Dreamcatcher 3.03 Feedback Thread


As an FYI, each unit that is shipped is tested for packet reception with an outdoor MK1 and 50-feet of RG6.


I’m running my DC 3.02Q with Power Over Ethernet and a USB-Ethernet dongle. Perhaps that’s another option if signal loss through the RG-6 cable is the culprit (and not a lack of supply voltage). Works very well for me. Yes, Ethernet is limited to 328 ft. But it can be stretched to 500 ft.

–Konrad, WA4OSH


I “was” going to take a few measurement on the DC 3.02q before I tried it on the 3.03

I plugged usb pwr DC into a different power supply – smelled hot

What did U1 marking 734600 GH12E do?

The EDUP dongle was also hot.

The U1 blistered my skin when I touched it…




I don’t have a board in front of me. If it has three legs on one side and one on the other, this is a GH12E voltage regulator. If your input voltage is too high, this part will probably get very hot.

–Konrad. WA4OSH


Yeah it’s the case style SOT 223 LDO voltage regulator vin 4.02vdc, vout 3.26

Still trying to restart after everything is cooled down

Now after “touch screen” arrives, the backlight is flickering badly, I think something is toast


The good news and the bad news. My v3.02q is down for more troubleshooting

The good news. v3.03 that arrived and is working, To answer @kenbarbi with 30’ coax
feeding mini-dish w/freesat lnb. this is the lnb status


Yup, that’s what I expected to see in my mind’s eye. You may not need to LCD display to run things entirely by remote control.

Before trying it again, I would definitely see if your wallwart is delivering 5.0 volts (or hopefully NOT more :frowning: ) before it destroys more boards. Don’t hook it up to your new board before you check it out.

Assuming it’s running at 5.0v out, I would check the key supply lines on the board with an VOM to make sure everything is stable. If fed correctly, your voltage regulator chip should not be overheating (Obvious).

Fortunately, there are not that many chips on the DreamCatcher that on may be able to trouble-shoot it down to the component level. As long as you uP is running, you can fix most of that on your bench.

–Konrad, WA4OSH


I think the bottom line… I tried to use a different untested power wall wart. Who knows what transients it introduced. So my v3.02q may be use-able for armbian/linux testing but not much else.

The lesson learned is for OUTERNET… you may consider supplying a power supply if you want to start to warranty any of future products.

I wonder if the “fast charge” different voltages may been in this new wall wart. Or is that only for usb-c?

Edit-- I guess I’ll have to order a second v3.03 so I experiment with DreamcatcherChat.


Fixed - - Not a Cable Length Issue

The short answer is my Sadoun Power Tech DG-280 Motor 500 feet away in the antenna farm at the base of my LNB/dish antenna presents what looks like a DC short to the DC 3.03’s LNB connector in the house causing an immediate cut in Bias T voltage. However, it is not a resistive short when measured with an ohm meter. I have never had this problem before with any of the other DC boards, so I guess you all can think about it, and propose work a rounds (other than not using the motor). As it appears, when I start up the DC 3.03 with no antenna connected, I get a full 14 vdc Bias T. As soon as I carefully connect the 500 foot connection that goes thru the Sadoun Power Tech to the LNB out in my antenna farm, Bias T goes to zero.

Measurements taken with the DC 3.03 inside my house and with the DC 3.03 at the base of the antenna where all the same - - my sj1 voltages were always higher than the jp1 voltages. In a working configuration sj1 reads 4.78 to 4.88 vdc and jp1 reads 4.84 to 4.91 vdc. The current draw to the Maverick or my 500ft away LNB/dish is 0.1 amps dc. Output Bias T voltages were on the order of 14 vdc inside, and 500 feet away on the order of 13 vdc. I wonder what other red balloons are experiencing? Ken


U1 is a linear LDO (thus generating some heat on transistor loss) making 3.3V it getting warm, thats normal. EDUP dongle is extremely small, and also getting warm due to the small heat dissipation surface. Despite its proven in RPis etc. and good, reliable one generally.


My guess is some blocking diodes. But for the life of me I can’t figure out how it worked for the v3.02 but not the v3.03 both using the same Skylark OS to control when/how the bias is activated.

Both system appear to a +dc voltage into the coax.


Yes, I thought of that when I measured the resistance, and didn’t see any diodes. Ken


Linear supplies are inefficient, hot, but nice and RF quiet. It is tiny, so it can get very warm.

But none of this stuff should be painful touching it more than 5 seconds. If can’t stand touching it for 5 seconds, it’s hotter than 45 degC or 115degF.

Yup, good part.

–Konrad, WA4OSH


After several hours of operation. No parts on the v3.03 new board are above 100 deg F. So all is well.

@Syed I am interested in purchasing a second v3.03 with no lnb, no tft screen, no wifi, no coax pigtail.


I’m wondering what project you are working on @ac8dg


Well I need two lora based transceivers to be able to talk to each other using the app DreamcatcherChat.

I fried my v3.02q somehow, but today got the new v3.03. So the plan is to get a second one (I can salvage most stuff from my old development v3.02q board). The goal is get two units to work “network” together probably starting using the ISM radio band around 915MHz (hopefully several miles apart). Legally only spread spectrum modulation schemes are allowed in the US. But my radio operators license gives me a lot of flexibility.

So… I had a bit of a set back today when my development v3.02 burned itself up.



Don’t want to disturb you from problem you’re working on. But want to say: A great 20 page User Manual ! :grinning:

Just received the latest DC in mail. I’m at page 10 where is says “You are now ready to take your device outside …”. Not today, for very cool and overcast. Going to put LNBF onto regular DirectTV dish and see what happens. So far, so good. Thanks Outernet Team.

Don (donde)


Works good with a direct TV dish, I get at max 4db, on an overcast and rainy day I get about 1db.


LoRa is Chirp Spread Spectrum. No worries.

–Konrad, WA4OSH