Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Russell Ault]

I ran into an unexpected issue recently, and I'd love to hear people's thoughts: in a nutshell, it appears that overloading the RF input on a QLXD RX can cause additional and unexpected (yet easy-to-calculate-and-avoid) IMD products.

Now, obviously the "solution" to this problem is to follow best practices and not overload the inputs of RXs, but I wanted to mention this anyway because (a) accidentally overloading the RX inputs isn't particularly difficult when there's an IEM TX antenna nearby and (b) other than these extra IMD products, the RXs appear to be performing normally (i.e. the only reason I "discovered" this issue is because one of my coordinated frequencies happened to land on one of these unexpected IMD products).

Figure 1 is a comparison of a scan made with a QLXD RX (H50 band) in blue vs. one made with an RF Explorer (plugged into the same antenna distro) in red. The three large peaks that appear in both scans within TV channel 26 are three IEM TXs with, shall we say, less-than-ideal antenna placement (FWIW, although the peaks exceed the vertical scale of the graph, the highest-recorded level in the data from the RF Explorer was -18 dBm, so the peaks aren't actually "off the charts" by much). Note that the blue QLXD trace shows several peaks within TV channels 34 and 35 that are absent from the red RF Explorer trace; these appear to only exist with in the RX itself, and are more than capable of causing "co-channel" interference. (At this point it's probably worth mentioning that I did do scans with a couple different RXs just to make sure this wasn't a unit-specific issue, but I got the same results with all of them.)

Figure 2 is a series of three QLXD RX scans, each with a single IEM TX carrier present. This more clearly shows the relationship between actual carriers (located within TV channels 24 and 26) and the IMD products (located within TV channels 31, 34, and 35 respectively). If my math is right, the relationship between carriers and IMD products implies that these are third-order products between the overloading carriers plus some sort of internal carrier at ~492.250 Mhz.

I'm tempted to start adding an extra generic device to my H50 QLXD coordinations at 492.250 MHz, which should theoretically help to avoid these IMD products even if the antenna placement gets a little sloppy, but I'm curious to hear what other people think?

-Russ

[Jonathan Woytek]

I ran into an unexpected issue recently, and I'd love to hear people's thoughts: in a nutshell, it appears that overloading the RF input on a QLXD RX can cause additional and unexpected (yet easy-to-calculate-and-avoid) IMD products.

I suppose I'm not immediately surprised by this, as I've seen this happen plenty of times in some of my amateur radio gear, where strong signals on the receiver can result in some ghosting. On that equipment, I can usually attenuate the RF into the receiver to help mitigate the issue, but that's not really an option here. Some good notch/band pass/band reject filters on the receive antennas could definitely help if one is in a situation where one needs to support cohabitating TX and RX within IMD product distance. Fixed filters are not a great option for gear that travels and needs to accommodate tuning changes, of course, and tunable band pass/band reject filters are not cheap, but they could certainly help solve the problem.

I have some QLXD receivers in V50 in a show I'm doing this coming week. I'll see if I can at least duplicate what you've observed.

Side note of interest: I'm small potatoes and wanted to start investing in good wireless to use for the musicals I do, so I decided on QLXD and planned to add about two units every year. When I completed my first four-unit fly rack, I mounted a PoE-powered network switch inside, and brought a couple of RJ45's out to the rear panel on the rack. This would be easy for me to plug into my existing control network, where I had a switch that could supply PoE, and I could supply an injector for use when I didn't have my full control network on site. My first show with just the first two units and no switch had no issues. Next year I added the next two units and the switch, and suddenly I had a TON of issues. I was seeing a high noise floor across the band and was now worried that I sunk a decent figure into a wireless system that was going to be unusable. I didn't have a good specan at the time, so I started methodically debugging as I do, and discovered that unplugging the switch from the control network cleaned up the band. I was discussing this in another forum when someone tickled a piece of long-forgotten knowledge: Ethernet gear can create a decent amount of VHF hash. A prosumer switch mounted immediately next to a set of receivers generates enough hash to make them unreliable and can render them unusable. Pro tip: Don't be me. Put the switch outside of the box. Live a happier life.

[Scott Helmke]

I was discussing this in another forum when someone tickled a piece of long-forgotten knowledge: Ethernet gear can create a decent amount of VHF hash. A prosumer switch mounted immediately next to a set of receivers generates enough hash to make them unreliable and can render them unusable. Pro tip: Don't be me. Put the switch outside of the box. Live a happier life.

With VHF (in our case ULXD) Shure recommends using shielded Ethernet cables.  And all the jumpers they include with receivers are in fact shielded!

[Russell Ault]

I suppose I'm not immediately surprised by this, as I've seen this happen plenty of times in some of my amateur radio gear, where strong signals on the receiver can result in some ghosting. On that equipment, I can usually attenuate the RF into the receiver to help mitigate the issue, but that's not really an option here. Some good notch/band pass/band reject filters on the receive antennas could definitely help if one is in a situation where one needs to support cohabitating TX and RX within IMD product distance. Fixed filters are not a great option for gear that travels and needs to accommodate tuning changes, of course, and tunable band pass/band reject filters are not cheap, but they could certainly help solve the problem.
{...}

For that particular rig I'm actually considering getting a pair of fixed-frequency HPFs for the RX side (which, as you say, should largely mitigate the issue), since I typically try to keep the mics near the top of their operating range (to keep them as far away as possible from the IEMs) anyway. The complicating factor is that we cheaped out on the antenna distro and used a pair of unity gain cable TV splitters instead of the usual 50 Ω stuff, which means that I'll either have to look for 75 Ω filters, figure out the full impacts of using 50 Ω filters in an otherwise-75 Ω system, or convince the group to shell out for a "proper" 50 Ω distro system (which, to be fair, is the long-long-term plan anyway).

I think what I find most annoying about this problem is that there's literally zero indication on the RXs that they're being overloaded (other than the unexpected co-channel interference from "non-existent" sources, natch).

{...}
Side note of interest: I'm small potatoes and wanted to start investing in good wireless to use for the musicals I do, so I decided on QLXD and planned to add about two units every year. When I completed my first four-unit fly rack, I mounted a PoE-powered network switch inside, and brought a couple of RJ45's out to the rear panel on the rack. This would be easy for me to plug into my existing control network, where I had a switch that could supply PoE, and I could supply an injector for use when I didn't have my full control network on site. My first show with just the first two units and no switch had no issues. Next year I added the next two units and the switch, and suddenly I had a TON of issues. I was seeing a high noise floor across the band and was now worried that I sunk a decent figure into a wireless system that was going to be unusable. I didn't have a good specan at the time, so I started methodically debugging as I do, and discovered that unplugging the switch from the control network cleaned up the band. I was discussing this in another forum when someone tickled a piece of long-forgotten knowledge: Ethernet gear can create a decent amount of VHF hash. A prosumer switch mounted immediately next to a set of receivers generates enough hash to make them unreliable and can render them unusable. Pro tip: Don't be me. Put the switch outside of the box. Live a happier life.

Wow, that's pretty wild! Were the antennas also near the switch, or was it spewing so much noise that the RXs themselves were being interfered with?

As a slight topic swerve, while I've been quite happy with QLXD for music, I've tended to discount their use for theatre because of their not-actually-diversity system. How have they been working for you?

With VHF (in our case ULXD) Shure recommends using shielded Ethernet cables.  And all the jumpers they include with receivers are in fact shielded!

That would make sense, given what Jonathan was mentioning; one of the few "legitimate" uses for shielded Ethernet cables, I guess!

-Russ

[Jonathan Woytek]

For that particular rig I'm actually considering getting a pair of fixed-frequency HPFs for the RX side (which, as you say, should largely mitigate the issue), since I typically try to keep the mics near the top of their operating range (to keep them as far away as possible from the IEMs) anyway. The complicating factor is that we cheaped out on the antenna distro and used a pair of unity gain cable TV splitters instead of the usual 50 Ω stuff, which means that I'll either have to look for 75 Ω filters, figure out the full impacts of using 50 Ω filters in an otherwise-75 Ω system, or convince the group to shell out for a "proper" 50 Ω distro system (which, to be fair, is the long-long-term plan anyway).

I think what I find most annoying about this problem is that there's literally zero indication on the RXs that they're being overloaded (other than the unexpected co-channel interference from "non-existent" sources, natch).

Funny, I don’t know if I ever actually noticed the lack of an overload indication. I have a set of the VHF log periodic panels with the LNAs on board, and those do have an overload indicator that I’ve tripped a few times when checking mics at FoH before taking the shoe racks backstage.

Regarding the filters, since you’re inserting these on the receive side, I don’t know that you will see much difference between 75 vs 50 ohm filters. I haven’t looked recently, but there used to be a bunch of 75 ohm filters of various types available to help deal with television reception issues. That said, it might make sense to just get the 50 ohm devices, since that’s the direction you’re headed in anyway.

Wow, that's pretty wild! Were the antennas also near the switch, or was it spewing so much noise that the RXs themselves were being interfered with?

As a slight topic swerve, while I've been quite happy with QLXD for music, I've tended to discount their use for theatre because of their not-actually-diversity system. How have they been working for you?

That would make sense, given what Jonathan was mentioning; one of the few "legitimate" uses for shielded Ethernet cables, I guess!

It was spewing so much that I was seeing hash even with antennas disconnected. I got one of the VHF paddle antennas, which helped to improve SNR, but didn’t actually change the observed noise floor. I did try a couple of shielded patch cables with the other receivers disconnected, but I was still seeing a lot of noise. Moving the switch out of the box seemed to solve the issue, even without shielded cabling, so I figured (and later confirmed) that the switch was creating most of that noise.

I’ve been really happy with the QLXD on lavs. I’m usually using them alongside stuff like Sennheiser EW. I work mostly with high school theatres, so we aren’t talking high tech budgets. Maybe put up against higher quality units, I’d see some issues. I’m usually running Countryman B3 elements in hair rigs. I have found that they sound a little better than the EW units, especially when pushed.

[Steve-White]

With VHF (in our case ULXD) Shure recommends using shielded Ethernet cables.  And all the jumpers they include with receivers are in fact shielded!

Stuff like this is why I peruse this site daily.  All my QLXD stuff is V50 band.  Racked up in pairs with UA874 distribution amp, ethernet router and digital mixer.

Guess who's going to be changing CAT6 cables in racks?

[Don Boomer]

Regarding the filters, since you’re inserting these on the receive side, I don’t know that you will see much difference between 75 vs 50 ohm filters. I haven’t looked recently, but there used to be a bunch of 75 ohm filters of various types available to help deal with television reception issues. That said, it might make sense to just get the 50 ohm devices, since that’s the direction you’re headed in anyway.

My experience with passive filters would suggest they would be very different (unless you are talking about active filters) since they are very much tuned to the impedance that loads them.

[Russell Ault]

{...} It was spewing so much that I was seeing hash even with antennas disconnected. I got one of the VHF paddle antennas, which helped to improve SNR, but didn’t actually change the observed noise floor. I did try a couple of shielded patch cables with the other receivers disconnected, but I was still seeing a lot of noise. Moving the switch out of the box seemed to solve the issue, even without shielded cabling, so I figured (and later confirmed) that the switch was creating most of that noise. {...}

That's...concerning. I don't see VHF mics very often around here, but it's something I'll keep an eye out for!

{...} I’ve been really happy with the QLXD on lavs. I’m usually using them alongside stuff like Sennheiser EW. I work mostly with high school theatres, so we aren’t talking high tech budgets. Maybe put up against higher quality units, I’d see some issues. I’m usually running Countryman B3 elements in hair rigs. I have found that they sound a little better than the EW units, especially when pushed.

Given the unusual way QLXD handles gain-staging, I would expect them to sound a lot better than the EW (or any FM wireless) when pushed. That said, my biggest hangup with using QLXD (and its ilk) for theatre is the number of times I've had to spread out the RX antennas to ensure coverage, which works pretty well for traditional diversity schemes and causes serious problems for "'intelligent' switching diversity" systems. Maybe this isn't such an issue with the shows you're doing?

My experience with passive filters would suggest they would be very different (unless you are talking about active filters) since they are very much tuned to the impedance that loads them.

This would have been my ASSumption, but it may not matter either way since I'm not sure that the kind (i.e. steepness) of filtering that would be useful for this application is even readily available (or, at least, cost-effectively so), and my limited understanding of UHF LC filter design is enough to suggest that I'm not qualified to DIY my own.

-Russ

[Henry Cohen]

My experience with passive filters would suggest they would be very different (unless you are talking about active filters) since they are very much tuned to the impedance that loads them.

Impedance matching, although very important mathematically and in high powered limited bandwidth applications (LMR, cellular, PtP links, satellite, etc.) is for the most part moot in our puny world of entertainment production wireless mics and IEMs.

First, antennas are transducers, which means their impedance varies with frequency, and our industry wideband antennas (at least a couple of hundred MHz in the mid-UHF band, or about 40% bandwidth) are anywhere from 37 -112 ohms depending on frequency.

Second, a 75/50 ohm mismatch is a VSWR of only 1.5:1, a loss of less than .2 dB, which is far better than the VSWR of a typical wideband antenna found in our industry (typically 2.5:1 average across the BW of the antenna).

Lastly, the front end of wireless mic receivers are not 50 ohms across the tuning range. They vary from 50 -75 ohms, if not more.

So, whereas it's best practice to try to match nominal impedance across components, it's not as important as the benefits proper filtering provides.

[Russell Ault]

{...} So, whereas it's best practice to try to match nominal impedance across components, it's not as important as the benefits proper filtering provides.

That's very reassuring!

Do you have any advice on where to find a pair of HPFs with a cutoff frequency no higher than 572 MHz and meaningful attenuation below 536 MHz (or is that as crazy an ask as I think it is)?

Thanks!

-Russ

[ProSoundWeb Community]