Sound cutting out with TTS 37 decoder

[HST43]

Hi there,

I have TTS fitted to my old Hornby ring field class 37 which also has a stay alive capacitor fitted, locomotive runs fine but after about 6 laps of the layout the sound cuts out but still have full control of the locomotive, once I reboot RM the sound comes back on again...I'm running on RM elink with latest software.

Could this be a thermal overload protection   with a old locomotive ?

[ColinB]

I had this issue with a Duchess TTS and that was a factory fitted TTS. I tried everything to stop it, in the end I came to the conclusion that there was some weird reaction between the TTS and the motor. I had two other spare decoders (class 66 and Merchant Navy) and they did the same with the same loco. The only thing I thought it could be is that the TTS sees a surge in current and shuts down the sound, but doesn't have any logic to switch it back on again. It didn't even seem to be related to speed it just happened sometimes after one circuit, sometimes more. In the end I used a Zimo sound decoder in the offending loco and moved the TTS to another loco, where it worked perfectly. So you are probably right. I had a similar setup as you HST with 5 pole ringfield motor, I also has "stay alive" fitted. The TTS worked perfectly for 6 months but then died. So they should work ok, I am pretty sure it died because I had "Heat Shrinked" the circuit and it was very hot in the loft on that day. 

[HST43]

Thanks for the info , yes I might swap the TTS in this old 37 mine is also 3 pole type ring field motor I've heard stories that they tend to pull alot of current somtimes if there not serviced well and can blow the decoder if running over 500 milliamp. All my other locomotives are newer dcc ready and work fine.

[96RAF]

The motor side of a TTS decoder is self protecting, but the sound amp and function outputs are bot, so care must be taken bot to overload them. You should also be aware of the total decoder limit of 800mA so if the motor is pulling to that limit of 500mA or more it may be starving the sound side.

[ColinB]

Yes RAF96 you are absolutely right, as I had a broken class 66 TTS decoder that had one function/port broken and I knew it going back to Hornby anyway I tested it on a Lima class 66 loco. Sure enough every so often the loco stopped because of the overload protection (I know it was that as when I replaced it with a normal decoder with a higher current limit it worked perfectly). The reason it had a function/port blown was there was a fault with the lights I fitted, so yes I agree no protection on those. There does appear to be some arbitary overload protection on the sound though, which causes the effect that this post is about. When my Duchess did exactly the same as this post I spent a week trying all sorts of things to stop it happening but as I said I could not fix it, but it only happens with certain locos. So I don't know if it is noise on the "back emf" measured that causes an issue, but it seems to be something unique with the motor. Take the TTS decoder out of the loco put it into another loco and no issues. Perhaps if you are ever talking to your mates in Hornby you could ask if they have ever noticed it. I doubt Customer Services are ever going to be able to answer a question like that.

[Fishmanoz]

You could avoid the issue and speculation about it by doing as recommended a DC stall current test before fitting a decoder. Always a good idea to do this with all X03/X04 or Ringfield locos as they can often exceed the 500mA limit of a TTS.

 

In this circumstance, well maintained includes remagnetising as weak magnets lead to high current draw. 

[ColinB]

@Fishmanoz in this case a measurement of the stall current doen't help. I must admit I do a stall test on my old locos and measure the armature resistance to get absolute maximum value but in my Duchess case it was well within specification for a TTS decoder. As it was bought new with the TTS decoder fitted by Hornby. No, this fault is an odd dynamic one, I have read a couple of posts on this forum identifying the same fault. It is a weird dynamic fault with how their system works, I don't know what is actually happening but RAF96 did mention that it might be the sound part having some cut off due to it thinking it is sourcing too much current. What I do know is that I checked the software version on the 3 that I tested and they ranged from early versions to the the class 66 one which was the latest.As I say took the offending TTS decoder out put it into another Hornby loco and it didn't miss a beat. I put a Zimo sound decoder in the problem loco and that works perfectly. What I should have done is changed the motor to see if the fault disappeared but I didn'twant the agro and the risk of breaking something.

[96RAF]

I have the spec of the TTS decoder amplifier chip so maybe that will tell me something.

 

You are right about BEMF Colin as TTS steam uses this to regulate the chuffs and work out if its pulling or coasting, so there mst he feed back from the basic decoder into the sound decoder part.

 

There was talk of making decoders self calibrating for motor characteristics as opposed to the RM hand cranked method of timing and fiddling with settings. I think some ESU decoders have this incorporated. You set a CV and the loco shoots off at max speed for a set time, then stops - job done. It would have been interesting to see if self calibration could have ‘cured’ these apparently motor associated TTS problems. I do know for sure that there are differences between the FS 3-pole and 5-pole motor settings.

[Creeping Jane]

One thing that came to mind is that you could fit two decoders set to an identical address. Use the TTS purely for the sound and another budget decoder for the motor?

[96RAF]

One thing that came to mind is that you could fit two decoders set to an identical address. Use the TTS purely for the sound and another budget decoder for the motor?

 

That could only work for diesel as steam needs to ‘feel’ the motor bemf to control the chuff/coast as noted earlier in the thread.

[ColinB]

Someone else had the idea of using two decoders, one TTS and one normal. I did have a think about this at the time and did wonder if you connected the TTS one to the motor via say two 10 kiloohm resistors, so that basically the TTS was just monitoring the back emf whether it would work. Trouble is without knowing what the circuits and the strategy inside the TTS are, it is hard to say whether it would. Perhaps one day I will have a play.

[96RAF]

All decoders are essentially the same architecture Colin and a block diagram of one would equally apply to another make.

 

Sound decoders and those with built in stay alive simply have extensions to those basic building blocks.

E.g. bAsic plus stay alive. Sound would add processor and amplier blocks.

/media/tinymce_upload/5581772acec05c06204a6413ede7ebfc.jpg

 

You would gain as much by running a simple bench check monitoring bemf across a pair of decoders that you can afford to lose if it all goes belly up.

[ColinB]

Yes RAF96 thank you for the circuit. It is the bit where the Microprocessor samples the back EMF that is the unknown. I assume it must somehow feed it back into the analogue to digital converter on the micro to do the closed loop control. As I say a task for when I am bored. 

[96RAF]

A two decoder setup Colin - Zimo and  Soundtraxx with smoke.

 

https://images.app.goo.gl/wAJwSWW3GUi4Tdcp9