How about 3 decoders?

I had the same problem with the Atlas O twin motor drives.  I tried a lot of different brands and combinations of decoders.  The best performance I was able to get was with 3 decoders.  One for sound, not connected to either motor.  And two identical *good* back-emf decoders with indentical settings for the two motors.  You want the best back-emf decoders, not the averages ones.  Lenz, Loksound, and Zimo all have superb motor control - far better than anything else (at least that was the case a couple years ago when I tried them all).  I went with zimo.  And of course soundtraxx for the sound.

One trick for multiple decoders in a single engine - set all the long addresses to be the engine number, set the short address to be all different.  That way you can use programming on the main to change which address is active, and to program each individual decoder on the main in short address mode.

Good luck!

- Ken

Wire motors in series?

When Bachmann came out with their first version of the 44 toner it had two motors, one for each truck.  They had a problem with those early models running rough due to the motors fighting each other.  Bachmann solved the problem on later versions by using one motor and drive shafts to the two trucks.  Modelers with the early versions discovered that those models ran much better if the two motors were wired in series rather than in parallel as Bachmann had built them.

Intrinsic motor or gear problem?

I run many DCC locomotives with two, three and four motors and have had no problems except with units with motor and or drive problems. So back to basics. Have you tested this unit with linear and PMW DC power packs? If there are low speed or surging problems you are dealing with drive problems which must be resolved before installing the decoder. Mechanical surging can drive a BEMF system and the operator to dispare.

Also measure the current(s) when running the DC tests. Although you may not see any physical surging, you may see the current surging which is also indicative of a drive problem. What are the currents? Several of my Tsunami install have proven very sensitive to over voltage and current. Although the TSUs are specified for 27 V (If properly specified that is the max peak to peak voltage of the DCC wave form), some have had problems at 21 Vpp. What voltage are you using? I have also seen problems when pushing the drive current. Even with can motors, O units can easily pull 0.5 Amp per motor or 1.0 Amp for two motors.The TUS 750 is only rated for 750 mAmp and tends to misbehave when over driven.

Also check the stall currents, the odds are that with two motors you will exceed the TSU 750 spec. You migh also want to run the DC test with the shell off and see if one motors is getting hotter than the other an additional sign of motor and or drive imbalance.

Logger01, I'm sorry, I neglected to include...........

...........the fact that the unit crawls and basically runs like a watch on straight DC, so it's not a mechanical issue.

Peter

Can you slow them down with JMRI.

I have heard that there are problems with BEMF when trying to run more than one loco.  Your 2 motor loco is the 2 locomotives in one body.  If you turn off the BEMF, can you slow them down with JMRI to get your speed down to the slow speed operation you desire

I still suspect drive problems or decoder incompatibility

As noted TSUs have proven sensitive to drive problems. In one case traced to commutator noise. After discovering the noise the brand new unit was run around a loop for about half an hour in each direction with the decoder removed (It ran beautifully while polishing the commutator and the brushes). With this break-in the noise temporally cleared up, but returned shortly after the decoder was reinstalled. Since there was little hope of getting a warranty replacement, we replaced the motor and have yet to have another problem. Note that these types of problems will be covered up by most DC power packs and only discovered when the motor fails.

Speed matching is often a pain, but the existence of this type of problem will also cause problems with speed matching for the two or three decoder approach.

If you want to shotgun an approach you might try, as posted, the motors in series. This will limit the top speed (not usually a problem for a little switcher) but may snub motor noise. Also try running the unit with a beefier decoder with and without BMF. Almost any inexpensive 1.5 Amp device should work. If the locomotive runs OK then you have a compatibility problem.