Kato N scale FEF TCS DCC problem

[jdcolombo]

Question mostly for Pete and John... any thoughts about putting a bipolar electrolytic across the motor leads? If the motor really works best on DC, then feed it DC. Yes, a cap is old-school if not brute-force, but it's a thought. A 10uF/35V bipolar is 5mm x 11mm... is there room?

Got no clue.

John C.

[peteski]

At this point the advantages of the coreless motro have been clearly identified so I won't reshas them.  But to add to the information, the weight is not really lost by the lack of iron armature. A coreless motor will have a similar mass to a regular motor of similar external dimensions.  THe iron "core" is still there, usually in the form of the iron/steel sleeve (the motor housing itself)  the inside there is a magnet which is also adds mass.  The magnet is larger than a magnet(s) used in conventional motors of equivalent size. So there is no weight loss by using a coreless motor.  Actually, since a physically smaller coreless motor can be used, that frees up more room for the solid metal chassis of the loco.

I don't think adding a cap in parallel with the motor will be that easy.  Plus it would probably have to be much larger than 10uF.  The problem is that the load on the motor output now becomes an LC circuit (motor winding is the L and the cap is the C).  I am a bit rusty now on the theoretical aspects of LC circuits but I'm pretty sure it would totally mess up any BEMF feedback and it might put undue stress on the motor output transistors (since the PWM pulses have an AC component in them).  But it would be worth trying that theory as I don't think that anything will melt or blow up (at least right away). 

But I think I would start with  a 100uF cap.  A bipolar cap can be made with 2 standard polarized electrolytic caps of the same value series connected by the negative leads.  So you end uop with a cap with 2 positive leads (and half of their nominal capacitance). So 2 220uF caps will become one 110uF bipolar cap.

[C855B]

Yeah, thanks for the reminder about building a bipolar cap. It's something I did a few decades back when messing with speaker crossovers.

The presumption here is having already identified one villain as BEMF, that it is going to be turned off regardless and not a factor. All that's left is what we are assuming, the "incorrect" frequency in the PWM in relation to the inductance. The large cap or even the "mere" 10uF might be enough to keep the waveform from hitting the floor and (in my head at least) allowing the flux field to collapse.

If I was inclined (...I'm not, too many other irons in the fire...), I'd put the loco on a running stand, drag out the o'scope and look at the output of the TCS versus the Zima (or whatever).

[peteski]

If I was inclined (...I'm not, too many other irons in the fire...), I'd put the loco on a running stand, drag out the o'scope and look at the output of the TCS versus the Zima (or whatever).

+1  Like you, lack of time is the factor.

[TrinityJayOne]

Thanks John, Mike & Pete for the information, I will add it to the ol' knowledge bank It had not occured to me that such a motor would be superior for DC applications with no BEMF feedback loop.

John, we are agreed on your sentiment about low speed running. As much as I like TCS' products if my brother's FEF was my FEF I'd be ripping the TCS out and trying something else. I will bring up the Zimo solution with him but I doubt he cares enough! I also know a few people in my club who would be happy even if their trains only had 'stop' and 'mach 1' settings....

Anyhoo, I have a damaged decoder that I need to send back to TCS shortly so I'll query them on the FEF & BEMF issue then - their customer support guy has been very helpful with my technical questions in the past.

[peteski]

Thanks John, Mike & Pete for the information, I will add it to the ol' knowledge bank It had not occured to me that such a motor would be superior for DC applications with no BEMF feedback loop.

 

BTW, coreless motor sar superior in both, straight DC application and in circuits where they are powered with PWM signal from a properly designed circuitry (with or without BEMF feedback).  BEMF (again, in a properly designed circuit) is just as beneficial in conventional and in coreless motors. The purpose of BEMF is to compensate for the load on the motor shaft and keep the rpms at the constant level (for the desired speed step selected by the throttle).

[Nick Lorusso]

I found this earlier http://powersteamguy1790snewjjje.blogspot.com/2014/12/the-dither-factor-in-tcs-decoder-for.html?m=1

[peteski]

Good find Nick!  Funny, dither CVs were mentioned in the 3rd post in this thread but nobody seemed to mess around with those.  Now I wonder: since FEF-3 is such a smooth running (low-friction) locomotive - why not just turn the dither off and see how it behaves?  Just set CV57=00

[nstars]

Good find Nick!  Funny, dither CVs were mentioned in the 3rd post in this thread but nobody seemed to mess around with those.  Now I wonder: since FEF-3 is such a smooth running (low-friction) locomotive - why not just turn the dither off and see how it behaves?  Just set CV57=00

I know dither and have used it in several locomotives to improve low speed running. But it has always been in locomotives with conventional motors. I haven't tried it yet with a coreless motor. I could be wrong, but my impression of Dither was always that it's a kind of low frequency PWM. Coreless motors don't like low frequency PWM and can even be destroyed by it. If you try dither, than please check if the motor is not humming very loudly.

Marc

[TrinityJayOne]

So not good news with my brother's unit. At the club it was randomly surging as if you had cranked the throttle to max speed and then wouldn't respond to throttle input until you stopped it completely, or occasionally it would return to normal speed (and control) on it's own after a few seconds. Today on the phone I had him try various dithering settings with BEMF off and he claims none of them really made much difference. This included setting CV3 to 0, CV6 to 20 & CV5 to 40, and setting both CV56 & 57 to 0. He claimed that made it slower, but not acceptably so. As an extreme I said to try setting CV56 to 50 and he he said this made it slower still, but overall still not an acceptable creep. No further testing occurred at this point - while trying SS1 the loco decided to do its surging trick (even with CV61 = 0) and went straight off the edge of the programming track falling 2' onto the carpet. Cab flew off and tender disconnected, but is apparently otherwise undamaged and nothing looks bent. I got off the phone at that point as he was cursing up a storm...

[jdcolombo]

So not good news with my brother's unit. At the club it was randomly surging as if you had cranked the throttle to max speed and then wouldn't respond to throttle input until you stopped it completely, or occasionally it would return to normal speed (and control) on it's own after a few seconds. Today on the phone I had him try various dithering settings with BEMF off and he claims none of them really made much difference. This included setting CV3 to 0, CV6 to 20 & CV5 to 40, and setting both CV56 & 57 to 0. He claimed that made it slower, but not acceptably so. As an extreme I said to try setting CV56 to 50 and he he said this made it slower still, but overall still not an acceptable creep. No further testing occurred at this point - while trying SS1 the loco decided to do its surging trick (even with CV61 = 0) and went straight off the edge of the programming track falling 2' onto the carpet. Cab flew off and tender disconnected, but is apparently otherwise undamaged and nothing looks bent. I got off the phone at that point as he was cursing up a storm...

Ouch!

Order a standard DC board from Kato USA ($5; they ship promptly by USPS) and a Zimo MX621.  Use Peteski's illustration in this thread to hard wire the Zimo:

https://www.therailwire.net/forum/index.php?topic=34884.0

Set the Zimo's CV2 to 20, CV9 to 51, and CV 56 to 154.  That should take care of this (you can do additional tweaking of CV's 9 and 56 if needed; you might need to reduce the value of CV20 to get the slowest creep).  Total cost: $40 street, and about a half-hour of time thanks to Kato's easily-removable boiler shell and nice open space under the rear of the DC board in the frame to stash a small hard-wired decoder.

John C.

[ntex]

In case this hasn't been brought up yet, here is the response I got from TCS tech support.

"Hello Steve,

We are aware of the issue and are very close to a solution. Unfortunately it will require the decoder be sent in to us for reprogramming. Our motor control has always been known for our slow speed control with almost any DC motor including coreless motors but the tinny size of the coreless motor in the FEF has found a chink in the armor. We apologize for this inconvenience.

This will be covered under our TCS Goof Proof Warranty. Please go to the Warranty Page of our website at:http://www.tcsdcc.com/Customer_Content/Customer_Info/Warranty/warranty.html and fill out the online Warranty registration. Then send the decoder to us. Filling out this online registration will help us to expedite the return of your Warranty replacement."

[peteski]

Now that is more like it!  Reprogramming the firmware on the decoder sounds like a very creative solution.  That is a nice feature of the microcontrollers (brains) of the decoder - they contain their programming in a built-in Flash memory which can be reprogrammed. 

[C855B]

Now that is more like it!  Reprogramming the firmware on the decoder sounds like a very creative solution. ...

Hmm. Didn't I say this a couple o' weeks ago?

https://www.therailwire.net/forum/index.php?topic=34879.msg406202#msg406202

(Seriously... I sent an e-mail to TCS with a link to this thread. Appears it paid off.)

[EspeeGoldenState]

In case this hasn't been brought up yet, here is the response I got from TCS tech support.

"Hello Steve,

We are aware of the issue and are very close to a solution. Unfortunately it will require the decoder be sent in to us for reprogramming. Our motor control has always been known for our slow speed control with almost any DC motor including coreless motors but the tinny size of the coreless motor in the FEF has found a chink in the armor. We apologize for this inconvenience.

This will be covered under our TCS Goof Proof Warranty. Please go to the Warranty Page of our website at:http://www.tcsdcc.com/Customer_Content/Customer_Info/Warranty/warranty.html and fill out the online Warranty registration. Then send the decoder to us. Filling out this online registration will help us to expedite the return of your Warranty replacement."

So how do you select the K6D4 decoder from the drop down menu?

I'm going to send mine in for it to sit later, as I still plan on installing a Tsunami in it.

Chris