Correlator Noise floor inflation by Legacy Signal Interfering with DCS

rad400 posted:
John, Adrain if you need any readings/testing from our layout to help you guys out, let me know.

It'd be neat to take a multi-meter with capacitance mode and measure between center rail and earth and outside rail and earth and let everyone on here know what the numbers are on your giant layout. If our rough math is close it should be a few nF. My old radioshack DMM can do 0.1nF resolution but I don't know how common that is.

 

rad400 posted:
   Does anyone know why the "add engine" function seem to be harder to perform than many of the other functions?

If you guys want I can do a separate focused post on this with the details found. I have lots of data on the delay time vs different DCS voltages, estimated error rates and all that.  It'll take me a few days to write it up cleanly.

rad400 posted:

A temporary fix for the issue of adding an engine to a DCS handheld (no problem when using the DCS app or when the handheld is tethered to the TIU), I have installed a tether cable from the TIU to the perimeter of the layout so people can connect their handhelds directly to the TIU for adding an engine.   Does anyone know why the "add engine" function seem to be harder to perform than many of the other functions?

Bob, given that you're talking about the handheld remote to TIU communication here, I'd consider something like this for each TIU.

Adrian has posted a nice tutorial about adding a reverse-SMA connector to the TIU.  Add a nice 900mhz antenna and you should increase the range of the remote/TIU communications.  There are also many 900 mhz amplifiers that might extend the range, but they mostly seem to be targeting cell traffic.  OTOH, I'm not sure how they'd know the difference, so they might also help if a hi-gain antenna doesn't do it.

 

For the single Legacy command base for the layout, this would be a dirt simple addition to boost the range of it's remotes.

 Sunhans Sh-2500 2500mw Wireless Signal Repeater 33dbm Wifi Signal Booster 2.5w

 

Can I ask an experiment be conducted? See if either layout owners have some isolation transformers enough to plug everything into them instead of directly to the line. Yes, this should make NO difference. And that is the experiment. Would line isolation affect the problem? The other thing bothering me here is the "capacitive imbalance". Won't such imbalance always exist since there are 2 outer rails and only one center rail? 

I'm ham radio guy (N3RHT) so I have just enough knowledge to be dangerous to myself and those around me. 

Don Merz

Don M.

gunrunnerjohn posted: There are also many 900 mhz amplifiers that might extend the range, but they mostly seem to be targeting cell traffic.  OTOH, I'm not sure how they'd know the difference, so they might also help if a hi-gain antenna doesn't do it.

 

GRJ! I thought about this too.... but once I started draw a schematic it stoped making sense. The antennas on the TIU and remote have to both transmit and receive. So which way does the amplifier point? If it points outwards then it can transmit but not receive. If it points inwards it can receive but not transmit. If you put amplifiers pointing both ways its an oscillator.

The TIU radio board has a switch inside that switches between TX and RX. It's controlled by the FPGA on the motherboard of the TIU. The FPGA knows when it's transmitting so it switches to the TX port, and sits on the RX port otherwise.  Poor us outside have no knowledge of when should be which way. You'd have to lift the switch on the radio board, steal the control signal and wire to an external switch after you added amps. It's a lot of work.

Don Merz 070317 posted:

Can I ask an experiment be conducted? See if either layout owners have some isolation transformers enough to plug everything into them instead of directly to the line. Yes, this should make NO difference. And that is the experiment. Would line isolation affect the problem? The other thing bothering me here is the "capacitive imbalance". Won't such imbalance always exist since there are 2 outer rails and only one center rail? 

I'm ham radio guy (N3RHT) so I have just enough knowledge to be dangerous to myself and those around me. 

Don Merz

Yeah I think there's always capacitive imbalance practically speaking.... like maybe 1nF and 1.1nF or something. It's only going to be an issue if the difference leads to a differential voltage up in the 1-2V range, which would have to be a major mismatch.

Adrian! posted:
rad400 posted:
John, Adrain if you need any readings/testing from our layout to help you guys out, let me know.

1) It'd be neat to take a multi-meter with capacitance mode and measure between center rail and earth and outside rail and earth and let everyone on here know what the numbers are on your giant layout. If our rough math is close it should be a few nF. My old radioshack DMM can do 0.1nF resolution but I don't know how common that is.

 

rad400 posted:
   Does anyone know why the "add engine" function seem to be harder to perform than many of the other functions?

2) If you guys want I can do a separate focused post on this with the details found. I have lots of data on the delay time vs different DCS voltages, estimated error rates and all that.  It'll take me a few days to write it up cleanly.

1) I don't have a DMM that reads capacitance but I will but out an email to the club members to see if anyone else has one.  If not, it might be a good excuse to buy a new toy.   Let me see what I can do.

2) A separate post on DCS voltages would be beneficial.

Thanks,

Bob D

 

 

 

 

gunrunnerjohn posted:
rad400 posted:

A temporary fix for the issue of adding an engine to a DCS handheld (no problem when using the DCS app or when the handheld is tethered to the TIU), I have installed a tether cable from the TIU to the perimeter of the layout so people can connect their handhelds directly to the TIU for adding an engine.   Does anyone know why the "add engine" function seem to be harder to perform than many of the other functions?

Bob, given that you're talking about the handheld remote to TIU communication here, I'd consider something like this for each TIU.

Adrian has posted a nice tutorial about adding a reverse-SMA connector to the TIU.  Add a nice 900mhz antenna and you should increase the range of the remote/TIU communications.  There are also many 900 mhz amplifiers that might extend the range, but they mostly seem to be targeting cell traffic.  OTOH, I'm not sure how they'd know the difference, so they might also help if a hi-gain antenna doesn't do it.

 

For the single Legacy command base for the layout, this would be a dirt simple addition to boost the range of it's remotes.

 Sunhans Sh-2500 2500mw Wireless Signal Repeater 33dbm Wifi Signal Booster 2.5w

 

John - I already added the 900mhz antenna that Adrain had in his post to TIU # 1 which is located near the office (see below).  I took a new TIU and upgraded it to accommodate the 900 mhz external antenna and tested it out before adding it to the layout.  In testing, I was able to go ~ 80-100 feet away and have good communication (including add engine) with an engine on the test track in the office.  When I added the upgraded TIU to the layout, I was not able to perform the "add engine" function and I was only one foot away from the TIU??????  Start, shutdown, horn, whistle, direction, etc, etc all worked well  but not "add engine". Any thoughts?  TMCC was OFF when I did the testing. 

I am not sure by adding an additional signal booster would help me at this point, there is something basically wrong which I am missing and need to find.  Hopefully in Adrain's new post, I can better understand how the "add engine" command differs from the other commands.

Thanks, Bob D

 

IMG_5168IMG_5170

 

 

 

Attachments

Photos (2)
rad400 posted:
gunrunnerjohn posted:
rad400 posted:

A temporary fix for the issue of adding an engine to a DCS handheld (no problem when using the DCS app or when the handheld is tethered to the TIU), I have installed a tether cable from the TIU to the perimeter of the layout so people can connect their handhelds directly to the TIU for adding an engine.   Does anyone know why the "add engine" function seem to be harder to perform than many of the other functions?

Bob, given that you're talking about the handheld remote to TIU communication here, I'd consider something like this for each TIU.

Adrian has posted a nice tutorial about adding a reverse-SMA connector to the TIU.  Add a nice 900mhz antenna and you should increase the range of the remote/TIU communications.  There are also many 900 mhz amplifiers that might extend the range, but they mostly seem to be targeting cell traffic.  OTOH, I'm not sure how they'd know the difference, so they might also help if a hi-gain antenna doesn't do it.

 

For the single Legacy command base for the layout, this would be a dirt simple addition to boost the range of it's remotes.

 Sunhans Sh-2500 2500mw Wireless Signal Repeater 33dbm Wifi Signal Booster 2.5w

 

John - I already added the 900mhz antenna that Adrain had in his post to TIU # 1 which is located near the office (see below).  I took a new TIU and upgraded it to accommodate the 900 mhz external antenna and tested it out before adding it to the layout.  In testing, I was able to go ~ 80-100 feet away and have good communication (including add engine) with an engine on the test track in the office.  When I added the upgraded TIU to the layout, I was not able to perform the "add engine" function and I was only one foot away from the TIU??????  Start, shutdown, horn, whistle, direction, etc, etc all worked well  but not "add engine". Any thoughts?  TMCC was OFF when I did the testing. 

I am not sure by adding an additional signal booster would help me at this point, there is something basically wrong which I am missing and need to find.  Hopefully in Adrain's new post, I can better understand how the "add engine" command differs from the other commands.

Thanks, Bob D

 

IMG_5168IMG_5170

We got exactly that antenna on our 5 AGHR TIUs. $9.99 on Amazon right? 

Engineer-Joe posted:

as per help from Barry in the past adding engines, I erase other TIUs from the remote and only have a number 1 tiu added. The remote spends much... much less time looking for the new engine.

Do your remotes have multiple TIUs listed?

Yes, all the club and personal remotes have between 4 -6 TIU loaded and all are in super mode.  4 TIUs for the passenger & Freight and 2 TIUs for subway cars.  I was doing the testing with 4 TIU loaded in the remote.

Does the search for engines start with TIU 1 or is it random?  All my testing has been done on TIU #1.  

 

 

 

The way it sped up the adds, I'd have to believe that the remote starts with TIU #1.

Many users also tether the remote to add directly to #1 I've read. Maybe that would help here while you're in super mode?

I believe there was a software update to help out when the TIUs aren't powered and the remote was waiting for a response.

where's Barry???

" on Sour mash and cheap wine " ??

Why go back to DCC when I have DCS!

Adrian! posted:
rad400 posted:
gunrunnerjohn posted:
rad400 posted:

A temporary fix for the issue of adding an engine to a DCS handheld (no problem when using the DCS app or when the handheld is tethered to the TIU), I have installed a tether cable from the TIU to the perimeter of the layout so people can connect their handhelds directly to the TIU for adding an engine.   Does anyone know why the "add engine" function seem to be harder to perform than many of the other functions?

Bob, given that you're talking about the handheld remote to TIU communication here, I'd consider something like this for e 

 

 

We got exactly that antenna on our 5 AGHR TIUs. $9.99 on Amazon right? 

FYI- I used the antenna you recommended but used a different connector which combined the bulkhead and the shielded wire into one item. $6 for 2 of them and easier to install.

 

DHT Electronics 2PCS 20CM SMA female to U.FL IPX IPEX WIFI Cable for U.FL Mini PCI Card

 

 

 

Adrian! posted:
gunrunnerjohn posted: There are also many 900 mhz amplifiers that might extend the range, but they mostly seem to be targeting cell traffic.  OTOH, I'm not sure how they'd know the difference, so they might also help if a hi-gain antenna doesn't do it.

 

GRJ! I thought about this too.... but once I started draw a schematic it stoped making sense. The antennas on the TIU and remote have to both transmit and receive. So which way does the amplifier point? If it points outwards then it can transmit but not receive. If it points inwards it can receive but not transmit. If you put amplifiers pointing both ways its an oscillator.

Pretty sure some of these have intelligent routing that detects when you're keying transmit and switches modes.  I know the WiFi model I used for my old house even stated that in the documentation.  I use this specific model, the one I previously listed was a newer model, but Amazon automatically linked my order history to the new model.  This has no connection to the internal WiFi logic in the router, it just figures out the direction and switches automatically from transmit to receive.

Sunhans SH24BTANP

A key operating characteristic is...

Operation mode: Bi-directional, half-duplex, time division duplex

I'm pretty sure many of the 900mhz buffers would have similar logic, though I haven't actually used them. In any case, they can't really transmit and receive on the same antenna without having a significant frequency difference between the transmit and receive frequencies.  This option is certainly worth a look if the hi-gain 900mhz antenna doesn't do the trick.

Okay guys. At AGHR (2700 ish ft)  we got 74.8nF outside rail and  58.6nF center rail measured as is. Certainly more than the raw capacitance than I expected from the rail structure itself . I tried adding cap on the outside rail and DCS (about 13V excursion) is happy until we get to about 90nF .... so roughly 2:1 cap imbalance is where the problems seem to start.

Adrian! posted:

FEM is usually about 1-2% accurate to real world. I'm using HFSS 14 for this.

Well, I looked at this, and then I look at that...

Adrian! posted:

Okay guys. At AGHR (2700 ish ft)  we got 74.8nF outside rail and  58.6nF center rail measured as is. Certainly more than the raw capacitance than I expected from the rail structure itself . I tried adding cap on the outside rail and DCS (about 13V excursion) is happy until we get to about 90nF .... so roughly 2:1 cap imbalance is where the problems seem to start.

I guess in the real world Finite element modeling has it's limitations.

gunrunnerjohn posted:
Adrian! posted:

FEM is usually about 1-2% accurate to real world. I'm using HFSS 14 for this.

Well, I looked at this, and then I look at that...

Adrian! posted:

Okay guys. At AGHR (2700 ish ft)  we got 74.8nF outside rail and  58.6nF center rail measured as is. Certainly more than the raw capacitance than I expected from the rail structure itself . I tried adding cap on the outside rail and DCS (about 13V excursion) is happy until we get to about 90nF .... so roughly 2:1 cap imbalance is where the problems seem to start.

I guess in the real world Finite element modeling has it's limitations.

I guess that means the metal track to ground cap isnt dominant over the electronics cap. 

RE:Capacitance readings on a large layout

I had a chance to take capacitance readings (center rail to outside rail) on the NJ-Hi railers layout  and came up with an average reading of 25uf which is quite a bit higher than the above readings, but we run ~ 7,700 feet of track and probably 3x that amount of wire below the layout.  

Bob D

 

 

 

I have to wonder how valid those readings are.  The reason I mention this is that the TMCC Buffer uses a series .22uf blocking capacitor on the output, see below.  If this were trying to drive 25uf, that's a 100:1 ratio, there wouldn't be any signal at all on the track!  Of course, it would be even worse with the Legacy base direct to the tracks, they don't have nearly the drive capability of the TMCC Buffer.

Attachments

Photos (1)

I just realized by reading again that you were measuring center rail to outside rail.  That's a bogus value, we need to know the capacitance between the outside rails and earth ground!  Trying to measure center rail to outside rail is probably almost impossible, surely there are things connected across those like signals, etc.

gunrunnerjohn posted:

I have to wonder how valid those readings are.  The reason I mention this is that the TMCC Buffer uses a series .22uf blocking capacitor on the output, see below.  If this were trying to drive 25uf, that's a 100:1 ratio, there wouldn't be any signal at all on the track!  Of course, it would be even worse with the Legacy base direct to the tracks, they don't have nearly the drive capability of the TMCC Buffer.

I will take readings again next week when I am at the club again. I was using a Fluke 189 meter and took several readings around the layout, to capture different TIUs and all the readings read in uf.

I took my measurements between outside rail and center rail, but when I was rereading Adrain's post above he talks about an outside reading and a center rail reading (At AGHR  we got 74.8nF outside rail and  58.6nF center rail), are these reading referenced to earth ground?

 

 

 

rad400 posted:

John _ I sent off my updated post before I read your updated post.  I will take the reading next week with reference to earth ground.

Bob D

Hi there!

I'm pretty sure everyone knows this but just in case.... this measurement should be done with the entire layout (TMCC & DCS & Power) totally off. If there's any voltage to Earth ground (even inadvertent paths)  it will give funny numbers in a capacitance measurement.

Okay I finally had a chance to finish bread-boarding that DCS:Legacy switch module we talked about like 1000 posts ago on this thread that I've been working out for my SD3R colleagues.

The schematic is posted up there somewhere although it's evolved a bit (I'll clean it up and post it later). Its the filter and 74HC123 triggering off the DCS waveform, using the Broadcom HCPL-9000 galvanic isolator to translate the voltage reference from layout to Earth ground, then 4-parallel of a 74HC4316 switch to switch the legacy on and off.

Here are some overall results:

Overall view of DCS packet and Legacy "holes"

test1

Start of the packet:

entry

End of the packet:

exit

Finally here are actual measured DCS results comparing the "add-engine" time (average of 5 tries) with and without the module as the unbalanced legacy voltage on the track is slowly increased from 0 to 3V. The switch has leakage so it's not flat, but its a lot better than no switch.

real_measurements

The only thing left is I haven't seen if there's any effect on a legacy train (I don't have one). I may test this out at AGHR for a few weeks and see what happens before I print out a final PCB.

Anyways it looks like it does what it's supposed to.

Attachments

Photos (4)

The Legacy commands are repeated, and it would be way more complex to find a null point in the TMCC/Legacy signal stream, so I think that's a needless complication.  I suspect the TMCC/Legacy commands will get through just fine. 

Since the TMCC/Legacy carrier is constant and uses frequency shift encoding, it would be fairly complicated to find a hole.  Also, if you were trying to synchronize the two, it's likely you'd be missing DCS commands.  If there is a conflict, who gets priority?  Since the DCS commands only use a very small amount of the bandwidth, it's far more logical to interrupt the continuous carrier of the Legacy signal as opposed to the DCS packets.

rad400 posted:

John _ I sent off my updated post before I read your updated post.  I will take the reading next week with reference to earth ground.

That will be a far more useful measurement.

Adrian! posted:
...snip...

Finally here are actual measured DCS results comparing the "add-engine" time (average of 5 tries) with and without the module as the unbalanced legacy voltage on the track is slowly increased from 0 to 3V. The switch has leakage so it's not flat, but its a lot better than no switch.

 Interesting.  I guess one measurement we need is the amount of Legacy signal differential between the outside and center rail on a typical and a larger layout.

Adrian! posted:
...snip...
The only thing left is I haven't seen if there's any effect on a legacy train (I don't have one). I may test this out at AGHR for a few weeks and see what happens before I print out a final PCB.
 
Anyways it looks like it does what it's supposed to.

 Looks great, I'm eager to see how the Legacy reacts to the interruptions.  This may be an ideal companion box for the TMCC Buffer.  Sticking this in front should do the trick.

Add Reply

Likes (4)
Dave GarmanJTrainsTed SDJ's Trains
The DCS Forum is sponsored by


OGR Publishing, Inc. PO Box 218, Hilliard, OH 43026 330-757-3020
www.ogaugerr.com
×
×
×
×
×