Hornby 2023 - Bluetooth decoders and control system

[Will77]

Hi,

     I ve got an electronics background and am working on a PCB with multi function control of RGB LEDs for red white mainline and shunting activities as the model is so short on space Im probably going to spin my own daughter board which will also fit the HM7000 21 pin.  I ve heard alot of people getting frustrated by the lack of keep alives avaible at the moment so was thinking I could add in the adition circuitry to my PCB then disperse the caps throughout the loco as normal.

I ve not got a Hornby R7377 so cant take one to pieces to look at the electronics has anyone done so or better has a photo of whats inside it other than the three 2.7V Supercaps?

Any help would be much appreciated.

 

 

[RAF96]

9 hours ago, BrakeCoach said:

Has anyone put one of the sound versions into the Hornby Railroad Duke of Gloucester?

 

Thinking of getting a few of these for my models but not sure if the old ones without tender circuit board housing would fit. I would have to put both the decoder and the speaker into the boiler.

My DoG TXS installation was a straight swap for the previous TTS instal. Originally used a Brit profile, then loaded DoG onto it.

I have TXS in a MN with the decoder in the smoke box and the smallest enclosure sitting on the chassis step.

[Michanglais]

I'm sure this has been asked a hundred times before but is there any way of resetting the decoders back to factory settings so that I can sort out the mess I'm in with assigned decoders not running in correct models. 

 

I've tried on the four decoders I've got. I don't really want to write them all off.

 

Any help would be gratefully received!

 

Cheers,

 

Michael

 

[RAF96]

14 minutes ago, Michanglais said:

I'm sure this has been asked a hundred times before but is there any way of resetting the decoders back to factory settings so that I can sort out the mess I'm in with assigned decoders not running in correct models. 

 

I've tried on the four decoders I've got. I don't really want to write them all off.

 

Any help would be gratefully received!

 

Cheers,

 

Michael

 

Standard reset procedure, write value 8 to CV8 or do it from the app. I can't show a screen grab as I don't have any decoders alive at present.

If the problem is just sound related then value 5 resets the sound but leaves other CV settings as is.

[BrakeCoach]

As per the Hornby website and in the PDF called "Available Profiles for HMDCC", it seems like sound files for the Class 91 and the APT is also in the works, with the former being set to Q1 2024 and the latter in Q2 2024. Good stuff.

 

This was last updated in mid-March of this year, so I wonder how far they have gone through them.

Edited April 25 by BrakeCoach

[BrakeCoach]

Here's the page I am referring to by the way.

https://uk.Hornby.com/hm7000/hm-dcc-guides

[leonk]

Can anyone help with which pads are C- and C+ on the 21 pin TXS decoder? 

On the 8 pin they're clearly marked, but on the 21pin nothing is clearly marked.  I can't tell which pads to solder the stay-alive wires to.

[NotofthiscenturyTim]

1 hour ago, leonk said:

Can anyone help with which pads are C- and C+ on the 21 pin TXS decoder? 

On the 8 pin they're clearly marked, but on the 21pin nothing is clearly marked.  I can't tell which pads to solder the stay-alive wires to.

Funnily enough I was also puzzling over that yesterday. Looking at high res photos of the board, the conclusion I came to was that there aren't any pads on the 21 pin version as it's a different board to the 8 and 18 pin ones which share the same layout. 

[RAF96]

7 hours ago, leonk said:

Can anyone help with which pads are C- and C+ on the 21 pin TXS decoder? 

On the 8 pin they're clearly marked, but on the 21pin nothing is clearly marked.  I can't tell which pads to solder the stay-alive wires to.

pins 16 and 20 according to the famous and comprehensive never read manual.

Edited April 30 by RAF96

[NotofthiscenturyTim]

1 hour ago, RAF96 said:

pins 16 and 20 according to the famous and comprehensive never read manual.

Does connection to pins 16 and 20 use the same onboard diode and resistor as the power bank socket and the C+ and C- solder pads? 

 

The (much-read in my case) manual is silent on that. 

 

It's also silent on the electrical characteristics of said diode and resistor or what values of capacitor could be connected to C+/C- and whether any components other than capacitor(s) are needed when making a direct connection to them.

 

I haven't been able to locate a teardown of the official Hornby power bank but the odd comment implies that it contains components other than just capacitors, despite the decoder containing an onboard diode and resistor which you'd typically find as part of a powerbank connected to decoder positive and ground/pins 16 and 20. 

 

Also unclear from the manual whether the decoder's clever tech to prevent the powerbank interfering with programming will work if a powerbank is connected across pins 16 and 20 rather than the socket or C+/C- pads. 

 

Rather than providing some useful factual specifications alongside a "you will void your warranty if you tinker" caveat, instead Hornby's "comprehensive manual" resorts to pure backside-covering: 

"We can not advise re “third party” or “home brewed” types of “Stay Alive” solution…. We do not guarantee their operation and cannot officially endorse their use.
If you decide to fit a solution you may invalidate the guarantee on the decoder for the reasons above. "

[leonk]

6 hours ago, RAF96 said:

pins 16 and 20 according to the famous and comprehensive never read manual.

The famous manual I got the C+ and C- references from?

 

Besides, I asked about the pads on the board, not the connector.  The locos motherboard doesn't break those pins out as far as I can see.

[Michanglais]

I was wondering whether anyone else was having the following issue:

 

I've currently got Next 18 Bluetooth decoders in both powered and dummy car of a TT HST.

 

The powered unit makes all the right noises and functions perfectly. However, the dummy remains obstinately silent. The lighting functions are all fine. I've tried a few different speakers in both the side of the decoder and in the special extension socket provided next to the speaker. Nada. 

 

Any ideas for what's going on/fixes, please?

 

Cheers,

Michael

 

[WIMorrison]

Does the dummy car have any load on the motor pins of the dummy car decoder? 

[Michanglais]

3 hours ago, WIMorrison said:

Does the dummy car have any load on the motor pins of the dummy car decoder? 

 

Good question. I'm not quite sure how I'd find out, tho. Would I not need some kind of measuring device?

[RAF96]

TXS decoders need a load to provide the ACK for programming.

Why should Hornby provide details of their power bank circuit. No one else does. Any competent electronics guy could reverse engineer it anyhow.

As previously stated several times in forum the decoder MCU controls a delay to prevent the power bank charging initially to allow programming, and this is the purpose of the transistor in the power bank. When triggered it opens the charging circuit. The diode and zenner diode are standard circuit design for stay alives and can be found in other makes.

[SteveM666]

On 01/05/2024 at 02:39, leonk said:

The famous manual I got the C+ and C- references from?

 

Besides, I asked about the pads on the board, not the connector.  The locos motherboard doesn't break those pins out as far as I can see.

I think you just answered your own question.

[meatloaf]

I decided to reflash my class 73 with the class 20 sound file. Its very good, better than the TTS class 20.

 

Very frustrating tho, that after almost 12 months of the decoders being available- that the class 73 profile still has issues writing to the decoder. Now on the 3rd attempt......

EDIT : Third time lucky :)

Edited May 3 by meatloaf

[lapford34102]

Hi,

Not certain if this has been mentioned before but have hardwired a Lais stay-alive to an 8 pin sound chip in an 08 and it works fine. However not yet tried to read it via my sprog but will do.

 

Stu

[RAF96]

11 hours ago, lapford34102 said:

Hi,

Not certain if this has been mentioned before but have hardwired a Lais stay-alive to an 8 pin sound chip in an 08 and it works fine. However not yet tried to read it via my sprog but will do.

 

Stu

There have been reports of alien stay alives being used successfully, some with plugs and some hard wired, providing they have charging circuitry included.

There does not appear to be any problem with programming as the decoder MCU enables the charging circuit after a time lag which is longer than a programming burst.

 

Class 73 profile is the largest file so any glitch in the loading process is more likely with this one, whether the download from the Hornby server to the app or the bluetooth transfer to the decoder.

Edited May 4 by RAF96

[meatloaf]

3 hours ago, RAF96 said:

There have been reports of alien stay alives being used successfully, some with plugs and some hard wired, providing they have charging circuitry included.

There does not appear to be any problem with programming as the decoder MCU enables the charging circuit after a time lag which is longer than a programming burst.

 

Class 73 profile is the largest file so any glitch in the loading process is more likely with this one, whether the download from the Hornby server to the app or the bluetooth transfer to the decoder.

Ive got one of the AE models stay alives in the 73 ands it works just the same as the Hornby one does

[ColinB]

On 04/05/2024 at 08:32, RAF96 said:

There have been reports of alien stay alives being used successfully, some with plugs and some hard wired, providing they have charging circuitry included.

There does not appear to be any problem with programming as the decoder MCU enables the charging circuit after a time lag which is longer than a programming burst.

 

Class 73 profile is the largest file so any glitch in the loading process is more likely with this one, whether the download from the Hornby server to the app or the bluetooth transfer to the decoder.

Stay Alive should have no effect on the reprogramming operation for the class 73, we are talking bluetooth so the Stay Alive will be fully charged  no matter what make you use. The only issue you get with "Stay Alive" is in the programming mode in DCC, because the programming time is a lot less than the charge time of the Stay Alive. The important bit with the "Stay Alive" is whether it has intelligent charging, signified by a zener diode and a transistor/fet, the cheaper ones don't.  

[ColinB]

On 02/05/2024 at 08:46, RAF96 said:

TXS decoders need a load to provide the ACK for programming.

Why should Hornby provide details of their power bank circuit. No one else does. Any competent electronics guy could reverse engineer it anyhow.

As previously stated several times in forum the decoder MCU controls a delay to prevent the power bank charging initially to allow programming, and this is the purpose of the transistor in the power bank. When triggered it opens the charging circuit. The diode and zenner diode are standard circuit design for stay alives and can be found in other makes.

I hate to do this but your post is adding an awful lot of confusion. Now if you are saying transistor in the powerbank delays charging, that is nothing to do with the MCU because there is no separate feed to it, that is good old analogue electronics. Now you could control it by switching the return path of the "Stay Alive" using a port on the processor and a suitable high current driver. Now if you look how Zimo do it on their decoders (the ones with the facility) they do true control with the MCU because all you add is a suitable capacitor. Perhaps we should just accept it works and LaisDCC and AE ones do as well.

[cravensdmufan]

1 hour ago, ColinB said:

Stay Alive should have no effect on the reprogramming operation for the class 73, we are talking bluetooth so the Stay Alive will be fully charged  no matter what make you use. The only issue you get with "Stay Alive" is in the programming mode in DCC, because the programming time is a lot less than the charge time of the Stay Alive. The important bit with the "Stay Alive" is whether it has intelligent charging, signified by a zener diode and a transistor/fet, the cheaper ones don't.  

Probably most folk already know this but when prompted by the app to remove the loco from the track when programming a decoder, if it is fitted with a "Powerbank" (or presumably any "stay alive"), it has to be off the track for at least 60 seconds before going to the next stage. I gave mine a little longer. I had  no problem loading the 73 file.  

[ColinB]

41 minutes ago, cravensdmufan said:

Probably most folk already know this but when prompted by the app to remove the loco from the track when programming a decoder, if it is fitted with a "Powerbank" (or presumably any "stay alive"), it has to be off the track for at least 60 seconds before going to the next stage. I gave mine a little longer. I had  no problem loading the 73 file.  

I still don't understand why the guys at Hornby didn't incorporate a reset command in their software. What we used to do when we reprogrammed a vehicle module is to issue a reset command at the end of the reprogramming session, on most micros you can do this, so all you do is create a command to tell the micro to do this. It saves on switching things on and off, the controlling program does it when it has finished . I suppose they just didn't have the experience to know it is a good idea.

Edited May 5 by ColinB

[NIK]

1 hour ago, ColinB said:

I still don't understand why the guys at Hornby didn't incorporate a reset command in their software. What we used to do when we reprogrammed a vehicle module is to issue a reset command at the end of the reprogramming session, on most micros you can do this, so all you do is create a command to tell the micro to do this. It saves on switching things on and off, the controlling program does it when it has finished . I suppose they just didn't have the experience to know it is a good idea.

Hi,

 

This is just a guess but perhaps they didn't incorporate a reset command as it might have been used as a fast way in to hack the decoder/programmer system without leaving a physical trace on the decoder?.

 

As an example if the Nazis had incorporated a 'Say Again' operation into the process for communicating the new wheel and patch board settings for Enigma machines and the Allies found out about it they would have been able to gain info about the wider system.

 

Regards

 

Nik

Edited May 5 by NIK
Adding explanatory info.