Potrail2378 Posted November 9, 2020 Share Posted November 9, 2020 Hi, I have a class 37 which I have converted to DCC. I have put a standard Hornby decoder in, and it works perfectly. I have now put a sound decoder in, which works fine, apart from it being 'jerky' while setting off and at low speeds. I put the standard decoder back in and it was fine, but when I put the sound decoder in, it's still 'jerky' I have set the acceleration and deceleration to zero on both decoders with my Hornby Select. Any advice please?Thanks in advance. Link to comment Share on other sites More sharing options...
ColinB Posted November 9, 2020 Share Posted November 9, 2020 You have to change cv150 to a value of 1. One of the other guys will tell you why. Link to comment Share on other sites More sharing options...
JemBobs Posted November 10, 2020 Share Posted November 10, 2020 Would this also work on a standard 8 pin hornby decoder? Link to comment Share on other sites More sharing options...
JemBobs Posted November 10, 2020 Share Posted November 10, 2020 I should also add that it is a Frateschi engine that I have converted to dcc. Link to comment Share on other sites More sharing options...
Flashbang Posted November 10, 2020 Share Posted November 10, 2020 CV150 is only available on the TTS decoder. The basic R8249 decoder does not offer these adjustments. Link to comment Share on other sites More sharing options...
ColinB Posted November 10, 2020 Share Posted November 10, 2020 I think CV 150 in the TTS decoder selects a different motor lookup or algorithm. I assume CV150 = 0 is for steam based locos, whereas CV150 = 1 is for diesels, seeing on every disel I have fitted TTS to (Bachmann, Hornby, Heljen), I have to change CV150. Obviously the motor map is one of the things that you would think Hornby would change when they load the sound map, as they know what the sound map applies to, saving all these posts. I just assume that it must be in a different part of the code so not that easy to change. Link to comment Share on other sites More sharing options...
Potrail2378 Posted November 10, 2020 Author Share Posted November 10, 2020 Hi guys, the sound decoder is a Hornby TTS Digital Sound Decoder, R8102, Class 37 if that's any help?I have fitted a 'can' motor as a kit from the guys at Strathpeffer. The standard Hornby decoder and the R8102 both operate the cab lights and the directional lights perfectly, the sounds on the R8102 work perfectly. The standard decoder works perfectly at slow speed and setting off, it's just jerky at slow speed and setting off with the R8102 fitted. I have a Hornby Elite in the attic, but I am currently in the workshop using a Hornby Select.I hope this can shed a little more light on the subject. Link to comment Share on other sites More sharing options...
ColinB Posted November 10, 2020 Share Posted November 10, 2020 As I have said before change cv150 to a value of one, it will be set at 0 on a new decoder. That will fix it. It has on all my class 37, class 47 and class 66 locos that I have fitted TTS to. You might need to use your Elite to do it, it depends on your software level for the Select to do it. Link to comment Share on other sites More sharing options...
Potrail2378 Posted November 10, 2020 Author Share Posted November 10, 2020 ColinB - thanks, unfortunately I work shifts so won't be able to do this until tomorrow morning, but I will certainly do as you advise. Link to comment Share on other sites More sharing options...
96RAF Posted November 10, 2020 Share Posted November 10, 2020 TTS CV150 is set at value 0 which suits most Hornby locos. The value 1 setting was initially provided as a quick fix for B’mann locos, but of course that was based upon their motors at the time and ultimately depends upon which decoder is fitted as to what is adjustable. The TTS CV15x algorithm changes the PID characteristics of the motor (Google it and be confused), however the TTS algorithm for simplicity only alters the P and I characteristics, so as it is a bit of a black art rather than a science you have to fiddle with the rising voltage settings then adjust the falling voltage settings to get a reasonable balance for any particular motor. E.g. if you look at the multiplicity of FS variants it may have a 3-pole or 5-pole motor which each have separate CV15x settings according to RR or Mainstream TTS factory fit. Hornby was working to provide (via higher range CV15x settings) an automatic calibration for any particular motor but I think this went back-burner for bean counter reasons. We live in hope for future decoder development. I do not pretend to understand how you procedurally calibrate CV15x but if you have the patience and work to a published sequence of adjustment a good motor control can be achieved for any loco fit. Link to comment Share on other sites More sharing options...
ColinB Posted November 10, 2020 Share Posted November 10, 2020 PID is Proportional Integral Derivative is the control loop for controlling a motor ( I had forgotten all this stuff), I assume in this case they are using the requested speed and the back emf as the parameters. i assume back emf is the feedback parameter. If you are into cars, it is used to control the idle speed control. I wrote a ton of software to do exactly that. That explains why it judders. Link to comment Share on other sites More sharing options...
Fishmanoz Posted November 10, 2020 Share Posted November 10, 2020 It's not just setting CV150 to 1 to get motor algorithm 2. While that works to fix things most of the time, you can also adjust the P parameter with CV153 and the I parameter with CV154 (similar for CV150 = 0, except with CVs 151 and 152) for finer settings. There is a post in here somewhere with a methodology for doing those adjustments to get the best outcome, rather than adjusting them randomly and hoping for the best. But again, you usually only have to set CV150 to 1 and the default settings for CVs 153 and 154 do the job, no further adjustment needed. Link to comment Share on other sites More sharing options...
ColinB Posted November 10, 2020 Share Posted November 10, 2020 Yes, in automotive a tuner/calibrater sets the values. In theory Hornby should set them for every model they make sound for, because only they know the characteristics of their motor. It is all to do with overdamped and under damped systems. If it is underdamped it oscillates which is why you get the juddering. I must admit setting cv150 to 1 seems to cure it on all my diesels, I was trying to think if I had to set it to 1 on my class 66, which is the only Hornby diesel I own with TTS. I would imagine with a steam loco, the valve gear acts as a damper, hence why it is not such an issue plus originally when it was setup that is probably what Hornby used to set the values. P is probably the amount of proportional feedback you use from the back emf, I assume there must be a cv for I and likewise D, but generally you can spend hours trying to get it perfect. I am just happy to set cv150 to 1. Link to comment Share on other sites More sharing options...
Potrail2378 Posted November 13, 2020 Author Share Posted November 13, 2020 Hi, and thanks for all your information and suggestions. I have changed CV 150 to a value of 1, but there is no difference.Do I now have to start playing around with CV 153 and 154 as suggested earlier? Thanks in advance. Link to comment Share on other sites More sharing options...
Fishmanoz Posted November 13, 2020 Share Posted November 13, 2020 Yes, your next step is to experiment with CVs 153 and 154 with CV150 on 1, or you can put it back to 0 and do similarly with CVs 151 and 152. Here are some links to where this has been covered previously: https://www.hornby.com/uk-en/forum/getting-smooth-running-from-tts-decoders/?p=1 https://www.hornby.com/us-en/forum/hornby-class-60-decoder-problem/?p=1/ https://www.rmweb.co.uk/community/index.php?/topic/129827-tts/page/3/ The lower CV of each pair is your P value and the higher your I (Colin, there is no D). The main adjustment is P and you will see from the links, you start with it at a low value, increase in increments of 10 until it stops being jerky then continue on steps of 10 until it goes jerky again. Now decrease in steps of 5 until non-jerky again. This is now your optimum P value. You can now do similarly with I using CV 152 or 154 (depending whether CV150 is 0 or 1). If I'm reading this one correctly, you again start at 0, increase until it gets jerky again, then back off to smooth. For the technical, what you are doing here with P is to have the change in motor speed to be critically damped when you adjust the controller. Then as you increase I, this gets you to the critical damping point more quickly (steeper leading edge on the speed curve). Putting it another way, the first order change you make is which is better of motor algorithm 1 (proportional) or 2 (linear). Then the second order is the optimum P for the better algorithm. Then the third order is the best I value for optimum P. Colin, D would be fourth order if it was there. Trust this helps. Please read my words in conjunction with the links provided. Link to comment Share on other sites More sharing options...
96RAF Posted November 14, 2020 Share Posted November 14, 2020 Without referring back to the TTS leaflet I am sure it explains in there that only the P and I characteristics are supported. Changing D is not supported. Link to comment Share on other sites More sharing options...
Fishmanoz Posted November 14, 2020 Share Posted November 14, 2020 It certainly does Rob, by not having any CVs for it. Given my understanding, and as I said above, it would be 4th order in any case. Link to comment Share on other sites More sharing options...
96RAF Posted November 15, 2020 Share Posted November 15, 2020 Your PI adjustment explanation Fishy is the easiest to understand that I have seen so far. Maybe we could introduce the theory of motor cogging now to totally bewilder folk. Link to comment Share on other sites More sharing options...
Flashbang Posted November 15, 2020 Share Posted November 15, 2020 He is a simple one to carry out.... Try removing all the suppression capacitors from the motor. Snip them off where soldered and discard them, they are not needed on DCC and may be the cause of the jerky running? Well worth a try anyway 😎 Link to comment Share on other sites More sharing options...
ColinB Posted November 15, 2020 Share Posted November 15, 2020 The D parameter makes it unstable if you don't know what you are doing, so I can see why they would not allow you to change it. If there is no D then it is just a PI loop, not PID. I spent hours years ago, setting 3 of them up on a "two stroke" Ford Fiesta. Anyway, I think there is something more fundamentally wrong with his setup, I have used TTS in my HST and I don't remember it being jerky and I didn't change cv150, mind you I was using a 5 pole ringfield motor so I suppose that helped a lot. TTS use in Ringfield motors is a bit dodgy anyway, a TTS max average current is 500 milli Amp whereas a Ringfield can consume 600 milli Amp, so if he is using a 3 pole ringfield then that may well be the issue, it is hitting the current limit ( I assume TTS has one). Funny thing is I have just fitted a TTS decoder to a brand new Bachmann class 66 and I didn't have to change cv150, which I normally have to. Link to comment Share on other sites More sharing options...
96RAF Posted November 15, 2020 Share Posted November 15, 2020 Yes Colin TTS does have current limits - all clearly set out in every TTS leaflet, including how the functions draw affects the oversll availability of current within motor and function individual limits. Link to comment Share on other sites More sharing options...
ColinB Posted November 15, 2020 Share Posted November 15, 2020 Of course it does, it is what it does when they are exceeded which the leaflet doesn't explain. Does it blow up. fold back and retry? I read a lot of the current limit data as I was thinking of fitting one to a Lima loco, when I noticed that the maximum current for the Lima was in excess of that for TTS. If it is folding back that would explain the jerky behaviour. Link to comment Share on other sites More sharing options...
Potrail2378 Posted November 15, 2020 Author Share Posted November 15, 2020 Hi, as I said in my (second) comments, I have fitted a 'can' motor as a kit from the guys at Strathpeffer. Does this use less current? Link to comment Share on other sites More sharing options...
ColinB Posted November 16, 2020 Share Posted November 16, 2020 I don't honestly know, but I would assume it would as it is a more modern motor. That would explain why your TTS is still working ok, forgetting the judder and you got it to work on a Hornby decoder. My ringfield based loco fried the Hornby decoder, the first time it stalled on a point, I now fit Zimo decoders. I tried a few of my really old ringfield based motors and they were drawing current awfulling near the TTS maximum value. That stilll doesn't explain the judder. I am wondering it is something to do with the back emf generated by the motor. There is an awful lot of undocumented issues with TTS. I wonder if the TTS has an issue with your motor and basically is jumping in and out of current limit, but unfortunately their current limit strategy is not documented in any of the documentation. It just says you mustn't exceed the maximum current but not what happens if you do. Link to comment Share on other sites More sharing options...
Potrail2378 Posted November 17, 2020 Author Share Posted November 17, 2020 Hi, and thanks for all your information and support. Fishmanoz - I followed the links you posted earlier on, concerning CVs and I had a lot of success! I setCV 150 to 1CV 153 to 80CV 154 to 25It works brilliantly now, thanks again. Link to comment Share on other sites More sharing options...
Recommended Posts
Please sign in to comment
You will be able to leave a comment after signing in
Sign In Now