Powermax SG9120B Teardown and Rebuild

[gpflepsen]

Earlier I'd made a post lamenting the poor performance of my SG9120B motor. It acted like it was binding, I could not adjust the backlash from the motor and it allowed too much movement to the dish. See this and this.

I decided to take the dish down and investigate the motor, and it would be a good time to methodically get a good install and aim on the dish. I was having troubles getting good tracking since first installation. I took the motor apart to see what was going on inside, mostly because the "M4" and "M3" adjusting screws didn't seem to be doing much.

This is what the interior looks like.




Two M4 cap screws held the motor and gear reduction to the case and worm shaft. After pulling the motor, you can see the worm drive to the main gear. The worm shaft is supported by a bronze bushing at each end, and the adjustment mechanism is opposite the worm shaft's spur gear.


Here's the motor and reduction gearing. You can see the motor gear and several more gears into the gear train are plastic, then they are made of metal for the remainder of the drive. This means the higher speed/low stress gears are plastic, then they change to metal when the speeds get lower and the stresses higher.


Next to come off is the two screws holding the plate on the adjusting machanism. It was apparent some spring pressure was there as the plate was backed off. I was just hoping to prevent one of those "snakes in a can" surprises! What I found under the plate was two springs, one apparently mangled during the manufacturing process.


I pulled the springs and the wedge that they were held in...


...and the worm shaft assembly. You can see how the wedge will take up the end-play of the worm shaft. The smaller "currency" M3 adjusting screw prevents the wedge from seating too deep and binding the shaft from too little end play. This adjustment is important to prevent motion of the mounting tube that results from the worm shaft moving axially.


I needed to replace the mangles spring, so I grabbed two el cheapo Bic pens to rob the spring from. These springs were about 1/32 too large in diameter to seat into the wedge, so I drilled out the holes to accept the new springs. One hole was about 5 degrees from vertical, so that was corrected when boring out the holes.


I also took the wedge to a sander to polish up the surfaces that are expected to slide when being adjusted. They were pretty rough and not really finished during manufacturing. 320 grit on a random orbital did wonders.

Here is a picture of the bell crank that the larger M4 adjusting screw pushes on. The bell crank pushes the whole bushing/wedge assembly and acts to adjust the worm shaft/worm towards the final drive gear, effectively setting the teeth engagement between the worm gear and final drive gear.


Here's a shot of the worm shaft back in the housing with the wedge, springs and bell crank in position.


You can see the gap in the worm shaft with the adjusting wedge up.


With the wedge down, the gap goes away! This is one motion to adjust out to eliminate excess dish rotation via the M3 screw.


Everything went back together in reverse order, with some light machine oil (3-in-1) on the bushings. I put some grease on the wedge, but I'd probably not do that next time. The grease keeps the wedge from responding (sliding) freely, but it will move. Just have to have some patience when adjusting, as the spring pressure is all that is pushing to make it tight. The screw limits the wedge motion.

With the motor assembled and on the table, I tried to measure the current by connecting jumpers between the microHD and SG9120B. This just would not work. I don't know if the ammeter blocked or attenuated signals, but it just didn't work. I started adjusting (by ear) by putting the M3 screw all the way in to give the most end play to the worm shaft. Slowly screw in the M4 screw until the drive starts to labor a bit then back off 1/8 to 1/4 turn. Then unscrew the M3 until the wedge doesn't back out any more. I did this by backing out the screw 3 turns then running the motor end to end then running the screw back in until the wedge stopped moving. I left the wedge screw putting pressure on the wedge and called it good.

The motor now ran from end to end without binding and the backlash was practically zero. I could not detect any backlash when twisting it by hand. It felt like a great improvement over what I had before. Also, the end-play on the output shaft was non-existent. I couldn't feel anything when pushing/pulling. What an improvement!

Next up, putting this bad boy back on the pole. I'll follow up with what I discovered when mounting, with some tips that maybe some, both new and old, could benefit from. For now, let's just say it was easy to hit 30W through 125W when I was done, just like I thought it should be!

 

[Anole]

Excellent work!
We want might want to it put over in the FAQ section for easy access.
Similar info probably applies to other motors.

 

[Tron]

Nice surgery Gpflepsen! I had to take one of my 9120Bs apart awhile back when I accidentally loosened one of the two F-fittings on the motor. It seems to be a well-constructed motor, at least compared to some others. I've never had one DiSEqC H-H motor that didn't have backlash problems (I've worked with several SG-2100s and my two SG-9120Bs)...

 

[gpflepsen]

Back out to the mounting pole I went... with the refreshed and hopefully improved motor in hand.

I double checked the j-pole for plumb. When I initially installed, I used the little bubble that is provided in the GSP 90cm dish. There's probably nothing wrong with this piece, but the accuracy is nothing compared to what a real level can provide. My suggestion is to let the kids play with it and use a level for setting the pole plumb. I've setup countless fixed dishes before and nothing beats the simplicity and accuracy of a real level.

I wasn't going to take anything for granted anymore with this motor after knowing the latitude scale was wrong, the elevation scale was unproven and the back lash was screwed up. When the mounting tube was off the motor, I measured the angle; it was 30 degrees.

My plan was to mount the motor rotated to the 30 degrees elevation which means the tube should then be perfectly plumb. The elevation scale is correct but the latitude scale was off 5 degrees. My recommendation is to forget the LAT and EL scales and adjust the elevation to make the tube plumb. Make sure the motor is set to 0 degrees, that's very important.

With the elevation set to make the tube plumb in that plane, the tube was out of plumb is the other plane. The following pics show how bad it was. Note the frame seemed to be plumb at the j-pole but poor manufacturing doesn't allow the plumb pole to translate to a plumb tube.

The motor frame at the pole.


The same side at the motor tube. I didn't measure the amount off, but I estimate it to be a bit more than 1 degree. That's not acceptable.


The tube was made plumb by adjusting the j-pole to account for the motor bracket. When this was done, the dish went onto the tube. To ensure the dish and motor are aligned together, I used a trick picked up from a post here by SatelliteAV. The trick is to measure from two known symmetric points on the dish to two symmetric points on the motor. I used the bolts for the lnb support arms and the top ears on the motor housing. I didn't want to measure back to the motor frame or j-pole due to the motor bracket being "off". Make these measurements equal and you should be in alignment to 0 degrees through the motor-dish.

I've previously measured my due-south satellite and know the lnb arm is 73 degrees and I have a marker (stake in the ground) off in the distance that I know points to 97W. So I motored over to the satellite w/ USALS and set the elevation to be 73 degrees on the lnb arm and adjusted the dish/motor EZ to look over the marker... Bam! Signal! Just some fine tuning on elevation and I was good enough for the morning. I motored over to 125W and tweeked a bit, went over to 72W and had signal. Time to go in and watch some FB! and rebuild the channel list.

I took the dish over to 30W just to see how it would go with DiSEqC 1.2. I was surprised the transponders lit up, many over 70% on the microHD. I didn't expect this to work that well down that low. My elevation to 30W is 9.4 degrees and probably 50' above the neighbors house. He probably thinks it's pointing right at his house!

To sum up the tips, ensure the following;

1) Make sure the pole is plumb using a real level and not the little fisheye bubble level that comes in the kits.

2) Set your motor elevation to the tube angle and check and adjust the tube to be plumb in all planes. Don't trust the motor mount to be true.

3) Measure the dish to motor to ensure it is symmetrical and that they are in alignment.

4) Put a marker out away from the dish location that points to the southern sat. Use a stake, birdbath, dog house or rock. This makes it easy to hit it on any future maintenance or equipment upgrades.

This is how I hit 30W to 125W without really trying! Well, being methodical and not trusting anything along the way made the difference.

 

[Wescopc]

Great instructions - When one has it clear in their mind about the geometry needed, it makes it easy to double check the elevations and declination settings. Good job!
Bob

 

[Cham]

Some great pointers, I have my 9120B on the bench in pieces right now waiting for a spare hour or two to replace some electronic components. Mechanically I think mine has a slightly bent worm shaft I think I can bend back into shape after I work on the electronics and get that part working again. The motor was only operating in one direction, likely excessive load caused by the bent shaft caused component failure.

 

[gpflepsen]

Some great pointers, I have my 9120B on the bench in pieces right now waiting for a spare hour or two to replace some electronic components. Mechanically I think mine has a slightly bent worm shaft I think I can bend back into shape after I work on the electronics and get that part working again. The motor was only operating in one direction, likely excessive load caused by the bent shaft caused component failure.

What electronic components are you replacing?

I'd be sure you have a bent worm shaft before trying to straighten it, I'm talking about measuring runout down to the thousandths. It's easy to mess it up instead of fixing it. Mine acted like there was serious runout somewhere in the gear train, but after doing what is posted above, it's a good runner now. I was so close to ordering another motor last week on two occasions. I'm glad I didn't now.

 

[Cham]

So far looks like a diode and a resistor, and possibly a relay. Need to spend some more time tracing the circuits and looking at some unused circuits boards replacement parts.
Motor only turned one direction and would almost stall once per revolution. Not sure how to measure the runout other than installing it and measuring turning force by listening to the motor loading up etc. Once the motor works both directions I'll look closer into the mechanical issues and pay some extra attention to the wedge adjustments.

 

[Cadsulfide]

Nice work!

 

[Cham]

Thought I would resurrect this thread rather than start a new one:

9120B play in shaft assembly:

Noticed some extra play in the motor shaft lately. Thought the gear backlash or worm shaft end-to-end play was the issue, seems my previous adjustments are still holding. Play is in the connection between the inner shaft and (bent mast) where the dish attaches. I removed the bolt and re-installed (seemed very tight to remove). Couldn't see any major issues with the connection between the two parts of the shaft or bolt. Re-assembled the shaft components and tightened the bolt (7mm hex I think) as much as I thought required, about 20ft/lbs without stripping or breaking the bolt. There is still a bit of play though.

I am thinking the nut (fits into a hex shaped tube to hold it in place) is not tightening as the steel motor (inner) shaft is threaded and only the hex head is actually tightening. Over time the bolt hole in the mast has likely worn a bit. Maybe installing a washer under the hex head of the bolt would help? Any pointers to remove this play?

Otherwise the motor has been working well for the last few months.

Tnx!
-C.

 

[Magic Static]

Been there before When you get it tight enough the play goes away. I used a normal hexhead bolt with a custom made spacer under it. Then I could get enough muscle to it to make it tight.

 

[Cham]

Sounds good. Will look for a longer grade 8 bolt and a spacer. Not sure if it is metric or standard, likely have something around that would work from the extra bolt pail..

 

[Magic Static]

for the spacer, I ended up grinding the corners off a nut that fit the bolt.

 

[gpflepsen]

Here's an update on this SG9210B about 2+ years following this post. The motor still works, but when aligning it today and doing a reset on the motor, the elastomer covers over the direction buttons were aged to thee point they were hard and brittle. Not a very good selection of material for the elements encountered outside. Grade "F" for this as I had to break the covers out to actuate the button switches.

 

[Cham]

Mine failed too. I just used some "Gorilla tape" (heavy duty black duck/duct tape) over the holes to try and keep moisture out.
I have seen some black rubber plugs somewhere that might fit... now to find them in my junQue box...when it warms up.