Re: [PSES] Unreasonable Design Penalties on Receivers
Re this: “I do want to add that nature, as it is with so many other aspects of life, is indifferent to our MDS. And the environment is what it is and, unfortunately we are given the environment as a black box with the methods of MIL-HDBK-235 left mostly a mystery. But I know you weren’t saying otherwise. “ CS103/104/105 are evaluated using an MDS -level for the discernable signal. But if you are approaching an intense EME, it means you are close to the sources. For a carrier deck, that also means that any desired communications with the carrier are at much higher levels than MDS. Since radio link intelligibility is proportional to signal-to-noise ratio, if the desired signal is higher than MDS, the ability to reject out-of-band signals should go up somewhat proportionally. You would certainly expect that a signal coming in 20 – 40 dB above MDS would result in out-of-band rfi rejection to increase proportionally. You could verify this yourself in the lab. Now if the source of rfi is not also the source of your comm link, the above doesn’t work. If you are being jammed by an unfriendly, you could be very far from the friendly signal source. As noted heretofore, CS103/104/105 were not designed for that scenario. These requirements facilitate co-location of antennas on a single platform, as is a sort of companion requirement, CE106. Finally, the book to which you refer has been out-of-print for over a quarter-century. I am supposed to be working on a second edition, before I forget everything and get put out to pasture… -- Ken Javor (256) 650-5261 From: Michael Viau Reply-To: Michael Viau Date: Saturday, May 30, 2020 at 11:14 AM To: Subject: Re: [PSES] Unreasonable Design Penalties on Receivers Ken, your last point is an excellent one. And you are absolutely right that I need to reconsider the burden with respect to the individual equipment itself and also that it is probably significantly less costly to attenuate the signals outside of the restrictions placed by the receiver itself. I do want to add that nature, as it is with so many other aspects of life, is indifferent to our MDS. And the environment is what it is and, unfortunately we are given the environment as a black box with the methods of MIL-HDBK-235 left mostly a mystery. But I know you weren’t saying otherwise. Before I lose you Ken, any chance you know where I can buy a copy of your book? I’ve been looking for a while now and keep coming up short. Thanks again for weighing in. Patrick, I appreciate you saying all of that. Often the “problem” (if you really want to consider it one) is that you never know when a well designed system is going to break. It passes all of the tests (which is a great thing) but a successful susceptibility test only tells us when equipment works, not when it doesn’t. From an integrators perspective we like to be able to predict risk from DC to daylight and most performance measurements are limited in scope. >From a budget (time and money) perspective, and in the opinion of literally >everyone else involved, it’s a waste of resources to have that information >(unless you need it). I try my best to empathize with both sides and strike a >balance whenever possible. Thanks again fellas, Michael On May 30, 2020, at 10:07 AM, Patrick wrote: Hi Michael I'm glad to hear more folks are thinking through the whole system, and whole environment, not just piece-parts. A component approach to requirement allocation makes sense when you procure a component. When you procure or integrate the entire system then the standard approach needs new thinking. Sometimes that's simply asking "why?". The why question brings out history and evolution. I find these to be invaluable in determining next- steps. Very glad to open these questions and have input from folks like Ken who helped drive years of evolution. Our company also worries about end-to-end performance. We delivery receivers and antennas, sometimes integrated, sometimes as piece parts. We carry the whole burden of RF front end performance. Questions like the ones you ask matter to us. They are carefully considered and discussed in our team meetings. We find our "customer" appreciates that we are thinking more about helping the war fighter accomplish their mission and return home than about pushing requirements out to someone else. Glad to know we are not the only ones! Keep up the good work! I've enjoyed this thread and look forward to more. -Patrick. On Fri, May 29, 2020, 9:08 PM Ken Javor wrote: Well, you may be in the uncommon position of procuring a brand new IFF transponder for your aircraft, and if that is the case, then as an integrator you can levy any spec you choose, but in general, the military buys certain models built by outfits such as Rockwell Collins (or whatever they are called today) and the avionics manufactur
Re: [PSES] Unreasonable Design Penalties on Receivers
Ken, your last point is an excellent one. And you are absolutely right that I need to reconsider the burden with respect to the individual equipment itself and also that it is probably significantly less costly to attenuate the signals outside of the restrictions placed by the receiver itself. I do want to add that nature, as it is with so many other aspects of life, is indifferent to our MDS. And the environment is what it is and, unfortunately we are given the environment as a black box with the methods of MIL-HDBK-235 left mostly a mystery. But I know you weren’t saying otherwise. Before I lose you Ken, any chance you know where I can buy a copy of your book? I’ve been looking for a while now and keep coming up short. Thanks again for weighing in. Patrick, I appreciate you saying all of that. Often the “problem” (if you really want to consider it one) is that you never know when a well designed system is going to break. It passes all of the tests (which is a great thing) but a successful susceptibility test only tells us when equipment works, not when it doesn’t. From an integrators perspective we like to be able to predict risk from DC to daylight and most performance measurements are limited in scope. From a budget (time and money) perspective, and in the opinion of literally everyone else involved, it’s a waste of resources to have that information (unless you need it). I try my best to empathize with both sides and strike a balance whenever possible. Thanks again fellas, Michael >> On May 30, 2020, at 10:07 AM, Patrick wrote: > > Hi Michael > > I'm glad to hear more folks are thinking through the whole system, and whole > environment, not just piece-parts. A component approach to requirement > allocation makes sense when you procure a component. When you procure or > integrate the entire system then the standard approach needs new thinking. > Sometimes that's simply asking "why?". The why question brings out history > and evolution. I find these to be invaluable in determining next- steps. > Very glad to open these questions and have input from folks like Ken who > helped drive years of evolution. > > Our company also worries about end-to-end performance. We delivery receivers > and antennas, sometimes integrated, sometimes as piece parts. We carry the > whole burden of RF front end performance. Questions like the ones you ask > matter to us. They are carefully considered and discussed in our team > meetings. We find our "customer" appreciates that we are thinking more > about helping the war fighter accomplish their mission and return home than > about pushing requirements out to someone else. Glad to know we are not the > only ones! Keep up the good work! > > I've enjoyed this thread and look forward to more. > -Patrick. > >> On Fri, May 29, 2020, 9:08 PM Ken Javor wrote: >> Well, you may be in the uncommon position of procuring a brand new IFF >> transponder for your aircraft, and if that is the case, then as an >> integrator you can levy any spec you choose, but in general, the military >> buys certain models built by outfits such as Rockwell Collins (or whatever >> they are called today) and the avionics manufacturer is building to a single >> spec – MIL-STD-461 – regardless of what the end item platform might someday >> be. Few receivers are built for one particular platform, if only because the >> point of a radio is to establish a communication link, so that it ends up >> on multiple platforms. >> >> >> >> In your example of IFF, with transmit and receiver frequencies close >> together and no inherent filtering from the antenna itself, a certain amount >> of isolation has to be built into the radio itself. Much like a radar has to >> be able to reject any close-in reflection or crosstalk of the transmit >> signal into the receive path, in order to be able to receive the desired >> reflection from far away. >> >> >> >> But in term of setting a limit, the required CS103/104/105 reject ratio in >> the radio’s tunable range but outside the receiver bandwidth is 80 dB above >> MDS. That is pretty good performance. And out-of-band to the radio, with a >> limit at 0 dBm, you are looking at on-the-order-of 100 dB of rejection (MDS >> ~ -100 dBm). It seems to me that it is reasonable to place any required >> isolation beyond that on the integrator, either through the out-of-band >> rejection of the antenna (if the threat is far enough away from the >> antenna’s design frequency range), and/or if a broadcast pattern comm type >> antenna, placement so as to achieve antenna-to-antenna shading, or the use >> of antennas at higher frequencies with some sort of waveguide, even if it is >> just waveguide-to-coax conversion, providing for rejection of lower >> frequencies through waveguide-beyond-cutoff performance, or through >> operational controls or blanking. The integrator has various tools at his
Re: [PSES] Unreasonable Design Penalties on Receivers
Hi Michael I'm glad to hear more folks are thinking through the whole system, and whole environment, not just piece-parts. A component approach to requirement allocation makes sense when you procure a component. When you procure or integrate the entire system then the standard approach needs new thinking. Sometimes that's simply asking "why?". The why question brings out history and evolution. I find these to be invaluable in determining next- steps. Very glad to open these questions and have input from folks like Ken who helped drive years of evolution. Our company also worries about end-to-end performance. We delivery receivers and antennas, sometimes integrated, sometimes as piece parts. We carry the whole burden of RF front end performance. Questions like the ones you ask matter to us. They are carefully considered and discussed in our team meetings. We find our "customer" appreciates that we are thinking more about helping the war fighter accomplish their mission and return home than about pushing requirements out to someone else. Glad to know we are not the only ones! Keep up the good work! I've enjoyed this thread and look forward to more. -Patrick. On Fri, May 29, 2020, 9:08 PM Ken Javor wrote: > Well, you may be in the uncommon position of procuring a brand new IFF > transponder for your aircraft, and if that is the case, then as an > integrator you can levy any spec you choose, but in general, the military > buys certain models built by outfits such as Rockwell Collins (or whatever > they are called today) and the avionics manufacturer is building to a > single spec – MIL-STD-461 – regardless of what the end item platform might > someday be. Few receivers are built for one particular platform, if only > because the point of a radio is to establish a communication link, so that > it ends up on multiple platforms. > > > > In your example of IFF, with transmit and receiver frequencies close > together and no inherent filtering from the antenna itself, a certain > amount of isolation has to be built into the radio itself. Much like a > radar has to be able to reject any close-in reflection or crosstalk of the > transmit signal into the receive path, in order to be able to receive the > desired reflection from far away. > > > > But in term of setting a limit, the required CS103/104/105 reject ratio in > the radio’s tunable range but outside the receiver bandwidth is 80 dB above > MDS. That is pretty good performance. And out-of-band to the radio, with > a limit at 0 dBm, you are looking at on-the-order-of 100 dB of rejection > (MDS ~ -100 dBm). It seems to me that it is reasonable to place any > required isolation beyond that on the integrator, either through the > out-of-band rejection of the antenna (if the threat is far enough away from > the antenna’s design frequency range), and/or if a broadcast pattern comm > type antenna, placement so as to achieve antenna-to-antenna shading, or the > use of antennas at higher frequencies with some sort of waveguide, even if > it is just waveguide-to-coax conversion, providing for rejection of lower > frequencies through waveguide-beyond-cutoff performance, or through > operational controls or blanking. The integrator has various tools at his > disposal, and the placing of 80 – 100 dB of isolation on the receiver > itself seems quite sufficient a burden. > > > > Also if the EME is from a carrier deck, consider that at the same time the > EME is that intense, the signal to be received from the ship is much higher > than MDS, so that just plain old-fashioned frequency-independent > attenuation can and will be employed by an AGC circuit to keep the intended > signal in the linear range of the IF strip, and maybe even out front of the > mixer depending upon how sophisticated the design. > > > > If someone is trying to impose an untailored CS103/104/105 on a radio in > terms of the desired signal being MDS, while tailoring the reject band > limit based on a stringent -464 EME, I’d say they need to sharpen their > pencils some. > > > > -- > > Ken Javor > > (256) 650-5261 > > > > *From: *Michael Viau > *Date: *Friday, May 29, 2020 at 6:22 PM > *To: *Ken Javor > *Cc: *"EMC-PSTC@listserv.ieee.org" > *Subject: *Re: [PSES] Unreasonable Design Penalties on Receivers > > > > Hey Ken, thanks for jumping in. > > I realize that I’m talking to someone who knows more about this subject > than I ever will, so please read this as if someone with the humility of > Piglet wrote it. > > > > For some clarity, I am an integrator and we usually have a good idea of > the end product we are integrating into when we are procuring the receiver. > As well as at least a rough idea of the antenna characteristics and cables > at play. So it’s not too difficult to come up with the expected rough > estimate for the EME that the receiver would need to perform in well before > the receiver is on contract. > > > > What I’m really trying to ask is why the MIL-STD consid
