## Re: [Jmri-developers] Splitting Straws?

 Re: [Jmri-developers] Splitting Straws? From: Paul Bender - 2010-08-24 18:08:40 ```On Aug 24, 2010, at 12:47 PM, Walter Thompson wrote: > The designer of the decoder now has to make a choice, How do I handle the odd number speed table CVs? They have to have a whole number that is either 4 or 5 greater than the preceding number in the throttle speed. > > Manufactures handle this dilemma in different ways! > > This means that when JMRI plots a straight line speed curve the Loco is not going to always respond with nice smooth speed changes as the throttle is changed from speed to speed. This is frequently true of interpolation functions, especially when the function is evaluated at discrete intermediate points. This isn't something we need to worry about in JMRI, because we only set the defined 28 values for 28 speed step mode. In 28 speed step mode, our linear curve is truly linear. Paul ```

 [Jmri-developers] Splitting Straws? From: Walter Thompson - 2010-08-24 17:47:43 ```Is this something we are not really aware of? About 128 speed step mode. 128 speeds are from 0-127. We think of how many speeds, so Zero is the first one, so the 128 is really a speeds of 127 to us. Speed of 1 sent to a decoder is special. That tells the decoder to do an “Emergency stop”. So that to our mind that leaves 126 speeds that we can tell the decoder to do. We have 28 steps in the speed table that define different points in the 126 speeds. Divide 126 by 28 and you get 4.5. So speed table step 14 should have maximum affect on throttle setting of 63, but what throttle setting do step 13 and step 15 affect? For 13 there is no throttle of 58.5 so its assigned to either 58 or 59. The same for step 15, its either in full affect for 67 or 68 speed on the throttle. The designer of the decoder now has to make a choice, How do I handle the odd number speed table CVs? They have to have a whole number that is either 4 or 5 greater than the preceding number in the throttle speed. Manufactures handle this dilemma in different ways! Digitrax increase the throttle steps by 4 as the speed steps increase, Speed step 1 affect throttle speed of 1, step 2 affects speed of 5, step 3 affects 9 and so on until step 27 affects throttle setting of speed 105. That leaves Throttle speed of 109-127 to be affected by step 28. Step 14 (often thought of as mid point) affects throttle setting of 52. Throttle mid point of 64 is most affected by step 17 Soundtraxx took a different approach they increase the throttle speed affected by 4 and then the next step affects the throttle speed by 5, that way Throttle mid point of 64 is affected by Step 14. NCE and QSI are also different than other Manufactures. This means that when JMRI plots a straight line speed curve the Loco is not going to always respond with nice smooth speed changes as the throttle is changed from speed to speed. Walt Thompson 1405 Portland Ave Saint Paul, MN 55104 Ph 651-644-8673 Moderator, Yahoo JMRIusers group NMRA, TLR, TCD NRMA #136859 ```
 Re: [Jmri-developers] Splitting Straws? From: Paul Bender - 2010-08-24 18:08:40 ```On Aug 24, 2010, at 12:47 PM, Walter Thompson wrote: > The designer of the decoder now has to make a choice, How do I handle the odd number speed table CVs? They have to have a whole number that is either 4 or 5 greater than the preceding number in the throttle speed. > > Manufactures handle this dilemma in different ways! > > This means that when JMRI plots a straight line speed curve the Loco is not going to always respond with nice smooth speed changes as the throttle is changed from speed to speed. This is frequently true of interpolation functions, especially when the function is evaluated at discrete intermediate points. This isn't something we need to worry about in JMRI, because we only set the defined 28 values for 28 speed step mode. In 28 speed step mode, our linear curve is truly linear. Paul ```
 Re: [Jmri-developers] Splitting Straws? From: Andrew Crosland - 2010-08-24 19:05:30 ```On 24/08/2010 18:47, Walter Thompson wrote: > We have 28 steps in the speed table that define different points in > the 126 speeds. Divide 126 by 28 and you get 4.5. So speed table > step 14 should have maximum affect on throttle setting of 63, but > what throttle setting do step 13 and step 15 affect? For 13 there > is no throttle of 58.5 so its assigned to either 58 or 59. The same > for step 15, its either in full affect for 67 or 68 speed on the > throttle. Internally, some decoders have 256 or more speed steps, so it is possible to have step 58.5, etc. Some command stations know this and will send alternating speed steps of, say, 58 and 59 to get an effective step of 58.5 in an effective 256 step mode. I'm sure this is documented in the RPs. Andrew Crosland ```
 Re: [Jmri-developers] Splitting Straws? From: Andrew Crosland - 2010-08-24 19:08:07 ```On 24/08/2010 20:05, Andrew Crosland wrote: > On 24/08/2010 18:47, Walter Thompson wrote: >> We have 28 steps in the speed table that define different points in >> the 126 speeds. Divide 126 by 28 and you get 4.5. So speed table >> step 14 should have maximum affect on throttle setting of 63, but >> what throttle setting do step 13 and step 15 affect? For 13 there >> is no throttle of 58.5 so its assigned to either 58 or 59. The same >> for step 15, its either in full affect for 67 or 68 speed on the >> throttle. > > Internally, some decoders have 256 or more speed steps, so it is > possible to have step 58.5, etc. > > Some command stations know this and will send alternating speed steps > of, say, 58 and 59 to get an effective step of 58.5 in an effective 256 > step mode. I'm sure this is documented in the RPs. > > Andrew Crosland PS this also applies in 28 step mode where an effective 56 step scheme can be used. Andrew Crosland ```