Tsunami Motor Control

[Sdynamo]

DCC gurus:

What makes the motor control on the Tsunami's (e.g. the TSU-750) so bad?  Is it the overheating of the decoder?  SBS4DCC has a heat sink for this - does this fix the issue?

Is it something else?

I've been using LOK and have had great success, but I like the steam sounds on the Tsunami's a lot.

Thanks for the feedback.

~Wiley

[jdcolombo]

No, it's not the overheating, which I've experienced only once (I returned the decoder for a replacement, and have had no problems since).  My experience is that overheating is not a common problem, although it does happen occasionally.

Instead, it is the inability to get smooth slow-speed "creep" from the engine.  I have tried, twice, to get excellent "crawl speed" performance with the Tsunami alone, and have followed the extensive directions on how to set the BEMF parameters that Bruce Petrarca has posted on his "DCC University" web site.   All to no avail.  One time it was with my Athearn Challenger; the other was with one of my Walthers/LL Berks.   After that, I gave up.  But I'll freely admit that I am VERY picky about slow-speed "crawl" performance in my engines.  YMMV, and I haven't tried this lately.

My second complaint has to do with matching the chuffs to the driver rotation.  This is another area that I am VERY picky about.   Standard two-cylinder steam locos have four chuffs per driver revolution.  Not 2, not 3, not 4.5 or 6.  Four.  Exactly.   Nothing drives me crazier than not having the chuffs match to driver rotation.   I've found that it is easier to match these two with the Tsunami by using a separate motor decoder, but it is theoretically possible to do it with a speed table in the Tsunami alone.   Nevertheless, it doesn't have to be this hard.  ESU, QSI and Zimo (to name a few I am familiar with) use their BEMF circuits to sense the motor rotation (which obviously is related in some way to driver rotation!) and time the chuffs accordingly.  On the LokSound, there are two CV's you adjust, one at speed step 1 (CV57) and the other at speed step 4 (of 28) (CV58) to achieve perfect chuff/rotation sync.  Once you do this, the decoder adjusts the chuffs at other speeds automatically via the BEMF sensing.   It works like magic, and takes five minutes, instead of hours (e.g., with the Tsunami).   If you are less obsessive about this than I am, it may not matter to you.

I do like aspects of the Tsunami's sound better than ESU's steam files.   I think Soundtraxx nailed the sound of the UP Hancock long 3-chime whistle on the FEF, Challenger and Big Boy, and that ESU's implementation is a poor second cousin.  The Tsunami has an sound EQ system that is VERY useful for tuning the sound with the small speakers we use in N scale; ESU lacks this feature.  The Tusnami also has automatic blow-down of the cylinders on stop, another thing ESU lacks.   On the other hand, the ESU's Big Boy file has the "open cocks" hiss at startup that the Tsunami lacks.

I wish the perfect sound decoder existed; it would have ESU's motor control, the open-cocks hiss at startup, and the BEMF chuff timing circuit along with the Tsunami's better sound samples (in some cases), it's EQ, and it's automatic blow-down on stop.   So far, it doesn't exist in a small N-scale decoder, although ESU's sound files are downloadable, which means they can be altered and improved - I know Matt Hermann is working on a dedicated file for the FEF, which might solve many of these issues.  But you can't add the Tsunami's EQ circuits to the ESU LokSound.   So I cobble together whatever seems to work best for a given application and try to be happy.

John C.

[Sdynamo]

Thanks for the great info John.

How are you matching the chuffs between 2 decoders?  If the Tsunami requires endless tweaking to get the chuffs to match I don't see how the second decoder is helping.  Is the second/motor decoder providing anything more than the increased slow speed control?

[jdcolombo]

The Tsunami does not implement CV's 5 and 6.  You can't do a 3-step speed curve with it, and often I find that with a separate motor decoder, you can get a good chuff match by using CV's 2, 5 and 6 instead of doing a complete speed table.

I've gotten good speed matching with a 3-point curve using a Lenz Silver Mini for motor control and the Tsunami for sound in about 20 other steam locos, including 10 Walthers/LL Berks, 4 Kato Mikes, 2 Bachmann Consolidations, 2 Bachmann EM-1's, the Kato GS4, and the FVM Hiawatha 4-4-0.   It did not work as well in my Athearn Challenger (I had one of the original ones, that came with a horrible MRC sound decoder; I ripped that out and put in a TSU with a Lenz, but I could never quite get the chuffs to match to my liking, and I eventually replaced my two-decoder setup with a single ESU LokSound using the BigBoy sound file).  Interestingly, this has all worked well with the Lenz/Tsunami combo; it seems to work less well when I use other motor decoders, perhaps because of the way they implement CV's 2, 5 and 6.   For example, in one of my original EM-1 installs, I used a TCS decoder (it handled the motor just fine, by the way).   But I just couldn't get the chuffs lined up as well with that combo, so eventually I took it out and replaced it with the Lenz. 

I tried mightily to get a good match between the Zimo MX621 that I ended up using in the FEF and the Tsunami using the 3-point curve, but it just didn't work, so I ended up having to do a complete speed table in the Zimo, which in turn was hard because it took me several hours to realize that CV5 also affected the 28-step speed curve.  You find these oddities in different decoders, which is why I kind of like to stick with what I know well if it works. 

In my diesels, I finally just replaced everything with ESU; trying to get good speed matching between the ESU LokSound I have in many of my diesel units and either the Lenz or TCS non-sound decoders using a 3-point curve was an exercise in frustration, so I simply ripped them all out, and put in ESU LokPilots, which have the same motor control circuits as the LokSound.  I redeployed the Lenz and TCS decoders in my fleet of FVM NS Heritage Gevo's and SD70's, where they seem to play well together.

Sigh . . . it's never easy.   God knows what the Bachmann Berkshires will need when they arrive (they are factory sound-equipped with what I understand is a stripped-down Tsunami board).   My fervent hope is that they run superbly out of the box, and that any "tinkering" will be limited to replacing the speaker.   My gut tells me this is a pipe dream.

John C.

[nstars]

Sigh . . . it's never easy.   God knows what the Bachmann Berkshires will need when they arrive (they are factory sound-equipped with what I understand is a stripped-down Tsunami board).   My fervent hope is that they run superbly out of the box, and that any "tinkering" will be limited to replacing the speaker.   My gut tells me this is a pipe dream.

John C.

Don't expect too much of the value Tsunami, it will probably lack all the adjustments you need .

Marc

[reinhardtjh]

This may give an idea of what will be in the N Scale Berkshire/Kanawha.  It's the Tsunami "Sound Value" decoder that is in the Bachmann HO 2-6-0  http://www.soundtraxx.com/factory/OEM_pages/bachmann/ho_2-6-0_sv.pdf

Other Bachmann relevant info is here: http://www.soundtraxx.com/factory/OEM_pages/bachmann/bachmann.php