zero position on SG2100

[B.J.]

As mentioned in other post, I had my SG2100 motor apart a couple times today, and I just happened to look more closely at the workings, and I found that I am either wrong, or seriously confused about how the darn thing works.

When I got my SG2100, I quickly noticed that when I'd send a goto zero or goto reference command, that instead of going to zero, that it would instead go to about 2.5 degree. This was also coupled with the fact that all my USALS positions were also off by about this same amount, so that if I adjusted my location longitude to be off by this same 2.5 degrees, that only then would USALS work.

Aftr a while, I finally got fed up with this, and I went outside, manually moved the motor to it's actual zero position, and did a HARD RESET. After that, the GOTO ZERO command would send the motor to the actual zero position. However periodically, the motor would seem to get it's memory scrambled, and the zero position would somehow get changed. Once the zero position was off by somewhere in the 10-15 degree range. But each time the zero needed reset, I could fix it with a hard reset after manually moving the motor to the real zero. ALL of this had me convinced that there was no physical zero sensor in the SG2100, and that the motor's zero must just be in ram or something.

Anyway, when I took the motor apart today, and was sending it goto commands from my laptop, while the motor was sitting in my lap, I looked at the LIMIT wheel, and there is definately a zero indicator on the wheel, and if I send a USALS command to go to the same longitude as my position, it goes to that point on the wheel, and stops there. So the hardware zero position seems to be functional. So I was wrong that there was no hardware zero position. However the behavior that I've observed since I bought this motor makes it pretty clear that the motor seems to not use this zero when doing USALS calculations, and doesn't seem to use it when you send it a goto zero command either. I'm starting to think that this hardware zero sensor is only used when you reset the motor to factory conditions, but that isn't clear either, because the instructions tell you to manually move the motor before doing a hard reset, suggesting that the zero position used by USALS and GOTO zero is a memory location, not determined by the hard zero sensor.

But anyway, I'm real confused relative to just how the hardware zero sensor position is used, and when it is transferred into the motor's memory.

Possibly related to this, my SG2100 has seemed to have lost it's capability to remember diseqC-1.2 positions for the past couple months. It used to be that the diseqC-1.2 positions would hold pretty much indefinately, regardless if power was lost or not. But recently, I can set a diseqC-1.2 position, and it then seems to be functional for a while, but if I come back the next day, and try to use it, generally the position has been lost. This made me think that perhaps there was a battery in the motor which retained the position info. But when I had the motor open today, I looked carefully, and there doesn't seem to be any battery in there at all. I even looked for one of those super capacitors, which can function pretty well as a battery provided that it gets power occasionally, but I didn't see one of those either. So I'm confused relative to why my motor seems to be able to remember diseqC positions for a little while, but then forgets. Strange. I like to use diseqC-1.2 rather than USALS, but if it can't remember the positions for more than a day, it's not very useful.
Anyway, I'm confused at the motor's memory lapse. Only thing that makes sense is that it's getting corrupted signals from the receiver which is changing the settings, but that doesn't seem to fit the symptoms either.

 

[AcWxRadar]

In some of the equipment I service at work, they have what is called a BBRAM (Battery Back-up RAM). A lithium Ion battery is actually built into the RAM IC. Now, I rather doubt that these motors are using this type of a device as they are not designed to last extremely long times. In the equipment we have them in, they last about 5 years.

My gut opinion would be that they are utilizing an EEPROM (Electrically Erasable PROM). This would make the most sense as it would not require a battery to hold the memory of the satellite positions. I would think that this is how I would design the motor's brain.

Next, we have other equipment (servo motors) that have internal encoders that count pulses to know where it is at precisely at all times, but they need a ZERO position signal or a HOME proximity sensor to act as a reference point. If it gets out of sync, you have to perform a "reset home" command if for some reason the system loses some encoder pulses and doesn't go where it is supposed to.

When the motor goes into its reset home program, It does a slow search for that home prox or limit switch, and then resets the encoder count to zero at that point. Perhaps, if it doesn't always read the encoder pulses correctly, it slowly goes off course.

Now, just possibly, they are applying somethig quite similar within our satellite motors. If yours is often losing its position, I would think that the encoder reeader is losing pulses or there is a piece of debris on the encoder wheel itself that is screwing up the count.

Just some food for thought, eh?

RADAR

 

[B.J.]

.....

My gut opinion would be that they are utilizing an EEPROM (Electrically Erasable PROM). This would make the most sense as it would not require a battery to hold the memory of the satellite positions. I would think that this is how I would design the motor's brain.

I had assumed that they used EEPROMs, but if so, I didn't understand why the motor would keep forgetting positions after a while. But perhaps I was wrong about them being programmed in the first place. Perhaps the EEPROM, which is probably built into some multi function chip, is defective. I used to program the UV erasable EPROMS with home made devices, and a UV florescent bulb I kept inside a brief case to protect my eyes. Sometimes those things would get flakey after a while, and sort of lost the capability to be programmed right. I tried to upgrade my home-made device to program EEPROMs , but I think they requirred a higher voltage to program, which the chips I was using couldn't handle. I don't know if the current EEPROMs still require the higher voltage, but if so, maybe the long COAX run is giving me enough voltage drop that the motor works, but the storage of locations doesn't work. I'm not sure if I'm using solid copper coax or the copper coated steel type. I know that there isn't a significant voltage drop when just the LNB is drawing current, but I'm not sure how much current it takes to program EEPROMs.

...
...

Now, just possibly, they are applying somethig quite similar within our satellite motors. If yours is often losing its position, I would think that the encoder reeader is losing pulses or there is a piece of debris on the encoder wheel itself that is screwing up the count.

Just some food for thought, eh?

RADAR

I think that the issues I'm having now, are largely related to my motor's EEPROM or whatever it uses for memory. However my question relative to the zero position is really related to how it's supposed to work, not relative to some malfunction in MY motor.
Ie my manual describes the hard reset thing, and tells you to move the dish to it's physical zero then do a hard reset. Back when the motor was working properly, this would change the zero position, so that a goto zero actually went to zero. So my confusion is that doing this hard reset seems to set the zero position into memory, and then, until you do another hard reset, it must use the memory zero position, even for the goto zero command. So I'm really confused relative to when it actually uses that cam thing on the limit wheel.
If I have time this morning, I may experiment a bit with the motor in the house, before I put it back up.

One other curiosity about the motor's operation is I was sitting here sending it USALS goto commands, and when the motor receives a command, the motor first starts to move, but only moves a very short distance, then it starts up again, and goes continually until it reaches it's intended position. This seemed odd, but now, on thinking back I'm wondering if perhaps it's going to zero before continuing on to the goto location, since I was always doing this from near the zero. But that wouldn't be logical if you were far away from zero going to another location far from the zero. I may experiment with this too, before I put it back up.

EDIT: I just did a couple experiments.
I manually moved the motor to 10 deg. Then told it to go to USALS zero (the program I was using doesn't have the real goto zero command).
It turned on, move a short distance, maybe to 9, turned off, then back on and went to zero. So the odd behavior was NOT it going to zero, it just seems to apparently first see if it's capable of moving, maybe for safety reasons, then continues on. Strange.

I then manually moved it to 10 again, and did a hard reset. Manually moved it to 20, and told it to go to zero. It went to 10, ie the zero I had set with the hard reset. SO this goto zero doesn't send it to the zero set by the cam wheel, but to the zero stored in memory.

I then sent it to 20 deg, and it went to 30, as expected. I'm going to try to send a REAL goto zero via a manual command.

EDIT2: Tried sending a manual goto zero via TSREADER's custom diseqC function, but I couldn't get it to work. I'll try that later when the motor's connected to my Diamond again.

 

[Vorg]

I had my 2100 apart and it has a small round object on the back side of the motor, likely magnetic. They use HALL effect to detect rotation. It has nothing to tell where center is. If it misses pulses, it would be off and stay off.

Most micro controllers have some sort of flash or eeprom to store data when power goes off. If they have a low power detect to dump when power goes down, then that memory is only updated at power loss. If they update every time the motor moves, then the memory is getting updated ALOT and they have a limited number of writes. Basicly, the memory wears out.

 

[B.J.]

I had my 2100 apart and it has a small round object on the back side of the motor, likely magnetic. They use HALL effect to detect rotation. It has nothing to tell where center is. If it misses pulses, it would be off and stay off.
...

Yes, there is a hall effect sensor there for the counts, but there IS a center detector. There is a wheel with 3 switches. The wheel has two bumps which activate the limit switches, but the wheel isn't round. Half way between the limit switches, the wheel changes diameter right at the zero point, and that activates the center point switch. Of course there are different flavors of SG2100, so I guess it's possible that they have different hardware, but that's the way mine is.

 

[AcWxRadar]

Yes, there is a hall effect sensor there for the counts, but there IS a center detector. There is a wheel with 3 switches. The wheel has two bumps which activate the limit switches, but the wheel isn't round. Half way between the limit switches, the wheel changes diameter right at the zero point, and that activates the center point switch. Of course there are different flavors of SG2100, so I guess it's possible that they have different hardware, but that's the way mine is.

I would say a cam type of mechanism. Three switches, one for CCW overtravel or a hardware limit, one for CW and the third for the HOME or ZERO position.

The HOME or ZERO position sensor or limit switch is probably only used if you command the motor to reset its home position. After that, it (hopefully) remembers how many pulse counts it has traveled from there, in either direction. If it keeps track of the pulses accurately, then it always knows where it is positioned. But, if it loses count at some point, then you need to RESET the HOME or ZERO once again.

That is when it goes back to look for that HOME or ZERO position switch. When it finds it, it resets its encoder count to zero.

I am not certain that my analogy is entirely correct, and it is probably far from perfect, but I am betting that I am in the ballpark. What do you think?

RADAR

 

[B.J.]

....

The HOME or ZERO position sensor or limit switch is probably only used if you command the motor to reset its home position. After that, it (hopefully) remembers how many pulse counts it has traveled from there, in either direction. If it keeps track of the pulses accurately, then it always knows where it is positioned. But, if it loses count at some point, then you need to RESET the HOME or ZERO once again.

That is when it goes back to look for that HOME or ZERO position switch. When it finds it, it resets its encoder count to zero.

I am not certain that my analogy is entirely correct, and it is probably far from perfect, but I am betting that I am in the ballpark. What do you think?

RADAR

What you describe makes sense, however it doesn't seem to fit all the observations I've encountered, most of which have been mentioned above.
The problem is that if I do a HARD reset using the button out on the motor, it doesn't use the zero sensor on the cam wheel at all. Ie when I sent the motor to 10 degrees, and did a reset, it never moved off 10 degrees, but it set the USALS zero to the new 10 degree location. So it doesn't use that wheel when you do a hard reset. Maybe a software reset might be different.

The other thing that's confusing me is that yesterday, when I did a GOTO Reference on my STB, it DID seem to use the cam wheel sensor, as it went to the actual zero, not the USALS zero I had just set, however in the past, I have seen dozens of times when I used that same GOTO Reference command, and it would go to some other location, not the actual zero, so that made me think that it was NOT using the sensor. It is possible or even likely, that since my motor seems to have lost it's memory, and since I had to remove power to connect it to the STB after setting the USALS zero at 10 deg, that perhaps it forgot where it's zero was, and reverted to using the cam wheel sensor.

Anyway, I have ordered an new motor, also an SG2100, so when that comes next week I'll experiment with that to see if it behaves differently from my motor that's lost it's memory.

 

[AcWxRadar]

B.J.

That is buggy.

Normally, I would say that the switch would fail in a manner where it wouldn't make contact when it reached the cam. But, what if the switch was intermittently making contact (shorting internally) in other positions? That might explain some of this odd behavior.

I guess it wouldn't actually have to be the physical switch that was the problem. It could be a short across the circuit traces or whatever electrical component that reads the switch. It probably goes directly into an IC.

RADAR

 

[B.J.]

One other probably unrelated factoid... I did find water inside the motor yesterday. I don't know just how watertight the thing is supposed to be, perhaps it just assumes that water will get in, and drain out the bottom or something. But there wasn't any gasket or seal between the two halves of the case, so unless they had some kind of waterproof glue that sealed the two halves, I'm not sure how it would keep water out. I was tempted to run some electrical tape around the thing, or perhaps put some teflon tape between the halves before I put it back together. Wasn't a whole lot of water, but it was just from 1 day in light rain.

 

[brex2001]

B.J.

That is buggy.

Normally, I would say that the switch would fail in a manner where it wouldn't make contact when it reached the cam. But, what if the switch was intermittently making contact (shorting internally) in other positions? That might explain some of this odd behavior.

I guess it wouldn't actually have to be the physical switch that was the problem. It could be a short across the circuit traces or whatever electrical component that reads the switch. It probably goes directly into an IC.

RADAR

How about condensation on the control board traces or the EEPROM leads?
I had trouble after heavy rains with my DG380 until I sealed the case halves with Felpro blue gasket maker silicone.

 

[geekt]

With most electronics, I've found that after you fully dry them out, they work fine again, unless something has literally been fried by a short...or a trace has been damaged...such as in a keyboard...

 

[AcWxRadar]

This might be a little interesting...

http://www.satelliteguys.us/broadband-forum/228983-wildblue-tria-nonstandard-repair.html

RADAR

 

[Vorg]

Mine has weep holes along the bottom edge and no seal. But the top shell should shed any outside water around the unit. Even where it comes together the top overlaps the bottom lip so water would not run in. What you want to see is if the board is getting wet and if so, how. Also look for signs of water having been on the board. If the water had minerals, they would be left behind on the board and that will cause problems.

 

[B.J.]

With most electronics, I've found that after you fully dry them out, they work fine again, unless something has literally been fried by a short...or a trace has been damaged...such as in a keyboard...

Yeah, I've found the same thing.... although sometimes if the device has been used when wet with impure water (like coffee or a coke...{personal experience}) then the device can be damaged.
I used to work at a corporation that had a department that designed and manufactured sophistocated (for the time) electronics, and they used to have a guy who would clean circuit boards by basically dunking them in a tub of some cleaning solution. I can't remember if it was a solvent or just soap and water, but it surprised me, and on a couple occasions after that, when I had electronics that were exposed to weather, or otherwise got wet, I just rinsed them off, and everything was OK. But I suspect that the success might depend upon the components used in the device. I can't imagine it working with the old fashioned paper capacitors.
I even had a SLR camera once that got involved in a canoe upset, and got completely soaked in a river. It was over a week before I got home. I then took it apart, and rinsed it off. I partially put it together enough to check it out, and everything worked fine, however there was a mechanical part that wasn't quite dry, and being a chemist, I decided to quicken the drying process by rinsing with acetone, which unfortunately dissolved a plastic part, which essentially glued the film advance mechanism. I was just a half an hour of waiting for it to completely dry away from fixing the thing, but ended up destroying the camera.

 

[Wescopc]

I repaired a cell phone by disassembling and running through the dish washer without soap, blow drying and reassembled.
Bob

 

[B.J.]

Mine has weep holes along the bottom edge and no seal. But the top shell should shed any outside water around the unit. Even where it comes together the top overlaps the bottom lip so water would not run in. What you want to see is if the board is getting wet and if so, how. Also look for signs of water having been on the board. If the water had minerals, they would be left behind on the board and that will cause problems.

I didn't notice the weep holes, but that makes sense. I didn't notice the circuit board being wet, but I did notice water dripping off one of the non-electriconic parts, although I can't remember what part that was. But it was clear that water had gotten inside.

If there are weep holes, that makes me feel better, although I might put electrical tap over the top, just to minimize what gets inside, but I won't put any tape on the bottom half.

That weep hole thing reminds me of some BUD actuators. Some actuators come with grease fittings, some didn't, but I think most or all came with holes for water to drain out. People used to think that "more is better" when it came to grease in the darn things, so they would pack or pump the bottom end so full of grease that it would block the natural drainage of the thing, and their actuators would fail quickly. I used to take mine apart completely, and just grease the screw by hand, using a minimum of grease, and seldom had problems.

 

[geekt]

I would never clean electronics with acetone. LOL I sometimes clean electronics with alcohol, but even that can be slightly dangerous. The reason is that alcohol cleans pretty well, evaporates quickly, and doesn't leave any residue. I wouldn't clean it with soap, because it would leave residue likely.
I wouldn't worry about rain water though, because rain water generally has less particles than tap water.

As for grease, with motors I usually grease the shaft, turn, grease, turn, until it seems to move easily enough.

 

[melbfla]

Speaking of grease anyone has a good recommendation on what kind to use on the 2100. I have mine apart on my bench it had quit . I took it appart the electronics are ok and I have it turning just wondering if there is a brand/kind of grease that is more desired.
Thanks

 

[AcWxRadar]

Speaking of grease anyone has a good recommendation on what kind to use on the 2100. I have mine apart on my bench it had quit . I took it appart the electronics are ok and I have it turning just wondering if there is a brand/kind of grease that is more desired.
Thanks

The grease used for fishing reels is good. Any non-conductive grease would be preferred, obviously. I would be cautious about mixing different greases, however. Sometimes their chemistry can react together and change the two from a grease to a concrete or a really sticky, pasty glue, or worse yet, to something caustic.

RADAR

 

[melbfla]

Thanks Radar I will get some . Planning on taking all apart and cleaning it . Has been doing fairly good for about 3 years now. I think it just wanted to get out of the sun....lol