Atlas SDP-35 conversion to DCC

Is there any way you can disconnect the motor body from the positive brush connection? Te terminals are connected to the command board in the 3-rail units It surprises me that they'd be using a different motor for the 2-rail engines.

Hi Matt

I was just about to post.  To my joy I've found that upon removing the truck from the body and testing it again, the motor wires are actually not connected to the motor body. 

This means that the short is via the mounting screws.  They have fibre washers which are supposed to be doing the job.

Enquiries (as they say in the police shows), are continuing.

Cheers

I was very surprised to read that the motor leads were connected to the frame because I installed DCC in an Atlas GP15 and did not have that problem. I hope you can work out the short circuit with relative ease.

At York there was a vendor called Harbor Belt Lines. They had a SP GP35 on a short stretch of track. I asked who made the model and the man told me it was an Atlas GP35. He then proceeded to tell that he installed the LokSound V4.0 decoder in it. He said it was the same decoder that the new ATlas F7s will have. He also had the decoder for sale. I think he said it was $170. The locomotive sounded great to my ear but what was most impressive about it was it had the best slow speed performance of any "China Drive" locomotive I ever saw. The guy was able to make it crawl. It was very impressive. I don't know if he changed the CVs or it is something with the decoder itself but I am dying to try out of these decoders myself.

Max I look forward to progress on your install and please let us know how it sounds and runs when you are done.

Thanks, guys.

I've replaced the fibre washers with 0.5mm styrene.  All seems to be well.  And guess what?  One screw had no washer! 

I've fitted dozens of Loksounds to HO locos, so I'm pretty confident that I can make it OK.

Interestingly, the Select L has provision for servos - could be a possibility for coupler opening control. 

When you start your first install send me an email if you need to ask any questions, Phil.

Cheers

Blimey Max, did you walk under a ladder recently? I swear you are jinxed with these locos.  

Anyway, you have it figured out correctly of course. I've done a couple of Atlas GPs and SDs and I guess I was lucky. Quick polarity check of red and black pickup leads was all it took. Motors were fully isolated. But when I took one apart (just because I could) I was surprised to see those little white plastic washers that were the only thing between life and death! 

The only issue I found was that the blue and yellow motor wires are sometimes not the same polarity per loco. One thing I highly recommend is to wire the motors in series not parallel. That way you get better starting and slow running with finer speed control at the expense of lower top speed. Also there's a chance that one truck can spin its wheels under load which diverts power from the non-slipping truck as you lose the "limited slip diff" effect of parallel wiring. But for me with only 15-20 car trains on level track, never been an issue. So it's orange > motor 1 +ve > motor 1 -ve >  motor 2 +ve > motor 2 -ve > grey for me.   

I used Tsunamis on the first 3 I did and they take far too much tuning to get good slow running imho. I should not need to learn about PID, KD KP, KI, and don't even go there with droop...       The good news is that Loksound and TCS WOWsound seem to have great motor control settings right out of the box. 

Looking forward to seeing it run!

Pete

Hi Pete

It's just like going back to square one, changing to O scale. 

I like to post everything on a project, so hopefully peeps coming after can get some benefit.

I am thinking of wiring the motors in series, from experiences others have had.  I plan to hook the motors up to DC to make sure they both turn the same way, first.  We're doing a benefit concert for the victims of the recent bushfires on the 19th, so rehearsals are taking up some of my time.

Loksound have a self aligning function which sets the decoder up for the optimum BEMF.  CV54=O  then press F1.  The loco takes off along the track for a few seconds and then stops.  BEMF is all set for you.

Another thing I'm doing as I go along is rewiring to the correct NMRA colour codes.  No rehearsal tonight, so I hope to get some photos up.

For now, off to the Dentist. 

Cheers

LOL Max! Sad but true. I've only modelled in O scale, after dabbling in 7mm steam era back in the UK, 20+ years ago. I bought a Red Caboose GP9 kit in 1992, my first Digitrax system in 1994 and never looked back. 

I do marvel at the fine, accurate detail of modern HO rolling stock though. And ironically today's HO locos seem to have better starting and slow speed running than the equivalent O scale units. I am envious of the HO world where you can run a highly detailed and accurate DCC sound equipped model straight out of the box. I have never bought an O scale model that didn't need at least some work before I would consider it "layout-worthy". But still I wouldn't trade the mass, momentum and sense of being trackside that I get from O scale.

Besides, with my aging eyes and lack of dexterity, I can't imagine how I'd work on HO equipment.      

Hope the dentist was in a good mood.

Pete

Max, Well done.  Would be interested in seeing a video and hearing the sounds from the Loksound Select L board in the SDP35.

He was, Pete - after I paid my bill.  His son is safe in private school next year. 

Thanks, Mike.  I hope to make a YouTube extravaganza. 

Cheers

Pete M posted:

So it's orange > motor 1 +ve > motor 1 -ve >  motor 2 +ve > motor 2 -ve > grey for me.   

Pete

That looks like yours are connected in series, Pete.  When I connect my two motors on the bench in series, one takes all of the power and spins; while the other doesn't move.

However, when I connect them in parallel (Orange M 1 + Orange M 2  > Red (+ 12 VDC)  and   Grey M 1 + Grey M 2  >  Black (- 12 VDC), both motors run at approximately the same speed.  (It could be exactly the same speed; I can't tell for sure - but it's very close).

I know that it's counter-intuitive, but that's what's happening.   

Cheers

Yes, mine are in series. But what you're seeing is normal.  Hence my comment above "...Also there's a chance that one truck can spin its wheels under load which diverts power from the non-slipping truck as you lose the "limited slip diff" effect of parallel wiring. ..."

When the motors are in series, they are dependent to some degree on each other's resistance: electrical in themselves and mechanical in their respective truck assemblies. This setup also relies on all wheelsets seeing some coupled mechanical resistance to work properly i.e. on the same track, not in free space or on separate rollers. In series, unless both motor/truck packages happened to present exactly the same load, the package with the lower load will be the path of least resistance and all the power will go there.     

In parallel wiring, both motor/truck packages are electrically independent so behave similarly to each other with or without load. One of the reasons dual motor drives wired in parallel don't do well at smooth starts and slow running is precisely because these cheap motors aren't exactly the same. One tries to move before the other causing binding and jerks at the start.

So in my "car guy" analogy, parallel wiring has a "limited slip diff" effect, while series is the opposite. Worse than an open diff in fact, encouraging slip in one truck once it starts. But I prefer the "voltage spread between 2 motors" effect of series to get smoother starts and more subtle slow speed control.    

Try your tests on the track so it loads up the trucks together and see which you prefer. One thing you will see in the series setup will show how poorly made these Atlas motors are. When you open the throttle step 1, you will see that the flywheel on one motor rises before the loco starts to move. That shows you which motor has the lower resistance and allows the commutator to rise due to too much shaft end float. Nothing we can do about that, other than fit high-quality ball bearing motors I guess.  

Please note that I am not an electrician and my attempts to describe what I see happening in my locos may be way off the mark!  

HTH

Pete     

That all makes a lot of sense, Pete.  Thanks.

I'm going to keep the motor wiring just twisted together until I can run the thing on the track.

There's no rehearsal tonight, so I'm hoping to get the daughter board and decoder fitted.

Then we will see what we will see. 

Cheers

Pete M posted:

The only issue I found was that the blue and yellow motor wires are sometimes not the same polarity per loco.

Well, mine were fine - except that to run correctly, they need to turn IN OPPOSITE DIRECTIONS.  I couldn't understand what all of the drama was about until I took a torch and checked the rotation OF THE WHEELS. 

Once I got both of the trucks running the same way, the loco performed quite well.  Motors wired in series.

Here's my home made decoder tester . . .

Made up from bibs and bobs.  You can see how easily the decoder can be removed if necessary.

Here's a side shot of the completed installation . . .

. . . and the Starboard side . . .

Plenty of room to mount a couple of servos up the back (for opening the couplers).  Later, maybe.

I decided against my better judgement to keep the grain of wheat globes in the front and rear lights.  Of course they are subject to flickering - unlike the LEDs.  Later, maybe.

I used a couple of old paper type HO speakers I had lying around - and they don't sound too bad.

It runs a bit lumpy (cogging), which might go once I've cleaned the wheels.

Off to make a movie of The Beast.

Cheers

Very nice. Since you series-wired the motors, did you use H.O. sound decoders for this conversion?

Thanks, Matt.

No.  I'm using Select L

I don't expect people to sit through all of it. 

The slow speed running is actually smooth, not jerky like in the video.

Enjoy.

Very nice Max!  So that slow running was Select L out of the box plus the 1-step back EMF calibration? That's at least as good as my Tsunamis after hours of "buttering about"    with 5 motor control CVs based on 3 gurus' step-by-step Tsunami tuning guides I found online. But in the end, I fear the dual motor issues prevent the "ultimate" smooth start and slow running performance we ought to be able to expect in O scale.   Of course, there's always the Weaver/P&D single motor drive, which as you know drops right in and runs super smooth right out of the box...  

Anyhoo, onwards and upwards. I am sticking with O scale no matter what!  I'll try to grab a quick video of my Tsunami locos over the weekend. It'll be interesting to compare.

Thanks so much for all the detailed posts, pics and vids of your journey. I'm learning every day! 

Pete      

That's very impressive performance. Watching the performance of DCC engines I've been re-thinking what type of command control to install on some of my old conventional locomotives. As there's no technical prohibition against using a 3-rail command system (TMCC or DCS) on a 2-rail engine, there is also no technical prohibition against using DCC on 3-rail can-motored engines.

Thanks, Pete.

Yes. Straight out of the box.  I went to the ESU site and found the EMD 567 D3A sound file, downloaded it and basically uploaded it straight into the loco.  Having a Lokprogrammer makes life easy, but all that I had to do was load the loco number into the file and change the horn to a multi-chime.  A couple of clicks.

The secret with ESU decoders is the auto-tune function.  Enter 0 into CV 54, escape and press F1.  The loco runs forward about 2 feet and stops.  The decoder has read the back EMF and adjusted automatically.  There is a tiny bit of cogging, but that's in the drive train.  Both trucks float about in their mountings, so any slack in the gears is exaggerated.

I'm happy with it - specially for a first attempt.  Thanks for your input.

Thanks, Matt.

I regret not changing the grain of wheat globes to LEDs.  Already the rear light has blown.  A job for another day.

For a couple of old paper speakers from my handy box, I'm impressed myself with the sound from the bigger decoder.  I had it on full for the video, but I'll turn it down to where we can talk over it, now. 

Thanks for your interest.

Max, Very nice work!  I like the sound out of the 567 prime mover and was imagining a heavy load of cars being pulled behind the locomotive where you could really get the high amp, notch 8 effect to really get the train moving and then back it out.   Volume wise hard to tell, prime mover vs horn vs bell, but the horn seemed a little low compared to the other sound effects?     Definitely worth replacing the bulbs with LED's as now days the golden white LEDs produce the correct lighting, or close enough, for a locomotive of this era.

Extremely well done, thanks for sharing!

MaxSouthOz posted:

Here's my home made decoder tester . . .

Max, Like the homemade decoder tester! 

Thanks, Mike.  Yes.  I cranked everything up for the movie, but the motor was a bit on the loud side.  I'm going to drop it down under the horn and bell, as they have no headroom left.  My wife complains that she can hear it at the other end of the house.   

Thanks.   The decoder tester joins a pile of ones I've made when I was into HO.  The beauty of this one is that the L decoders just plug in.  I've only got two locos (so far  ), but it's really handy for cross testing.

Cheers

Max, I appreciated seeing the Seal that was posted at one of your street crossings. :-)

Dan Weinhold

Lancaster, PA

Nice! Really wish Atlas would do an SDP45...

Thanks, Dan.  I wondered if anyone would notice.  The seal will be incorporated in the dockside part of the layout as I go.

Thanks, Steven.  I enjoyed the ride on your layout. 

Cheers

hi max:

i'm pretty sure the motor case can contact the truck and not short out. my guess is that your original problem was where the motor power solder lug bends over the top of the can. chances are it was touching the case, and in your handling you inadvertantly "fixed" it. if this is the case, watch that it doesn't creep back into the picture. 

i have just "found" the loksound select L. WOW! almost everybody over here wrote off loksound when they required 83.384766 ohm speakers. anybody brave enough to check again will find loksound listened and makes real nmra dcc docoders with 8 ohm speakers. i have only played with a couple alco diesel sounds. are the steam sounds just as good?

at work, most of my collegues have english as a second language. all of you multi-lingual people amaze me, but we are going to have to have a lesson here for you max. it is too scary for us to read that you took a torch to your trucks to see the wheels. use flashlight next time to prevent heart attacks..

calvin.

Hi Calvin

It's in the text somewhere, the problem was where the truck screws on to the chassis.  I replaced the fibre washers with styrene.  Problem solved.  See the photos.

Modern ESU V4 decoders will run 4Ω to 16 Ω speakers.  Yes.  The steam is also good.  Matthew Herman from ESU LLC is a real dynamo - working  with their decoders/ECos et al to keep at the head of the pack.

Ha ha.  I get it.  In Oz we call a flashlight a torch - same as UK.

Cheers