Problem with longer Cat5 cables


 

Darren L

TVWBB Member
Whenever I use a 40ft or longer cable to the RD3 the servo goes nuts. It works great with shorter cables. What I found is that the voltage is dropping due to the cable length. A 12ohm resistor in series with the blower solves the problem, so I added a switch to the RD3 to select the resistor when required. Has anyone else had an issue with longer cables? I'm measuring 4.8v at the HM and 3.7 at the RD3. Blower voltage drops to around 10.5. mA draw for the HM is reduced by about 200mA with the resistor. I've checked all of the solder joints, grounds etc. without finding any issues.
 
As you are aware, I am able to run my HM on a CAT5 cable up to about 50ft, being long tested (years) with a many cables.

I am not sure why you are having issues, if you do not see the voltage drop with a short cable but do with a long cable I would suspect the following:

Poor CAT5 jack on either the HM or RD. If the pins are corroded or bent so they do not provide solid connection that could be an issue. Or could be solder issues on the CAT5 jack at the HM.

Poor connection to the punch down knives in the RD CAT5 jack. You can cut away the plastic and solder wires to the knives to assure 100% solid connection there.

Poor quality CAT5 cable or bad crimp/ends.

Weak power supply or issue with the HM power regulator circuit, or bad solder joints on the HM board. If the PS browns out under load, or the Vreg circuit has issues, your HM may buckle under a load. I would check the soldering on the entire vreg circuit, check component values, particularly the capacitors and the polarity of electrolytics.

I make my own cables, have been using old school solid core CAT5e wire and the ends you can get at Home Depot. If you pay for shipping I would be glad to make up a long-ish CAT5 cable and send it to you for free, I have lots of this stuff sitting around. I will test it here with my HM first so we KNOW it works, then we will see what it does with your setup. Just shoot me an email and let me know what length cable you need and I will crimp one up and send it to you.
 
Poor quality CAT5 cable or bad crimp/ends.

My first thought. Are you making your own cable, or are you using a commercial cable? Whole buncha years ago (less than a year after 10BaseT was ratified by the IEEE,) I got asked to diagnose some cabling problems. Problem there was the hand made cables didn't have the individual conductors inserted all the way into the plug. Secondly, is there any chance you're using stranded twisted pair cable with sold wire plugs (or the reverse?)
 
my cables are all un-shielded and work fine up to about 50ft. Cat6 cables offer more isolation between pairs (if they are true CAT6, some are not), but that isn't required here.

Big difference between Cat5 and Cat6 are the bonded pairs (gives you equal length conductors in each pair,) along with more twists/foot. Yes, Cat6 is miserable to terminate compared to Cat5, especially if you want to maintain 100mbps and 1gpbs speeds (pairs must remain twisted up to something like 1/2" from the cable end,) but that's not much of an issue here. For these purposes here, even Cat3 would probably be sufficient.

Shielded will give you some protection against induced stray voltage, say, from running too close (maybe parallel) to power or other energy sources.
 
Since he is measuring a significant voltage drop at the end servo I don't think interference is the issue, if his 5v was solid I would be more convinced he may need some shielding.
 
Shielding will only prevent exterior influence.... where's it coming from? I've even left a 15' cable coiled up on top of my power supply with no issues, that should be not far from worst case.

My bet's on a questionable plug or jack.

I've got a cable tester kit lying around here some place, and a LONG cable (home built) downstairs. I'll take a quick look & see what I get for a voltage drop. This cable has ran at 100mbps speeds, so I'm fairly sure it's a good test.
 
Why not use 22-awg cable instead of standard cat 5e 24-awg? They make 22-awg cat5e cable for long POE runs.
You could also use 22-4 solid as used in standard alarm house wiring.

22-awg has twice the current carrying capacity of 24-awg and should definitely clear up any minor voltage drop issues.
 
Frankly I dont think it is a cabling issue, as I have used many long CAT5 cables and never had a problem. Though I make my own CAT5 cables and have no idea what the quality of the pre-made cable may be, which is why I offered to send a tested cable to him.
I suspect a bad CAT5 jack is making a poor connection, or a bad solder joint somewhere....
And I never asked the question, are you running one servo or three?
 
Just one servo. I need to clarify that the voltage is not dropping on the HM side. Only at the RD3 side. I soldered the wires to the RJ45 punch down connectors as Ralph so wisely suggested. The cables I have are 24ga. With the fan running 100% and no resistor I get 4.93v at the HM and 3.6v at the RD3 side using their respective grounds. If I use the HM ground I get 4.39 at the RD3. With the RD3 ground I get 4.14 at the HM. With the resistor in series I get 4.95 at the HM and 4.73 at the RD3. The main point to note is that the HM voltage at the jack is not browning out. It's clearly a voltage drop across the cable. I like the idea of 22-awg cable even if the gauge is not the root of the problem. I think Ralph has a point concerning the quality of the cables I have. These are just some I pulled out of a used parts drawer because they were available.
 
Darren, I get your point, Voltages are good at the HM cat5 jack but drop at the RD end of the cable, so voltage dropped across the cable, pretty clear. (though it could be the cable or either CAT5 jack causing the issue frankly)
I am going to send you a CAT5 cable for free, will test it here first, then lets see what it does on your system. If it doesn't work for you then you have a bad CAT5 jack or other issue with your system.
 

 

Back
Top