Measuring the performance of an MFJ-264 1.5 kW UHF dummy load with the NanoVNA, and explanation of what S11 is and how it relates to line and load impedances and to SWR.
Thanks Bryan ! I think I've figured out what the issue in the dummy load is. Will try to show that in the next vid (how to get it to work to 700 MHz+). Maybe within the next week hopefully. Busy bringing a new computer on-line today and this weekend :-)
Hello, thanks for the video. I have this same dummy load but it measures 63 ohms and when I measure it with the nanovna it measures a swr less than 1.28. How can I bring the dummy load to 50 ohms or would it not be necessary?
Interesting. Maybe try an Ohmmeter on the Carbon coated cylinder inside. Perhaps the contacts are bad (if it shows 50 Ohms, then there is 13 Ohms of contact resistance somewhere). If the Carbon-coated cylinder is the problem, I suppose its possible it got overheated at some point and the sheet-resistance went up. If that is the case, then for most work, I think SWR of 1.28 is still not too bad. That would translate to a 18 dB return loss (S11 = -18 dB), which is very good. Only about 1 or 2 percent of power is reflected.
@@ce4ufc630 Hi. The answer is 'yes' and 'no'. There is a webpage associated with these videos and more. In fact, it contains, in addition to the video series, detailed notes from an entire course on Radio Design. It's a bit rougher than the associated videos, but maybe OK for anyone who really wants to dive in deep. Sadly, it does not have an active posting feature enabled yet. I'm still learning the website hosting software. Anyway - it's here: ecefiles.org . Hope that helps some. 73's
Greetings .. Greetings.. Thanks for video. Just a comment.. I have two of the MFJ-264 and both have a similar issue. Really low return loss / high swr between the 200-230 Mhz range. Using a NVNA from 1 Mhz to 650 Mhz sweep with 401 sample points selected gives more resolution. I believe you can see a small bump on your test around the 200Mhz mark. Would be interesting to see if you MFJ-264 has any issues at the 200Mhz range as well. Thanks.
I just tried narrowing the range and it looks OK. There is a peak around 250 MHz, but it's -14 dB S11 (14 dB return loss). Even that could be due to my cables/connectors. Have you tried using a completely different run of connectors/cables/adapters to see if maybe the problem is in the interconnect?
@@MegawattKS Thanks.. I used a DSA815 Specturm Analyzer with a return loss bridge, and a MFJ-259C using a point-to-point connection on both MFJ-264 I have the the SWR etc was like 9:1 around 200Mhz.. Of course I could be bumbling 🙂
Yes. At DC (or very low frequencies), the reflection will be 42dB lower in amplitude than the signal launched onto the coax by the transmitter (or measurement device). It won't be that small at higher frequencies of course. As shown later in the video S11 degrades with increasing frequency. And this calculation is only taking into account the dummy load resistance value. The coax is assumed to be exactly 50 Ohms here - and in general it's characteristic impedance will be off somewhat from 50 Ohms as well...
Sorry - can't really advise since I don't have experience with the 250x. Looks like it's oil cooled - so that's a thing. On the flip side, the 264 might have more usability limitations (seems to only allow about 10 seconds at a time for 1.5 kW, and I wouldn't even go over that at 1kW since one always wants some margin). Not sure what the cooldown time is between tuning bursts. All in all, 10 seconds sounds kinda short ?
The multimeter shown at timestamp 1:18 is just set on Ohms. For this DMM, this is the "auto-range" setting for measuring DC resistance. I think they implement auto-range by trying each range and finding the lowest range that gives a non-overload reading (this is done to maximize precision in the reading). In this case it responded very quickly because they probably start at the lowest range (1000 Ohms) and it didn't overange there - so it doesn't have to try any more ranges and they just report that reading (51.0 initially and 50.9 later). In practice, I would just say it reads "51 Ohms" because the last decimal place won't affect the dummy loads ability to do what it needs to. It just has to present something resembling 50 Ohms to the DUT (Device Under Test).
Good question. Yes - I think so. For a case like this, I usually do a quick "response-cal", to normalize out the cable loss. (I've found the adapters don't matter much if the accuracy goal is below a tenth of a dB or so on S11).
