Відео

Thank you. Just taking advantage of modular design principles ;-)

Thank you very much! I just tried to add some useful functionality that you typically do not see in a booster. And of course, providing around 50 Amps from one booster (although 6 modules) might be sort of unique.... Better make sure you have real powerful DC supplies!

Thanks. Glad you enjoy it! More to come ;-)

Unfortunately the Node RED server that generates the messages is down right now. After moving to our new home I was running into a problem starting it up again, I probably need to set it up from scratch, but did not have the time to do so (it runs on a Raspberry Pi and I am not that familiar with Linux, so it needs some time to do it). It will come back, but I don't have a date right now.

In theory yes, there are a few points you need to verify though. First the IO pins for the DCC signal. The Tinkerface Shield feeds them to pins 2 and 3, but you can remove the jumpers and feed the input to the pins your motorshield needs. If the motorshield does not support driving the H Bridge from two inputs, you can use only one, but in this case you can not generate the cutout for RailCom, probably a minor problem. The second point to verify is that the motorshield can't exceed the maximum analog voltage when you do a short circuit test, it should have a zener diode protecting the current sensor pin.

Watch videos #133 and #137 for information including schematics of the Tinkerface Shield. For more information about the PowerShield, including schematics, watch video #109

Yes. Any DCC command station will work. The Plus version will also be able to receive commands to control the outputs via DCC or Loconet switch commands. With Loconet, it is also able to send back some track status information encoded in switch commands.

Yes, in this example I am using a standard motorshield and power it from VIN so the VIN trace remains intact. VIN, when using the DCC AUX shield can be powered with up to 18 Volts and it provides 5V to the Arduino board as well, so the onboard LDO of the Arduino is bypassed and does not heat up. Watch video #130 for more information how the AUX shield works.

You are correct in your observation that the PurpleHat stops the profiling process if the locomotive is not moving after the first two or three step increases (indicating a problem with the loco setup or hardware). The same is the case if the next speed step does not result in a higher physical speed (indicating the maximum speed of the loco has been reached), or if the end of the test track is reached before a next higher speed step can be tested (indicating the test track is too short). So, in your case I think you should try to find out why your locomotives are not moving at low speed settings. Unless something is wrong with the decoder itself, I would look at the track voltage as well as mechanical problems inside the loco's power train (motor, gears, grease, dirt) and also verify the CV settings of the decoders. Since you are saying that several locos with the same decoder have different running characteristics, I would first checking the power train and the CV settings. It is in your best interest that all locos run smooth starting with the first speed step, and at least in my experience, that can be achieved. Once you have that right, speed profiling should no longer be a problem.

Yes, if you use the SilverHat along with an IoTT Stick you can control the polarity of the reversing output using a configurable turnout address. You can use the event handler on the Stick to listen to a block detector event and then issue the turnout command to set the reversing section polarity. Watch video #136 for an explanation of the configuration options.

Thank you. As mentioned in the video, it is a re-run I did because the original video got a music license write-up. It's funny, I used the same license free music for more than five years, and all of a sudden, it's no longer ok to use it for monetized videos ;-(

@@IoTT Oh no!. Well, glad it finally made it out.

Or combine it with a good IMU and you can actually use it to draw your track plan. Watch here: ua-cam.com/video/xracbWHlD6M/v-deo.html

I assume you mean Linux? If you are running JMRI on Linux, it is no problem for the Stick to communicate with it.

The Stick can connect to JMRI either via Loconet over TCP or WiThrottle protocol. Simply activate the respective server in JMRI and configure the IoTT Stick accordingly. Both protocols can to speed profiling. To write the CV values of the speed table into the decoder you have the choice: When using Loconet over TCP you can directly write it using Ops or Service Mode programming. WiThrottle only supports Service Mode. Alternatively you can save the table data in a file and import it into JMRI to do the programming of the decoder.

That's what clocks and timers are for ;-)

Yes, you should. It's really nice if you have an idea at what scale speed a train is going based on your throttle position!

Assuming you cut the VIN trace on the motorshield between the two pads next to the Arduino header, you can simply restore it by soldering the gap. That is good for more power than the standard motorshield typically can deliver. After restoring the connection, you can use a single power supply the the DCC AUX DC jack und supply it with up to 18V for your track voltage. The DCC AUX shield then takes care of supplying 5V and VIN to the Arduino board, bypassing the LDO on the Arduino board. Just make sure all pin header pins are correctly connected to the board below.

@@IoTT Thank you very much! That's what I needed as info

The PowerShield is using two BTN7960 H-Bridge chips. that is the same chip as on the pretty popular IBT-2 boards.

Thanks for sharing. I think I have one of the PM42s sitting around somewhere. I guess I need to give it a closer look!

So far, the update file has support for Windows and Apple. I'm pretty sure it would be possible to do a script for Linux, I have just not much experience with that OS. But if somebody can send me a Linux script to run the update utility I am happy to integrate it in future updates.

Thanks. I'm currently thinking about doing it. No promise, but good probability ;-)

The SilverHat can do both, acting as auto reverser, and of course over current and short circuit protection. But to do that, it relies on the current measurement capability of the PowerShield. With only Arduino and Tinkerface shield you can't measure the current and you also have no way to interrupt the power delivery. If you add a PowerShield, it is just a complete SilverHat booster, which btw I plan to officially introduce in one of the next two videos.

Thanks. I was not away, it's just this video was a nightmare to do. It was originally in the makings as Video 140, but then it was ways too long so I decided to only do the PowerShield part. Then I planned it for a week ago, but in the process of making it I figured I needed some more testing, so it was back to square one. Finally it's here now.... The circuit breaker on the power supply side of the command station is really only for the wiring between the PS and the CS. The command station could do without. But for sure it improves safety, i.e. reduces fire risk, in case you have a problem in the wiring. 10 Amps or more can be quite intense...

You need a PurpleHat kit and a Stick. On your end you need a car to install the kit on and a digital command control system that can receive Loconet or WiThrottle commands, e.g. via JMRI, which has servers for both protocols. That's it.

Yes, the CVs for speed and acceleration are the same. Some decoders do not support values of 0 for mid and high speed in non-table mode, but you can use default values like 128 and 255 to prepare the decoder for speed profiling. Check your decoder manual for the best values.

If it shows New_IoTT that is when it is ready to do the setup. If it shows IoTT_Stick.... it means the setup timed out and it initialized its own access point. You need to restart to get back into the connection mode. It seems that the Stick Plus2 is somewhat more difficult to connect to than the older versions. I have been working on some changes which will come with the next release. What I found that helps is if you walk away from the Stick maybe 10 feet or so when you try to open the captive portal for the setup.

@@IoTT Hans, I tried moving away from the Stick and didn't have any luck. I also noticed that the Broadcast SSID (New_IoTT-Stick_3029699340) was no longer visible in the available networks list on my phone. I will look forward to your changes in the new version.

If you can't set the values to zero, you should be able to set them to 1, 128, 255 which should also give a "neutral" speed curve. If you do that, you also need to set CV 29 to no longer use the speed table. Doing a configurable delay to allow for longer accel time after direction change would be possible. The problem is that it increases the needed test track length or you have to go back and forth more often. The standard way to deal with the problem is to set the accel and decel CVs to zero for running speed profiling, so that the loco has immediate reaction. What scale are you modeling? Are you using very heavy locomotives?

@@IoTT Thanks for the reply. I'm running HO locos and the ones I'm testing with are Scale Trains, which probably are on the heavier side. I have a 10ft track to run them on. Even with the Accel/Decel set to zero, they still don't start up to the previous speed really quick. If I watch the graph, it dips each time it changes direction, with the dip larger the faster it's going...

@@stephenwinter3759 I checked the code. At this time, the wait time after direction change of reaching speed is 500ms plus 4ms per speed step, so if the speed step is 20, this would be 580ms before the measurement. I could make the 4ms multiplier configurable. Can you try to do an estimation what delay times would be needed? Or can you maybe send me some screen shots from the graphs you are getting?

@@IoTT I did a screen capture during one of the runs that shows the dips and backtracking of the speed. ua-cam.com/video/NsQiECYsQ28/v-deo.html If you need more, I can grab the loco back form the club and setup my test track again.

The built in upgrade conversion normally only work one version step at a time, so upgrading so many steps is difficult, particularly if you try to load the entire stick configuration. You have a much better chance if you try to save and load only one category at a time, e.g. the RedHat configuration, and store it in a separate file. To do that, click the save button on the RedHat page and store it in e.g. rhcfg.json. Then try to load just this file into the stick with the new software version. Chances are it works. And if not, it is much easier to compare the old file with the new and make the changes in a text editor. Hope that helps.

I did not particularly test for it, but did not notice any difference for various rolling stock. I tend to think the usually used plastic material is reflective enough even in black color (steam engine shells typically are not made from carbon... ).

If it is because of my accent, click CC to read the captions. If it is because you have no idea what I am talking about, oh well, then this video is not for you.

Thank you. The advantage if downsizing. Move to a considerably less expensive home and you have some money left to build a shed ;-)

Check videos 17, 18, 19 for an example how to define and use an array to store positions etc. They also show how to program the status machine to control several servos simultaneously.

Thank you. On the way soon...

It's just a shed... But thank you anyway ;-)

The captive portal does not really work with the PC browser, it is best used from a smart phone. Phones have the tendency to change the access point if they get a better signal somewhere else. What you can try is to configure your phone to not automatically connect to your standard AP while you are connecting the Stick to the AP. You can do that by disabling the auto connect check box for the default AP in your wifi settings AP list.

@@IoTT Thanks Hans but my phone will not connect. I have a Samsung A13 which seem to be very slow is scanning the available WiFi networks. Then as I said before it tries to connect but by which time the stick has created it's own access point. I then tried an old phone which will connect but doesn't come up with the captive portal. How do I connect to the captive portal if I don't get any notification box up when I connect? I am using the new Plus2 Sticks which has been programmed with the M5UpdatePlus2 and I have tried two of them with the same result. I did try loading the non Plus2 code and it connected and came up with the notification box but when I touched it a message was shown saying no web page. I am at a loss to why this will not work.

@@davidtwist3659 Some phones take a little longer to detect the stick. In the latest software version I have therefore increased the timeout to 4 minutes before it sets up its own AP. What I also noticed is that sometimes it is better to walk a few meters away from the Stick when doing the wifi setup.

@@IoTT Hans thank you for the reply. I reprogrammed the M5StickC Plus2 again and removed the network from the phone. Placed the stick on a long power lead to remove it away from the computer noise. I did manage to get connected and I got the notification. I clicked the notification and I got the message Web Page not available. I reprogrammed another Stick Plus 2 and I got the same Web Page not available message.

Hans, I have an update. I managed to set up the WiFi by using an older version of the Stick software and then setting up the WiFi with no issue. Then I reprogrammed it to the latest and for the Plus2 Stick. It would look like there is something wrong with setup web page on the latest version.

After observing this for some time, I think the value of train identification has been over-estimated and that is probably the main reason that it is not being used in a broad scale. What is the convincing use case? In manual operation you identify the train by looking at it, so you don't need it. In automated operation, the system needs to keep track of the location of each train anyway and that is quite easily done using the track model and simple occupancy detection (the degree of freedom for trains is quite limited, the location only depends on the travel distance and some turnout positions). So, train ID in the best case would serve as confirmation. The 3rd use case would be convenience functions like automatic assignment of a loco to a throttle, but that is technically not possible because a decoder needs to be refreshed (therefore assigned to a throttle) before it can send the ID. So, what value proposition remains? I don't see any.

@@IoTTIn operations, it would be identifying trains to the dispatcher on the panel. For automation, train identification would allow for a train to go outside of the detection areas under manual control, come back, and be identified when it got back to detected territory, or a train to show up in the automation software when it is placed on a staging track. Somehow, the LoDi system and ESU detect a locomotive when it is first placed on the system, even if it is not being controlled by a throttle. Also, for fault recovery, block-tracking systems go completely haywire as they lose track of where everything is.

Hi John, I ship USPS. The postage you pay is shown when you place the order. I don't know if Canada charges VAT for shipments from the US, but that would be on top. Hans

Sorry for that, I am still in moving mode and everything is very temporary. Will get better sometimes in May.

Hopefully you spend your time printing rolling stock ;-)

It really was. But we have to go with the facts....

I like to add that all the tabs are easy to open, except the PurpleHat. It always takes several attempts

If the SpeedMagic tries 3 speed steps and the speed is not increased, the process is terminated. This is the case e.g. if you try to create a profile for a speed the loco can't reach. The same might be true in this case. You should make sure that the loco starts moving at the first speed step. Most decoders have a kick start feature that helps breaking loose the anchor of the motor. For further analysis it would be helpful if you could have a look into the loconet commands. Check the Loconet viewer of the RedHat. This would help to better understand what's going on.

@@IoTT Thanks Hans, Since your site is temporarily down, I couldn’t read your manual. Instead I watched the videos several times. I tried to stick to your instructions and one of them was to zero everything out before the test. Thank you for clarifying the test requirements for me.

Thank you. Regarding the instructions, the idea is to have them on the web page. Unfortunately, the old page got corrupted when I tried to upgrade the CMS, so I have to rebuild it. Of course you can message me in discord or send me an e-mail if you have questions I can help with.

@@IoTT Great, thanks Hans!

Thank you. For the notebook power supply, the same principle applies: Nominal current of the booster plus 10-20%. So, if the nominal current of the PS is 4 Amps, set the trip current of the command station to max. 3.5A. As for the noise, I am still filming in a very temporary setup in the garage after moving. To avoid echo, I had to open the door half way up, so you can hear that stupid bird out there and an occasional car going by. And at one point it started raining. The good news is I got the building permit for my shed, so things will get better over time ;-)

Thank you.

Great idea! Much safer!

Many thanks!

The principles are the same, so there is not really a need for it. Connect a strong enough DC power supply to the power shield(s). The AUX shield can than be powered from a separate power supply or from on of the power shields using the jumper that feeds the voltage to VIN (as long the voltage is not more than 18V).