Naughty Mods I: DG-380 Separate Drive Power

[pendragon]

I'm in the process of revamping my FTA systems to accommodate a bunch of new dishes and LNBs. To take advantage of the opportunity, I will be sporadically posting documentation of some of the odd changes I'm making along the way.

First up is the DG-380 HH motor used to drive my 1.2m that has a Invacom QPH-031 LNBF (orthofeed for H&V FSS and L&R DSS). I originally set this up to feed all four inputs of a powered 4x8 switch, which in turn feeds 4x1 DisEqC switches, one for each receiver. The other DisEqC ports went to my other dishes, all of which are motorized. One of my receivers (actually a USB box) was designated as the master to control the motors. The switching matrix for all of the dishes is located a substantial distance from the 1.2m, so when I installed it I used a power inserter and a F tee connector to split off the commands and power for the motor from the LNBs. That meant I had five lines, 4 from the LNBF outputs and 1 to the motor, running between the 1.2m and the switch matrix. This seemed a better approach than running a cable to the motor and back to the switch, or requiring the master to always drive the same band/polarity to move the dish. It also kept the LNB outputs from going through the motor.

Afterwards I had one of my new USB boxes fry while driving the 1.2m, a few minutes after first hooking it up. The manufacturer chalked this up to infant mortality and sent a replacement box, which has worked fine for the past half year. Nevertheless I wasn't thrilled about the load placed on the USB box, given my experience and its relatively lightweight power supply. What I wanted was to have the motor power for the DG-380 separated from the receiver commands. That way I could use a beefy power supply to turn the motor and relieve the receiver from supplying any power at all (the 4x8 switch already provides independent power to the connected LNBs. The receiver switches voltages for H&V and 22 KHz on/off, but at a miniscule current). Because the motor power is a constant 18V (or whatever you want it to be), the drive speed doesn't drop if a V polarization is selected.

The best place to make this mod seemed to be in the DG-380 itself. It already has two connectors, so no external changes would be required. This voids the motor warranty, but I could care less. The mod can probably be done in a similar fashion to other motors. The steps are:

1. Loosen and remove the three screws on the bottom of the motor that hold on the top cover.

2. Remove the top cover.

3. Loosen and remove both F-connector hex nuts that mount the circuit board to the case (see "DG-380 Opened" photo). A deep socket worked for me.

4. Disconnect the ribbon cable tying the circuit board to the motor.

5. Cut the two traces indicated in the "DG-380 Circuit Mod" photo and add the jumper wire shown. You should check that the traces are broken with a meter, but be aware there is a nonconductive coating over the board, so you may have to puncture it or scrape some off to make a valid test.

6. That's it. Reassemble the case.

The photo shows which connector accepts the power for the circuit and the motor (right side when looking from the back when the motor is mounted on the mast) and which connector accepts DisEqC/USALS commands (left side). Note the circuit board photo is from underneath, so don't get crossed up.

Other observations:

When I was testing the mod with my USB box, the receiver would not send commands when directly connected to the motor's command port. I could see this on a scope. It apparently wants a little current draw to operate. I simply hooked a spare DisEqC switch in between the motor and the USB box and all was well. That pretty much mimics how this will be set up in real life.

Because both F-connector ports are used up, the receiver commands have to be split from the line(s) to the LNB outputs. Of course this is exactly what I wanted to do in the first place. There are many ways to do this, but a power inserter, F-tee connector and an optional terminator makes a clean install. I put the tee in-line with the receiver cable, and used a M-M F connector to connect it to the inductor-only side of the power inserter. I put a 75 ohm terminator on the capacitor-only side. The port common to both drives the receiver command line.

 

[Techfizzle]

looks like there is melted transistors on the board

 

[pendragon]

looks like there is melted transistors on the board

That's the non-conductive coating applied to the entire underside of the circuit board.

 

[FTABman0]

Very nicely done pendragon!

Who cares about the warrantee if you know what you’re doing! You needed it to do what you wanted and you achieved it! Well done!

I enjoyed reading this thread and will be following along with the rest of your mods!

That would sure take the load off the USB device for sure! Constant 18V even vertical!?!

Well now, got to give it to you, innovative indeed! Invention of other concepts for the DG-380 motor use! I like it!

This is more than a hobby to most of us, it is about achievement on what we want and set out to do!

The country can learn a lot from us!

Again, nicely done! :up

K E V I N

 

[Anole]

easy, and I like it:

I was with you right up to the final paragraph.
Had to read it 8 times to figure out what you were saying.
You might want to include a simple drawing to clarify.

Another tool which would work for the task, is a common diplexer.
Doesn't even need power passing to the TV Antenna side, in this application, so any off-the-shelf unit should work.

OR, you might put a 1uh choke inside the motor on the diseqc command in connector.
I just got in a little shipment from DigiKey, and will send you one if you like.
They were less than the price of the stamp to mail it. -

A traditional power supply for a motor, when you do this, is a laptop supply.
They put out 18 to 22 volts, at several amps, and should last forever in a job like this. - :up

 

[pendragon]

I was with you right up to the final paragraph.
Had to read it 8 times to figure out what you were saying.
You might want to include a simple drawing to clarify.

I've attached a drawing of the wiring from the receiver that demonstrates how everything gets split.

Another tool which would work for the task, is a common diplexer.
Doesn't even need power passing to the TV Antenna side, in this application, so any off-the-shelf unit should work.

Perhaps I'm missing something, but we have to pass receiver DC and 22 KHz to the switch so it can choose polarity/band. We also have to pass 22 KHz (DisEqC/USALS) to the motor. Many splitters/diplexers have diodes wired inside to isolate the ports, which won't help in this case, because they're pointing the wrong way. They also tend to cutoff AC below a few MHz.

OR, you might put a 1uh choke inside the motor on the diseqc command in connector.
I just got in a little shipment from DigiKey, and will send you one if you like.
They were less than the price of the stamp to mail it. -

The motor already has the necessary chokes inside. Cutting the traces isolates the two paths so I can run two cables to it. Why do I want to run two cables for the motor plus however many run to the LNB outputs? Because my switches are 100' from the dish/motor. Doing this saves a lot of cable (600' vs. 900' given 8 receivers, 4 LNB outputs and a powered 4x8 switch).

 

[pendragon]

There is also one other minor benefit of doing this that I forgot to mention. Because I have virgin 18V heading to the motor over RG6, I put another F "tee" in that line by the dish and connect the open end to a ground block with a weather cap on it. When I'm battling the elements on the roof with a signal meter in hand, I don't have to depend on the meter's internal batteries or risk blowing a switch to get power for the meter and the LNBs. I simply plug into the power tap. I ended up doing this for all my dishes.

 

[Anole]

good stuff:

Now that I see your drawing, I realize it was only your use of the term "power inserter" that caused my confusion.

When I said diplexer, I didn't mean any other sort of device nor splitter.
The diplexer common side would go to your receiver.
The high band or satellite side would go to your LNB.
(that is normally power-passing)
The low band or TV antenna side would go to your motor's command 'n control port, carrying the 22khz signal.
(on units which do power-pass, the 22khz gets through - it's possible that is not a universal truth for non-power passing diplexers)
This would prevent the motor from loading down the LNB signal.
The results would be identical in function to your drawing. - :up
I don't believe diplexers have any diodes, but if so, they'd be biased properly to function.

As for using a choke, the one printed on the motor's board is insufficient for its intended purpose.
If we were using the motor unmodified, then the diplexer above, your power inserter, or an additional 1uh choke inserted in series with the printed inductor would be desirable (based on previous reports and measurements).
Wiring up your power inserter as shown, accomplishes the same end result.

Anyway, just more ways to skin the cat.
And I think we all agree as to the need... -

 

[pendragon]

When I said diplexer, I didn't mean any other sort of device nor splitter.
The diplexer common side would go to your receiver.
The high band or satellite side would go to your LNB.
(that is normally power-passing)
The low band or TV antenna side would go to your motor's command 'n control port, carrying the 22khz signal.
(on units which do power-pass, the 22khz gets through - it's possible that is not a universal truth for non-power passing diplexers)
This would prevent the motor from loading down the LNB signal.
The results would be identical in function to your drawing. - :up
I don't believe diplexers have any diodes, but if so, they'd be biased properly to function.

Sorry, my mix-up. I read diplexer and thought splitter.

I just now rummaged through my junk box and found only one TV/SAT diplexer (I don't know why I have it because I've never needed one). The SAT side passes DC as expected, but the TV side cuts off below around 50 MHz. That makes some sense, although it may not be representative of such diplexers in general.

I agree this can be done many ways, some better and some worse for any given installation. I have an 18V power trunk RG6 that goes to all of my dishes, so that part comes for free. I'm also on a rampage at the moment to simplify wiring, because I want to move a lot of it inside, and will have less room for a 40-50x8 switch matrix (plus room for expansion) than I currently employ outside for a 10x8 matrix. Given my constraints, two trace cuts and a jumper is easier, cheaper and less cabling than other options. I also like the reliability factor of there being an air gap between the receiver's DC out and the power supply. With a different problem, the optimal solution might be completely different.