Servo motor drive voltage

This is the place for any magazine-related discussions that don't fit in any of the column discussion boards below.
Post Reply
ProboticsAmerica
Posts: 8
Joined: Thu May 19, 2005 1:01 am
Contact:

Servo motor drive voltage

Post by ProboticsAmerica »

The chip driving the motor inside a Hitec 422 and 300 series servo never sends zero voltage (short to ground) to the motor. At a full stop it provides (given a five volt source) +3 (relative to the input ground) to one side of the motor and +3 to the other side of the motor. Since there is no difference between the voltages there is no movement. When the servo is commanded to move or needs to counteract force against its arm the voltage on one side of the motor goes up to five volts (in pwm pulses) and the voltage on the other side goes down eventually to very close to ground. <p>Why?<p>Why not just set voltage = ground on both sides of the motor when at rest. Does anyone have a design for a circuit to 'reset' the voltage to be 0-+5 volts on one side for movement (grounding the other) one way and flip that for movement the other way?<p>Your help is appreciated.
Gorgon
Posts: 325
Joined: Wed May 04, 2005 1:01 am
Location: Norway
Contact:

Re: Servo motor drive voltage

Post by Gorgon »

The reason for the midlevel voltage in idle mode, is the fact that this is the 'zero' level for the motor drive.<p>5v is positive activation and 0v is in fact negative activation.<p>If the idle level was at 0v, both the (-)drivers would be activated. To activate the (+)drivers you first need to turn the (-) off, delay a bit and then activate the (+). The reverse procedure is needed to turn it off.<p>If you didn't insert the delay, the outputs on that side of the motor would short the +5v and 0v a moment, with undesirable consequences. The lesser one would be to bleed the battery.<p>The reason for this is the use of saturation of the driver transistors. To turn it off you need to remove the excess energy stored in the base, and this takes some time.<p>So, to make the circuit less prone to failure, you only activate one side at a time, (+) or (-).<p>--<p>I hope you understand my English. It is with some anguish I write my reply, since English is not my primary language and the grammar discussion is going on in other threads on the forum. ;) <p>TOK
Gorgon the Caretaker - Character in a childrens TV-show from 1968. ;)
rshayes
Posts: 1286
Joined: Tue Mar 04, 2003 1:01 am
Contact:

Re: Servo motor drive voltage

Post by rshayes »

Gorgon:<p>Your English is quite understandable. The criticism in the other threads that you have been reading is not directed at people using English as a second language. It was directed at people whose primary language is English, but are too lazy to write it at above a fourth grade level.
Gorgon
Posts: 325
Joined: Wed May 04, 2005 1:01 am
Location: Norway
Contact:

Re: Servo motor drive voltage

Post by Gorgon »

Thanks Stephen,
It was really a lame quip at a safe distance from the heavy sluggers. :D <p>To be honest, some of the postings do have the air of SMS around them. I'm sure some are intended but all are a bit on the tough side to decode. :confused: <p>TOK
Gorgon the Caretaker - Character in a childrens TV-show from 1968. ;)
ProboticsAmerica
Posts: 8
Joined: Thu May 19, 2005 1:01 am
Contact:

Re: Servo motor drive voltage

Post by ProboticsAmerica »

Thanks Gorgon!<p>Just to make sure I understand. In all places not stated voltage measurements are relative to common (non-motor) ground:<p>Condition 1) Input 1 to motor is +3, Input 2 to the motor is +3 <p>Command to move one direction comes in:
Average positive charge rises on side one, the driver transistor passing the current is already saturated. Average positive charge stays the same or drops on the other side - that driver transistor is also already saturated. Motor begins moving<p>Command to reverse comes in: Average positive charge on side one immediately drops below 3v without a delay. The average positive charge on side one boosts immediately without delay. Both driver transistors remain saturated because it is a change in voltage rather than a move from 'off' to 'on'.<p>There is never a short and the drivers transistors never turn off or reverse. <p>Is that right?
Gorgon
Posts: 325
Joined: Wed May 04, 2005 1:01 am
Location: Norway
Contact:

Re: Servo motor drive voltage

Post by Gorgon »

To be honest, I'm not sure I follow you. :) <p>In an H-bridge motor drive, you have 4 transistors, 2 on each side, one pulling towards 5v (+) and one pulling towards 0v (-).<p>If both transistors on one side of the motor are active, you get a short.<p>In idle mode (motor off) none of the transistors are active. The motor will 'hover' in the middle between +5v and 0v. Like the +3v you measure.<p>When you activate the (-) transistor on one side, and the (+) transistor on the other, the motor runs in one direction. Using PWM you can adjust the speed by just turning those transistors on and off.<p>Normally you don't reverse a running motor on the fly. There are some physics involved in stopping the motor and reversing it. This takes a lot of energy to do, so the best is to just pause and then reverse. When pausing, the drivers return to idle and the question of the saturation is solved.<p>You are then starting from idle again, in the other direction.<p>I hope this made it a bit more understandable. It did for me, at least. ;) <p>TOK
Gorgon the Caretaker - Character in a childrens TV-show from 1968. ;)
ProboticsAmerica
Posts: 8
Joined: Thu May 19, 2005 1:01 am
Contact:

Re: Servo motor drive voltage

Post by ProboticsAmerica »

Ah - ok, The circuit I'm working on is not necessarily an H-bridge, but it probably contains similar either circuitry or programming. Its the chip inside a Hitec servo. The chip directly drives the motor movement. Unlike an H-bridge, the voltage applied to the motor's positive and negative poles is never zero (relative to ground) while the servo is plugged in. For some reason - and it might be transistor saturation related - "non moving" for a Hitec servo means apply 3 volts to both poles of the motor. When the chip wants it to move it drops the voltage on one side and boosts the voltage on the other. It uses PWM while doing it. <p>The mystery is 1) Why is 3 volts applied to both motor poles to make the motor not move (instead of ground applied to both) and 2) what circuit or components could I use to translate voltages like:
Side One Motor Side Two Motor My application
+5v +2.3v + magic black box = +2.7v side one +0 side two
+2.3v +5v + magic black box = +0 side one +2.7v side two
L. Daniel Rosa
Posts: 400
Joined: Tue Dec 25, 2001 1:01 am
Location: Bellingham, WA
Contact:

Re: Servo motor drive voltage

Post by L. Daniel Rosa »

To ground both terminals of the motor, the lower arms of the H-bridge mest be energized- which takes power. And yes, there is alost certainly an H-bridge involved even if it is entirely within the IC.<p>As for why 3V in idle mode instead of 2.5V, that's an invitation to guess. Perhaps there is more leakage current on the upper arms which biases the (idle) output slightly ablve the midpoint. Maybe it's not idle but actively maintaining position, and the manufacturer found that centering the PWM at 60% duty cycle gave better results than 50%.
Gorgon
Posts: 325
Joined: Wed May 04, 2005 1:01 am
Location: Norway
Contact:

Re: Servo motor drive voltage

Post by Gorgon »

1)I think that was answered by:<p>" In idle mode (motor off) none of the transistors are active. The motor will 'hover' in the middle between +5v and 0v. Like the +3v you measure. "<p>2) That depends on how much power 'your application' needs to function?<p>Are you using the output from the servo driver without the motor? Do you drive it in parallel with the motor? Or, do you only use it as a logic or low power analogue output?<p>Does your +2.3v value indicate that this is the low (-) driver output to the motor when activated?<p>TOK
Gorgon the Caretaker - Character in a childrens TV-show from 1968. ;)
ProboticsAmerica
Posts: 8
Joined: Thu May 19, 2005 1:01 am
Contact:

Re: Servo motor drive voltage

Post by ProboticsAmerica »

I checked the oscilloscope and its actually not a 50% pwm - its a hard 3v. When the motor kicks in it sends PWM pulses up (more positive) on one side and down (toward ground) on the other. It does this when the motor is in the circuit or removed. <p>What does 'hover' mean in this context? I took it to mean 'floated' which would suggest that if I put a load between the measured voltage and ground there would be little if any current. When I actually did this there was substantial current flow and the 3v was maintained. <p>As for application, right now I am using the output from the servo chip in parallel with the motor to drive a modified H-bridge to control a 12 volt 2A motor. One servo motor lead drives one side of the bridge and the other drives the other. When the chip makes one lead higher voltage = movement that direction. While performance is ok, I would like it to be better. The H bridge is transistor based rather than an IC as I have yet to understand the PWM coming from the servo well enough to convert it into PWM in a format that could be fed into a commercially available H-bridge IC (which brings us to these posts). It seems to me working with the servo would certainly be easier if one side of the servo motor was given 0-5 volts while the other was given ground and then that setup was switched to move the other way (like a standard H-bridge), but that's not apparently the way its working. <p>I've also noticed a fair amount of heat generation and was wondering if I have an issue with the 3 volt idle voltage turning on both sides of my bridge at the same time.
Gorgon
Posts: 325
Joined: Wed May 04, 2005 1:01 am
Location: Norway
Contact:

Re: Servo motor drive voltage

Post by Gorgon »

<blockquote><font size="1" face="Verdana, Helvetica, sans-serif">quote:</font><hr>What does 'hover' mean in this context? I took it to mean 'floated' which would suggest that if I put a load between the measured voltage and ground there would be little if any current. When I actually did this there was substantial current flow and the 3v was maintained. <hr></blockquote><p>Sorry about the wording. I think you can compare this to an op-amp on each side of the motor. Maintaining the +3v is also functioning as an active brake on the motor between the pulses. I agree with you that this is not an H-drive in a normal way.<p>The RC-receiver is not putting out normal PWM either. The pulse train is based on 1-2mS pulses each 20mS. The servo driver is comparing the length of the pulse to the position of the servo. When there is a difference, it pulses the motor the correct direction. In this way you can have up to 50 pulses each second (Sometimes more if the 20mS is shorter. I think it can be as short as 12mS, depending on the radio system.)<p> <blockquote><font size="1" face="Verdana, Helvetica, sans-serif">quote:</font><hr>I've also noticed a fair amount of heat generation and was wondering if I have an issue with the 3 volt idle voltage turning on both sides of my bridge at the same time. <hr></blockquote><p>Since I don't know your circuit solution I can not say what is causing your heat, but you need to create some sort of 'dead-band' for your driver around 3v. Like: Turn on and OFF your low (-) driver at 2v and your high (+) driver at 3.5v, or something like that.
A couple of comparators on each side will do the trick.
By the way, how is your protection for reverse induction voltage from your motor? You need to bleed this off with some diodes. If not you will destroy your transistors.<p>TOK :)<p>[ May 21, 2005: Message edited by: Gorgon ]</p>
Gorgon the Caretaker - Character in a childrens TV-show from 1968. ;)
Post Reply

Who is online

Users browsing this forum: No registered users and 32 guests