Britannia Builder

I have a Hornby Class 47 TTS which I'd planned to upgrade to TXS when the sound profile became available, but in view of the above comments I decided to download the 47 profile to another TXS loco temporarily so that I could compare the TTS and TXS sounds side by side. I thought the new TXS horn sounds both on F2 and F3 were very feeble compared with the TTS low-high and long-high horns. No doubt both sets are authentically recorded from real Class 47s, but the TTS sound recorders caught one on a better day! I'm in two minds now as to whether to upgrade this loco - better motor control and F28 etc weighed against frankly rather feeble horns. Regards, John

I can't find an export profile option either? One limitation of the CV restoring that I described above is that the CVs relating to auto-calibration can't be restored manually - they are read-only on the CV screen in the app - so it is necessary to re-run auto-calibration after a profile update. Regards, John

Updating a sound profile appears to reset all the decoder CV values to the default for that profile (except that the DCC address in CV1 is not changed). This is a nuisance if the decoder is heavily customised, eg with a complex speed curve. I've found that if you copy the CVs from the decoder to the app before updating the profile (by clicking the refresh button on the app's CV setting screen), this preserves the customised values in the app and you can then quite quickly recover the values by clicking on each modified CV in turn and just pressing exit without rekeying the value. It only takes a couple of minutes to recover all the CV values this way. It would be even better if there was a reverse refresh button in the app to copy the CVs from the app to the decoder. Regards, John

I've updated my DMU running the Class 31 profile to the latest V2.0 version and the constant braking distance function now works well. Previously it would accelerate on entering the ABC section of track before stopping in a rather random distance. Now it decelerates smoothly and stops at the same point irrespective of the starting speed, before auto-reversing. I tried a variety of slow to medium speeds, but didn't dare try a high speed because my ABC section leads up to a buffer stop. Regards, John

I think page 106 is incorrect where it says that the range of CV27 is 0-2 - Appendix 8 says the range is 0-3, 3 being both rails (ie either rail will trigger the action), and that's what I've been using. Regards, John

you can program stopping distance and I’m assuming the question is what would be the prototypical to program?how many rails need to be cut? The answer is one. However, it seems from the manual that the detection is rail sensitive based on CV27 value so you could have cut sections at different places on different rails and be able to program different actions Although you can choose which rail to detect on, I don't think that there's any way to program different actions in a single loco depending on which rail has been detected. I was thinking of setting my DMU up to shuttle between platform sidings in my two stations, but the rail cuts would have to be at different distances from the buffers since the power pickups are in one end car of the 3-car DMU. Another issue, at least for RailMaster users, is that the ABC reversal doesn't change the direction shown on the RM screen (obvious when you think about it) so the controls get out of sync. I suspect that the app would be the same. Regards, John

I tried ABC yesterday on my 00 gauge layout. I made up a circuit board with 4 1N1007 diodes in series in one direction, and a single one facing the other way in parallel to these, then inserted this into the track feed on one rail of my programming siding - I used this siding because it already had insulated joiners and a separate track power feed. I set the CVs on one of my locos (which has a power bank) to detect ADCC on both rails, and to stop in a constant distance and then reverse after a 5 second delay. It did work as advertised - it appeared to react a second or so after the pickup axles were in the isolated section, and then stopped in a short distance irrespective of the initial speed. I left the stopping distance CV at the default which I think from memory is 25, although it doesn't say what the units of this are. So it appears that you can program the stopping distance, but I would suggest doing some tests before ballasting everything down. It was interesting to try this, but I probably won't use it further. Regards, John

I decided to try this, drove the DMU onto the programming track using the old decoder, reinstalled the TXS decoder and found that I could read and write CVs, so the decoder wasn't entirely blown up - I then set CV8=8 (twice, as recommended) and lo and behold it all works again! I was really surprised because a short circuit seemed like a hardware rather than software fault - I wonder where all the current was going? Anyway, many thanks for your help Rob, you're a genius. Regards, John

I can't change cvs via the app because the power won't stay on, but I suppose I could try it on the programming track if you think that might still work in the face of the apparent short circuit? Is it worth reinstalling the decoder and trying this, do you think? Interesting to hear about the transient spikes and snubbers. Regards, John

My Hornby DMU with 8-TXS decoder derailed at slow speed over wrongly-set points yesterday, and since then it has had a permanent short circuit. The eLink trips whenever the DMU is on the track. Prior to this the decoder had worked perfectly for 3 months or more. I removed the decoder and fitted a blanking plate, and the motor ran with a 9v battery against the wheels. I then reinstalled the old non-sound decoder and the DMU runs nicely with that. I guess the decoder must have blown, although it seems strange that a simple track short could have caused that. I'll contact Hornby and see if they will replace it under warranty, unless anyone has any other ideas? Regards, John

Thanks Rob, that's very helpful. So it looks as though the input resistance is 110 ohms (3 x 330 ohms in parallel) and so the initial inrush current is 12v/110 ohms = 110 mA? Regards, John

How does the static charging work when the app is not being used, ie on pure DCC operation? Do all the power banks start charging simultaneously as soon as the track power is switched on? I don't think we can determine the peak inrush current without knowing more about the control circuitry. There is presumably a resistor in the charging circuit to limit the peak current. The famous manual says on p120 that the power bank takes approximately 2 minutes to charge. The individual supercapacitors are 1.0 Farad and 2.7 volts, so charging a single one at a constant current 'I' amps would take 1.0x2.7/I seconds, so for 120 seconds I=1.0*2.7/120, ie 22.5 milliamps. Three capacitors in series are presumably 1/3 of the capacitance and three times the voltage, so the resulting current would be the same (I think!). I suspect that the charging circuit will be constant voltage rather than constant current (or perhaps some more complicated regime), so the initial inrush current could be higher, perhaps substantially higher, but they must have designed it to not overload the decoder. EDIT: the previous two posts were made while I was toiling over this calculation! Regards, John

Yes, I had something similar a couple of weeks ago. I think it was the flange squeal or perhaps the brake - certainly something very annoying. It was the Black 5 profile on a Bachmann 3MT Tank. I think I might have triggered it by using the sound test buttons in Railmaster's loco function button setup. I couldn't get rid of it, but after powering everything down it had stopped when I next ran. Regards, John

I've tried the iPad and it allows CV149 to be changed to zero even when the loco is under Bluetooth control, so this must be a difference between the iOS and Android versions of the app. Haven't checked whether the auto calibration will actually run when under Bluetooth control, as I didn't want to upset my carefully tuned loco! Regards, John

Yes, I'm sure that Bluetooth operation will use either the simple or complex speed curve, according to how bit 4 of CV29 is set. Regards, John

Bit 4 of CV29 should be set to zero, to use the simple rather than complex speed curve. It will be zero by default, if you haven't set a complex speed curve. The manual page 53 incorrectly states that the default for bit 4 is 1, just to confuse matters. The default value of CV29 as a whole is decimal 2 - setting bit 4 to 1 for a complex speed curve would make it decimal 18, which is what I had when I experienced the problem described at the start of this thread.. Regards, John

I'm using an iPad, but perhaps CV149 is greyed out if you are under bluetooth control at the time you try to change it? I haven't tried that. Regards, John

I'm a bit puzzled by the assertion that auto calibration only works under DCC - I used the app to set CV149 to zero. It may well be that the loco was under DCC rather than bluetooth control at that time, but I didn't have to use the old fashioned programming track method to change the CV. I also disagree with the manual's assertion that a general decoder reset is the easiest method of preparing for tuning - as I explained in the first post of this thread, using the default simple speed curve seems sufficient, and reloading the profile seems to me to be overkill. Regards, John

Yes, I've always done it when installing. It's very easy and it seems to help. Despite the warnings it doesn't really need a very long straight - 3-4 feet seems ample and if it goes a bit round the curve as it decelerates I don't think it really matters. I just got caught out by doing things in the wrong order after having to reset the decoder. I've achieved excellent motor control on my 4 TXS locos without having to do the stage 2 (PA) or stage 3 (PID) adjustments described in the manual. Incidentally I did try the PID mode out of interest - I set CV149 to 1 and tried with the default values, and nothing seemed to change - I then tinkered with the P value but again it made no difference. Then when I ran the loco the next day it behaved dreadfully - it looks as though you need to power cycle the decoder after setting CV149 to 1 in order for it to notice the change! I reverted to CV149=2 and power cycled and all was well. Regards, John

A couple of weeks ago I had to reset the 8TXS decoder in my Bachmann 3MT Tank following a loss of DCC control, as reported previously on this forum. I then set it up again by laboriously entering all the modified CVs, including my complex speed curve, and re-ran auto calibration. Since then the loco has 'hunted' at medium speeds, ie the speed regularly oscillates above and below the average - not too badly, but enough to be noticeable. I hadn't noticed this previously. After a lot of head scratching I consulted the famous manual and saw on page 112 a warning (in red type!) that the auto calibration must be done with all the speed curve CVs set to default values. So I set CV29 bit 4 to use the default simple speed curve (without changing my complex speed curve) and re-ran the auto calibration. This seems to have cured the 'hunting' problem. CVs 145-147, set by the auto calibration, were 49/15/4 after the erroneous calibration, and 25/11/3 after the second one. It's a bit surprising that the speed curve should affect the auto calibration - I assumed that it just applied maximum voltage to the motor and monitored the back emf to see the acceleration - but perhaps it uses speed step 127 of the speed curve to decide what voltage to use? This would tie in with the acceleration being about half as much with my complex speed curve, which has a much lower top voltage. I can't really fault the manual, and I had read it from cover to cover at the outset, but it's easy to overlook these things when one becomes familiar with the procedures. The interaction between auto calibration and speed curves is not intuitively obvious. Just thought I'd point this out in case anyone else gets caught out by it! Regards, John

I have a Bachmann 3MT Tank, which appears from photos to have a smaller diameter smokebox than the K3, and I fitted both the decoder and the smallest speaker in the smokebox, with the power bank in the cab. On my Hornby Peppercorn A1 (Tornado lookalike) I fitted decoder, speaker and power bank in the smokebox. Regards, John

Thanks Rob. I'm reluctant to start swapping decoders between locos because it can be fiddly to get the 8-TXS wiring harness into the bodies, and I don't want to break anything by unnecessary dismantling. I will try the Black 5 on Bluetooth again in the next few days to see if the problem has really gone away. I've been using the same iPad to run the app all the time. It would be good to hear from others who have previously reported this Bluetooth loss of control on this forum to understand if their problem got resolved, and if so how. Regards, John

I tried my Black 5 under Bluetooth again this morning and, strangely, it all worked perfectly! I haven't changed anything in the setup since Friday as far as I can see, and I tried it a number of times on Friday when it always lost Bluetooth control as soon as the loco started to move. I've now got the CVs as I want them so I probably won't be running it under Bluetooth again. Incidentally, on the related but opposite problem where locos sometimes lose DCC control but continue to operate fine under Bluetooth, necessitating a CV8=8 reset of the decoder, which I've experienced at least 3 times, I now believe that it happens a few minutes after the loco has been run under Bluetooth, probably when the iPad goes to sleep. Regards, John

Thanks Rob, that's good to hear. Do you have any more details on the nature of the bug? It hasn't affected my other 3 TXS locos, one of which is also running the Black 5 profile. I can live with it because I normally use Railmaster to operate my layout, as long as I know that it will be fixed in due course and isn't a faulty decoder. Regards, John

I installed my 4th 8-TXS decoder today, in a Hornby Black 5 which previously had factory-fitted TTS. I installed the Black 5 profile successfully. I didn't install a power bank, since it has pickups on loco and tender and never lost power with TTS. The sounds work fine under bluetooth, but when I set the loco moving it stops after a few seconds and the sounds also stop. The bluetooth light on the app remained blue, so it hadn't lost power. Cycling the track power restored the sound but the same problem occurred as soon as the loco started to move. It all appears (at least on cursory testing) to work fine under Railmaster/DCC. I think I've seen this problem reported more than once before on this forum, but I don't recall seeing a solution. Do I need to send the decoder back for a replacement? Regards, John