farawayI am sorry for the discussion triggered by my remark about the humming sound of ESU decoders. They do "work as designed" for those of us that hear them, eg. like me. There is nothing you can do. I discussed it lengthly with ESU developers a long time ago (this part of the hard- and software has never changed). I am glad most do not hear it or do not notice it and are happy with an extreme flexible sound decoder. I love the concept of customer sound uploading. I and my locomotives will join you again when ESU reworks it motor control. There are rumors they will do so with version 5. Have a great Christmas season!
I had the same discussion with ESU, Atlas, and IM; and they all acknowledge the buzz exists and that there is nothing they can really do about it. I'm simply surprised that so few others notice it. Hopefully the upgraded version will address it.
I too like the ESU motor control, and hope others enjoy their decoders as they seem to.
My new Christmas Athearn GP15 has the Tsunami2, but the slow speed performance hasn't been upgraded like I had hoped it might.
- Douglas
Reinhard
OK, that's another one that is in the v4 manual but not Select, but should work on either. It's another bit-mapped CV like CV29, it controls multiple things. Per the v4 manual, the default value is supposed to be 24, not 0.
One of the things it controls is the motor startup delay, that's bit 2 in the CV. But the one that is probably meaningful for the motor buzz is bit 4. The choices are "adaptive regulation frequency" or "constant regulation frequency". If the total value for CV124 was 0, then prime mover delay was off and it was on adaptive regulation. Changing the value to 24 turned on the constant regulation (and another setting that should only affect Marklin AC power) while keeping the prime mover delay off. Now, it doesn't tell you what those two things mean, and it's not a specific term, but based on what they call it I can guess it has something to do with the way the PWM motor drive signal is generated. It may be something like using a lower frequency PWM to get started (more torque, but also more buzz) vs. always using the high frequency (lower starting torque but no buzz).
Again, the v4 manual says this is the default, but any sound project creator can set the default CVs however they want, so this may not apply to all decoders. I'll have to check mine where I hear no buzzing - I'd guess they are all set to 16 or higher (bit 4 is value 16, so if the CV is at least 16, the constant frequency setting is enabled).
--Randy
Modeling the Reading Railroad in the 1950's
Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.
CV29 doesn't control BEMF. Most of the Euro decoders DO have a CV which you adjust for the 'type' of motor - this can have a drastic effect on the operation, since the settings vary from 3 pole open frame motors to coreless can motors and everything in between, and the optimal PWM drive is vastly different for each type. Loksound actually doesn't have this, but the manuals for the Select and V4 give some recommended settings for the BEMF parameters for different motor types, plus they have the automatic adjustment.
Sometimes activating BEMF (CV29) will cause a locomotive to emit an slight audible buzz. I only had that happen with one locomotive and I can't remember which one, or whether it was outfitted with a Loksound decoder.
Tom
https://tstage9.wixsite.com/nyc-modeling
Time...It marches on...without ever turning around to see if anyone is even keeping in step.
Randy, there is a CV that will help that buzz noise. Right now I can't put my hands on my notes. CV ? is changed from a value of 20 to 24 . See if I can't locate the CV later. Thought is was with my Lok Sound info> found it ! CV- might be set to a value of 4, change that to 20 or24.
So you're saying this buzz is even louder than the prime mover on a low volume setting? I always set my locos on a fairl ylow volume level, the stock 3/4 or full blast is WAY too loud to be realistic. The other side of the room is supposed to be many many miles away, you never hear a prototype that far away. I am a firm believer in low volume that can be heard near the loco only - not only is the quality often better, it also allows you to hear the sound increase as the loco approaches your viewing spot and trail off as it goes away.
Now, you did mention about the quality of the locos - the switchers I have are either Stewart/Bowser or the older P2K Alcos. The Bowsers always have been dead quiet, best and quietest running locos I have, hands down. They're still quiet with TCS motor only decoders (T1's from about 7-8 years ago, after they added BEMF) and the one with a Loksound. The P2Ks have TCS also, and are fairly quiet, certainly nothing objectionable. The worst loco I have in service is probably the Bachmann 44 tonner, though I do have an MDC boxcab sitting here on my desk and most of us know how loud they are - bit it's getting completely reworked, including shell modifications, to better resemble the loco it's going to be. The T-1 4-8-4's when muted are a bit noisy with all the valve gear and lots of wheels making noise, the motors are rather quiet. I used to run those for hours at a time looping the layout at step 1 for background sound while working on something.
I've got a raft of Atlas/Kato RS3's but thus far the only ones I've put decoders in are from years ago and have either NCE D13SRJ or old TCS T1 (prior to BEMF) and they all ran quietly as well, excepto for one that turned out to have a bad motor - somehow there was the glue used to hold the commutator segments in all over one of the segments, so it was very noisy when it ran but usually stopped on the glue insulated segment and then wouldn't self-start. Scraping the glue off fixed it up just fine. Kato motors are good, but the Canon in the Bowsers is even better.
Since I have some time off, maybe I'll go get the Bowser with Loksound and set up a small test track and see if I can hear this buzz. I've operated it at slow speed in close proximity many times, although not in a totally quiet room and never heard it.
I know what I'm talking about. I just don't know the cause or the remedy, unless its just a matter of settling for whichever brand of sound decoder is less offensive.
I watched JLWII2000's review of the IM GP10. Its incomplete, as most reviews on the net are.
Any review posted on the net should include a part where the loco runs at speed step 1 through 5 on F8 mute. Especially switchers.
Switchers pull few cars and run slowly, which is different than an SD90M which would pull long trains at high speeds....creating a bunch of ambient noise. Switcher operations create low ambient noise and the volume must be turned down accordingly, which reveals any noises being generated by the loco.
Including traditional drivetrain noises.
Being the owner of 2 IM GP10s w/Loksound, I can tell you that JWLII2000's review did not cover everything one needs to know before buying a switcher.
For example, I just watched the review JLWII2000 just posted on the IM cab forward. I hear no buzz. The thing just creeps along perfectly in step 1, both forward and reverse.
wjstix Doughless CNR378 Doughless rrinker If CV2 is already 0, you can't make it any slow with any sort of speed table, you need to work with the Back EMF CVs and adjust the motor control. --Randy Randy, can you be more specific in terms of CV adjustments? CV 2 at 1 provides acceptable speed for me, but the stops are really abrupt. Do you have any magic combination of CV values to do your BEMF suggestion? For abrupt stops or starting you should be looking at momentum CV3 for acceleration and CV4 for deceleration. Start increasing CV4 until the stopping is what you want or expect. Peter I've tried the deceleration momentum, which is CV4. Cranked up the value to near its maximum. It works and looks good when taking the throttle from, say, speed step 5 of 128 down to 0. The deceleration over a longer period masks the abrupt stop when going at low speed. But going from step 1 of 128 to 0 looks abrupt. CV4 doesn't seem to matter. Loksound and even old QSI both tend to have a softer landing at that speed even with no momentum. Is it the same way when starting? Does changing CV 3 affect that? Assuming you've set CV2 to zero, and haven't inadvertenly plugged in some odd speed curve, not sure how much more can be done? I'd try setting CV 5 down a ways (say to 100) and see if that helps. Reducing the top speed reduces the amount of each speed step, so should make the transition easier. Some decoders have a "brake" function, usually one of the function buttons. Maybe make sure that it isn't "on" somehow. Some decoders have a "switcher" setting (another function button) that cuts the top speed down, and reduces or cancels momentum settings. I'd check that out...if you had to put CV 4 to it's max to notice a change, it might be the problem. Usually a setting of 15-20 makes a big difference. Otherwise, if it's a new engine, maybe it's just not broken in yet? When I get a new engine, I like to run at around 2/3 full speed forward for 10 min. or so, then in reverse for another 10. The engine almost always runs better after that.
Doughless CNR378 Doughless rrinker If CV2 is already 0, you can't make it any slow with any sort of speed table, you need to work with the Back EMF CVs and adjust the motor control. --Randy Randy, can you be more specific in terms of CV adjustments? CV 2 at 1 provides acceptable speed for me, but the stops are really abrupt. Do you have any magic combination of CV values to do your BEMF suggestion? For abrupt stops or starting you should be looking at momentum CV3 for acceleration and CV4 for deceleration. Start increasing CV4 until the stopping is what you want or expect. Peter I've tried the deceleration momentum, which is CV4. Cranked up the value to near its maximum. It works and looks good when taking the throttle from, say, speed step 5 of 128 down to 0. The deceleration over a longer period masks the abrupt stop when going at low speed. But going from step 1 of 128 to 0 looks abrupt. CV4 doesn't seem to matter. Loksound and even old QSI both tend to have a softer landing at that speed even with no momentum.
CNR378 Doughless rrinker If CV2 is already 0, you can't make it any slow with any sort of speed table, you need to work with the Back EMF CVs and adjust the motor control. --Randy Randy, can you be more specific in terms of CV adjustments? CV 2 at 1 provides acceptable speed for me, but the stops are really abrupt. Do you have any magic combination of CV values to do your BEMF suggestion? For abrupt stops or starting you should be looking at momentum CV3 for acceleration and CV4 for deceleration. Start increasing CV4 until the stopping is what you want or expect. Peter
Doughless rrinker If CV2 is already 0, you can't make it any slow with any sort of speed table, you need to work with the Back EMF CVs and adjust the motor control. --Randy Randy, can you be more specific in terms of CV adjustments? CV 2 at 1 provides acceptable speed for me, but the stops are really abrupt. Do you have any magic combination of CV values to do your BEMF suggestion?
rrinker If CV2 is already 0, you can't make it any slow with any sort of speed table, you need to work with the Back EMF CVs and adjust the motor control. --Randy
If CV2 is already 0, you can't make it any slow with any sort of speed table, you need to work with the Back EMF CVs and adjust the motor control.
Randy, can you be more specific in terms of CV adjustments? CV 2 at 1 provides acceptable speed for me, but the stops are really abrupt.
Do you have any magic combination of CV values to do your BEMF suggestion?
For abrupt stops or starting you should be looking at momentum CV3 for acceleration and CV4 for deceleration. Start increasing CV4 until the stopping is what you want or expect.
Peter
I've tried the deceleration momentum, which is CV4. Cranked up the value to near its maximum. It works and looks good when taking the throttle from, say, speed step 5 of 128 down to 0. The deceleration over a longer period masks the abrupt stop when going at low speed.
But going from step 1 of 128 to 0 looks abrupt. CV4 doesn't seem to matter. Loksound and even old QSI both tend to have a softer landing at that speed even with no momentum.
Is it the same way when starting? Does changing CV 3 affect that?
Assuming you've set CV2 to zero, and haven't inadvertenly plugged in some odd speed curve, not sure how much more can be done? I'd try setting CV 5 down a ways (say to 100) and see if that helps. Reducing the top speed reduces the amount of each speed step, so should make the transition easier.
Some decoders have a "brake" function, usually one of the function buttons. Maybe make sure that it isn't "on" somehow.
Some decoders have a "switcher" setting (another function button) that cuts the top speed down, and reduces or cancels momentum settings. I'd check that out...if you had to put CV 4 to it's max to notice a change, it might be the problem. Usually a setting of 15-20 makes a big difference.
Otherwise, if it's a new engine, maybe it's just not broken in yet? When I get a new engine, I like to run at around 2/3 full speed forward for 10 min. or so, then in reverse for another 10. The engine almost always runs better after that.
Starting from step 0 to 1 is not as abrupt as stopping, and I can live with it. I've done a lot of trials with different locos and different speed settings. I reset the CVs frequently.
Momentum has no effect on the abrupt stops. It has a lot of effect on all other performance, as do many other motor control CV settings.
OEM PROTOs and Athearn's as well as Tsunami Genset all exhibit the same abrupt stopping characteristics.
I've thought about breaking in, and have done that. Maybe they need more. But other locos quipped with QSI, ESU, and TCS don't have that problem with the same run time.
Maybe I'm being picky, but all of my locos are GPs and switchers and the move from step 1 to 0 is a common move for me in operation. Again, heavy momentum makes decceleration from step 4 to 0 look smooth enough, but I don't spot cars at industry doors that way.
rrinker I must be the only one who doesn;t hear the buzzing from Loksound. And my favorite to operate is my Baldwin switcher, at very slow speed and very close to it while switching cars. Maybe the Canon motors in them are not as affected by the motor drive pulses? But my old Loksound 3.5 PCM steamer doesn't buzz either. Though I think they used Canon motors in those, too. --Randy
I must be the only one who doesn;t hear the buzzing from Loksound. And my favorite to operate is my Baldwin switcher, at very slow speed and very close to it while switching cars. Maybe the Canon motors in them are not as affected by the motor drive pulses? But my old Loksound 3.5 PCM steamer doesn't buzz either. Though I think they used Canon motors in those, too.
I hate to mention brand because the net can be a source of misinformation or confusion, and I don't want to impugn a company. Lets just say that the newer stuff, even nonsound decoders, buzz a lot on my layout. 21 pin stuff from several manufacturers.
Only my QSIs and NCE DA-SRs are truely silent. Tsunamis are silent also. And they are the old tech guys by today's standards.
I use an NCE Power Cab factory out of the box and my locomotives are all Atlas, Intermountain, and Walthers Proto from about 2004 vintage to the newest vintages. As a general observation, the older vintages are quieter than the newer vintages. None have been run very long.
faraway The new Tsumani 2 behaves out of the box similar to Loksound without the nasty humming sound.
The new Tsumani 2 behaves out of the box similar to Loksound without the nasty humming sound.
Thanks Reinhard. It seems that each decoder has its imperfections in one way or another and compromises have to be lived with. I was hoping that replacing my buzzy Loksounds with Tsunami 2s will not impair ultra slow speed performance.
I'll try one and see what happens.
farawayThe new Tsumani 2 behaves out of the box similar to Loksound without the nasty humming sound.
That's encouraging news to hear, Reinhard.
I think Peter is on the right track, I'd suggest setting CV 3 and 4 to say 10 and see how it reacts. Increase by 5 or 10 if it's not enough difference and try again. The momentum CVs will smooth out the transitions between speed steps.
I suggest to go to the "Advanced" page of the decoder in Decoder Pro. The values of Motor Kp Coefficient and Motor Ki Coefficient should be set to values 75 and 5 as a first try. Stay for Kp values above 25 and Ki between 3 and 10. The unusal low valus of Ki is key for slow speed.
The speed table shall be "User defined speed tabel" start at 1 slow increasing up 10 in increments of 2 and than straight to the max. Forward and Reverse Trim set to 40 is fast enough for most engies.
Give it a try and the older Tsunami 1 will behave quite nice.
Arto Anyone have any good CV settings to make a Tsunami equipped diesel start/stop slower/smoother without getting into all the speed table stuff? Step 1 on the throttle makes an abrupt start/stop. Step 1 needs to be slower. I tried changing the Throttle Speed Limit (%) on the JMRI Roster Tab to 75% (from 100%) & it didn't seem to make any difference.
Anyone have any good CV settings to make a Tsunami equipped diesel start/stop slower/smoother without getting into all the speed table stuff?
Step 1 on the throttle makes an abrupt start/stop. Step 1 needs to be slower.
I tried changing the Throttle Speed Limit (%) on the JMRI Roster Tab to 75% (from 100%) & it didn't seem to make any difference.
Arto,
One of my biggest beefs about the Tsunami decoders and why I prefer Loksound or TCS Wow!. The motor-control for the Tsunamis is so-so and definitely inferior to the latter two brands. While I do enjoy sound (in reasonable doses), motor-control is WAAAAY more important to me than sound - especially in switchers, which make up the majority of my roster.
I've never been able to get a Soundtraxx decoder under 2 or 3sMPH - even with tweaking CVs. With either ESU or TCS decoders, I can consistently get <1sMPH - right out of the box.
I believe the newer Tsunami 2 decoders have made improvements in the area of motor-control. However, I haven't been able to confirm that yet. Until I do Loksound and TCS will remain my go-to sound decoders.
woodone Which Tsunami are you working with? The TSU-1000's or the newer Tsunami 2 ? The TSU-1000 do not support CV's 2, 5 and 6.
Which Tsunami are you working with? The TSU-1000's or the newer Tsunami 2 ?
The TSU-1000 do not support CV's 2, 5 and 6.
Not quite right, they only support Vmin (CV2)
Are these the decoders where the momentum CVs don't do anything? I have a couple of those.
To prevent jackrabbit starts, I adjusted Vhigh. I don't think it's the CV it's supposed to be, but the Soundtraxx documentation does tell you what it is. You may need to play with Vmid, too. This does limit the top speed of the locomotive, but I never run my engines at warp speed anyway.
It takes an iron man to play with a toy iron horse.
Mark Gurries has this page for tuning soundtraxx decoders
https://sites.google.com/site/markgurries/home/decoders/decoder-motor-tuning/soundtraxx
Arto And thanks CNR - that's good to know. So that Roster Throttle setting would work with an Android running through JMRI? Or just from JMRI?
And thanks CNR - that's good to know. So that Roster Throttle setting would work with an Android running through JMRI? Or just from JMRI?
Sorry Arto, don't know the answer to that part. If I remember I'll try it next week at the club.
Setting CV 2 to 1 or 0 is about the best I can do provided the loco doesn't vibrate or shutter at that low of setting. On step 1 of 128, take off and overall speed is acceptable, but stopping is still way too abrupt.
I dont have JMRI and figuring out BEMF adjustments to yield better results is above my pay grade.
Thanks Randy. That's what I was afraid of. I have some ATHG F45 with the same that I made a lot of BEMF changes to. I guess I'll copy those to the U50 and see what happens. The F45 aren't where I want them yet but better than default.