View Full Version : CIM Motor Capacitors
ProgramLuke
22-03-2011, 13:53
Where is a good place to buy capacitors for the DC CIM Motors allowed under:
A one microfarad (1 μF) or less non-polarized capacitor may be applied across the power
leads of any motor on your ROBOT (as close to the actual motor leads as reasonably
possible)
Where is a good place to buy capacitors for the DC CIM Motors allowed under:
Locally? A Radio Shack. You may be able to find another electronic component store near you as well.
Online? Mouser, Digikey.
May I ask why you're doing this. Nothing wrong with it, but I'm just curious if you're having problems with the cRIO or the Dlink bridge cutting out.
Al Skierkiewicz
22-03-2011, 14:45
Luke,
I know that many teams do not think of moving the motor controller closer to the motor it controls but this is a good way to reduce motor noise and allow the motor to be connected directly to the controller without having to add wire to the motor. The CIM motor has not been known as a noise producer as compared to the FP for instance.
ProgramLuke
23-03-2011, 12:38
May I ask why you're doing this. Nothing wrong with it, but I'm just curious if you're having problems with the cRIO or the Dlink bridge cutting out.
We are indeed having issues with communications. We are using all 4 CIMS on our drivetrain but if we don't have a battery that idles around 13V We lose comms if we try to accelerate quickly. What happens is that the Dlink drops out and we have to wait for it to restart to get communications. Are other teams having issues like this and do you have any solutions?
DarrinMunter
23-03-2011, 12:44
Do you have the power converter (12v in, 5v out) connected to the 12v output at the bottom of the PD board? And then the 5v from the converter feeds the radio.
ProgramLuke
23-03-2011, 12:54
Yes the Dlink is powered by the converter.
DarrinMunter
23-03-2011, 12:55
I found a PDF that shows how and where the converter gets connected.
DarrinMunter
23-03-2011, 12:59
Can you block up your robot so you can run the motor and keep the robot in place?
You might want to take a voltage measument from the 12v coming from the PD board. If its going below 12V you have a problem with the PD.
That output has a special built in power supply that keeps the voltage steady at 12v - no matter what the battery is running at. ~ unless the battery drops way low, 7-8 volts.
billbo911
23-03-2011, 13:01
Do you have the power converter (12v in, 5v out) connected to the 12v output at the bottom of the PD board? And then the 5v from the converter feeds the radio.
Yes the Dlink is powered by the converter.
Luke,
Darrin's specific question included "Do you have the power converter (12v in, 5v out) connected to the 12v output at the bottom of the PD board?" for a very good reason.
If it is connected there, the regulator in the PDB will provide +12vdc even when the battery voltage dips below +12vdc. If the converter is not connected there, but is connected to the standard Wago connections, you will loose the +5vdc out of the converter when the voltage dips.
So, do you have the power converter (12v in, 5v out) connected to the 12v output at the bottom of the PD board?
ProgramLuke
23-03-2011, 13:15
That's one of the issues. Unless the battery is absolutely fully charged and reading 13v+ we will get a drop to 8v or less. At that point we get loss of communication. We just did a run reading 12.8 volts and it immediately dropped to 8v- and lost communication.
ProgramLuke
23-03-2011, 13:19
Thanks for the guidance. We had it coming out of the PD through a regular circuit. We will get that changed right away and see if it makes a difference. Really hoping this is the issue.
Thanks again!
That's one of the issues. Unless the battery is absolutely fully charged and reading 13v+ we will get a drop to 8v or less. At that point we get loss of communication. We just did a run reading 12.8 volts and it immediately dropped to 8v- and lost communication.
You're dropping too much. I'd look at the main power path from the battery all the way to the Power Distribution board. Bad crimps, looks nuts, nicked wire, undersized wire. You have too much resistance somewhere, or else some bad batteries.
A clue would be to feel all the wires from the battery terminals to the PD board. If any are warm, then there's your issue.
Also, 6AWG wire is a tiny bit small for the loads you might be putting on it. If you're running more than 2 feet of battery cable, I'd recommend bumping it up to 4AWG.
Also check for mechanical issues. Can you rotate your wheels pretty easily? Are they tough to spin? If they're tough, you might be binding too much and pulling too much current.
DarrinMunter
23-03-2011, 13:21
Bad battery? We've had to get rid of about 2 of ours this year, because of that.
If your using a good battery, then you have excess load somewhere.
Run your motors one at a time and measure the battery voltage, the motor with the largest battery voltage drop is your problem. It could be the motor itself, or a speed controller going bad.
DarrinMunter
23-03-2011, 13:23
Thanks for the guidance. We had it coming out of the PD through a regular circuit. We will get that changed right away and see if it makes a difference. Really hoping this is the issue.
Thanks again!
We did the same thing two years ago! We were at the BMR, when a mentor from WildStang help us out.
Thanks for the guidance. We had it coming out of the PD through a regular circuit. We will get that changed right away and see if it makes a difference. Really hoping this is the issue.
Thanks again!
Yep, that's the main issue. The 12V coming out of the bottom of the PD board is a boost circuit. It is supposed to operate down to a battery voltage of 4.5V.
http://usfirst.org/uploadedFiles/Robotics_Programs/FRC/Game_and_Season__Info/2011_Assets/Kit_of_Parts/Power_Distribution_Board.pdf
Page 6 of 10.
theprgramerdude
23-03-2011, 14:35
We are indeed having issues with communications. We are using all 4 CIMS on our drivetrain but if we don't have a battery that idles around 13V We lose comms if we try to accelerate quickly. What happens is that the Dlink drops out and we have to wait for it to restart to get communications. Are other teams having issues like this and do you have any solutions?
Yes, the solution is to not accelerate so quickly. Motor noise isn't causing the problems you are experiencing, it's safe to say. What's happening is that while trying to accelerate quickly, your send out far too much current to your motors, which results in a massive voltage drop on the system. This voltage drop will cause things like the radio, and possibly Crio, to shut down. You probably either need to instead tell your drivers to be easier on the throttle, implement a software safeguard against reckless behavior, or some combination of the two.
Additionally, partially charged and old batteries just like to exacerbate the problem described above. While a new one may be capable of putting out enough current to keep the voltage sufficiently high, a bad one will surrender.
Al Skierkiewicz
23-03-2011, 14:38
We did the same thing two years ago! We were at the BMR, when a mentor from WildStang help us out.
Me?!?
DarrinMunter
23-03-2011, 15:12
Me?!?
It was Raul,
He said that he seen a couple of teams make the same mistake.
Blew us away, because everything was working on the test frame. When we moved everything over to Lunacy robot some how that got changed.
Lesson 1: Move over stuff one thing at a time.
Lesson 2: Take picutes before hand. Use them as a reference.
Lesson 3: Labels! What to where.
You can learn a lot from making mistakes - even more from fixing them.
a new one may be capable of putting out enough current to keep the voltage sufficiently high
A healthy battery can put out enough current to trip the main breaker without browning out the cRIO.
Al Skierkiewicz
23-03-2011, 17:56
A healthy battery can put out enough current to trip the main breaker without browning out the cRIO.
Not exactly, the Crio power supply on the PD will continue to output 24 volts until the battery voltage falls below ~4.5 volts. The analog jumper in module 1 monitors the battery voltage and when it falls below 5.5 volts, the Crio output is inhibited to allow the battery voltage to rise with no load. Sort of a no-battery backup battery.
Not exactly, the Crio power supply on the PD will continue to output 24 volts until the battery voltage falls below ~4.5 volts. The analog jumper in module 1 monitors the battery voltage and when it falls below 5.5 volts, the Crio output is inhibited to allow the battery voltage to rise with no load. Sort of a no-battery backup battery.
How much current does it take to trip the main breaker.
theprgramerdude
23-03-2011, 18:09
How much current does it take to trip the main breaker.
A lot. It can withstand 600 amps for a second, if I recall. I'll try to find the curve.
Al Skierkiewicz
23-03-2011, 18:09
In a few seconds the main breaker would take ~720 amps to trip. The NP18-12 battery is only speced at 500 amps out into a short circuit at which point (at full charge), it is dropping 5.5 volts just across the internal resistance of the battery ignoring any other losses internal to the battery. 500 amps will drop another volt across the #6 wiring ignoring the resistance of the main breaker, Anderson connector, crimps and terminals.
In a few seconds the main breaker would take ~720 amps to trip.
If you have a link to the spec for that could you post it please.
The NP18-12 battery is only speced at 500 amps out into a short circuit
So it is. Interesting. The 2009 battery spec says 720 amps for 5 seconds.
bobosalad
23-03-2011, 19:24
What are you running besides your four cims? Did this problem show up during Duluth? If it did, you guys managed it pretty well. After the communication dropped, were you ever able to get it back?
ProgramLuke
23-03-2011, 21:21
We are also using 2 of the smallest banebot's, a window motor, and the compressor.
Al Skierkiewicz
23-03-2011, 22:16
The main breaker sheet is on CD already. You just have to search to find it. The actual spec is 600% over current for a few seconds and several minutes at 200%.
Tristan Lall
23-03-2011, 22:28
If you have a link to the spec for that could you post it please.The main circuit breaker is Terminal Supply's CB3-SM-120 (http://terminalsupplyco.com/Store/Product.aspx?pc=CB3-SM-120), which is actually a Cooper Bussmann 18X series breaker (probably a 185120F). Their characteristics are here (http://www.cooperbussmann.com/pdf/fdee239d-3e78-4df3-832f-fea4236a26bf.pdf) (from the manufacturer, see page 32 printed/34 PDF), or here (http://connectorconcepts.com/products/cb_trans/_files/18x.pdf) (directly) or here (http://www2.usfirst.org/2005comp/Specs/120acb.pdf) (for an older spec sheet from FIRST).
According to that diagram, some breakers may last for up to 3 seconds before tripping at 720 A.
The main breaker sheet is on CD already. You just have to search to find it. The actual spec is 600% over current for a few seconds and several minutes at 200%.
The Terminal Supply P/N CB3-SM-120 that AndyMark markets as their P/N am-0282 is spec'd by the manufacturer Cooper Bussman at 440 amps nominal for 5 seconds.
I see Tristan beat me to it. Nominal 440amps, range 352..528 amps for 5 seconds.
The Terminal Supply P/N CB3-SM-120 that AndyMark markets as their P/N am-0282 is spec'd by the manufacturer Cooper Bussman at 440 amps nominal for 5 seconds.
I see Tristan beat me to it. Nominal 440amps, range 352..528 amps for 5 seconds.
Does this alarm anyone?
We're using this thing as a switch, not a breaker.
Does this alarm anyone?
We're using this thing as a switch, not a breaker.
Yep, that explains why our "off switch" has a red reset button. :P I guess it's just another safety feature built into the system.
This is a great thread.
I'd like to add a note that 1712's and several others converters 12v - 5v for the Dlink, failed.
It failed by providing only 4.1 volts.
Once you have tried all of the above advice, try tracking the 5v in at the Dlink and make sure its OK all the time.
Dropouts are happening in other areas of the robots this year and they are hard to diagnose, so look at the threads on Jaguars and encoders next as well.
theprgramerdude
26-03-2011, 15:01
Does this alarm anyone?
We're using this thing as a switch, not a breaker.
It is a breaker - a short circuit breaker. The other breakers down line from the main one are the real ones that do all of the work to prevent component damage. Ever wonder why we end up having like a pound worth for the whole system?
One thing I learned early on, in robotics what you get is what you don't see.
Use meters to reveal things.
Voltage is easy.
DC Current is thought to be hard, but check out the Kussmaul Electronics 091-8 strap on ammeter. (inductive ammeter)
If you are overloading the main or subcircuits, this will pinpoint the problem for less than $20 and without disconnecting any wires.
Snap-on Tools has similar meters, but they are pretty expensive.
Daniel_LaFleur
26-03-2011, 17:58
Does this alarm anyone?
We're using this thing as a switch, not a breaker.
Not really.
The snap action breakers are the real protection circuits while the 120 amp breaker acts as a single point emergency switch for anyone who needs it and as a catastrophic short protection.
In the 9 years I've been in FIRST, I've never seen one trip, not have I seen a need for it to trip (except this years 1519 minibot ... which has no breaker :P )
Its required by the rules.
What I would like to say here is that in my 11 years with FIRST I have seen three of these fail.
Only one was a complete open and you could feel that it was broken beacuse the lever was floppy.
The other two would fail to 'snap' on, although they looked to be on,
about 2 out of 5 times, which created havoc until we figured it out and replaced it.
Onboard diagnostics ? Voltmeters ? Operators watching the displayed voltage ?
Al Skierkiewicz
27-03-2011, 06:49
Guys,
The breaker is rated for switch use, so in our application it is fine. Please remember where it is in the circuit. If there would be a short on the PD, or any wiring feeding it, the main breaker is intended to trip to prevent fire. While I have a had few reports of teams tripping this breaker, I have not personally witnessed any that tripped on the field. I have seen several over the years that were damaged in manufacture and needed replacement.
vBulletin® v3.6.4, Copyright ©2000-2017, Jelsoft Enterprises Ltd.