RS-485/Modbus networks can use a couple of termination methods to ensure signal quality and lower the risk of signal reflection. Your case may be able to get away with the 100 Ohm resistors you have if you are using CAT-5 cable since the impedance of the cable fits that well. If you are using other types of cable, you may need 120 Ohms or some other value depending on the impedance of the cable. Even then, your 100 Ohm resistor may work out okay, but it will depend on the cable length and it will also dissipate more power with its lower resistance. You will probably have decent signal quality, but again that depends on your cable length and type.
As an alternative, you could consider AC coupled termination where an R/C network in series across the pair is used as a terminator instead of just a resistive load. I don't have the formula off hand for calculating the required resistor and capacitor pair, but I'm sure you can find an online calculator relatively easy. AC coupled termination will lead to reduced signal quality in some cases, but it also dissipates less power in the network which could help tremendously if you have long runs or topologies with a high number of nodes. Again, this will depend on your particular situation and requirements.
But if you want to stick to the general rule, just parallel the four 470 Ohms resistors and get really close to the standard 120 Ohm termination value. As in all analog electronics, your mileage will vary as passive components can have wildly different tolerances leading to wildly different measured values. Measure your 100 Ohm resistors with a multimeter. You may find one of them reads on the high side and closer to the 120 Ohm "ideal" value. As an old mentor of mine said, analog electronics is a total crapshoot. If it work, it works.
Edit: Just saw your edit update. I think you'll be just fine with the 100 Ohm since your run is short and you are using CAT-6 cable.
On second thought, as per the reply from I'll go with this little contraption. https://pic8.co/sh/0SDhKl.jpg
One less thing to troubleshoot is worth a lot. Might try the 100 Ohm one later, after everything is set up.
And for extra science, put a variable resistor and see how far away from 120 ohms you can go in either direction before you get signal issues, see how much of a margin you have on either side.
Holy wall of text, batman. Thank you so much!
I'll try with the 100 Ohm (reads 98 on the meter) one.
Time to get one of those resistor sets in through hole. Only have SMD booklet type sets. Suppose they won't stand up to the mechanical stress of converting them to wired ones. But I'll try, if I'm desperate.
Thanks again.
Time to get one of those resistor sets in through hole.
When I've ordered the cheap Chinese ones they are all blue as if they're metal-film, with an extra band reading a higher precision of value, but when testing they are still 5-10% tolerance (which is fine for most things). Just know they lie about them being 2% metal film resistors.
Glad to help out. It's not too often I get to put on my old EE hat and think about some classic electronics stuff so this was fun.
Yeah the is probably where I would go if I were in your situation just simply because it's going to be a closer fit and you get to make a Frankenresistor in the process. But contrary to prairie's thoughts on the matter, I'd be wary that the Frankenresistor will become a problem long term and will make troubleshooting more of a pain if it is forgotten about or someone new comes encounters it. But it should be good for now. :)
'Frankenresistor' :D It's a quality build, measures right at 115. Used Sn60Pb40 solder, no worries.
We have factory shut down until January 9th. After that I'll be able to borrow the correct value. Glad we could have some fun. Let's see how it goes.
Frankenresistor will become a problem long term
Higher inductance? I'd agree, use it until you get a proper 120 ohm to drop in place.
(post is archived)