Balancing power to PS3 trucks, help.

I see that a lot when the engine is on it's side. I did try to cure it once by actually moving the tach to the other truck. It wasn't worth it and I expect it would be tough for you to try that here as well.

 Sometimes brand new trucks just need a break in period and the low speed performance will smooth out. If the engines on the track do you notice one truck dragging the other or the engine stumbling at very low speeds?

Normal operation, the truck with the tach will get enough power to run at the command speed, the other truck just gets the same voltage and very often doesn't start running at low settings.  It has no effect when you're actually running on the rails.

You can clean motor commutator area with electric motor cleaner. Make sure no issues with gears and lubricate.  If still off or does this when coupled to track, replace motor. G

That is what I wanted to know. So, there is no means to balance voltage to the trucks.  I have since taken the motors apart and lubricated the bearings and I found some bumps on the fitting of the mating housing and since I did that, the motors have come closer to the same speed, starting and stopping equally, but the two speeds are slightly off.  I would be nice to be able to tune the two trucks to the same rpm.  Thank you very much for this response. Bob.

It's a rare locomotive that will run the two trucks at the same speed with the wheels turning freely.  It takes very little energy to turn the wheels with no contact, and the slightest difference in the drivetrain friction or the motor characteristics will make one run faster than the other.  With them both running, you can very easily stop the drive motor that doesn't have the speed sensor.

IMO you're trying to answer a question that probably should be asked unless you see actual operational issues while running under load on the track.

I am listening to you guys, so I have yet to get this loco on the tracks under actual loads.  I will update everyone once that happens. Thanks again for all of your input, I appreciate it. Bob.

I think you'll see when it's on the tracks, you won't see that.

Note that there is a potential traction issue with virtually all twin motor locomotives that have tach based speed control.  If the speed controlled motor loses traction and starts spinning, it will have it's voltage dropped to maintain the set speed (in the air), and the other motor will have minimal voltage as well.  I see this happen on the entry to poorly designed grades, or even a dip in the track.  We had one module on our club layout that had a dip, and several locomotives would lose traction there and just stop with the one truck spinning.  Yes, we fixed the module.

OK, I am just dreaming now, but since I am a multi engine pilot, synchronizing engines is a matter of independent throttle control.  What if an additional tach. was put on the other truck and just measured rpm and not power so that one could see the relative rpm's of both trucks and then be able to independently control that one to match the other.  If that was possible, the traction would be greatly improved, because once one of the trucks is not pulling the same as the other, the efficiency is greatly reduced.  Once one truck is trying to go faster or slower, the effectiveness of the other truck is greatly effected.  Am I wrong?, I know sometimes I think too much. Bob.

I notice this when I clean wheels, its just like JRJ said,  normal without load. All my MTH O gauge diesels do this.

machiningfool posted:

OK, I am just dreaming now, but since I am a multi engine pilot, synchronizing engines is a matter of independent throttle control.

Funny thing, so am I.

machiningfool posted:

What if an additional tach. was put on the other truck and just measured rpm and not power so that one could see the relative rpm's of both trucks and then be able to independently control that one to match the other. 

That's quite possible, and it wouldn't be all that hard to do, just another sensor and obviously some additional software.  However, I have never seen or heard of a case where that's been done on a model train.

Yes, I think it is possible and I also think that battery time would be increased quite alot.  I might also try just running one powered truck and see how that works.  I would probably have to add some weight on top of the truck, I might just try it.  I am trying to imagine what happens when one truck is not pulling the same as the other.  If it is slower, it is putting drag on the one with the tach.  If it is faster, it might just be spinning and therefore exceeding the traction point and then also causing more drag.  I don't really know, but I am thinking about it. Bob.

It really balances out when they're on the rails.  I can tell you with some personal experience, having a single powered truck is not as effective as having both powered trucks.

You're looking for problems where none exist.

My asking questions is for the express purpose of finding out something that I don't know. So, what you are saying is that the voltages and the traction point happens automatically when in operation as if it already has synchronizing ability even though differences in voltage applied, differences in bearing friction, will result in the two trucks turning at the same rpm.  If they are not turning at the same rpm, one would be pushing or pulling the other, wouldn't it?  So I guess what you are saying is that even if it is not turning at the same speed, even though  one might be slipping, it still offers more push than it does drag? Is that correct?  Bob.

GGG posted:

You can clean motor commutator area with electric motor cleaner. Make sure no issues with gears and lubricate.  If still off or does this when coupled to track, replace motor. G

Above advice is sound. CRC has a brushed motor cleaner. I use as shown:

Air dry after application. All my engines (ps3 and ps2) are smooth at scale 2smph and 1smph most of the time.

Thank you very much, I appreciate it. Bob.

machiningfool posted:

So I guess what you are saying is that even if it is not turning at the same speed, even though  one might be slipping, it still offers more push than it does drag? Is that correct?  Bob.

Actually, what I'm saying is the forces equalize when there is a load on both trucks.  Neither truck spins or slips unless it loses traction with the rails.  All the major makers use the same scheme for cruise, one motor is speed controlled, the other truck gets the same motor voltage.

machiningfool posted:

... even though differences in voltage applied, differences in bearing friction, will result in the two trucks turning at the same rpm.  If they are not turning at the same rpm, one would be pushing or pulling the other, wouldn't it? ...

But the voltage applied is the same.  The motors have published speed-torque curves for a given applied voltage.  At one end of the curve is the stall torque (maximum torque at 0 speed).  At the other end is the no-load torque (0 torque at full speed).  From manufacturing variation, motors of same model have slightly different speed-torque curves giving rise to the observed scenario - in other words different speeds for the same applied voltage.

In normal operation, and helped by rubber traction tires, both wheel trucks are spinning at the same rate.  This means both motors are spinning at the same rate.  Since the same voltage is being applied to both motors this means that the two motors are generating slightly different amounts of torque.  But it is still more torque than one motor by itself with the same applied voltage. 

The "fun" comes when you think through the various scenarios...and why they converge to the case of BOTH motor trucks spinning at the same rate on the track.  For example:

Engine stopped. Now commanded to some speed.  The motor with the tach (call it A) starts to spin first.  The gear transmits the torque to the track and the traction tires transfers "all" of it to the track.  But the engine does not move because the other truck (call it motor B) is still locked because its motor is providing no torque.  So the motor control increases voltage to BOTH motors A and B.  At some point either the A motor's torque will be able to start dragging the B truck and/or the B truck will start to spin and the control voltage can be lowered.

The key is A and B can be delivering different amounts of torque and yet spin at the same RPM.  I don't know of a simple analogy but two people, say an adult and a child can be pushing a disabled car and be providing different amounts of force.   Even if the child is not pushing as hard, the force provided still helps.  Or a tandem bicycle still benefits even if one person is pedaling harder. 

Well, that's my view of the situation anyway.  There is a related but separate discussion on electro-mechanical efficiency (e.g., Horsepowers of pulling power output vs. Watts of electrical power input).  At first glance you'd think efficiency is maximized if A and B are perfectly matched and operating at identical points of their speed-torque curves.  But that's at first glance.  Then you go down the rabbit-hole of MU consists with mis-matched engines, etc. etc.

 All I can add to this conversation is what I see more of when upgrading other brands of G scale to MTH command. At very slow speeds (1 or 2 MPH), one truck does seem to drag the other. The engine will make a jerky motion attempting to move. If you advance the speed from 1 or 2 MPH to say 5 MPH, they seem to smooth right out. Then I'd agree with the example of both spinning and maybe one is doing slightly more work than the other, but it's not noticeable to the eye watching the engine's movement.

 It was so pronounced on a particular engine I converted, that I removed traction tired wheels and replaced them with solid metal ones. By adding weight, the traction is good enough yet allows one truck to slip and equalize any low speed differentials. I don't like tires on engines with cheaply made plastic gears anyways (weaker, less quality).

This is a common concern, it's natural to check out the loco unloaded, or 2 locos running light a few apart to see how they match.

Under actual conditions of pulling a train, the motor running faster will be more highly loaded and slow down a bit and equalize with the other motor.

So, unfortunately the natural tesst for how a loco runs is probably the least representative of the actual running conditions.

Oh, and by the way there are DCC decoders that monitor the rpm and bemf of multiple motors in a loco. Better control if there is a severe mismatch. They do it electronically without a tach.

Greg

Right.  By removing the traction tires you remove the constraint of both motors spinning at the same rate.  The same speed-torque curves apply (since the voltage is still the same to both motors).  But now the two motors can spin at different speeds and hence deliver wildly different torques.  This can actually be less stressful on the transmission as you're providing a backdoor for the excess torque.  This is where those anti-skid or ABS systems in cars show their stuff.  Clearly the model train manufacturers could add electronics to detect slipping wheels and then back off the voltage etc. etc.  It would make an interesting high-school science project but I just can't imagine any manufacturer doing this! 

Try this With the engine in a cradle upside down, hook power to it and run with the remote, get it started about 35 to 45 MPH put you fingers on one set of wheels the secondary motor you can bring to a stop and the lead keeps turning while still applying pressure to the secondary squeeze the lead wheels till they slow down and the secondary will start spinning more,  the more you load the main wheels the faster the secondary will spin. Don't know if that helps ,but it's my observation 

Clem

Then, if I understand correctly, the slave motor, because of differences in windings, bearing friction, etc., will be helping the force, even though it might not be the same force as the truck with the tach and also not slip and cause drag.   The ideal setup would be what Greg mentioned using a synchronizing circuit to ensure that both are turning at the same rpm. regardless of the differences in the mechanical differences  of the two trucks, because the circuit would be rpm driven and provide the independent voltage necessary to each truck to match the rpm.  My thinking about this is not so much for someone that provides track power, but like me using battery power.  More efficiency would provide longer runs for a given battery charge and smoother more accurate starting.  Greg, tell us more about the auto synchronizing system that you mentioned.  That sounds interesting. Boy, I think this subject is interesting, maybe a subject for Mythbusters.   Thanks for all of the input.   Bob.

I think once both motors spin at all, this becomes less of a factor. Yes I run track power. I can run ten engines handling a train and they will all behave. That's on 2 rail O scale with fine flanges on cars that are in between these engine groups. The power draw for ten engines is just a few amps.

 In fact, the slave boards don't use a tach. So when an engine like the G scale MTH Triplex is asked to do a job, the tractive effort is huge. Yet the power draw is very low. It's only looking at one motor. All the others fall in line once the engine moves.

Looking at each engines start and stall might be beneficial if you continuously run one battery powered engine at 1 MPH (and you went to a decoder with that capability). I think it would need someone to tweak it who actually knows what they're doing.

Here is what I think.  Just because, after the initial acceleration, say from level ten and up, the wheels appear to be turning at the same rpm, they might be, but one truck might be not exerting the same torque as the other, therefore, one is working more than the other and losing efficiency.  The only way that I see to ensure that the rpm's are the same on both and applying the same torque, is to have an rpm driven type of electronic device and vary the torque independently to each truck to arrive at the same rpm.  Greg said that there is such a circuit with DCC and I would like to know more about that. Bob.  

It's in the DCC decoder and not available as a separate circuit to control motors. Something like this could be built, but it would have to be able to control the power to the motors independently.

Some of the more sophisticated DCC decoders use the BEMF (which has been around since the late 40's) and basically count motor rpm. Even with 2 motors in parallel, you can tell if they are running different speeds.

Unfortunately in MTH the simplest thing would most likely be 2 decoders, and a tach strip on each motor/truck.

Greg

Greg Elmassian posted:

...

there are DCC decoders that monitor the rpm and bemf of multiple motors in a loco. Better control if there is a severe mismatch.

Please identify some specific DCC decoders that monitor the speed of multiple motors and use that info to better control the engine.  I'd like to read about them.  Do they independently control the motors?  Thanks.

stan2004 posted:

Greg Elmassian posted:

...

there are DCC decoders that monitor the rpm and bemf of multiple motors in a loco. Better control if there is a severe mismatch.

Please identify some specific DCC decoders that monitor the speed of multiple motors and use that info to better control the engine.  I'd like to read about them.  Do they independently control the motors?  Thanks.

 

 I too am curious. Would a PS3 in dcc mode have the ability to adjust each motor?

Well Greg, this whole conversation is encouraging. I am going to dive into this a little deeper, thanks everyone, I like it.  Bob.

willygee posted:

stan2004 posted:

Greg Elmassian posted:

...

there are DCC decoders that monitor the rpm and bemf of multiple motors in a loco. Better control if there is a severe mismatch.

Please identify some specific DCC decoders that monitor the speed of multiple motors and use that info to better control the engine.  I'd like to read about them.  Do they independently control the motors?  Thanks.

 

 I too am curious. Would a PS3 in dcc mode have the ability to adjust each motor?

The MTH PS/3 in DCC mode does NOT have any capability to adjust each motor independently.  Both motors are still wired in parallel just like any other MTH PS/2 or PS/3 locomotive.

I know that the QSI DCC decoders do look at the BEMF "pulses" for 2 motors that are wired in parallel, you can tell this because it counts motor segments and applies a multiplier for "auto chuff". I do not know what algorithm they use in controlling the motors, I would suspect they look at the motors together somehow to do BEMF load correction, the BEMF settings have the standard 3 parameters and if I remember correctly have 3 speed bands that these 3 parameters can be set in.

The point here is not to pimp DCC or QSI, but that the technology exists to "look at" 2 motors that are in parallel. Clearly you cannot control them independently, but you could make some very smart decisions based on what the 2 motors are doing.

Clearly the best thing would be 2 independent motor drive circuits. DCS has taken the manual tachometer path, on one motor only, whereas using "electronic tachometers" would most likely be less expensive, perhaps even offsetting the cost of adding another independent motor control.

Of course this all depends on the electronics. If I was reading correctly, the DCS custom ASIC is no longer used, but an FPGA, which allows any amount of programming (but I may be mistaken about this being in the decoder, it might be in the TIU)

I am not a DCS expert, but I am an engineer and pretty familiar with these topics, having watched the great BEMF fiasco over the years.

Regards, Greg

I forgot to add, when running a loco like a 2-8-8-2 Mallet, I can see/hear the synchronization with the 2  "trucks", and my Aristo one, like many large scale Mallets, has one truck that does indeed start before the other, although, I can creep MUCH slower than 1 scale mile per hour, more like a tenth of that.

Greg

Greg Elmassian posted:

I know that the QSI DCC decoders do look at the BEMF "pulses" for 2 motors that are wired in parallel, you can tell this because it counts motor segments and applies a multiplier for "auto chuff". 

By that benchmark, the ERR Cruise products actively control both motors as well.  Since they use back-EMF and also commutator crossings at low speed, they seem to match your description of this operation.

Well, I appreciate this information and will investigate this setup and it looks like others are also intrigued by this possibility.  Thanks again for your trouble in enlightening a dummy like me and I will use these suggestions, and also, if you can think of anything else, that would be great. Bob.

I forgot to ask you this. You mentioned two DCS boards and a tach sensor in each truck.  So, would I control both boards with one DCS remote, and would that bring me closer to having the wheels turn at close to the same rpm? Bob.

Thank you GunnrunnerJohn, I will look into the ERR Cruise Products. Thank you, Bob.

I've never seen any locomotive with two tach sensors for speed control.  I haven't seen them all, but I've opened up hundreds of two-motor diesels from most of the manufacturers, and that's never been a feature.

I don't want to keep beating this drum, but if this was perceived to be a problem, some manufacturer would have equipped the units with dual tach sensors.  However, just the opposite is true.  The DCS units that run four motors have a master board controlling one set of two motors with a single tach reader, and a slave board (no tach sensors) that is simply slaved to the master's motor setting.  This is the scheme used in my UP Propane Turbine with four motors.  You'll note four motors, but only one tach encoding strip.

Greg Elmassian posted:

...

The point here is not to pimp DCC or QSI, but that the technology exists to "look at" 2 motors that are in parallel. Clearly you cannot control them independently, but you could make some very smart decisions based on what the 2 motors are doing.

Please share your ideas on the smart decisions when the motors are driven in parallel.

Also, do the QSI decoders extract motor-segments or commutator transitions and detect speed variations in twin-motor diesels?  I can understand doing so for single-motor steamers for chuffing feedback.  I'd think it quite the challenge to do so with two motor in parallel...and to wit, what would be the point of extracting this data if the motors cannot be independently controlled?  Or are you saying the QSI DCC decoders perform these smart decisions - in which case where I can learn more.

I think the point here is to determine what can be done to "balance power to PS3 trucks" per the thread title.  If there are some things that materially improve power balancing in twin-motor diesels by sensing motor speed I'm in.  As I've stated, during "normal" operation the two motors are at the same speed...perhaps delivering unbalanced torque.

The new MTH One-Gauge PS3 F7 AA set is using a tach equipped slave board in the trailing A unit...Just thought I'd throw that into the mix.

smart decisions....

"Please share your ideas on the smart decisions when the motors are driven in parallel...."

this ought to be good.

Casey Jones2 posted:

The new MTH One-Gauge PS3 F7 AA set is using a tach equipped slave board in the trailing A unit...Just thought I'd throw that into the mix.

They're really using two separate and independent boards in those, so it would logically have a tach reader.

gunrunnerjohn posted:

Casey Jones2 posted:

The new MTH One-Gauge PS3 F7 AA set is using a tach equipped slave board in the trailing A unit...Just thought I'd throw that into the mix.

They're really using two separate and independent boards in those, so it would logically have a tach reader.

Not really as can't run them independently. Both A units have to be plugged into each other not operated as a lashup.

I guess I'd have to see more details on how that one is assembled, it doesn't sound like any of the existing configurations that I've worked with for PS/3.

I don't mean to beating a dead horse either, but probably the reason manufacturers haven't matched the rpm's of the motors, is that it would add more cost  to their bottom line and they run smooth, so why do anything else, but I am after efficiency.  Now, since I don't have this technology, I can't really say whether or not this is worth the effort, but I am willing to try and just maybe there will be a benefit, maybe slower speeds, longer runs for battery power and more pulling power.  After all, I have learned a lot more than I new before beginning this post, and that is worth something. Casey Jones, tell us more about the new PS3 double tach setup, what is with that? That sounds interesting. Bob.

Well I'll be darned.

I don't wish to tear open my brand new PRR F7's w/ PS3. So I shined a flashlight inside the trailing A and I see a tach tape (painted flywheel) and what looks like a tach.

 Looks like you could talk to Ray about a PS3 slave. I heard they don't offer them, but it appears they do make them.

Apples and oranges and you guys are forgetting you need independent motor FETs to control two motors. So more than tach and software.  Ps3 slave does use tach but again only tach motor of slave controls voltage. G

What would be the best approach to solve this problem? Bob.

Sorry, the term "smart decisions" means that software could do something reasonably intelligent, even though somewhat hamstrung by 2 motors in parallel. I'm not saying that anyone is doing this much more that probably averaging the responses in the QSI.

And to the other question, I did already answer it, that running a 2-8-8-2 Aristo mallet, it has 2 motor blocks each with it's own motors and you can tell by the chuffs that it is reading each motor's commutator transitions...

If I was to design a system, I most likely would of course jump to separate drivers for each motor, and then let the microprocessor handle both motors, so you could do smart stuff when one motor block slips, including have the prototype option of that actually happening.

What's the point of monitoring both motors if you cannot control them independently? Well here is an idea, if either motor slips (easy to detect) then reduce speed and then ramp up again... just like a real engineer would do. I can think of a lot of things, but this is all what if stuff, nothing you can do to change the programming of the systems as they stand.

You know, some people on this thread are genuinely interested in exploring ideas, and some just want to confront and put down.

I'm in the hobby to have fun and try new stuff... if I wanted to stifle ideas, I could go to a retirement home.

Greg

machiningfool posted:

What would be the best approach to solve this problem? Bob.

But can you define the problem?  If you get 1% more battery running time is that worth $350 of electronics...less the value of a beer?   How about 2% more battery time?  Etc.

From reading posts on OGR for years there's not much discussion about power efficiency and need-to-know about where the Watts are actually going.  With growing interest in battery power I suppose this will change.  Here's my take.  The mechanical output Horsepower vs. electrical input Watts efficiency of the motors used in these diesels runs about 60-70% at peak efficiency.  For example, here's a Mabuchi RS-385 which one might see in an O-gauge engine.  With 12V applied, peak efficiency is about 65% delivering 80 g-cm of torque spinning at 8200 RPM.  Do the math on your engine (gear transmission ratio, wheel diameter) to get scale-MPH.  Note how the efficiency drops off dramatically when you go faster or slower than the sweet-spot zone.

Picking off more numbers this is 9 Watts of electrical power in, delivering about 7 Watts of motive power into the transmission.  So that's 18 Watts for two motors...and we know that you can easily be drawing an incremental 18 Watts of power off the transformer when the diesel starts pulling.  In other words for a typical O-gauge diesel pulling, say, 3 Amps (54 Watts at 18V AC command voltage), this motor would NOT be operating near optimal efficiency.  One would have to run some experiments but I'd guess that power wasted due to mis-matched motors is less than 10%;  if someone can provide a better guess I'd like to hear it!

In other words you're looking to save a few Watts.  You could save that much power just turning off smoke.  Or you could lower the volume and save a Watt or two.  And like improving MPG in your vehicle, acceleration/braking habits can improve efficiency more than 10%.  Are you using a mainline diesel for switcher duty?  The gearing is different to allow the motors to operate closer in the efficiency sweet-spot which I'd think could double running time rather than just get you 10% (or whatever).  I have no idea if manufacturers select motors to maximize efficiency...but you could.  If you have money and time on your hands perhaps investigate swapping motors.  It may require different electronics to drive, say, brush-less vs. can DC motors.  Or if you are the mechanical type you can swap in a different gear-box to operate at maximum motor efficiency for the speed that you most operate your engines.

So even if you had separate tachs and separate motor drive electronics (which itself consumes power that must be accounted for) I'm not sure you would see improvement in battery run-time. 

Is this PS3 on battery?  If that's the case and you're willing to muck around with the electronics, I'd think you could save several Watts of power by bypassing the AC-to-DC bridge-rectifier that's on PS3 boards.  Since PS3 boards can run on DC, it seems you could save the lost power in the diode drops.  So 3 Amps dropped "unnecessarily" across two 0.7V diodes is 4 Watts.  I'd think that is more power saved than a motor balancing scheme.

All this just my opinion of course...

Dah..... I just seek the truth! Dah.... no lies!

Why do U want to fight? I am a saint, I think, in my own mind. I lead DCC development. Which wire is positive?

I just want to forward toy train development. Now that we're discussing things (MTH), let me attack you.... OK?

If you defend anything, I'll get my boys to attack you too!

That's my legacy. Who am I?

Well, I am truly impressed by the technical knowledge presented on this forum and, like I said before, I am learning more things that, for now,  am capable of comprehending, but, I am learning.  I certainly appreciate the presentation of Stan004, that graph puts things into a logical perspective.  Maybe Gunrunner was right, I am looking for problems that don't exist.  When I get my layout finished, I will  do some tests, for my own amusement, to prove that,  I am looking for problems that don't exist.  Thanks for all of the input, and keep it coming.  Any kind of knowledge is good, whether it be positive or negative.  One great inventor said, "Even failures are good source for learning". Bob.

Engineer-Joe posted:

Dah..... I just seek the truth! Dah.... no lies!

Why do U want to fight? I am a saint, I think, in my own mind. I lead DCC development. Which wire is positive?

I just want to forward toy train development. Now that we're discussing things (MTH), let me attack you.... OK?

If you defend anything, I'll get my boys to attack you too!

That's my legacy. Who am I?

And I'm done here....

Well, I am back and still haven't any proof that one motor should be more efficient than two, but I just finished a 1:32 T1 and driving 8 wheels with one motor.  I used two gear boxes from Hudsons and the motor is out of an MTH one gauge F3 and weighs 11 pounds.  There won't be any differential power to the drivers because one motor is driving all eight mechanically, so as soon as I can I will perform a pull test and see how she does.