Paul Kimbrel's Musing

I started building the computer interface on my perf-boards and was tossing around the idea of purchasing a pre-made circuit board for it. I found that Far Circuits had the board available at a pretty decent price, so I decided to go ahead and get it. I also decided to continue the prototype on the perf-board while I waited for the circuit board to arrive. The board came in the mail and I looked at the convoluted mess on my perf-board and decided there’s no test like production.

So I started to solder parts onto the circuit board. I’m glad I did. It saved me a lot of time.

I got it all assembled and went to hook up the potentiometers (oh, I got another one of those ten-turn deals for testing) to the circuit board. But I couldn’t figure out how to get the potentiometer to give me a range of 0-5 volts. After clearing out the cranial cob webs, I found the answer was painfully simple. I had to refresh my memory on how voltage drop across a resistor works. Duh.

Anyhow, I figured it out and got the potentiometers hooked up. I also hooked up some LED’s in place of the future relays. Get the LED’s to light… and I can get the relays to close. I might leave the LED’s in there so I know which relay is being activated.

I started testing and found that all but the “left” direction was firing. After probing around on the voltages, I found that one of the quad-op amps was not putting out the right voltage on one of the amps. I pulled it and replaced it with another one. Let that be a lesson to you kids… always buy twice what you need!

Lo and behold, the left direction lit up. I tried different values on the parallel port and found that potentiometer positions match up perfectly with the requested settings. Yahoo!

Next step… rig up the relays. I’ll be etching my own board for this one!

I decided to monkey around a bit with this DAC so I could understand how it works before I through it into full use. I learned a lot tonight about the TLC7528CN as well as my parallel port and Windows XP. I started by hooking up a boat-load of hookup wire to a parallel port connector:

Now, this is just for the prototype. When I get this in a case, I’ll actually have a print port jack rather than a plug and use an extension cable between the box and the computer. Now, I simply hooked all that hookup wire into the DAC on my perf-board and put one the op-amps into service. I’m not quite sure what the op amp is trying to accomplish (U4A on the fodtracker schematic). It’s output is always 1 volt, but so is it’s input. So, yeah. Seems pointless.

Anyhow, the way the DAC works is this: You provide a “reference” that chip uses as a basis for a maximum value. Then, the input value is converted to range from 0 to that reference. My understanding is that this DAC is a current-mode chip. I’m assuming that it’s controlling the current relative to the reference. However, my sensors are going to be potentiometers, which will give me a varying voltage. According to VE2DX’s website, by hooking the reference voltage into the output terminals and measuring the voltage on the reference lines, you can use the DAC as a voltage-mode chip.

Wiggie. I had this brief flash of “aha” when I got to thinking about it and it made sense to me. However, I’m not sure I can put it into words. Perhaps someone else can explain it better. I think the op-amp feeding the DAC’s output is there to force the voltage to be 1 volt on the output lines all the time. If, by keeping the voltage to one volt there, only the current can change on the output. Then, since the input current on the chips reference lines is kept constant (by the chip, maybe??), then the reference line voltage must change relative to the output.

In other words, the DAC gets a signal for a value of 127. That’s about half way up the scale to the reference. Now, output current is changed to reflect that value. Normally, the voltage would change with the current, but the op-amp forces that the line to be at 1 volt all the time. So the current is changing with the value, but the voltage is constant. Now the reference lines are supplying the reference current, but by forcing the voltage to be the same on the outputs, the signal is force back into the reference lines in the form a voltage change that moves in relating to the outputs.

That’s my story and I’m sticking to it.

Anyhow, I got everything wired up. Now, the schematic calls for a 14v supply voltage. This voltage, I think, is supposed to come from a Yeasu rotor controller. The idea as that this interface will simply control the stock rotor controller (rather than the user hitting buttons manually). Well, I’m not doing that, so I didn’t have a 14V supply. I tossed around several ideas, and then I realized that all my radio gear runs on 13.8V lines! Eureka!

I got one of my radio power supplies out and got some Anderson Power Poles hooked up to my perf-board. I had my 13.8 volts. I confirmed that my 7805 was doing it’s job on the 5v line and that I had 1v off the op-amp. All was good.

Now, I had to figure out how to write to the parallel port. Unfortunately, Windows XP doesn’t let you just write to ports like previous versions did. Can’t have those viruses printing documents, now! I dug around in Google for a bit and found “UserPort.” I found it Here. It’s a device driver that opens up specific ports. I used it to open up my printer ports and the rest was academic.

I had downloaded the free Microsoft .Net SDK a while ago and it comes with a basic “C” compiler (my native language, really). It also supports embedded assembly language so I could poke at data ports directly. So I started writing values to the printer port and guess what?? NOTHING HAPPENED!

Whohoo!

Well, this DAC that I have is actually a dual-DAC. To tell the chip which DAC to use, I had to set a bit on the “auto linefeed” pin (remember, this is a printer port). Easy enough. I monkeyed around with that for a while and started seeing some results. But they weren’t consistent. I finally started taking the machine gun approach to programming and started setting all the control bits on the port high to get SOMETHING to happen. And it did… I was able to program one of the DAC’s. Give the parallel port a 0 - I get zero volts. Give it 255, and I got 1 volt. Give it 100, I got .3 volts (ish). It worked!

Until I tried switching DAC’s. Switching was pretty hit-or-miss. Until I figured out that the parallel port’s “strobe” pin is also hooked to the DAC. It seems that I can hit the A/B switch on the DAC all day long, but it won’t switch until you tap the strobe pin. That tells the DAC to read the “auto linefeed” pin and use it’s current value to select the right DAC.

Viola! I have two working DAC’s with a voltage from 0-1v.

Next up… hooking up the comparator circuit.

Okay, I hooked up the capacitor across the left and right terminals of the rotor and viola! It works!

Here’s the whole test assembly:

Here’s a close up of the capacitor:

I took the power off the transformer’s secondary leads and run it to an old two-position knife switch of mine (easier than soldering up to small switch). Then I used alligator clips to hook the knife switch to the left/right rotor contacts.

Here’s a close-up of the transformer:

Yep, those are 120 volt lines coming in there. Nice hot bare wires! I treated this setup with kid gloves. Speaking of kids… kids, don’t try that home. Or work. Or school. Or ever. Very dangerous. My next step will be to get that sucker in a box (and yes, I did unplug it for the night).

Here’s a video of the whole thing in action:

That “clunking” sound you hear when it stops is the motor’s axle dropping and hitting the workbench. It appears that it pops up a bit when it rotates. Cool!

Next up… starting the controller. I want to make sure that sucker works before I go through with the rotor mod.

Before I popped open rotor controller, I decided to try it out. It certainly sent a signal to the rotor and got it moving, but the motor in the controller didn’t seem to be turning. After opening and poking at it a bit, it turned out to be a cracked gear. I guess I shouldn’t feel bad about opening it up. I traced the signals and figured out where everything goes.

Once I had that figured out, I started pulling the transformer out. The transformer had an awful lot of leads and from the traces, the best I was able to figure out was this: There are three terminals in the middle of the transformer and two leads (red and yellow) coming out of the bottom.

After consulting with Doug again, I learned that the (as pictured above) two right terminals are the primary winding. The terminal on the left is the common for the secondary windings. Yep, windings. Plural. Turns out there are two secondary windings in parallel with each other. Both put out roughly 20 volts in phase.

Here are Doug’s notes that he faxed me (which I subsequently re-drew in Visio - faxes aren’t kind to pictures):

The idea is that one secondary winding is used to drive the rotor and the other is used to drive the motor in the rotor controller. Well, I’m not using a motor-based controller in the end, so I just tied the two together since they’re in phase.

I hooked everything up and couldn’t get the motor to turn. Note that there’s a capacitor between the two output leads going to the rotor in the notes above. It seems that’s rather important. Doug informed me that the capacitor is there to shift the phase one of the wires relative to the other. This phase difference is what makes the motor turn one way or another. I have a bit of experimenting to do with this before I fully understand what’s going on there.

Up next… I hook it all up (correctly I hope) and get the motor to turn!

So I ask my dad if he can bring our (okay… his) old rotor controller over. We’re sitting there eating dinner and I ask, “So how attached are you to that controller?” I can’t recall the answer, but it was about as direct as my question. So I ask again, “So how attached are you to that controller?” My mom pipes up, “What he’s asking is if it’s okay to break it?”

I’m not quite sure, but I think he said it was okay. Hope so!

I got a rotor controller for the sole purpose of harvesting the transformer. More on that in the next post. The other stuff I needed was the parts to the computer-to-rotor interface. From the FOD tracker website, I found the schematic of the interface: FodTrack Rotor Interface

I went through the schematic and pulled out the parts I needed:
1 - TLC7528 DAC (U1)
2 - LM324 Quad Op Amp (U2, U3)
1 - LM328 Dual Op Amp (U4)
1 - LM7805 5V Regulator
4 - 1N4148 Diode
4 - BC546 General purpose NPN Transistors

…and a bunch of resistors and capacitors, etc., etc.

Right. I also needed a potentiometer for the rotor. According to Doug’s project, I need a 10-turn 2k potentiometer. In other words, it takes 10 turns to go from 0 to 2k ohms. That number of turns is enough, if geared properly, to cover the range of the rotor. I got one (as well as the parts above) from mouser.com. The part number was #594-53411202 (Vishey 534-11-202).

I also needed a gear. Again, from Doug’s project…

From Small Parts, Inc. (www.smallparts.com) It is part #GD-4860, cost about $6. It is a 60-tooth 48-pitch plastic gear with a 1/4″ hole. Small Parts has a catalog full of cool stuff. Try to order some other stuff with the gear, to offset the shipping charge.

That was around 2000, so the price has gone up a bit. Not the part itself - but the shipping. $3.45 for the part (which is now GDS-48060-01) and $8.95 for shipping. FedEx no less (I’m a UPS kind of guy).

Next up… I need to get that transformer out of the rotor control box!

It was arduous, but I finally got the stinker of a rotor open. Doug’s pictures show it coming apart rather easily. Not my model! Before I continue, let me lay out some terminology. It’s not correct, I’m sure, but at least you’ll know what I’m talking about:

Antenna Shaft - The part of the rotor where the antenna mast attaches. This is the part on the outside that actually rotates.

Crank Shaft - The part inside the rotor that turns the antenna shaft.

Gear Assembly - The gears and associated shafts holding them together.

Drive Assembly - The motor and the initial gear that moves the whole thing.

Chassis - The surrounding metal casing.

My job would have been a lot easier if I had gotten a hold of an old Radio Shack rotor like Doug’s. And I may yet go get one (I can’t imagine it would be that hard). But I’ll try this anyhow since I have it.

Doug said he just took out the screws in the bottom of the rotor and whole gear assembly came out. Mine, however, seems to be permanently connected between the crank shaft and the antenna shaft. So I could get it pulled out to save my life.

I did notice that there was a retaining ring at the bottom of the crank shaft.

I figured - what the hey! Of course, I found I needed a new tool to do that (snap ring pliers). I got the tool, got the ring off and tried pulling it all apart. Still no dice. When in doubt - pull harder! “Chink chink chink”

Ball bearings start falling out. On my to do list - buy more bearings. It seems the crank shaft rides on some ball bearings in the the metal plate holding the drive and gear assemblies together. They’re not in a container of any kind. They’re just sitting there. That should be fun to reassemble!

But now I had the bottom out.

Here you can see where the ball bearing sat - where the crank shaft goes through (about 3/4 of the way down the picture).

Next up… acquiring the parts.

I’ve been tooling around with this project for a bit now, doing research, getting materials, etc., so here’s a quick update of how I got to today…

First of all… references. I was posting a question on QRZ around rotors and I got several great references. The best one was Doug Braun’s website:

http://www.dougbraun.com/rotor_mod.html

From here, I’ve gotten some great ideas and information on how to create my az-el rotor. First, I needed an azimuth rotor. That was easy. When we had our house re-roofed, I had the guys take our old TV antenna down. It had a perfectly good rotor on it.

I started looking for software to control the rotor and found this website:

http://ludens.cl/Electron/fodtrack/fodtrack.html

Now, I’m not entirely interested in the software (since I’m not running DOS or Unix), but I am interested in the hardware described that FOD Track controls. Several other software packages can control the same hardware and it’s pretty straight forward.

It works like this… An interface is hooked up to the computer’s parallel port and a couple of 8-bit signals are sent to the interface to control the rotor. The interface takes the 8-bit signals and runs them through a digital-to-analog converter (DAC) to get a voltage from 0-1 volts (so, for example, a digital number of 128 would be .5V).

The analog signal from the DAC indicates the desired position. Now comes the tricky part. Both the azimuth and elevation rotors are assumed to have potentiometers (variable resistors) on them that can be used to indicate their position. The higher or lower the resistance, the farther the rotor is in one direction or the other.

Then, the voltage drop across the potentiometers is compared to the voltage coming from the DAC. If they are not the same, the rotor is turned in the right direction to make them same. When the voltages match - the rotor stops. Some calibration is involved to get the right values to align.

The problem with my azimuth rotor is that it does not have said potentiometer on it. So… I’m going to add one!

Cue Doug’s site. Doug has done just that with a cheapo Radio Shack rotor. I’m going to do one with a Magnavox rotor (from Menard’s). I may have to deviate from the design a bit (the chassis are not quite the same), but I intend to chronicle this adventure on this blog.

Of course, once I get the azimuth rotor figured out, I’ll have to figure out the elevation rotor. But one challenge at a time!

I’ve decided to give amateur satellite work a serious go and decided to try and tackle building my own azimuth / elevation rotor. It’s essentially a rotor that can move in the horizontal (azimuth) as well as the vertical (elevation) plane. Then, I plan to build a computer controller for the rotor.

The end result will be a rotor that will take commands from a software program that knows the position of the satellites I want to work. The program will automatically position the antenna to follow a satellite’s path in the sky as it goes over. The same program will control my radio frequencies to compensate for the doppler shift associated with a satellite going over so fast.

Groovy stuff…