Defeating "RF out of range" with light RF engineering?

[PA] Here's a 100W 900 MHz LDMOS from Freescale in case anyone wanted to run their trains from a neighboring city.

Hello Adrian.............Welcome to the forum and thank you for all the great information. Most of which goes right by me. But I do enjoy reading your posts. This is all good, keep up the good work, just maybe MTH will learn something. By the way I live 10 miles from Adrian Michigan. and I enjoyed my visit at AGHR. Matt is a terrific ambassador of your club. 

Clem

Adrian: you have a real gift for technical writing!  Even though I'm a conventional runner, your posts are educational and very interesting. Keep 'em coming!

When the system displays a message like that, I believe there could be many reasons. It is or could be more than just the actual out of range that it states.

I'm sure you know this, have you tethered the remote? (take the antenna out of the equation)

Does that cure all of the problem?

When exactly are you getting that message?

I have good communication with the stock antennas. The first remote I bought, I enhanced the antenna and the TIU's antenna too, on the recommendation of someone else. Since then I have compared it to the stock set up and have found that placement of the TIU is more important.

Adrian! posted:

[PA] Here's a 100W 900 MHz LDMOS from Freescale in case anyone wanted to run their trains from a neighboring city.

I think you need an FCC broadcast license for that one.

Adrian,

Have you tried the following from page 162 of The DCS Companion 3rd Edition?

TIU Modification 

If everything appears to be in good operating condition and there are still sporadic problems with TIU to DCS Remote communications, it could be that the TIUs are physically placed differently than was suggested earlier (refer to Part III - DCS Implementation, TIU Channel Assignment and Placement earlier in this book). If this is the case, there's a relatively simple modification that can be made to the TIU that can increase its range for communicating with the DCS Remote. 

Note that this modification may void the MTH warranty for the TIU and should only be attempted by knowledgeable individuals or MTH authorized technicians. 

To perform the modification, proceed as follows: 

• Disassemble the TIU as described earlier in this section and locate the transceiver board. It is the small circuit board that's plugged into the left, front part of the main TIU board when viewing the TIU from the channel input side. 
• While gently but firmly holding down the transceiver board, raise the antenna (black wire attached to the transceiver board) so that it is standing straight up. Do not under any circumstances lift up on the transceiver board! 
• Locate where the wire would touch the top half of the TIU's case and drill a 9/16" hole in the top half of the case at that spot. 
• Replace the top half of the TIU's case with the antenna wire protruding through the hole. Ensure that the transceiver board does not lift up while replacing the case! 
• Reassemble the TIU as described earlier in this section. 

The range of TIU to DCS Remote communications should be improved.

This and a whole lot more is all in “The DCS Companion 3rd Edition!" This book is available from many fine OGR advertisers and forum sponsors, or as an eBook or a printed book at OGR’s web store!

Sounds great, will you update us when you give this a try?

Barry Broskowitz posted:

Adrian,

Have you tried the following from page 162 of The DCS Companion 3rd Edition?

TIU Modification 

If everything appears to be in good operating condition and there are still sporadic problems with TIU to DCS Remote communications, it could be that the TIUs are physically placed differently than was suggested earlier (refer to Part III - DCS Implementation, TIU Channel Assignment and Placement earlier in this book). If this is the case, there's a relatively simple modification that can be made to the TIU that can increase its range for communicating with the DCS Remote. 

Note that this modification may void the MTH warranty for the TIU and should only be attempted by knowledgeable individuals or MTH authorized technicians. 

To perform the modification, proceed as follows: 

• Disassemble the TIU as described earlier in this section and locate the transceiver board. It is the small circuit board that's plugged into the left, front part of the main TIU board when viewing the TIU from the channel input side. 
• While gently but firmly holding down the transceiver board, raise the antenna (black wire attached to the transceiver board) so that it is standing straight up. Do not under any circumstances lift up on the transceiver board! 
• Locate where the wire would touch the top half of the TIU's case and drill a 9/16" hole in the top half of the case at that spot. 
• Replace the top half of the TIU's case with the antenna wire protruding through the hole. Ensure that the transceiver board does not lift up while replacing the case! 
• Reassemble the TIU as described earlier in this section. 

The range of TIU to DCS Remote communications should be improved.

---

Hey,

Yeah our guys tried all the stuff from the book already but the bottom line is the board design itself is kinda weak. A10dBm transmit power and a -69.2dBm floor is kinda measly for the size of our layout (about 50 feet in a non-line-of-sight channel condition. If you look into the link budget for wifi router at 2.4 GHz, the typically transmit about 20-25dBm and have a Rx floor around -89 to -95 dBm.

Playing with the antenna around can buy you 3-4dB in propagation loss maybe, but from the RSSI pin I looked at, we're about 20dB in the hole on SNR, so big picture changes to balance the link budget are the probably best avenue... good old engineering common sense is to attack the root causes...

Engineer-Joe posted:

When the system displays a message like that, I believe there could be many reasons. It is or could be more than just the actual out of range that it states.

I'm sure you know this, have you tethered the remote? (take the antenna out of the equation)

Does that cure all of the problem?

When exactly are you getting that message?

I have good communication with the stock antennas. The first remote I bought, I enhanced the antenna and the TIU's antenna too, on the recommendation of someone else. Since then I have compared it to the stock set up and have found that placement of the TIU is more important.

Well I went straight to transceiver and can directly see the RSSI is too low in the TRF dinosaur chip (at least for our layout). It's saying you know like -70dBm range, so the SNR will blow since the input referred floor is right there.... You'd want this to be -50 dBm range for a reliable link. In some cases we're standing a good 60-80 ft away and behind a wall.

Again since WiFi is typically 15dB more output power with a 20dB better floor (like 2000x better link margin total) it's sort of intuitive that the radio is under designed... or more like designed with a smaller range in mind than we actually have with better channel conditions than we actually have.

JTrains posted:

Adrian: you have a real gift for technical writing!  Even though I'm a conventional runner, your posts are educational and very interesting. Keep 'em coming!

Thank you. Much appreciated! I really work hard to learn to write "well-ish". I'm not an english major... but my PhD advisor used to beat on me and make me revise the journal and conference papers like 100 times each, and I think my PhD dissertation went through 300-400 revisions. It really paid off though, especially when I'm preparing the lecture notes or writing proposals for NASA and stuff. I always say technical writing is the most important technical skill!

I worked in aerospace for many years.  We used to joke that the job's not done until the paperwork weighs more than the airplane.

Adrian,

Welcome to the Forum and thank you for your information on this subject. Well written, informable and entertaining all at the same time. Looking forward to your results.

I also enjoy reading your posts, but I'm built too low, it all goes right over my head.* Maybe some of this will rub off one of these days if I keep following long enough? Anyway, I look forward to seeing your future posts and project updates, I'll be following from below.

* Famous quote by Foghorn Leghorn talking about Henery the chicken hawk.

Adrian! posted:

Hey,

Yeah our guys tried all the stuff from the book already but the bottom line is the board design itself is kinda weak. A10dBm transmit power and a -69.2dBm floor is kinda measly for the size of our layout (about 50 feet in a non-line-of-sight channel condition. If you look into the link budget for wifi router at 2.4 GHz, the typically transmit about 20-25dBm and have a Rx floor around -89 to -95 dBm.

Playing with the antenna around can buy you 3-4dB in propagation loss maybe, but from the RSSI pin I looked at, we're about 20dB in the hole on SNR, so big picture changes to balance the link budget are the probably best avenue... good old engineering common sense is to attack the root causes...

The panStamp AVR2 radio that I use with my RTC program, uses the next gen TI CC1101. It replaced the TRF6901.

The CC1101 does work better. Its transmit level is 12 dbm and the receiver is spec'ed at -112 dbm.

I've never gotten it to work reliably much below -90 dbm. That's for a CC1101 to CC1101 link using a TIU emulator that I wrote. When you connect to a real TIU, you lose some of this range.

Not sure if this is useful to you or not.

Interesting post, that is one of the most annoying messages

I have gotten "out of range" messages when I'm right next to the TIU.

It all depends on what's wrong.

Engineer-Joe posted:

I have gotten "out of range" messages when I'm right next to the TIU.

It all depends on what's wrong.

Sometimes the "out of range" message has nothing to do with the RF signal. It just means that the communication timed out. I have a second TIU that is offline currently and I get "out of range". Well, of course, it isn't powered up. There are several scenarios with corrupt data in the remote as well.

George

Range can be a problem  on large layouts. We have TIU in super at 4 different locations with the antennae modification. (sticking it straight up- through the tiu) and placing the tiu above the layout. The room is 80 feet long and it works reasonably  well.

I also fly RC airplanes and can buy cheap transmitters and receivers that have a range of at least 1/2  mile or so. (2.4 )

  It seems the DCS remote needs a little work to increase the range but I would have to have it already installed. I could never replace an IC chip(or whatever is needed) unless it was a straight plug in....   Hope you're on to something.

George S posted:

Engineer-Joe posted:

I have gotten "out of range" messages when I'm right next to the TIU.

It all depends on what's wrong.

Sometimes the "out of range" message has nothing to do with the RF signal. It just means that the communication timed out. I have a second TIU that is offline currently and I get "out of range". Well, of course, it isn't powered up. There are several scenarios with corrupt data in the remote as well.

George

Yup many things can be wrong ... agree completely. For my case when I monitor the rssi it's definitely to low so at least my problem is definitely RF signal and SNR. If I have this issue I guess it's reasonable to assume at least some others have it too. 

Gregg posted:

Range can be a problem  on large layouts. We have TIU in super at 4 different locations with the antennae modification. (sticking it straight up- through the tiu) and placing the tiu above the layout. The room is 80 feet long and it works reasonably  well.

I also fly RC airplanes and can buy cheap transmitters and receivers that have a range of at least 1/2  mile or so. (2.4 )

  It seems the DCS remote needs a little work to increase the range but I would have to have it already installed. I could never replace an IC chip(or whatever is needed) unless it was a straight plug in....   Hope you're on to something.

My thinking is sort of the handheld remote is a world class pain to modify.... you need external power and all kinds of stuff so it's better to help the sensitivity of the dinosaur chip in the TIU. Of course sensitivity can't be improved infinitely since the noise floor of -174dBm per Hz is the bottom line of the universe at room temperature ... but the TI chip falls many many dB short of that.

Also im not insulting TIs work when I say dinosaur chip.... it's just very very old to me... they did the best they could with what was there at the time.... the first RF chip I designed was for WCDMA generation .... like iPhone3s era .... so it's all way before my time 

the reason sticking the antenna up helps is that you eliminate the ground plane condition. If you want to understand it totally.... In the book I posted earlier you can read about dipoles and monopoles radiating over ground planes... it's somewhere in there. That only buys you about 2-3dB (like a factor of 2) in SNR so it may not be enough...

Adrian

sent from my iphone7

Adrian,  A little snarky with the adjectives no?  It is an old design so it is what it is.  Most people do not have an issue with TIU range.  Certainly for large club layouts it can.  But your starting to add $100s of $$ to the TIU, which is ok for the Club, but Joe consumer won't pay $600 for a TIU, plus the fact that it has remained backwards compatible is good no?  You can take that transceiver out of a Rev G and use it in a Rev L.

So just the facts.  You don't want some one doing that to your designs in 20 years  You do have a budget at NASA I assume.  G

I think the bottom line here is there a reasonable mod that can be done to the existing TIU to improve the range.  Obviously, given the timeframe it was designed in, if you were starting over you could probably improve the range.  However most of us, me included, would like not to have to spend $10,000 to have better range.

GGG posted:

Adrian,  A little snarky with the adjectives no?  It is an old design so it is what it is.  Most people do not have an issue with TIU range.  Certainly for large club layouts it can.  But your starting to add $100s of $$ to the TIU, which is ok for the Club, but Joe consumer won't pay $600 for a TIU, plus the fact that it has remained backwards compatible is good no?  You can take that transceiver out of a Rev G and use it in a Rev L.

So just the facts.  You don't want some one doing that to your designs in 20 years  You do have a budget at NASA I assume.  G

Sure!

I know it's kinda costy to drop $100 mmics everywhere. I think I mentioned that at the top.

If I was to do the design from scratch i'd go with some 802.11.ac chip like the bcm43460 or something similar where the wireless link is rock solid. The post's thinking is more about making the best of the situation you have without a ground up design meaning you can only change the perimeter and not touch the core circuits... 

While there's some great stuff from the pre wifi era (like srrs) I find a lot of it is icky with outside filters, dual PLLs and too many conversion stages. I'm usually the biggest critic of the SoCs that I've served as the chip lead on.  We have a wall of "bad design decisions" on the lab wall actually. They serve as reminders for next time!

Basically whats being suggested is:

If the tiu is in front of you and the reception is poor in the condition you are operating it under, and you've tried the existing suggestions, and you aren't in a position to redesign the radio, then external amps to help a limited rfic are an option to consider

I guess I could have been clearer for anyone else reading this.

I only got "out of range" messages standing right next to the TIU when the problem was something else than reception. I rarely have reception issues. I didn't know that at first. I have learned that from troubleshooting.

How would a new home user know that the message generated you posted above, was from something else? I just wanted to share to new guys reading this post, that the message can pop up for other reasons. If they have only one TIU and one remote, they could follow Barry's book and try tethering the system and other simple fixes. I would bet that a very high percentage of end users, don't need to modify anything. Just mounting the TIU up so it's vertical, seems to get the same results as the TIU antenna mod in the book.

 I believe you could modify the antennas to eliminate reception issues. For most users, this would be unnecessary. A lot of readers here can get the false impression that this mod is necessary from your post. I fell for the same thing when I started with DCS. It ended up to be from the wrong wiring and lack of understanding how the system works.

 I now have four TIUs and three remotes. I have three separate layouts and could be four if I count the 3rail loop as another. I only modified the very first setup after seeing this message and not knowing better. Since then, they all stay stock. Each time I have had issues, something else has been the culprit.

 It seems to me that adding engines can cause this message when the reception is not the problem. Poor habits like just pressing "add" triggers it for me.

Many clubs have a separate TIU, test track, and tether the remote, just to add engines for use there.

My posts here are only my guess work and I am just a home user.

SanDiegoMark posted:

Adrian! posted:

Hey,

Yeah our guys tried all the stuff from the book already but the bottom line is the board design itself is kinda weak. A10dBm transmit power and a -69.2dBm floor is kinda measly for the size of our layout (about 50 feet in a non-line-of-sight channel condition. If you look into the link budget for wifi router at 2.4 GHz, the typically transmit about 20-25dBm and have a Rx floor around -89 to -95 dBm.

Playing with the antenna around can buy you 3-4dB in propagation loss maybe, but from the RSSI pin I looked at, we're about 20dB in the hole on SNR, so big picture changes to balance the link budget are the probably best avenue... good old engineering common sense is to attack the root causes...

The panStamp AVR2 radio that I use with my RTC program, uses the next gen TI CC1101. It replaced the TRF6901.

The CC1101 does work better. Its transmit level is 12 dbm and the receiver is spec'ed at -112 dbm.

I've never gotten it to work reliably much below -90 dbm. That's for a CC1101 to CC1101 link using a TIU emulator that I wrote. When you connect to a real TIU, you lose some of this range.

Not sure if this is useful to you or not.

Mark's idea is a really solid one too.

I just read the TI sheet in detail. That transceiver is pretty good for the signals the remote works with. I like the -112 dBm... thats close to ideal.

It's channelized to 0.5 MHz = 0.5x10^6 Hz --> 10log10(B) ->56dB... -174 dBm/Hz + 56 dB = -117dBm so a 5dB noise figure (difference between thermal floor and receiver floor) which is pretty good. The 12dBm is a bit tiny still but probably workable with such high sensitivity. An option for this appeared out of thin air too if you wanted to be absolutely sure never to have RF reception issues again:

By sort of accident in researching Mark's suggestion I also found this part (cc1190) which is a non connectorized version of what I was suggesting out of discrete amplifiers. It's a packaged chip with LNA and PA together with 0.5W (27dBm) of output power and a crazy good 2.9 dB noise figure. It's part of the simplelink product line from TI, which is designed by people that know stuff (I see them at IMS and in the IEEE-MTT research journal all the time). My first impression is that it has a look and feel of a CMOS chip, but deep down  I'd bet it's a MMIC (probably GaAs).  That 45-50% PAE is a big clue...way too good for a CMOS PA.

Anyways, to top it off TI offers an eval board with a combo of this booster and with the TRX chip Mark is suggesting (Combo board) so you don't even have to do anything except pull the IF signals out of the radio board and tap into the eval board. It's $59, so it won't break the bank either.

Again... I'm not saying everyone needs it, I'm saying if you really have reception problems like I do (I'm sure on my own situation anyways)... it's an option.

Also: TI wont ship that to the US, but their vendors sure will. The usual suspects (digikey, newark, mouser) have them for sale...

I bought those boards from Spain a long time ago and threw them on the back shelf somewhere.

Engineer-Joe posted:

I bought those boards from Spain a long time ago and threw them on the back shelf somewhere.

Because they blow?, because they are too complicated? or because you didn't need them?

What's the actual frequency of the DCS remote to TIU link?  I notice these boards talk about the 869 MHz band.  Of course, you'd also need one set of these for each TIU...

Mark's page has the info. I remember having to order with a note to set/change the frequency or something like that?

Why? Because I'm really over my head here with some of this stuff. Sorry.

  I found them and can't find the paperwork yet to say what I ordered! I don't even remember why 2 of each?

maybe someday I'll understand this jargon:

http://www.silogic.com/trains/OOK_Radio_Support.html

I read something about programming the set-up and never completed it. The original set-up by Mark works on the TIU it's plugged into just like MTH's WIFI box does. I bought the MTH WIFI that came out as I was building this stuff.

Mark's program actually reads the track much faster and is set up very well for running trains. The next step for me would have been to use it on a tablet or another computer for each layout. I would suggest anyone try it to use for people who are experiencing problems troubleshooting a large layout. Instead of having the MTH remote just use a generic message that there's a problem, Mark's program will read the track thru the TIU and show if there's an address conflict for example when adding engines.

So I use Mark's program and the TIU loader for trouble shooting when there's errors or to see what's wrong with the system or engines. It's nice to see the exact problem to help fix it.

anyone can also read Ray's pages for TIU antenna upgrade:

http://www.rayman4449.com/MTH_DCS_Info-Tips.htm

http://www.rayman4449.com/MTH_...Antenna_Modification

Adrian! posted:

GGG posted:

Adrian,  A little snarky with the adjectives no?  It is an old design so it is what it is.  Most people do not have an issue with TIU range.  Certainly for large club layouts it can.  But your starting to add $100s of $$ to the TIU, which is ok for the Club, but Joe consumer won't pay $600 for a TIU, plus the fact that it has remained backwards compatible is good no?  You can take that transceiver out of a Rev G and use it in a Rev L.

So just the facts.  You don't want some one doing that to your designs in 20 years  You do have a budget at NASA I assume.  G

Sure!

I know it's kinda costy to drop $100 mmics everywhere. I think I mentioned that at the top.

If I was to do the design from scratch i'd go with some 802.11.ac chip like the bcm43460 or something similar where the wireless link is rock solid. The post's thinking is more about making the best of the situation you have without a ground up design meaning you can only change the perimeter and not touch the core circuits... 

While there's some great stuff from the pre wifi era (like srrs) I find a lot of it is icky with outside filters, dual PLLs and too many conversion stages. I'm usually the biggest critic of the SoCs that I've served as the chip lead on.  We have a wall of "bad design decisions" on the lab wall actually. They serve as reminders for next time!

Basically whats being suggested is:

If the tiu is in front of you and the reception is poor in the condition you are operating it under, and you've tried the existing suggestions, and you aren't in a position to redesign the radio, then external amps to help a limited rfic are an option to consider

I understand your point, and what you  are doing to make the TIU better for large clubs.  Good News.  I still think you miss my point....  you called the design engineer lazy, you uses negative adjectives about the chosen hardware, as if  toy train manufacture creating something totally new and unique to toy train control, 2 decades ago, has/had the budget of a Government organization....like NASA.

Kind of like the comments people make about QSI ProtoSound.  A design from the late 80s that did more than many conventional trains can do still today, but gets held to today standard of Legacy and DCS Command control.  Really an Apples to Orange comparison.  Just tell us how to improve the TIU and Remote.  G

Engineer-Joe posted:

I bought those boards from Spain a long time ago and threw them on the back shelf somewhere.

The boards that Joe is referring to are not the TI CC1101-CC1190 eval boards but the panStamp radio boards that are made in Spain. These boards have an Arduino-like processor and a CC1101 radio. They come fully assembled and ready to plug into a PC USB port. My RTC progam uses them. How to order and details are on my OOK web page.

Engineer-Joe posted:

Why? Because I'm really over my head here with some of this stuff. Sorry.

  I found them and can't find the paperwork yet to say what I ordered! I don't even remember why 2 of each?

Joe, Probably sometime after you bought these panStamp radio boards, I convinced Daniel, the guy who sells them in Spain, to make fully assembled versions available. My OOK web page has details and pointers to his web page. The two smaller "boards" on the left are the actual panStamp radios. The two larger boards are panStick boards which have a place to solder the panStamp, a place for an antenna connector and a USB interface. Now available fully assembled.

Mark, thank you for responding. I got lost reading the page about the need to program the boards once built?

Engineer-Joe posted:

Mark, thank you for responding. I got lost reading the page about the need to program the boards once built?

Yes, since the panStamp boards contain an Arduino-like computer, you have to load some program code into it.  The loading procedure is described here:

http://www.silogic.com/trains/...Support.html#XLoader

The XLoader program that loads the panStamp and the program code needed, "RTCModem", are in this zip file:

http://www.silogic.com/trains/...nStamp%20Uploads.zip

Adrian,

Great information, and I think this looks promising for the folks that actually need more than the typical solutions offer.  I also don't see anything wrong with calling a dinosaur a dinosaur.  When you have a 16+ year old product that used a a 15 year old radio when it was brand new, I don't see anything wrong with offering an idea for improving it's signal, or calling it out for being plainly old.  Sadly there are folk that push back hard any time anyone suggests there may be a better way to do something than the way DCS does it, and it keeps many folk from participating in discussions on the DCS forum here.  Just know that many people do appreciate it when people come up with ideas to improve things.  

JGL