Title: Robotic Electronics
Description: Senior Design project
Ducc - December 5, 2004 05:03 AM (GMT)
Hey guys, here's the deal. Doug suggested I start looking for funding for my senior project (seniors have to do a project, and usually try to get a company or the school to pay for it). So I found a company that is willing to let me build a robot that tests brake parts built to micrometer tolerances. They have some machines that work ok now, but they're not very accurate.
So I (and friend) took the job, and it just so happens that the same company is offering an internship for a student to come in and write LabView software to control the robots they already have. The software they use is fine, except it doesn't offer much room for future expansion, so they'd like to convert everything over to LabView.
As luck would have it, the lab I worked in this summer had some equipment controlled by TERRIBLE LabView software. I finally got fed up with the software by the middle of the summer, learned LabView, and rewrote the stupid software. In the process, I got a pretty good understanding how to use LabView (especially data acquisition and control). Naturally, I have a LOT to learn, but I know my resources. Anyway, I took the internship at the company for two reasons: first, it'll help me learn LabView better. Second, it fits right in to our senior project; when I finish the software, it should be easy to edit to operate with our new robot.
Sounds good, right? Yes it is. I asked an engineer at the company if he could give me some sensors (LVDT’s) so that I could play with LabView and see if I could get it to work with their existing equipment. He gave me about $500 worth of random electrical stuff. Here's a list.
- An Advantech 32 bit I/O card (PCI-1730)
- SOLA Hevi-duty (their spelling, not mine) 24 volt, 2.5 amp power supply
- Random on/off switch
- Proxy sensor
- Power cable
- A board that hypothetically will connect to the Advantech card, to allow easy wiring. (note: I don't have the cable that connects the two, and I hear it costs about $150)
- Three Omiron G2R-2-S 24VDC relays
- One fuse of some kind (should protect the equipment, so he said)
- Some other random crap that you don't care about
- Two pneumatic valves made by IR ARO: one "four way" and one "two way"
(No LVDT’s because he doesn’t have a spare yet. More on the way.)
I have been doing some research on these things, and I've got a pretty good idea how they connect and what they do. We have a robotics lab and professor at school that knows a lot about this stuff (actually, every professor I've talked to seems to know what this stuff is, so it must be commonplace stuff in the engineering world).
In any case, I'm looking for a good basic electronics website that can explain some of these things in simple terms. I'm sick of pouring through electronic wiring schemes and trying to decipher what they do. Do you have any help websites for me?
I have one semister and one summer to finish the LabView project, and three semisters to finish the senior project. My goal is to finish this LabView thing by the summer, so I can get a mechanical engineering internship. I think I can do it, but it's going to be tough.
_Z_ - December 5, 2004 05:53 AM (GMT)
Ducc - December 5, 2004 02:17 PM (GMT)
As in, he handed me a relay and said "This is a relay" and I said "WTF is a relay?" (in my head). And as in figuring out later that "Woah! The power supply clips to the metal bar, just like everything else does! They must have planned it this way!"
You know, parts help. I just assumed the plastic things he was plugging stuff into were there to help clamp down wires, and it turns out I was right. But that kind of information would be nice to read somewhere.
I just assume play with things, but I can't do anything until I get the connector cable, so I may as well read everything I can. Thanks for any input you have.
_Z_ - December 5, 2004 03:05 PM (GMT)
A relay is nothing more than an electric switch.
A bar that conducts electricity is called a bus bar (usually ground). Not all electricity runs through wires.
Plastic wire clamps are mechanical, not electronic. -Or are you talking about a connector?
I really need more description. If you need help with component identification, or schematic representation, that's one thing. If you want to learn the basis of electric theory, that's another.
Feel free to take any part of this off-board if you want. You have my e-mail.
Ducc - December 5, 2004 04:10 PM (GMT)
Okay, I need help with component identification. And actually, more than that, I need help with knowing what the heck the component does, how it works, how it's used.
Your description of a relay is exactly what I'm looking for. It'd be nice to get into a bit more details, like describing how the coil makes a magnet that pulls the switch over, and that "a typical relay has the following wired connections...". See what I mean?
The plastic wire clams are electrical in the sense that they are used to wire up an electrical system. Even the metal bar that everything clamps to probably has a name that I'd like to know.
_Z_ - December 5, 2004 05:41 PM (GMT)
Any current through a wire emits a charged field. When that field cuts through a ferrous material, it generates magnetism. Hence, an electro-magnet.
A relay can be wired in a number of ways, depending if the coil is tapped, and how the contacts are arranged. There is no standard. This is where mfr's data sheets really come in handy.
The same goes for most other components. Take fuses for example. There's the plastic kind you find in your car, or the glass tubes with metal caps that you find in some consumer electric products. But I could show you a fuse that looks like a surface mount 'chip' resistor or capacitor. Or one's that look like a piece of wire insulation. Some fuses are as big as sewer pipes. Batteries are another good example. They come in all shapes and sizes, from one's barely larger than the head of a pin, to others as big as a storage trunk.
I saw this big black box once I could have sworn was a truck battery. Turns out it was a power resistor. And there are chip resistors small enough to put a couple hundred in a thimble.
As you can see, pointing you to a website for various types of component identification can be a daunting challenge. If you can give me examples (mfr & part number would be ideal), I can point you in a specific direction, and fill in any gaps left behind- as to how it works, what it does, etc.
Ducc - December 5, 2004 08:26 PM (GMT)
Well, that sucks that there's no standard. I may just have to play with them.
As I said, I think I'm at the point where I can wire this up and make it work. I just wanted to know if there was an electronics website that you've used in the past, or still use.
But now that I know you know things about this, I'll probably end up posting something here in the future. :)
_Z_ - December 5, 2004 08:43 PM (GMT)
:lol: When I took all this stuff, there was no internet! So it might surprise you I don't have any websites I use on a regular basis. I did it the 'old-fashioned' way- books. On that, I can recommend dozens of titles, if not hundreds.
That said, I can (and have) plumbed various websites through the more recent years, and can pretty much spot the good one's from the not-so-good. Trouble is- most are pretty specific. It would be silly to send you to a site describing say- deMorgan's law when your looking for h-parameter AC analysis.
Ducc - December 18, 2004 07:48 PM (GMT)
Okay Z, check this out.

I put those power cable wires into the power supply. Did I do that right? I have no idea in which order they should go. They're labeled ground, N and L. (I put the green cable into the N and white into L and black into ground).
_Z_ - December 18, 2004 10:33 PM (GMT)
DON'T PLUG IT IN LIKE THAT!! :o
Black is line (hot)
White is neutral
Green is ground.
Depending on the design of the power supply, you may very likely damage it!!
You have the line (hot) going to ground
The ground going to neutral
And the neutral going to line (hot)
That's a switching power supply. They don't usually like getting hooked up wrong.
_Z_ - December 19, 2004 01:29 AM (GMT)
Here's a
link to the product's webpage, and the
specifications (pdf)
Ducc - December 19, 2004 04:42 AM (GMT)
HOLY CRAP!
I did plug it in like that. Then I reversed the white and green and plugged it in again. *winces*
Jeez, your reply scared me. I swapped them to your recommendations, and it came on. :)
See why I need a reference website!? :)
Based on this experience, I feel I need to ask many more questions instead of just plugging things together and watching what happens. Especially since power is now involved. I already tried to take a wire from a positive and negative terminal, and touch it randomly to a solenoid, but it didn't do anything. I am fairly sure I need to connect the two "switch" leads at the top of the power supply to make it enable the output leads... is this correct?
_Z_ - December 19, 2004 05:18 PM (GMT)
| QUOTE (Ducc @ Dec 18 2004, 10:42 PM) |
HOLY CRAP!
I did plug it in like that. Then I reversed the white and green and plugged it in again. *winces*
Jeez, your reply scared me. I swapped them to your recommendations, and it came on. :)
See why I need a reference website!? :)? |
I think throughout the course of this project, you're going to need a LOT of reference websites! :blink:
| QUOTE |
| Based on this experience, I feel I need to ask many more questions instead of just plugging things together and watching what happens. Especially since power is now involved. |
I couldn't agree with you more. :thumbsup: I use the old carpenter's addage: Measure twice- cut once. If you're not sure how something's connected, find out before you arbitrarily connect power. And then double-check it. Make sure it's not wired wrong. At best, nothing will happen, or components will get damaged/destroyed. At worst, you could be writing your obituary.
| QUOTE |
| I already tried to take a wire from a positive and negative terminal, and touch it randomly to a solenoid, but it didn't do anything. I am fairly sure I need to connect the two "switch" leads at the top of the power supply to make it enable the output leads... is this correct |
No, the 'DC OK' terminals are a signal switch to control other devices. But it shouldn't hurt anything if you short them together.
Luckily, this PS is protected against open loads and short circuits. Switching power supplies usually need a load to operate. If there's no load, they latch at zero output. This condition is maintained until power is removed from the power supply, wherein it resets. You might want to try connecting a 1200-1800 ohm 1/2 watt resistor across one of the (+) and (-) output terminal sets to load the supply to keep it operating.
Give me some info (i.e. mfr/part#) on the relays, and I'll get you a hook-up diagram.
I also recommend (if you haven't done so already), go to a Radio Shack, or home improvement center and get yourself a basic multimeter. Analog or digital- doesn't matter. Make sure it has AC volt, DC volt, and resistance (ohms) capabilities.
Ducc - December 19, 2004 07:55 PM (GMT)
I saw one at Radio Shack and almost bought it, but it was 50 bucks. We have 5+ in the lab where I work, so I may borrow one for a while.
I got the power supply to make the relay switch, but then I realized that it was switching because I gave it a lot of power. So I hooked up a 50 ohm resistor and an LED, and it lit just fine. I calculated the resistance needed to power an LED, choose the correct nearest resistor, and it worked. But then I took out the resistor to see how bright I could make it, and it blew. Damn. :) I have more. No loss.
Anywhoo, now that I got the power supply working, I need to get the IO card working to control the relays. I assume here that I simply need to tell the card to send a 5 volt (or whatever it is rated for) output and wire it to the relay. Unfortunately, I'm still trying to figure out how the heck to do that. The manual for the card is confusing. I put a wire from IDI 0 to one relay contact and another from EGND to the other relay contact. Nothing happened. I'm still working with LabView to see how to make this work. Any ideas?
_Z_ - December 19, 2004 08:03 PM (GMT)
| QUOTE (Ducc @ Dec 19 2004, 01:55 PM) |
| Anywhoo, now that I got the power supply working, I need to get the IO card working to control the relays. I assume here that I simply need to tell the card to send a 5 volt (or whatever it is rated for) output and wire it to the relay. Unfortunately, I'm still trying to figure out how the heck to do that. The manual for the card is confusing. I put a wire from IDI 0 to one relay contact and another from EGND to the other relay contact. Nothing happened. I'm still working with LabView to see how to make this work. Any ideas? |
Need specs on the data card.
Are you wiring the contacts, or the coil? Wiring the card to the relay contacts isn't going to do much... wiring it to the coil will- if the card can sink the current. (?) Otherwise, you need a line driver.
(btw- LED's aren't current limiting. Like most semiconductors, you will blow them if you overdrive them. Follow the mfr's data on current drive.)
Ducc - December 19, 2004 09:17 PM (GMT)
Oh, the coil. I got the basic idea of the relay down.
I have an Advantech PCI-1730 32 bit I/O card.
_Z_ - December 19, 2004 09:52 PM (GMT)
Advantech has their pdf datasheet files under lock & key... I can't access them.
What might be happening is that the card can't source/sink enough current to pull the relay closed. -Just a guess. What kind of relays are you using?
Ducc - December 19, 2004 10:55 PM (GMT)
I posted the specs in the first post. The relays are Omiron G2R-2-S 24VDC.
You are exactly right about the card not supplying enough current. I can light an LED with the card, but not throw the switch. Of course, I don't REALLY know what I'm doing here, so I may have it hooked up wrong.
The only data sheet I could find for the Advantech card is here:
PCI-1730 I hope that helps?
So I'm going to need an amplifying circuit (or a "line driver"?) to make these relays work? What is a line driver?
_Z_ - December 20, 2004 08:24 AM (GMT)
| QUOTE (Ducc @ Dec 19 2004, 04:55 PM) |
| I posted the specs in the first post. The relays are Omiron G2R-2-S 24VDC. |
Ooops... :ph43r:
| QUOTE |
| You are exactly right about the card not supplying enough current. I can light an LED with the card, but not throw the switch. Of course, I don't REALLY know what I'm doing here, so I may have it hooked up wrong. |
Depends on what side of the card you're using. (more below)
| QUOTE |
| The only data sheet I could find for the Advantech card is here: PCI-1730 I hope that helps? |
Much, thanks. I couldn't manage it from this end. :shrug:
If I'm reading the spec sheet right, the 1730 has two types of outputs. One is TTL level (which is 5V logic, and won't drive the relays directly), and the other is what is called 'open collector'. This is the side you're interested in. Refer to Fig. 3-9 on page 32 for connection, paying attention to the cautions and limitations. That should work, the EGND can sink 50mA. What concerns me is what do you do if you want more than two relays energized. (?) Each one draws 21.8mA. (link to relay spec sheet below)
| QUOTE |
| So I'm going to need an amplifying circuit (or a "line driver"?) to make these relays work? What is a line driver? |
Only if you were to use the TTL outputs. You'd need to convert the 5V logic level to 24VDC operation through a buffer amp, or an opto-isolator, which is what it looks like the 1730 uses in it's isolated output side. A 'line driver' is nothing more than an amplifier to increase voltage/current to drive line loads- in this case, relays.
Relay data sheet (I'm assuming Omron, and not Omiron) (?)
Hope this helps-
Ducc - December 20, 2004 02:43 PM (GMT)
Sweeet. I'll look at this when I get home from work today.
I won't have the problem with the relays. I only have three, and I'm sure I won't be switching them all at the same time. I can stagger them if necessary.
By the way, I also have a fuse that says (on the fuse itself) "Time Delay 500VAC 10kA IR 1 Amp" And it looks like it's an Amptrap ATQ1. Should I put this before the power suply just to be safe? It says AC, so I assume it breaks the black input wire...
_Z_ - December 20, 2004 07:22 PM (GMT)
Yes, that's a high energy time delay fuse. It protects the power supply from high inrush currents.
Yes, it goes in series between the power plug and the 'Line'-side terminal of the Sola. Break into the black wire as you intend. You don't really have enough stuff connected to warrant this, but it's a VERY good habit to get into. I'd also recommend putting a 2.5 Amp fuse of some type in the output leg of the Sola (between the PS (+) terminal, and whatever you connect it to). Even though the Sola is ouput-protected, a fuse is cheap insurance against your 'wondering what happens if I do this' fascination. :D
PDF spec sheet for Amptrap