I'm having some odd issue where the serial data coming out of the LCS SER2 box is not recognized by the DZ-2500 switch machines. I can see the data on my 'scope, and it appears to be reasonable, but it's totally unrecognized by any DZ-2500C switch machine! I switched to the DZ-2001L data driver and that's working with all 24 switch machines, so I know something is amiss with the SER2 data output.
I also don't get any serial data out of the CSM2 boxes, another oddity that I can't explain!
Replies sorted oldest to newest
@gunrunnerjohn - I have a vague memory of reading comments that the DZ2500 did not work well with certain flavors of Legacy connected via LCS SER2. One notion was that the tx/rx voltages were slightly off and need resistors placed on the data lines. The only reason I remember that is because I had to buy a voltage shifter for a 5V to 3.3V serial connection. That may very well not be your case, but I thought I would mention. My personal thought would be to make sure the expected flavors of settings match (stop bits, parity/no parity, ...). A
Is this is how you have them wired? You can’t just use the ser2, you need to add the data driver, either Zstuff, or Lionel’s new data driver for 2500 switch machines.
If you have an older zstuff data driver you may have to change out a resistor. I forget which one. As Wells mention but I think it highly unlikely you have an older version.
I have no problem driving them with the DZ-2001L, it's the latest version. I figured out the CSM2 boxes, they need to have aux power or they don't work. Odd that the documentation appears to suggest the block control is optional, so I figured I didn't need relay power.
Other than the CSM2, I know of no "new" Lionel driver for the DZ-2500 switch machines.
Gary, the SER2 IS the box that won't drive the DZ-2500C switch machines. They totally ignore it's output. I can see it on my 'scope, and it looks fine, but they don't work. Connect them to the DZ-2001L and they work.
Grasping at straws... probably of NO use... but, Found this...
Lionel changed the output current on the Legacy command base. That has screwed up the Z-Stuff DZ-2001 data wire driver. Dennis Zander of Z-stuff has figured out how to fix the problem by changing a resistor on the DZ-2001 board.
And this... DZ2500 - If all was well with TMCC... hook it up in parallel with the Legacy base connect the common to the Legacy base rather than the TMCC. This way you could use both CAB-1 & 2 and operate the switches with the signal being converted from Legacy to TMCC.
I don't have a TMCC base in the picture, but there's no problem with the DZ-2001L data driver. I think there's some issue with the data coming out of the SER2 serial terminals.
I'm scouring the net John! Someone, Somewhere has to have run into this before and hopefully posted it!!!
Update: Nothing! This is from the same post I was cribbing from before... this is the entire thread ...
@gunrunnerjohn posted:I figured out the CSM2 boxes, they need to have aux power or they don't work. Odd that the documentation appears to suggest the block control is optional, so I figured I didn't need relay power.
this bit me also, Emailed Dave Olson and found out they have to have aux power. He was very helpful if you reread page 13 of the manual it’s in there but easy to miss.
John I am confused how would the SER2 box drive the switch machines? That’s what the CSM2 or the data driver are for. Maybe I’m missing a piece of the puzzle
John, are you saying that you are trying to use the SER2 serial output without any DZ-2001 in the picture? If so, then is the problem that the ground references for the serial data are different?
The DZ-2001 is optoisolated from the Command Base ground (which should be earth ground, pin 5), and uses layout common (outside rail) for the ground reference.
I think the SER2 is just getting earth ground from the Legacy base, so wouldn't it have a different ground than the DZ-2500s?
Ignore the above if you're driving the DZ-2001 from the SER2, which should work.
The SER2 has an opto-isolated serial data output, so I gave it track ground. When I looked at the serial data signal on my scope, it looks identical to the one coming out of the DZ-2001L that I have, and it has the same ground reference. That being the case, it escapes me why it doesn't work, it has to have that ground reference to work with stuff like the TPC, etc.
In any case, I did get the CSM2 serial data working, you have to also provide it a common outside rail ground and power as well. That was not obvious, but careful reading of the CSM2 document does make a passing mention of powering it. Did I mention that Lionel documentation sometimes is left wanting?
@zhubl posted:John I am confused how would the SER2 box drive the switch machines? That’s what the CSM2 or the data driver are for. Maybe I’m missing a piece of the puzzle
Serial data should be serial data, and it's the same polarity and amplitude as what comes out of the CSM2. That's what is confusing, I can't see any reason it wouldn't drive them.
@gunrunnerjohn that makes since, with that in mind though I do remember reading that when the machines go in any given direction they send out a serial signal. I don’t know why that could’ve been used to show the switch positions over a STM2
OK, here's an odd one that I'm still trying to figure out! I fired up the layout this morning, all was well, switches all working in command mode, etc. Fast forward about 15 minutes, and I hit the master power to turn everything off to tinker with my TT wiring (no shorts, just fixing something). Turn the master power back on, and none of the switches are recognizing TMCC commands!
NOTHING CHANGED EXCEPT A POWER CYCLE!
I can see the LED's blinking on all the CSM2 boxes for each switch command, so I have to presume the commands are going out. I even checked one of the switch serial data inputs on the CSM2 breakout board with my 'scope and the data is there for the switch command.
If I power cycle a time or two, eventually switch operation will come back! Once of the reasons for separating the serial data between three CSM2 boxes was for debugging, and to localize any DZ-2500 switch machine that might be stepping on the serial data. However with this failure, ALL the switches stop recognizing serial data no matter which CSM2 they're getting their serial data from!
Anyone else see stuff like this with this combination?
I have had such issues but always figured I did something wrong. In most cases I gave up and came back the next day to find the switch machines working
Well, I may be doing something wrong, but I sure don't know what it is! I was swapping switch machines until I realize I was chasing ghosts! Now I don't know what to think! 
John, you mention your problem comes and goes with power cycles. Could it be a problem with your power strips your using?
Doug
Well, I’ll tell you one thing I’ve learned here; that’s if you’re having trouble figuring this stuff out, I’ve got no chance in “he??”. Guess it’s manual switching for me... 
John,
This sounds like the problem I had when I implemented your signal booster into my wiring, but it wasn't "sporadic." No matter how low I turned down the output on the device, my DZ 2500's just "froze"! I tried trouble-shooting and disconnected all Dat wiring from every switch. As I rewired the dat wire to each switch, the red light on the DZ-2001L would grow dimmer and dimmer with each switch added to the chain. After about 3 or 4 switches were connected there was total interference of signal to the switches. You and I tried figuring it out, but were both baffled.
I tried reaching out to anyone else on this forum at the time who incorporated your signal booster with DZ-2500's to see if there were similar issues. I had no responses and figured I was the only one who had this issue because everyone else was using different switch motors.
Greg
Greg, perhaps not helpful but I have some DZ2500 's on my layout as well as Tortoise machines. The difference is my DZ2500's are actuated using ASC2's. Adding GRJ's signal booster had no affect on the turnouts with that configuration since they see only a contact closure, not serial data.
John, can you draw up a diagram on how everything is wired?
I have about 20 DZ2500s spread as far as 45' away from the RW I use to power the DZ2500s and the LCS SER2. I connected the DZ2500 data wire directly to the SER2 screw terminal block OUT, then I have the SER2 screw terminal block COM wired to my layout common ground terminal block - RW is also connected to layout common ground terminal block. Then have a pair of wires distributing power and common to the DZ2500s. Thus the DZ2500s and SER2 COM share a common ground with all the layout.
I have a DZ2001, and used it on my CAB1 base, and then with my CAB2 base. When I decided to get the LCS WIFI, I got the SER2 also. I assumed, maybe incorrectly, that the SER2 buffered the Serial data, thus I decided not to use the DZ2001, and go directly from the SER2.
DARN! If John, of all people is having *issues* what's in store for me? I'm building my first layout in 25 years (30+ years since the build) i.e., pre- TMCC. I went all in on Legacy and now I'm sitting here with a box full of new Tortoise Machines contemplating my fate!
@gunrunnerjohn posted:The SER2 has an opto-isolated serial data output, so I gave it track ground. When I looked at the serial data signal on my scope, it looks identical to the one coming out of the DZ-2001L that I have, and it has the same ground reference. That being the case, it escapes me why it doesn't work, it has to have that ground reference to work with stuff like the TPC, etc.
John, FWIW I just checked - the SER2 "COM" at the output is NOT isolated from the PDI ground. It is connected.
So if you were connected track common to that port, you may have been shorting track common to the TMCC earth ground (pin 5 from the command base).
I don't know if that's the reason you couldn't drive your DZ-2500s from the SER2, but that's the condition the DZ-2001 is designed to prevent.
@Professor Chaos posted:So if you were connected track common to that port, you may have been shorting track common to the TMCC earth ground (pin 5 from the command base).
I don't have anything connected to the SER2 COM terminal at this time, I abandoned trying to use that.
@MED posted:I have about 20 DZ2500s spread as far as 45' away from the RW I use to power the DZ2500s and the LCS SER2. I connected the DZ2500 data wire directly to the SER2 screw terminal block OUT, then I have the SER2 screw terminal block COM wired to my layout common ground terminal block - RW is also connected to layout common ground terminal block. Then have a pair of wires distributing power and common to the DZ2500s. Thus the DZ2500s and SER2 COM share a common ground with all the layout.
That's exactly how I tried to use the SER2 output, and I got crickets. The serial data was never recognized at all by any of the switch machines.
@trnluvr posted:John, you mention your problem comes and goes with power cycles. Could it be a problem with your power strips your using?
I deliberately bought industrial power strips with NO internal surge suppression for that very reason. They are plain-Jane power strips with nothing but a circuit breaker switch.
Back when all the LCS boxes came out it was discussed in length you still needed to use the Zstuff data driver with the SER2. Can’t remember why been years since this last came up. Hence my pic. Maybe a email to Dave Olson or Dennis of Zstuff is in order.
Gary, I get identical behavior with the DZ-2001L data driver powering all the switches, but I may hook it up again out of desperation!
Dennis,
I don't think you will run into the problem that John is having with dz2500 switch machines when using a tortoise machine. The wiring is totally different. I am using asc (older version) controller. As a help in reducing the amount of wire you need I put the diodes at the asc and then ran a power and a single wire off the diode pair to the tortoise machine.
Marty
The Tortoise machines will be rock solid, and I'm kinda' wishing I had gone a different route on mine! 
You have already thought of this - I would take a switch I haven't installed yet and hook it up to a transformer and SER2. I would suggest the common ground from transformer to SER2 to DZ25000. Eliminates as many variables as possible. Once I got this working okay, I would copy how I wired the test set up to the layout, but add only a couple switches on the serial data line. Since you used the CSM2, breaking the serial data path is not super difficult. Get that to work, add a couple more DZ2500s, repeat. As you can guess, you keep adding till you can't get it to work, and you may have identified a DZ2500 or CSM2 or wiring to them that is causing your problem.
My thought process behind the common ground between track and SER2 and DZ2500 power was that I needed a common ground between SER2 and DZ2500 for DZ2500 to can see the serial data, and common ground between track and DZ2500 so that the DZ2500 can detect a train when the train applies ground the green or yellow wire via isolated rail for derailment prevention.
Well, I broke out the DZ-2001L data driver and sky-wired it in, and this time so far it hasn't failed to work for a power cycle in about ten tries. That being the case, I'm going to say that this issue is very likely something odd about the CSM2 units, they would fail ever two or three power cycles consistently. I apparently wasted a lot of money on the CSM2 boxes, I wonder what the odds are that the other functions work?
The really odd part is the data coming out of the DZ-2001L and the CSM2 looks identical on the 'scope. AAMOF, even the data coming out of the SER2 looks the same as well, and I never got that to work at all!
I'm not sure what all this means, but I'm going to make some serial cables and wire the DZ-2001L in permanently unless it fails me in the next day or two. I'm tired of chasing phantoms!
Well, final word is in. 
I got to talk to Jon Z. for a spell about this issue, and it turns out it's not totally unknown that the DZ-2500 gets testy about talking to the CSM2. Jon talked to Dennis Zander about it, but there was no fix forthcoming. The speculation is that for some reason when powering up the DZ-2500 sees something on the serial input that makes it think it's in the wired remote control mode and it stops listening to TMCC commands. That would explain why all of them stopped listening to the serial data and not just some of them.
In any case, I think I may have one more card up my sleeve. I suspect that delaying the power being applied to the DZ-2500 switch machines for a short time (a second or two) would likely allow any serial data stuff to settle and not to confuse the DZ-2500 mode selection. I may try that before I throw in the towel totally on the CSM2 serial data.
@martind posted:Dennis,
I don't think you will run into the problem that John is having with dz2500 switch machines when using a tortoise machine. The wiring is totally different. I am using asc (older version) controller. As a help in reducing the amount of wire you need I put the diodes at the asc and then ran a power and a single wire off the diode pair to the tortoise machine.
Marty
@gunrunnerjohn posted:The Tortoise machines will be rock solid, and I'm kinda' wishing I had gone a different route on mine!
Thanks Guys!!!! My anxiety just dropped to 6.7!!!!
I am some what confused. (Nothing new there) If you can get it to work without the CSM2, what is the advantage to getting it to work with it. Seems like a huge waste of income if you can get them to work without it. Just asking as I am not fully understanding.
John, I wish I had 1/10th of your knowledge. I've definitely spent some time and money on my train obsession that did not pan out... did I waste it ...I don't think so. Education isn't free... I'm still learning and I'm hopeful that I will continue to learn until... well, we all know when *until* is.
ROCK ON JOHN! Let's have a bourbon!
@CurtisH posted:I am some what confused. (Nothing new there) If you can get it to work without the CSM2, what is the advantage to getting it to work with it. Seems like a huge waste of income if you can get them to work without it. Just asking as I am not fully understanding.
The data driver adds another component to the mix. Also, I'm using the CSM2 for switch positions, (and I plan to), I prefer the serial data come from the same source, it will eliminate any common ground issues.
@Dennis-LaRock posted:John, I wish I had 1/10th of your knowledge. I've definitely spent some time and money on my train obsession that did not pan out... did I waste it ...I don't think so. Education isn't free... I'm still learning and I'm hopeful that I will continue to learn until... well, we all know when *until* is.
ROCK ON JOHN! Let's have a bourbon!
I wish I had all the time and money that I spend chasing down rabbet holes, but that's another topic. Yep, eventually the lessons sink into even the densest heads, I'm a case on point. 
Make it a Maker's Mark Manhattan and I'm in Dennis.
@CurtisH posted:..... If you can get it to work without the CSM2, what is the advantage to getting it to work with it. ....
The CSM2 has a feedback loop that allows you to know the DZ-2500’s position using an iPad/similar device when using Lionel’s Layout Control System (LCS) software.
Yep, and I plan on using LCS on the system once I get the basic operation laying flat.
@CAPPilot posted:The CSM2 has a feedback loop that allows you to know the DZ-2500’s position using an iPad/similar device when using Lionel’s Layout Control System (LCS) software.
Thanks Ron. Now I finally understand why someone would want to add this costly device to the equation. For me, the issues John is having have reinforced my decision to go with the multiple data wires (for DZ-2500 debugging purposes). Layout control software or reporting of switch positions was never a requirement for me.
John - I hope you get a suitable resolution to this.
George
I believe I have a workable solution by using the DZ-2001L Data Driver, but I'd like to have all the switch functions consolidated in the CSM2. I can use the data driver and the CSM2 together and have it functional. Since talking to Jon Z. about the configuration, I believe there is a solution at hand using only the CSM2, I just have to pursue it.
Before the CSM2 there was discussion on how to use the STM2 for switch position sensing. You still needed to buy an LCS module (STM2) plus a DZ-1008 relay. Here was one possible way to do it:
Here was another:
The CSM2 was to make it easier.
Again, I have no problem using the DZ-2001L Data Driver with the CSM2, never said I did. I just would rather have it all integrated in fewer components. That also leaves me a spare data driver in the case that something does go amiss with a CSM2 box.
John,
I agree what you said earlier, that the tortoise would be a lot easier, other than putting them in upside down, and about half the price of the dz2500. I also tried the dz2500 and dz1000 with the asc1 and never had much repetitive success getting them to work.
Marty
I have set my switches to use the "BLUE" wire but I dont think I will.
Instead I will have a nice control panel with two LEDs for each turnout showing the position of the turnout and a momentary switch hooked to the DZ-2500C circuit board that will then connect to the DZ-2500C motor.
I feel that I can't control 100 or so DZ-2500Cs quickly enough on a handheld or iPad. Sorry too old!
So the panel will work for me.
I am planning a couple of "local" control panels to handle sections so I can quickly change switches in a section of the layout. My layout also is only 24x12, with 24 switches, so not nearly as complicated as 100 switches. I probably couldn't even find 100 switches on the iPad, how many pages of the LCS drawing does that take?
@gunrunnerjohn posted:... how many pages of the LCS drawing does that take?
I am finding that as an issue for my plan. Solution: for the switches that are for switching and within easy reach will now be manual using Caboose Industries ground throws. For switches not in easy reach will be controlled only by a DZ-2502 controller mounted on the layout's side closest to the switch. Still working on the best way to show them on the LCS app.
Switching that would be controlled by a central dispatch (iPad) for group operating sessions would have the LCS setup.
Access to this requires an OGR Forum Supporting Membership
