------------------------

[QUOTE=sayzer]

Try using another servo or at least make sure that your servo is ok.

Also, as you know, for most of the servos, the pulse range is from 100mS to 200mS.

> That's really new ...


> Mhhhh, I do not know who's inside ... but he has a lot to learn about R/C Equipments ...

Alain

PS: ALL servos accept the 40-50 Hz frame rate ...

---------------------------------------------
There must be some language in PicBasic to tell the PIC to which frequency to set the internal oscillator, but I haven't been able to find it.

What am I (still) doing wrong here?
---------------------------------------------

Yes, there is a PicBasic command... POKE

Use POKE to write values to ANY register in the datasheet. I usually use binary values so it's easier to see and change individual bits:

POKE OSCON, %10110001

(this is a made up value, I don't know what you will need exactly)

For Acetronics,

This is not the place for your egoistic posts.

Why don’t you go to a primary school and show the little kids how much you know about everything you think you know about?

You can top your ego that way.

Sayzer,

If you want to know, I do not do care about my EGO ... no time for that.

I'd just like you NOT to tell wrong things to those who do not know ...

It's difficult enough to learn pic's programming ... try not to add your own errors nor misunderstandings ... verify first !!!

Alain

Engineer, 30 years of DiY R/C, ... and always doubting ...

Ok, Acetronics I agree . People make mistake even the ones with 30yrs of experiences.

Could you pls tell us what happens if we send 250mS - 300mS pulse to an R/C servo at 50Hz?

Also, 350mS and 400mS at the same freq?

How would servo respond to these pulse(s) ?

Pls do not make it so complicated so that we, the learners without 30 yrs of experiences, can learn something from your valuable words. Before you post your answer, make sure that your answer does not conflict with your previous posts;

----------------------

Hi, sayzer

So, you've decided to be frankly ridiculous ... your problem !!!

250ms pulses @ 50 HZ gives .... a firm "1" ( or "0" ) level !!! you've just recreated the resettable monostable mutivibrator !!!

Who should go back to primary school ??? ( period = 1 / frequency ... no need a highschool for that )

your servo wil just have the little "jump" for amplifier settling at power up ... and won't move any more.

you shouldn't have tried .... you're really boring.

Alain

Finally you got it Acetronics,

That is what I have been saying.

If you look at the original loop, below, the code is sending pulses from 200mS to 400mS. In this case, the R/C servo would goto extreme edge and stop there at its max input pulse (let say 240mS). However, the code keeps sending larger pulses upto 400mS even though the R/C servo already stopped at 240mS.

-----Orginal code-----

loop:

for i = 200 to 400 'When i=240, servo stops, but loop keeps increasing the pulse.
low servo
pulsout servo,i
pause 18
next i

goto loop

---------------------


An engineer like you, age 49, should be in teaching mode rather then an egoistic mode. You should be using more explicit words to express your knowledge and have others understand what you are trying to say rather then using complicated engineering terms (an expression of egoistic unsatisfactory career to show others how much you know and others don’t)

I will stop here. This is going nowhere.
Let stop here Acetronics; it has no benefit to any of us.

-----------------------------

Eris's pulses are NOT 200 to 400 ms, as you tell,

but 200 to 400 x 5µS !!! this EQUALS 1 to 2 ms ...

read your manual more carefully before hoping to give lessons ...


Straight back to primary school !!!

Hope not to read you any more about that. CENSORED !!!

E.O.F.

Alain

Acetronics,

I would like to apologize for the things I said.

You were right and I was wrong. I missed some parts and behaved impolite.

I also sent you a private message to apologize.

Sayzer

---------------

[QUOTE=sayzer]Acetronics,

I would like to apologize for the things I said.

...

Hi, Sayzer

that's kind from you ,

the only important thing is you've learnt something useful ... I don' like this teaching method ... but I'm pretty sure you will always remember about PULSOUT resolution.
a good joke is my favourite method for students to keep things in mind.

That's the only real important thing....; may be ... life respect too, but it's not the right place to discuss it here.

Alain

I've tried all the tips you guys gave, and have also been trying any and all PicBasic servo control code I could find on the net, but something is still not right. I've tried four different servos that I know to be functional. I've tried a 16F690 and a 16F628A, with the internal oscillator and with an external 4MHz crystal. I've tried changing the pulsewidths and pauses to every reasonable and many unreasonable numbers. I've switched breadboards. I'm shopping around for a goat to sacrifice to the PIC.

Here's what I'm trying to use now, with a 16F628a:

trisb = 0
cont var byte
cont = 0
led var PORTB.1

hither:

low led
for cont = 0 to 50
pulsout PORTB.0,110
pause 18
next cont

goto yon

yon:

high led
for cont = 0 to 50
pulsout PORTB.0,190
pause 18
next cont

goto hither


It is a slightly modified version of some code from the rentron site, which was meant for a 16F84. My include file (not the whole thing of course) looks like:

device pic16F628A, xt_osc, wdt_on, mclr_on, lvp_off, protect_off


My LED does flash convincingly, but still my servo heads to one extreme and stays there. This is keeping me awake at night now. What mistakes might I still be making? Does anyone know of an example of a working servo controller written in PicBasic, including a circuit diagram and not intended for a 16F84? I just need something functional to build on.

You're probably sick of this thread by now, but any help would be appreciated.

Try this code.
Schematic and code are also attached.
------------------------------------------------------------------------


@ DEVICE intrc_osc, MCLR_OFF,WDT_OFF,PROTECT_OFF

DEFINE LCD_EREG PORTA
' Set LCD Enable port

DEFINE LCD_EBIT 7
' Set LCD Enable bit

TRISA=%00100000
' RA5=Input, Rest= Output

TRISB=%01100001
' RB6, RB5, RB0 = Input, Rest=Output

'--------Assign name to ports---------------
potvar var portb.0
'Pot is connected to RB0.

servo var portb.3
'Servo is connected to RB3.

Button1 var portb.5
'Button1 is connected to Portb.5 and increases the frequency.

Button2 var portb.6
'Button2 is connected to Portb.6 and decreases the frequency.


'---------Variables-------------------------
pulse var byte
'Variable to hold Pot value for servo Position

freq var byte
'Variable to hold Button1 and Button2 adjustments for Frequency

'--------Initial Setup----------------------
Freq=18
'Initial Frequency

'--------Main Loop-------------------------

loop:

IF button1=0 then freq=freq+1
'Incr freq

IF button2=0 then freq=freq-1
'Decr freq

IF freq>30 then freq=30
'Max. Low freq

IF freq<5 then freq=5
'Max. High Freq

pot potvar,255,pulse
'Read pot and store the value in pulse variable

If pulse>240 then pulse=240
'Max. Pulse 240mS

IF pulse<60 then pulse=60
'Min. Pulse 60mS.

pulsout servo,pulse
'pulsout the stored pot value (pulse) to servo pin.

pause freq

lcdout $fe,1,"Pulse:",#pulse
LCDOUT $fe,$c0,"Freq:",#freq


goto loop

--------------------------------------------------------------------------

Hi, Eris

Seem strange ...

There's one last thing we didn't had a look to : Your supply !!!

Could you describe it or better send a scheme.

On my side, I try your code on my 628 R/C testboard ... just to be sure!.

Alain

Hi, Eris

Bad news ... I copied and pasted your code, just changed portB N°s as required here ... put a 6LR1 cell ( there's a regulator on the board !!! ) and a futaba servo.

Every thing goes as you programmed it : the led blinks and the windscreen wiper does its job ...

The problem is not ( no more ... humour ) in your code ...

So, What to do ??? ...

First verify your servo supply ( good common ground w/Pic ... enough current to feed the servo ...)

Second verify the signal lead of the servo ... servo behaviour smells no received signal !!!
In that way I would recommend you !!! SOFORT !!! to add a 1k resistor between Portb.0 and your servo connector.

I keep the line
Alain

Many thanks to you smart and helpful people - I've finally got this working. I was spending all my time checking and rechecking the code, and it seems that the problem (or the final one, anyway) was that the servo and PIC didn't have a common ground. And here I thought that breadboard with two separate bus strips would be *helpful*.

I did, at least, pick up a lot of PicBasic tips while working on this, so I should be better equipped for the rest of my project. Thanks again!