CV 6 does not seem to be working.

[cudaken]

 I have a PK 2000 E-7 A with QSI sound decoder that I am speed matching to a PK 2000 E-6 B using a Digitrax 123 decoder.

 I am only changing CV's in the E-6 with the Digitrax 123 decoder. Simon 1966 originally did the speed matching. What made me want to re speed match was the B unit was pulsing  back and forth against the A unit.

 I found CV 2 (using Decoder Pro) was sat at 47, CV 6 - 121 and CV 5- 211.  My test track is 9.5 foot by 5 foot I will add. Change CV 2 to 47 and starting speed match was better. Top speed seems right on the money so I left CV 5 at 211.

 Mid range is another story. I start the engines about 10 inch's a part, set the speed step to 50 (Digitrax SEB) and start the run. One time around the 10 inches becomes 2 feet. I have sat CV 6 from the original 121 all the way up to 170 and the A unit still out runs the B unit. Could CV 5 have any effect at speed step 50? SEB is set to 128 speed step.

 I have also confirmed the CV changes have taken with Decoder Pro.

 They have seemed to stop fighting each other, getting ready to see what happens with a load.    

 Thanks for the coming answers!

              Ken

[CSX Robert]

Do the decoders have the 28 step speed tables enabled(what is the value of CV29)? If they do, then CV5 and CV6 will have no effect.

[cudaken]

 I will check, and thanks for the answer.

           Ken

[cacole]

Is BEMF turned on on either one or both decoders?  If it is, that's what might be causing the pulsing you mention.  Turn it off on both decoders.

[rrinker]

 Think of the start/mid/top settings as a piece of string. 0 at the bottom, and let's assume 12 votls at the top. With no settings in CVs 2,5, and 6, you have a tight piece of string in a straight diagonal line from 0 on the left to 12 on the right. Increase CV2 and the zero goes up (but the string is still held tight, no sags). Adjust CV5 and the 12 side comes down, again the string stays tight. Now add in CV6. Hold the string in the exact middle and adjust is up or down, while not moving the ends. If you pull it down, you have a gentler slope from 0 to the middle, and a steeper slope from the middle to the top. This affects the speed at EVERY step above step 1. Up to the middle (the display on a DT400 is 0-100 percent, not the speed step 1-128) there will be less change per step, after the mid point each step will result ina  greater change in the loco speed.

 If you reverse it, and lift up on the middle of the string, now you have a steep slope from 0 to the middle, and a gentle slope from the middle to the top. Exactly the reverse of the above. Each speed step  change past 1 will result in a large speed change, after you get halfway there each speed step will result in a lesser change. The actual speed at every step will be affected, NOT just the speed at the midpoint.

 Enter the 28 step speed tables - same concept but instead of 3 points you have 28, and since they generally only work in 28 speed step mode, each speed step gets a specific speed based on the value of that table CV. You can get two locos using 28 speed steps and 28 step tables to be exactly in sync for each and every speed step that way - the downside being the amount of fiddling to get it perfect and the fact that youw ill be limited to 28 speed steps not 128.

                                      --Randy

[CSX Robert]

rrinker

Enter the 28 step speed tables - same concept but instead of 3 points you have 28, and since they generally only work in 28 speed step mode...

This used to be true, but I believe most current decoders can use the 28 step speed table when operating in 128 step mode. I know Digitrax, Lenz, and NCE decoders can.