Understanding amp ratings.

[crchisholm]

So, I have five parallax standard servos. I want to run them at 5 volts and in some circumstances, all five may be called on to move at once. I have a five volt regulator (radio shack 276-1770) and a wall wart (24 volt - 550ma) - I also have smaller voltages but I thought it might be hand to have 24 volts lurking out there. Now being the rocket scientist that I am, I'm thinking that I can feed all five of these servos from the output power off of the single 5v regulator and since the specs say the regulator can handle up to 1a then as long as I don't allow the combined throughput to go over 550ma, I should be ok. Any comments on my thinking here, no matter how unflattering, would be welcomed.

At this point I have never burned my house down, melted my computer or sent to east coast of the US into darkness, but have had my share of parts frys. Just trying to find a way not to do that.

Thanks in advance

 

[Ian Rogers]

The 5 servos CAN equal 1A.. 200ma max in each... The biggest problem here is the 5 volt regulator will be working extremely hard regulating 5 volts down from 24v... My guess is that the regulator will barely be able to keep a constant 500mA let alone 1A... Make sure you have a large heat sink.

If you can swap the wall wart for a 7 - 9v output you may fair better.......

 

[crchisholm]

Thank you for your quick reply.

I also have a 9vdc/1000ma and a 15vdc/1.4a. If I drop back to the 15 so that I can also run a small 12v dc motor in the same circuit (12 volt regulator of course), would that be safe, or would you recommend I find another source for the motor and use the 9v/1000ma. Also, do I need to have a 5v regulator per servo, or can I safely get away with less. (thinking space rather than cost)

 

[Sceadwian]

The nominal current rating of a servo is not the only value you need to pay attention to. The full load current (what it will practical draw from your load) is even more important, and so is the stall current for fault conditions, personally I would test the servo's under a full stall (lock the servo arm in a vice to keep it from moving and measure the current draw under full deflection in both directions) I can guarantee that even 2 or three of these servo's under heavy load will draw more than the 5 volt regulator can put out. There is a HUGE difference between current draw when you're testing a servo on a bench with no load on it's output arm compared to when you put it into a device and actually start doing some real work other than overcoming the friction and mass of the servo arm and geartrain. You won't see it on a meter but even under light load the initial current the servo's draw is much higher than you might think.

24 volts regulated down to 5 volts is going to drop 19 volts. At 1amp that dropped voltage will be 19 watts of pure heat. Without a really good heat sink and active fan the regulator will overheat and fail, or more probably go into shutdown (if it's thermally protected)

Even if under load the current draw of the servo's is relatively low you have to take fault conditions into account, otherwise if something goes wrong the regulator is going to go up in smoke, or just as wasted go into shutdown all the time.

Thermal considerations are much more important than most people in electronics like to think about because it's something that has nothing to do with electronics directly, yet it is critically important to circuit lifespan and practical circuits in general.

 

[Ian Rogers]

All I can tell you is....... I use a 1mA linear rectifier on a .net board.... the current consumption is bordering on 500mA (complete with an LCD screen). If I use a 1A 12v transformer.... the screen will not boot up... If I use a 750mA 6v transformer all is well... the thermal properties of the regulator go out the window when it has to drop the voltage more than required....Remember that 7 volts at 500mA is being wasted... the regulator is basically a heater element... good if you need to heat the room


Sorry Sceadwain.... I posted this at the same time.

 

[Reloadron]

This is likely the data sheet for your servos. This is the part you want to pay attention to:

Technical Specifications
Power requirements: 4 to 6 VDC* (see Power Requirement Notes below)
Maximum current draw: 140 +/- 50 mA at 6 VDC when operating in no load conditions
15 mA when in static state
Communication: Pulse-width modulation
Dimensions approx 2.2 x 0.8 x 1.6 in (5.58x 1.9 x 40.6 cm) excluding servo horn
Operating temperature range: 14 to 122°F (-10 to 50°C)
Weight: 1.55 oz (44 g)
*Power Requirement Notes
Futaba specifies 4-6 VDC for this servo. However, we find that this servo is tolerant of a 9 V battery for
short periods of time when there is no load, as used in some activities in the Stamps in Class series of
tutorials. (Slight jittering may be observed when batteries are fresh; this does not cause damage). Do
not use this servo with an unregulated wall-mount supply, or a regulated wall mount supply exceeding 6
VDC.

This all goes back to exactly the point Ian makes. First a standard LM7805 (the Radio Shack Regulator) really will not be a good choice. Especially considering a 24 volt wall wart supply limited to about 550 mA. Note the currents in the data sheet are for the servo motor unloaded. I would be looking for a 5 volt maybe 3 amp regulator along these lines, as well as a beefier supply to drive it with a lower voltage out, like as Ian suggested around 12 volts.

Now if you want to roll your own Radio Shack does offer this 12 volt 3 Amp transformer. They also sell bridge rectifiers like these that would handle the load. Unfortunately I don't think they offer a 3 amp 5 volt regulator.

Anyway, I agree with Ian for the reasons I have mentioned.

Ron

 

[crchisholm]

I have learned a lot...or more accurately, learned that I don't know a lot. I have probably 30 wall warts and I am going to see if I can find a better fit (or buy one if I can't) I am also going to see if I can figure out now to test the load on the servos. (yes I am very green). I don't have any specific project that this is for except to help me understand what I do and don't know.

Bridge rectifiers is another vary dark area for me. But now I have a few slugs for my Google gun.

Thank you all very much. All 14 of my brain cells are clacking right along.

Charlie

 

[Reloadron]

Anytime!

Ron

 

[Sceadwian]

ReloadRon, your assumed data sheet is useless without a given stall current. If more than the rated current is drawn at any point from more than a couple servos, which WILL occur during stall or loaded conditions those datasheet asumptions are irrelevent. The regulator will go into overload, shut down and the entire circuit will restart after it's recovered until it fails again, repeatedly.

 

[Reloadron]

Sceadwian, parallax markets what they call a continous rotation servo which is this one. The OP also states quite clearly:

So, I have five parallax standard servos.

As far as I know and could find Parallax markets but one "standard servo" as they call it and that is this one and the one I linked to. I have one lying here on the table. These are small plastic servos that weigh in at a few ounces. My "assumed" data sheet it would seem is the data sheet. How much current do you think this thing will draw in a locked rotor or stall?

I agree with your suggestion:

The full load current (what it will practical draw from your load) is even more important, and so is the stall current for fault conditions, personally I would test the servo's under a full stall (lock the servo arm in a vice to keep it from moving and measure the current draw under full deflection in both directions) I can guarantee that even 2 or three of these servo's under heavy load will draw more than the 5 volt regulator can put out.

Though I hardly see need for a vise when you can easily just hold the thing still. Yeah, real servo motors I agree but this is a toy. No, they don't list torque specifications, travel speed or slew and they don't list locked rotor or stall current as it is a toy, not something anyone is about to toss into a heavy duty commercial application. I suggested a 3 amp circuit regulating off about 12 volts. The LM7805 or the regulator I suggested will shut down for over current or for that matter fuse the supply line. If I thought it was worth the time and effort this weekend I could run this thing and collect data but I don't expect to see much. This is a servo motor that Radio Shack retails for $14.99 and can be had online for a few bucks less.

While I agree with what you nicely covered I just don't see the need to beat this toy up collecting data that will likely be not really necessary. Granted, that is just my line of thinking on it. If you want one of these things to screw around with let me know and I'll send you one absolutely free.

Ron

 

[Sceadwian]

They list running current with a tollerance though Ron, not including the stall current is just silly. Stall current could be as much or more than 5 times the no-load, even on a toy.

 

[Reloadron]

They list running current with a tollerance though Ron, not including the stall current is just silly. Stall current could be as much or more than 5 times the no-load, even on a toy.

I don't doubt that for a heartbeat and a 1 amp stall current would not surprise me. Sceadwian, these things pour in from China by the millions. They are toys. The data sheets generally suck. They are not about to include all sorts of data like some of the parameters we mentioned. So winging it and short of testing I figured a 3 amp 5 volt supply for 5 motors would be adequate. Fused also. So while I agree not including the stall current is silly I don't expect to see a really good data sheet on a $13 USD (likely 3 bucks) made of plastic from China. I do expect to see it on a data sheet for a servo motor costing a few hundred bucks. Even if they drew an amp in locked rotor, I would not suggest a power supply of 5 amps for 5 motors.

Ron

P.S. Sending you some more snow!

 

[Sceadwian]

They are not about to include all sorts of data like some of the parameters we mentioned. So winging it and short of testing I figured a 3 amp 5 volt supply for 5 motors would be adequate.

No need to guess if you just test it, which takes only a moment with even a cheap multi-meter... They already wrote down the no load current.. It would have taken less than 5 seconds to measure the stall current.

My only assumption is an esoteric one; all assumptions of fact lead to a decrease in learning. The poster will have to determine what his power supply current requirements will end up being as he has no purpose. Hiding the dirty details of what all could be going on in any given situation will set someone up that's trying to learn to fail, if they can't absorb the information or research on their own as crchisholm has already demonstrated they are doing there is no amount of simplification that can alleviate the lack of knowledge.

You can't dumb down science, you can only try to explain it better without avoiding the pesky details which are in fact what it's all about.

 

[Reloadron]

OK, so let's assume for now the stall current is 1 amp. He has 5 servos. What would you suggest for a power supply?

Ron

 

[Sceadwian]

There is no basis to recommend a suggestion Reloadron, by crchisholm's own statements there is no application for what he's asking about, which means there is no power supply to suggest because there is no application. It's impossible to engineer something for every purpose, or something for no purpose, engineering is base on application.

The supply could depending on the application need to supply as little as the minimum no load current of one servo (and the rest would be powered off to avoid the 15ma idle current) or it could be as bad as the full rated stall current 100% of the time. It always depends on the application, if you don't know your needs, you can't find a way!

 

[Reloadron]

What I am asking is if a motor, any motor has a stall current of 1 amp and I have 5 such motors how much current should my supply be capable of delivering? Stall current is obviously the max current the motor can ever draw and it will only draw that till smoke comes out. So having 5 motors with a 1 amp stall current should I design with a 5 plus amp supply? Shouldn't matter what the application is. OK, let's use the vise. Let's anchor the motor in the vise. These little motors have arms extending from shaft center. So should I hang weights on the thing and see where it fails to lift the weights? This would be dead weight testing the motor. Going to need a small precision weight set. You would also need to know how much force the motor would deliver at what current.

Running the motor into a stall as you suggested and measuring the current will tell you one thing. It will tell you the stall current and that is it. No more and that is as good as it gets short of dead weight testing the motor and measuring the torque versus current. The data sheet on that motor is as good as it gets. They give no load currents which granted isn't much to work with. So you tell me how to calculate the needed supply? Short of using a mountain of equipment the average visitor here likely does not have. We have the no load current and we can get the stall current, so now what do we do? Engineering? Generally I look at the application, making damn sure I know the requirements of the motor and then I buy the motor suited for the application. I generally don't place the cart before the horse as it pisses off my superiors.

Ron

 

[Sceadwian]

Wow reload you must never have been a hobbyist on a shoe string budget. You can measure the no load current stall current and stall torque with a 15 dollar kitchen scale and a 5 dollar multimeter, using nothing but a coat hanger as a control arm. If you want to go advanced you could easily determine the speed slew rate both no-load and under load with the same coat hanger and a pocket full of change using a cell phone camera or a webcam. Wouldn't take more than an hour and then you have real data not guesswork.

Application is critical to answering the question not irrelevent. If all the poster wants to do is test servo movement and coordination then a 1amp regulator is fine because they'll never be loaded. If however they are for robot wheels in contious rotation you would rate it for stall load probably with some headroom for comfort.

 

[Reloadron]

Sceadwian, my entire early life I used kitchen table methods. Today with life being better I try not to look back. Maybe this weekend I screw with that little motor out of curiosity. I am curious about the $15 kitchen scale though? Use the arm to push down as in compression?

Hey, I never said the application was irrelevant. What I said was everything (all the numbers) about the application need to be known. Look, for the better part here I agree with you. My point was that I can't see the worry in using for example the 3 amp supply I suggested. Then too, hell, now that I think about it an old ATX form factor PSU would give a pile of 5 volt current. Cheap too.

Ron

 

[Sceadwian]

The three amp supply is fine, for practicality. But like the OP said he currently has no purpose in mind, so if he did use the 3amp one and there's more torque required the the 3amp supply will fail. Ballpark estimates of what might be okay are fine but should be qualified carefully to avoid sending him down the wrong path. That's why I think it's important to measure the stall current, no more guessing at that point.

Yeah the scale works nice, can be a little tricky to rig up but it's got a 1 gram resolution with a 2kg weight limit. for bigger motors just make the arm longer and stiffer. Just keep the axis of motion out of the plane of gravity and the arm doesn't effect the reading. The scale can be calibrated cheaply with freshly minted pennies. They're 2.500 grams each best guess I've been able to find for tollerance of the weight is .05 grams which is way more accurate than a hobbyist needs.