22 - Digitrax SE74 - Disappointments for basic turnout and signal control

[@jimworcester981]

I was also initially disenchanted with the SE74. My opinion has changed a bit now that I have about 6 of these devices installed between our club layout and my home layout. You are correct that the SE74 needs something to control the signal logic. I doubt that Digitrax could have come up with a device that would have all the variables anyone would want. JMRI does that well.....and for free! One point to make. The SE74 is essentially one half of the older SE8C board that did the same things (maybe a little better) but did twice as many. 32 signal heads and control for 8 Tortoises. I have come to appreciate that the SE74 doesn't do as many in that there was a lot of wiring that all had to come to the same location if using an SE8C. With only half as many circuits, the SE74's can be located around the layout, closer to the signals and turnout motors being controlled, thus cutting down on the wiring lengths and congestion!! As for the logic to make signals work, JMRI provides all the variety you need. There is no way that a pre-programmed circuit board could handle some of the signaling requirements found at the entrance to a station platform where you might have 3 or more turnouts in a row that needed to feed info to a variety of signal head configurations. We have that on the club layout and JMRI came through like a champ. One other comment since you mentioned the Smail as an alternative to the Tortoise. I use Smail's exclusively on my home layout and they are great. I like to be able to program the switch command code into the Smail so the number matches the number on my layout diagram. HOWEVER, the SE74, controlling Tortoises, provides feedback to JMRI as to the state of the turnout - closed or thrown - and you cannot get that from a Smail without some fancy circuitry (I use the Smail contacts, a resistor, and a channel on the BDL168 block detection board to indicate turnout position). On the negative side, it is almost impossible to get the desired operating switch number into the SE74 to control a specific Tortoise. I was so frustrated with the SE74 when I first started using it that I made a series of videos and put them on my channel to show all these applications. I also have a video on using the BDL168 and a resistor to monitor the position of a Smail. I feel your frustration that Digitrax didn't make a simple signaling controller, but it won't be the first time I've been frustrated with them. Once I finally resigned myself to learning and using JMRI, I found that many devices are available to accomplish our goals on the railroad. I am using Digitrax exclusively, warts and all!!

[@Vman7757]

You might look into LCC ( Layout Command Control ) to do what you want. And you don't have to change much of what you have already have done. It's relatively inexpensive. And takes traffic off the DCC bus.

[@gncascadedivision]

As I'd said in other comments, I've looked at LCC, and a bunch of other options. I can also happily program away in JMRI, I just don't want to do any of that :) The SE74 does exactly what it was designed to, sadly, which isn't what I want / expected / hoped it would do! I'm still trying to figure out a middle ground for signaling simplicity vs complexity. Iain

[@NEC_Amtrak]

I'll be using a few of those SE74s on my layout. I'm doing block detection, signals, etc. Didn't make sense for me to have both the DS74s and the SE74s. But you are correct, the SE74 is just two separate devices in one package. A DS74 and 1/2 of an SE8C.

[@gncascadedivision]

Yep, it seems like it's half of two devices, which is a bit puzzling other than Digitrax are moving away from raw boards like the SE8C and more into prettier plastic boxes that hide that away. I think there's value to it like I said in the video, and I will likely use a few elsewhere, but it could have been a lot more. I can make it work for my uses, or I can use alternative solutions, but for someone starting out or just wanting basic signal controls that doesn't have that extra knowledge, it's a shame the SE74 couldn't have fit their needs and that part of the market. Iain

[@jimworcester981]

As I said in my other comment I, too, am/was disappointed in Digitrax.Having done the work to install block detection and signalling on two large layouts, I must reiterate that signalling is VERY complex once you leave the isolated A1, A2, B and C element of a single turnout. Those can be controlled with the contacts on either a Tortoise or Smail, and done so far cheaper than any other solution. The problem is when you attempt to add signalling for situations beyond the simple thrown/closed turnout status. There would never have been a way to included all the possibilities on one simple circuit board without requiring considerable programming with CV switches, etc. The multitude of possibilities requires the use of a computer. Digitrax actually states that in their documentation. Once a person figures out JMRI, it is quite simple to set up all the variations you might want to use. If you are going to use JMRI at all, then DEFINITELY get into controlling the signalling with JMRI. Otherwise, just use the switch machine contacts!

[@ronjenkins549]

FYI there are about 30-50 signal logix. I don't believe that they could install all of that in the device, My signal is going to be different there yours. That why you use JMRI for the logix. This is my guess. Great video keep up the good work.

[@gncascadedivision]

Oh sure, but there are devices that can do that though, there's just a whole plethora of jumper switches you have to set to define what signals you have connected. My point in the video is that it wouldn't have been hard to go with basic signaling given Digitrax sell a set of signal heads and have a recommended layout for those around a turnout. They wouldn't need to have added 40-50 different types of logic, just one set to control that basic logic. Set an ops code to enable or disable it, and then use JMRI if you want to define your own and go beyond that basic logic. What they have with the SE74 is something that's trying to solve a valid problem and need, but is missing a part of the solution and requires a lot of extra work by the customer that purchases that device. Iain

[@WirenwoodModelRailway]

Hi Iain! I'm only posting this as a lot of what I've done on my layout was heavily influenced by what you did on Newvastle! But maybe look at some micro devices - ESP32 / Arduinos connected by Ethernet and using MQTT? That way all of your point motors etc can stay on DCC and you don't have to build a full layout diagram in JMRI, but when JMRI throws a turnout, it can send a message out on MQTT that it was thrown, or in the future that a block was occupied, and you can have a bit of logic in the ESP / Arduino to act on that - switch on LEDs etc. I've found Ethernet / MQTT to be far, far more predictable and reliable than the old RS485 bus. Regards, Chris

[@gncascadedivision]

Thanks, Chris. The whole point was to avoid all that though :) I'm certainly capable of going that route, and honestly, I was able to get the logic working in JMRI by using basic lights that alternate based on the turnout status via DCC. Wasn't that hard, and didn't need to do the whole layout diagram. But again, that's more than it needs to be, and then comes down the SE74 being an expensive device for still needing to that in JMRI. It's only really worthwhile IMHO if you _did_ make everything light up in JMRI including block detection, yellow caution signals, etc. The Tortoise and Smail switch machines can handle basic alternating LEDs for turnouts, so don't even need a microcontroller solution. Just adds a whole lot more wiring! The Digitrax solution looked like it was pretty slick with the signal driver cables and connector boards, but ultimately falls just short for the price. More to come soon. Iain

[@marsfromrexford]

Lain, have you ever looked at the Atlas ABS system? Seems like it more or less runs off turn out position but all this gets very complicated for me anyway, over my head at the moment. And it's very expensive. Great work! Mario

[@gncascadedivision]

Yes, the Atlas ABS approach is what I hoped / assumed the Digitrax SE74 would provide. Like I said in the video, there are devices that can provide this basic signal control, and Digitrax did the hard part that would take it 95% of the way there! It is an expensive device though for still needing to configure the signals in JMRI, but it's tough finding the balance between what it costs vs basic signaling by some other means. Iain

[@rayronvr]

Great video! I have a se74 and you are correct. It depends on what you want your layout to do . If you are a computer nerd like me, i want full contol of my layout from jmri. the se74 is a good tool but is limited if you want logic control outside of jmri. Have you thought about LCC? I know it can be pricey and adds another layer of complexity but does have logic control outside of a computer from my understanding.

[@gncascadedivision]

I mean, I have JMRI and used it extensively on my previous layout, and I have the signals working with the SE74 using JMRI, there's just no point in that. I don't want to spend that much time with JMRI when I can be building and running the railroad :) LCC is something I've looked at, but again, that then moves away from Digitrax equipment into a whole other realm when it's something that could have been implemented with the SE74 to make things easier. We make model railroading too hard and complex for people :) Iain

[@rayronvr]

@gncascadedivision yeap absolutely agree. havent seen a company be able to handle micro control and macro control flexablity that can be simple, affordable, and universal for DCC.

[@kevinbreger653]

Thank you for video. Anyone have trouble with using DS74 controlling routes and then causing loss of control of trains via loconet? (Trains stop responding to any commands from command station/throttle). Using all digitrax.

[@frankbeans289]

From my understanding "signaling" and "detection" logic go together, but control of the switch is supposed to be completely separate. You should not mix the two. So, it's not, in my opinion a "miss". Digitrax has exactly copied best practices of what is done in the real world prototypical rail signaling and control system. You are intended to pair the SE74 with a BDL168 or even better a BXP88...pairing detection with signaling. This is the main function and intent of this device. Since this device is already a DCC decoder they threw in some extra separate functions like being able to also "control" some motors/solenoids (i.e. the turnout/switch), and detect manual switch inputs as well, saving on wiring. The simple reason for this "best practice" is because if in the real world, the device simply set the signals based on the what it assumed to be the state of the switch, then if there is a mechanical failure of the switch, the signal could be in state B when the points are mechanically in state A. If that is what you want, then you could simply do that with the SPDT relays built into the Tortoise. However, this $120 device provides the ability to implement a lot more advanced prototypical signaling (i.e. ABS, APB, Interlocking, and CTC signaling). So, as you figured out, don't get this if you don't want to have advanced approach, limited approach, restricted, lunar, interlocking, and eventually CTC type signaling deployed in your layout. Just use the relays built into the Tortoise or whatever switch motor or solenoid you are using.

[@derekalexander4030]

Fortunately for us, JMRI is free. Unfortunately for us, Digitrax had the opportunity to create a stand alone signaling system and they didn’t. Smails can eliminate the need for the SE74 and they can run directly from the bus and they have direct wiring to push button control and eliminate a lot of extra wiring and the 12 volt bus for the tortoises. How are you doing in your recovery?

[@gncascadedivision]

Yes, as much as using a Smail with the SE74 is redundant IMHO, using the Smail by itself makes a lot of sense. The Tortoise won't let you use a push button, the Smail does. The Tortoise requires DC accessory bus power _and_ DCC for frog powering, the Smail can run just off DCC, etc. Follow up video coming soon that shows the Smail in action by comparison to the Tortoise. Recovery is slow, but getting there. Iain

[@QRCoal]

What it won't do is control slow motion DCC Concepts slow motion points. After many trials I eventually put my ten SE74 in a box and just left them their,

[@johnjohnston2962]

Model train technology has very user friendly controllers.

[@gncascadedivision]

They sure do, and like I said in the video, is why the SE74 is a bit of a disappointment and a miss from Digitrax. There are other products that can do the basic signaling, and the SE74 was 95% of the way there with the features it already is. So, I'll use them in some instances, but in the cases around the station platforms, it's a miss. Iain

[@marianosanchezramirez3352]

te agradesco que lo ayas praducido al español abemos mucha jente metida en los trenes en miniatura

[@gncascadedivision]

De nada!

[@woodalexander]

There are plenty of things to legitimately criticize Digitrax for on their newer products, but not having signaling logic in the SE74 isnt one of them. The only type of signaling that can really use on-node logic is ABS, and with LCC, its possible, although more complex than just doing it theough JMRI. Smails don't provide feedback to the system like the SE74 does. I'm not sure I'd buy much of anythijg Digitrax makes anymore due to their bugginess, poor documentation, and refusal to adopt NMRA standards of LCC abd RailCom, but your main criticism of the SE74 in particular just iant fair.