This video’s full transcript has been provided below for your convenience.
Mike — Three two. Okay. We are live. It says we’re live and hello, and good afternoon, everybody.
Tim — Yeah. We’re live.
Mike — We’re live. And we got Tim here as well. And so question, we get a lot and I know JP’s, I hope JP’s watching, or he’ll catch up with us in a minute. Why does the S-meter not work? Well, the S-meter works fine, but how does an S-meter work and why does an S-meter on a flex radio with a three kilohertz wide bandwidth read about S four or S five, right? Normally you plug in a dummy load or shorten it. It’s about the value. I know why, I sort of understand why as I was saying earlier, but I can’t really explain why, and I can know how to test to make sure it’s working correctly. So, Tim, and we can bring up SmartSDR. We’ve got an example. We’re going to show you here. And I’m going to learn something because I can’t wait for this. So, Tim, tell us about how we calculate an S-meter and what’s different maybe than what we grew up with in a super hit radio and why it’s different.
Tim — Sure. Okay. So first things first we’re sponsored today by Zaxby’s. This is one of these things that I do. I need to make sure that I have my drink available.
Mike — My boss is probably going to call me about that.
Tim — Yeah, probably. So just let him know that we’re getting some residuals for that.
Mike — You should know it here by now.
Tim — Yeah. So the S-meter is a really good question. The answer is it takes a little bit to get to the finish line, but we’ll get started with it. So the first thing is that the S-meter is not an arbitrary scale. I mean, you can go look this up in Google, and it will tell you that an S one is, and I’m doing this from memory, I believe is negative 137 dBm or something like that.
Mike — There’s a whole Wikipedia on it, isn’t it?
Tim — Yes, yes, there is. And so there is a defined scale S nine actually means something and that’s a good thing. But the problem that we get into is that the radios that you and I grew up with, there was no real easy way to measure RF power. And so the designers of these radios did a little cute trick. And what they would do is they would look at the AGC voltage in the radio, because these are all analog circuits and would derive an S value based on AGC value, AGC voltage.
Mike — Automatic gain control. And this is why, on older radios, when you turn the AGC off your S-meter pings to full scale.
Tim — Yes, exactly. And so this is the way they did it and they did a fairly good job of calibrating their AGC voltages based on this known scale and it worked okay. But then along came software defined radios. And the one thing about a software defined radio is that when you’re digitizing RF, you have a whole lot more visibility into the actual RF power because you’re digitizing it. You have that information available to you.
Mike — It’s all about the power and the past band. Right?
Tim — Right. And so one of the things that a lot of people don’t realize and I guess if you thought about it hard enough you would, is that our radios are nothing more than spectrum analyzers with a transmitter built into it. And so we calculate the S-meter value in the same way that a spectrum analyzer measures RF power.
Mike — Now, can I interrupt you, JP said, “You got to get a spectrum analyzer on this, because this is what makes sense. And we’re not really set to bring one in the video.” But yeah.; A spectrum analyzer by the way is, oh, your flex radio is a spectrum analyzer, but that’s…
Tim — Yeah.
Mike — Carry on.
Tim — So essentially how a spectrum analyzer works is that you take a unit bandwidth of spectrum. And then with an ADC, analog digital converter, we are digitizing the RF that’s within that unit bandwidth. And so when you’re looking at it on a display, what you’re actually seeing when you see all the little display jumping around, what you’re seeing is that signal after it’s been processed by a fast Fourier transform or an FFT, and we take the FFT and we break the spectrum up into lots of little chunks and little pieces.
Mike — This is the bins that we saw in the 5000 days in power SDR?
Tim — This is the bins.
Mike — Okay.
Tim — And so each pixel that is on the display represents an FFT bin and essentially all an FFT bin does, or all an FFT does is that it calculates the… It looks at the power based on the spectrum bandwidth. And so since you have a little slice of spectrum, and we’re measuring the power in that little slice of spectrum, and a unit bandwidth is a bunch of those little bins or those little chunks of spectrum all lined up against one another, or next to one another to show away form. What we essentially do is that we take, we do what’s called integrating the power of those bins. And essentially we are adding up the RF power that’s in each one of those little pieces of FFT, those little FFT bins.
Mike — Those little bins. How wide is a bin?
Tim — So a bin, really the width of a bin is going to be dependent on several things. One, in our radios, it’s going to be dependent on the radio itself. Our different models sample at different frequencies, have different sampling rates. And that’s part of it. The other part of it is how much you’re zoomed into a spectrum, because the more you zoom in, the smaller the FFT bin bandwidth size is. And so it depends. It can be down around the one, two hertz range, which is pretty good resolution. And then way zoomed out, it could be tens of hertz. So it really depends on the spectrum. So when we calculate an S-meter, after all this technical mumbo jumbo, when we calculate the S-meter, what we’re doing is… And I’ll just use a standard example.
If you had a single side band… If you had your radio configured single side band, and you had a 500 hertz receive filter, then what we are doing is… And let’s just make it… Let’s just say that we’ve got our radio configured so that one FFT bin is one hertz. So we have 500 little FFT bins that are now representing… No, excuse me. We have 5,000 little FFT bins that… No, I had said 500 hertz, right? So there’s 500.
Mike — 500 hertz.
Tim — Yeah. Doing the math in my head wrong. So there’s 500 FFT bins that are contained within that 500 hertz filter. And so we can very easily add up what the power level is because we have that information digitally. And so we add all of that up and we then display that as part of the S-meter. Because we add that up, the power level is done in dBm, and so we calculate that number, and then we convert dBm into an S number. And so this is the part of the play here where Mike’s going to go bring up the SmartSDR display. And we’re going to talk a little bit about this. And so, Mike, why don’t you go and change to a 500 hertz filter because that’s what I’ve been talking about.
Mike — All right. Well, what we’ll do is we’ll go to CW.
Tim — Yep.
Mike — And then we’ll pick up a 500 hertz filter. And do you want me to zoom in some more?
Tim — Yeah, sure. Zoom in a little bit more. There we go. That’s a pretty good size.
Mike — Yeah, here’s where the F one is.
Tim — Right. And so right now, you’re incredibly fortunate to look at a radio that is in a location that has an incredibly low noise floor. And we would all kill to have a negative minus 40 dBm signal on 20 meters at this time of day.
Mike — Actually hold on, just noticed something. Let me turn the tuner off, but it’s the same.
Tim — Okay. Yeah.
Mike — And we’re seeing an S one here, right?
Tim — So you’re seeing an S one level and everything. And so what we’re going to do is the S-meter has the ability to show S units and dBm. So we’re going to go, and we’re going to change this to dBm because these are, or this is a linear scale.
Mike — I clicked on it. That’s all I did. You can switch back.
Tim — Yep. And so what we’re seeing here is that within this 500 hertz filter, there is essentially a negative 116, 117 dBm of RF power within this filter. And so to prove that we are looking at power within the filter, I’m going to have Mike change this to a three K filter.
Mike — Okay. And Tim has a question, by the way, is why the noise floor in the spectrum scope moves up and down as you zoom in and out?
Tim — Why does the spectrum?
Mike — Why does the noise floor on the spectrum move up and down as we zoom in and out?
Tim — Oh, okay. So the answer to that is that there is some magic that we can do with decimation, with the DSP. And this is a little bit beyond the scope of what we’re going to talk about today. But as you zoom in and everything, the little S, the FFT bins get smaller. And so when you make those little bins smaller, there is less noise per bin. And so, because we’re integrating the RF power and RF power in that FFT bin is a combination of noise and signal. And so as you zoom in, there is less noise in each FFT bin. And as you add up those bins and you have less noise the representation of the noise for it goes down.
Mike — Okay. Now, you wanted me where? At 500? You wanted me to go to where?
Tim — Let’s go to three K.
Mike — Let’s go to three K. So we’re going to go way wide.
Tim — Yeah. We’re going to go way wide.
Mike — And clicking on the broadcast screen does absolutely nothing.
Tim — Okay. Yes.
Mike — And I zoom out one?
Tim — Nope, that’s fine. So you’ll notice that we made the filter a whole lot bigger. And our dBm signal went from a negative 116 all the way up to negative 108 dBm.
Mike — Almost DB. Eight to 10 DB.
Tim — And we also went from an S one to an S three. So go ahead and click on the DB. Let’s go back to the S-meter so we can see that difference.
Mike — You’re looking on the screen.
Tim — Yeah. We’ll get this right eventually, Mike. And so we’ve basically gone from S one to S three by doing absolutely nothing but changing the size of the filter. And so this basically reinforces the concept that what we’re doing is we’re looking at this RF power within a unit bandwidth on the spectrum display. Now, you’re going to get a lot of people, and we get a lot of talk about this is that when you disconnect the antenna, there should be zero S zero because there’s no signal. Well, that’s not true because we all live in the real world. And we are permeated by an electromagnetic field. It is everywhere. There is RF noise that is atmospheric. There’s RF noise that is manmade around you. There is RF noise that is cosmic. It comes from space. It’s everywhere.
And so the only way to actually reduce the amount of RF or RF signal that the radio can receive, and the radio is incredibly sensitive, would be to put the radio into a essentially a chamber that is RF immune. So we’re talking about a Faraday cage. And then, because there’s also electrical noise inside the radio. I mean, components generate noise. That’s always there too. Then you’re going to need to take the radio and you’re going to need to bathe it in liquid helium. And once you get the radio at that cold, then you may actually see an S value that is essentially S zero when you disconnect the antenna.
Mike — So simply put, a true S zero is a measurement of no noise present.
Tim — So, well, a true S zero, like I said, I haven’t looked at the numbers, but I believe an S zero is like a negative 141 or 142 dBm. And that’s pretty darn low to be able to get down to an S zero. Although we can simulate an S zero here. Mike, let’s go. Put it on 50 hertz and see what [crosstalk 00:15:04]
Mike — Okay. I got it, by the way. So, an S one is 0.2 microvolts RMS relative to 50 Ohms at minus 121. And according to the Wiki and I just quickly, hopefully I brought up the right one.
Tim — So let’s click on the 50 hertz filter and see how low we can go.
Mike — Back up? So let’s go back. So we’ll go to the 50. Okay.
Tim — And so let’s click on the S. So, hey, we hit S zero. And then, yeah. So we’re at…
Mike — 124.
Tim — Yeah. Somewhere in that neighborhood. So we’re pretty low. And to get S zero, and if you disconnect your antenna, if you put your antenna on transvert or port or an uninterrupted…
Mike — Flip over to RXA where on this radio, there’s no antenna.
Tim — It might drop a little bit more.
Mike — We dropped a little bit.
Tim — Yeah. So we’re down to near 130 dBm in a 50 hertz filter. So you can get close to, I mean, you’re reading an S zero there.
Mike — So what this is by the way. So this is a Flex 6700 and the antenna is a HF 60 vertical, nothing special at about the end of 80 or 90 feet of the equivalent of LMR 400.
Tim — Not bad.
Mike — With a crapload of radials all through a swamp and a mess.
Tim — It’s like I said, I’d love to have a negative 130 dBm noise floor on 20 meters at my QTH.
Mike — Right. And to show that we’re actually, there are signals out there, if by zoom out, you will see them popping.
Tim — Yeah. Let’s go find some signals and…
Mike — So let’s go here. This is typical. 20 meters could be dead, and then there’s here. And that might be 14,074.
Tim — Oh, I wonder if that’s FT eight. Yeah.
Mike — Oops.
Tim — I was going to say, if it disappears for 15 seconds, then comes back, then we know what it is.
Mike — Okay. So we’ll go back to up to 1,074 and actually we’re in CW. So we’ll change mode over to DIGU. And low and behold, we have energy being measured up to minus 109.
Tim — Yeah. Pretty strong.
Mike — Now, we didn’t prep for this, but since we’re looking at this and asking a question, one of the things we do better than your typical super hat is the ability to hear weak signals beside very loud signals. If I remember correctly, that’s called dynamic range. And that’s something that in an older radio is really hard to do because you’ve got this automatic gain control, trying to reduce your gain, at the same time killing your weak signal, correct?
Tim — Yes.
Mike — Yeah. Okay.
Tim — That is absolutely correct.
Mike — Because we have our bins, does that not give us the ability to deal with the loud signal here, but not clamp down the gain on the weak signal beside it?
Tim — Well, actually it does. The AGC algorithm that we use in our radios is very unique. It’s called a dual track algorithm. And so what we’re able to do is we’re able to analyze these signal levels and these FFT bins that are within a unit bandwidth. When I say unit bandwidth, I mean the receive filter. And so one of the things that you want to do is that there’s not just one type of signal that affects the volume or the gain within the receive pass band. And anybody who has operated during the summer, has heard lightning crashes. And these are very high energy, short duration, pulse type noise, QR Nancy and QRM.
Mike — By the way, I guarantee what we’re looking at is lightning crashes, because there’s one big honking storm headed our way and it’s [crosstalk 00:19:32].
Tim — Yeah.
Mike — Horizontal lines.
Tim — Yeah. I was going to say, it certainly looks like lightning crashes on the pan adapter. And so what our dual track algorithm does is that we are listening for changes in volume or changes in gain throughout the past band. And we’re looking for volume changes that are incredibly short duration.
And then we are looking for volume changes that are longer duration. So for example, if you’re operating side band, you’re on 40 or 80 meters and at night and somebody’s talking, they talk, there’s a lot of signal, but in between the… At some point you have to breathe, you have to pause for words and stuff. So it’s not a continuous signal. It’s a signal varying game. However, while you’re talking, there’s also a incredibly large lightning crash that hit where the signal jumped up twofold or something. And so our…
Mike — Like right here?
Tim — Yeah. So our AGC essentially is dual track. And what we will do is we are listening for those loud pulse noises and we’ll clamp those down without clamping down the volume of the rest of the signals in the pass band. And so that helps out a lot by… While you’re talking that the weak station that you’re listening to, the AGC doesn’t reduce the volume down, so you can’t hear that weak station during that momentary loud pulse of noise. Unfortunately, in order for the agency to work properly, you have to have…
Mike — I want to interrupt you right now. So, this is from lightningmaps.org. And if you haven’t looked at lightning maps, it’s just really cool. But you can see this stuff coming over like here on the Georgia Bay into lake, and then over. Georgia Bay is here, Georgia, and then into Lake Huron. And these are the crashes we’re hearing. I’m going to bring this back up, Tim, because I want to show this to you. And I had a question from somebody else as well, from Tim again, why is the noise floor on the spectrum scope moves up and down as you zoom in and out? We did that one. Oh, follow on, here, I’ll just post it. But wouldn’t even light cause some level of noise? Absolutely true, if it was in our receiver pass band.
Tim — Exactly. I was going to say the frequency of light is so much higher than what our analog to digital converter is digitizing that it’s insignificant.
Mike — The other thing I was going to say, I’ll bring SmartSDR back up in a minute, but if I was to pull up a TS eight 50 with a Kenwood with this going on, I would be seeing my S-meters slamming over all the time and being saying, “What’d you say? What’d you say?” Because your automatic gain control would clamp down the receiver, meaning shutting down its ability to receive. And then it would slowly decay. And again, it come back and then it’d get slammed again like a wave on a beach. And this is why, while if you’re listening to a net or something, and you’ll hear people always say, “What’d you say? The lightning crash took you out.” I don’t notice lightning crashes anymore. I mean, I hear them, they’re a bit of a click, but they’re not the same.
Tim — Yeah. So the follow on here is that with the AGC, the key thing is setting the AGC threshold correctly. And I believe we talked about this a little bit the last time we got together. Where you want to decrease the AGC threshold down to the point when you’re listening to a portion of the band that has no signal, you want to decrease it until you are no longer hearing the AGC increasing the volume of noise.
Mike — Great. And that’s a couple of places, right? Yeah. This is the AGC threshold here. I’ve got it set at 55. That’s actually… I know that it’s a little hot, you get to know your station and you can see that in two places, that’s one place. And the other place, somebody turned off the RX tab, is right here. You’ll see when I move it moves in both places.
Tim — Yep.
Mike — And I can’t imagine ever being above 50 on a low band.
Tim — So yeah. I very rarely am above about 45 on any band. I know that on six meters where I’m at, I’m fairly in a fairly quiet area for six meter noise, which is amazing. And I know that if I’m pointing my antenna west, which is lower noise than if I point it north, that my AGCT is going to be somewhere in the neighborhood of 23 or 24.
Mike — Right. And it’s in the manual, how to set that.
Tim — Yes. It’s in the manual how to set that. There’s also a nice article in the help desk that talks about that also.
Mike — Good. Okay. I think we sort of touched on everything. I always wonder how to do AGC and FTA when you’re not really listening. And I think you turned it off. No, you don’t actually. That’s old school. And that’s because of the gain control. The old school, in the super hit world, you turn the AGCT off that you got it all. Let WSJT engines manage the loud signals with the weak signals, which you did an okay job with. In the Flex world, the correct way is dial off the frequency a bit. Slide your AGC to zero and increase it until you hear the noise increase its 3 DB. But what that means is a noticeable increase. That’s the high level way of doing it. Pretty much right. Tim?
Tim — Yeah.
Mike — And then leave it there. Now, go back to whatever WSJT, FT eight, FT four frequency your on and leave it alone for a while.
Tim — Yeah. And I normally set the attack on the AGC when I’m operating digital modes to slow. So there is a fixed option and essentially the fixed option caps the amount of gain or volume that the AGC will apply to a signal. And that’s absolutely what you don’t want to do on a digital mode. So I normally operate with FTA, I normally operate slow. The only time I ever change that to a faster attack release is that if I’m on the lower bands and there is lightning strikes that are happening that might be kind of wiping things out. I want the AGC to react quickly and recover quickly because I know that’s the kind of noise that I’m wanting to get rid of.
Mike — There we go. So yeah. So old school, new school. Yes. When would we ever turn AGC off in a FlexRadio? I can’t think of a reason. Other than hacking around and testing something.
Tim — I mean, there’s some possibilities if you were using some outboard analog equipment modems or things like that, where you wanted to have a fixed maximum level audio gain. That is one possibility, but in general, I never used fixed or turned it off.
Mike — Okay. And Don NFSKT said, there’s a lot of misinformation. Well, yeah. Welcome to the internet. If you’re looking at community, you’ll get to know who you can trust. And I keep starting to say this a fair bit, that you’ll know A, in the community you want to watch for Elmers. They understand the technology. And if you don’t understand and some somebody says, “Do this,” don’t hesitate to challenge why. And if you don’t get an answer, you can email us. I actually have a banner here. So let me put this up, because email us at info and we’d be happy to help you out or provide direction with the radio, how it works.
The manual all 200 and X amount of pages is well worth. You can’t read it all. But as I said, it wasn’t Facebook, the other eight, print it off, take it to Kinko’s or something. It’s not expensive to print in black and white. It’s like 25 bucks or so. They do a nice Cerlox binding and you just leaf through it. Of course, I’m a geek. When I got my ham handbook. My first one as a kid, I must have turned every page. I didn’t necessarily read anything. And I went, “Oh, cool stuff about antennas. Oh, connectors.” You get the idea. And that’s where I used it as a reference to get going with stuff like that.
Tim — I was just looking in my bookshelf for my 1957 handbook. I can’t find it.
Mike — [crosstalk 00:28:50] I went to 74 I think. And I still have it and a bunch in between. So the handbooks are…
Tim — I was going to say, the only reason I have the 57 is because my first Rock Bound CW transmitter was a kit, or not a kit, but was a project described in that handbook. And I had to go replace the inductors in it to get it to work. So it’s here in my bookshelf somewhere.
Mike — Yes, you do use AGC comp. You don’t use any compression in digital modes if…. Here’s the question Ryan’s answering.
Tim — Okay. So that’s a good question. So the compression in our radios is actually a CSS. And when you’re using DIGU, which is the mode that you should always use for any digital mode, we turn off any of that audio processing. We turn the EQ off, we turn all the compression off so that there is no DSP signal manipulation done. What you’re getting is directly out of the modulator from the digital med software.
Mike — Okay. And at risk of opening a big can of worms for Jeff Wood who wanted to ask about all mode squelch. Yes. We’re well aware of it. In the meantime, if you run FR stack, there is a S-meter squelch built into FR stack that might help you for the time being, Jeff, and how to set it up. So, perfect. Cool. That’s all I have today. This was a 10 minute conversation, lasted 30. From the comments I’m seeing, I’m sure it helped out. And I don’t want to say go RTFM, go at the manual, but there is some cool stuff there that may…
Tim — It’s a great reference tool.
Mike — It’s written in English. I say cautiously, but it’s easy to understand. And we’re, like I said, the company’s here. If I can’t answer it, Tim can’t answer it, I can reach out to Eric, the VP of engineering usually answers me fairly quickly unless he’s up to his neck in something. And we’ll try to get you to the answers that help. And it was a challenge for me. I think I’m pretty technically smart in a global sense, but I jokingly say, I go to lunch with these guys when I’m in Austin and I have to sit there and just listen, because they’re in a whole new crowd about how to process RF. And I think what we hear and what we’ve got is, you have a lab grade piece of equipment that you get to talk on as well.
Tim — Yeah, yeah. I was going to say, yeah. When you’re in Austin, you don’t ask what time it is because you’ll get a dissertation on how a watch works.
Mike — I will you that. What a great bunch of people though. We have an awesome time together.
Tim — Yeah. Well, I was going to say, I had fun doing this. If you guys out there, if you want to hear Mike and I banter about on something else, send him an email at the email@example.com and we’ll have a couple of more of these impromptu get togethers and chat about how the radio works.
Mike — Perfect. Thanks, Tim. Have a good day. I hope everybody worked a bunch of guys on six meters yesterday. That was just one of the… If I was at a real job, a different job, I couldn’t really lie about… I got to test out the amplifier. It’s going to take me all day trying to heat it up. I worked some Europeans and I actually worked some new Canadian, I worked a Canadian province, but I don’t know where the BC guys are. Where are your V sevens? I could see you working, but turn the beam east.
Tim — Yeah, I’ve been… Whenever I see any double hop, skip come out on six I’m all over it. Because being on the east coast, the grids that I need now are CM and CN. So it’s few and far between, but I was able to get 10 new grids last week. So I’m up to 320 on six.
Mike — Oh, wow. And hey, Don, while you’re waiting for your six meter antenna, try loading up your 40 meter dipole. If it resonates, you’ll be surprised what you can work.
Tim — Yep. And I was going to say, and if you don’t want to spend a lot of money on a six meter antenna and you don’t have a big footprint and I’m going to give somebody some cred for this. There’s a company in North Carolina called Par Electronics that makes a stressed Moxon on six meters. It’s got an incredibly low footprint. That’s about the only kind of antenna I can use here on my little postage stamp piece of property. And it does incredibly well for a small two element, six meter antenna. Like I said, I’m up to 320 grids on six, and I’ve only been doing it for two years. So get on six. It’s fun.
Mike — And you got a little heater sitting behind you too though.
Tim — Well, yeah. That helps.
Mike — Thanks everybody. 73.