PS2 engine "jerky" at low speed

Jon,

Specifically, I refer to:

 

1.  Keep your finger away from the Protocast button.
2.  Purchase only new locomotives under warranty.
3.  Limit your maintenance activities to lubrication of gears and rollers.
4.  If the need for any further maintenance or a repair is thought to be required take the locomotive or electronic device to an MTH trained and authorized technician.

Those were not the most important of Ted's messages, as I read them. You need to read between the lines and understand what Ted was getting at.

You didn't pay any attention to what several of us were saying. When I posted a not-so-subtle "I told you so" after MTH told you exactly what I had posted 3 weeks earlier, you didn't have the good graces to even admit that you should have paid better attention.

Nearly every time someone posted an idea, you pretty much either dismissed it or paid it lip service. You were much more interested in telling us all what you thought. Even when we attempted to state how your suppositions weren't reasonable, you would a argue your point, and your arguments were based on ignorance of how DCS works.

You just had to be blaming everyone and everything for what turned out to be your own mistakes. When MTH reiterated what was the most likely cause of your problems, you criticized MTH for allowing a very minor, rather well-known bug to persist rather than take any responsibility for your own actions.

You also needed  to get the last word on any discussion. You have no idea how frustrating your posts were to those of us who were attempting to assist you.

I'm going to attempt to not respond further to this thread unless I am forced to do so.

You might also be best served by limiting your response to this post to simply recognizing what you've done and to changing your future behavior in this regard. However, if you simply can't stay silent, feel free to go for the last word.

Hello John,

These experiments might shed some light if you're still pondering what to do next.

1. Run the two-motor configuration in conventional with speed control disabled (Whistle-Bell-Bell).  This puts rectified track voltage across the motors and essentially takes the tach out of the loop.  I suppose you could also run a long pair of wires from a DC variable transformer directly to the motors (after pulling the 5-pin connector) to achieve the same effect. Adjust the voltage to the speeds of interest. You don't have DCS speed control but I figure you can differentiate between speed changes from load variations (curves, grades) vs. jerking.  

2. Disengage the front motor (non-tach) but re-attach the truck so the chassis can run.  The front motor will spin freely so you'll need to find a place to temporarily stash it unless you cut the wires and remove the motor. Then run the engine with JUST the tach'd motor using DCS speed control.  If not pulling a bunch of cars a single motor can drive the chassis at reasonable speeds. If it still jerks, then maybe there is something amiss in the speed control/tach loop.

As vinndiesel said the speed control algorithm has been known to go south requiring re-loading of the soundset (which contains some speed control parameters).

If you are going to make your own tach tape sticker anyway I wonder if it's practical to swap motors? So after applying the sticker to the front motor's flywheel, I figure this could be done by either moving the tach assembly to the front motor, or by swapping physical positions of the motors. I'd be interested in hearing if anyone with too much time on their hands has tried this!

Stan,

I think #1 sounds like a pretty good idea, I did run it on the bench in conventional mode on test blocks, the front motor didn't run until I had quite a bit more throttle than it took to run the rear motor.  However, from what I understand of ProtoSound2, I didn't remove the speed control from the mix using that technique.  I did remove a lead from each motor and run them conventionally from a bench power supply, it took twice the voltage to get the front motor to run.  The motor also ran very uneven until I got it spinning at a higher rate, mimicking the operation on the test blocks with speed control. 

As far as #2, when it's on the test blocks, the rear motor runs at a constant speed based on throttle position, whether it's DCS or conventional, so I've 1/2 tried that.  That is another good thing to try, just the locomotive running shouldn't have any problem with a single power truck.

Unplugging the motor plug would have been easier, didn't think of that.

I also reloaded the sound set a few times, including the one for an SD-90, but that didn't change the operation.  I have a tach tape coming, but I'm guessing the difference in the motor operation is more telling. 

I'm going to take a motor from one of my other diesels and stick it in, that's probably the direct route since I really do suspect the motor at this point.  If that doesn't change it, I'll try your suggestion #1.  I just wish my MTH spare parts box was as deep as my Lionel one, I have all the spare parts for TMCC stuff, including several spare motors.

Given the length of the harness in this one, moving the tach to the front motor isn't practical, though I guess you could transplant everything and put it in the rear.

Jon,

Given the length of the harness in this one, moving the tach to the front motor isn't practical, though I guess you could transplant everything and put it in the rear.

Can you unsolder the motor wires at the motor, remove the tach reader harness, swap motors, resolder, and replace the tach readers harness?

Well, that front-motor/truck assembly is looking guiltier by the minute.  If it runs jerky on a test-block with straight DC voltage applied, it would be a challenge for any speed control system to control it as the feedback comes from a different motor and is indirectly coupled through the truck-track-truck mechanical path.

Barry, that's something that I'll consider if the other motor doesn't work.  However, I'd have to use a new tach strip to do that, and when I get the one that's coming, I'd like to make sure it's going on a good motor.   That's what I meant when I said my MTH parts box is not as deep as my Lionel parts.

That's what I figured.  Also, there doesn't seem to be any friction when turning it by hand, if anything it actually feels slightly better than the rear motor.  I'll know when I try the replacement motor.  Certainly, all indications are it's the motor.

John

DCS, Proto2 and Proto3 are very sophisticated systems.  The diagnostic procedures that may have worked in the past, even with TMCC just a few years ago, may no longer be valid.  Like with modern medicine where exploratory surgery is virtually a thing of the past locomotive repair is in a new era and for the same reasons.  There are better and less invasive diagnostic tools available and the new methods avoid the risk of serious complications that comes with cutting open a patient.

Barry, Dave and Marty are three of the guys to whom MTH sends new products for beta testing.   They identify bugs so MTH can fix them before the products get sent to the public.  Barry in particular has done us all great favors on this forum and with the DCS users meeting at York.  Part of that work is to document problems so MTH can fix them.  If you have a problem they are unaware of you have a rare problem indeed and they will be happy to get to the bottom of it but only if you help them to help you.

Some of your problems seem to be self inflicted.  If you would take a moment to realize that your first guess in diagnosing a problem may be incorrect with equipment that is new to you and then follow the requests of our unpaid pros for more information it would likely save you time and money.

Following my item five would have been particularly helpful to all concerned in a couple of cases.  If you have any more problems post a video before taking things apart, that should cut down diagnostic time by a large margin.

On June 14th at post 7, the motor was identified as not working correctly, though it still runs.  So why are we still analyzing every other component of the system?

This is like having a tire pressure monitor say the tire is low, getting out and looking at a clearly flat tire with a nail in the side, and then going back to the dealer and stating you wonder if the tire pressure monitor algorithm is written correctly.

The horse is dead.  Replace the motor, or both if this has been used heavily, and save them for single motor units.

While sometimes entertaining, this is not a blog it is a forum, and the problem is this overly lengthy, repetitive and sometime contradictor analysis doesn't help the person a few weeks or months later who has the same problem.

Even if there is something else wrong with the tach reader, or speed control chip or sound file, you still have to replace the motor.  It is not working correctly.  :-)  G

I'm actually waiting for a new motor now.   The unit has very little operating time, the motors shouldn't die that fast, though it's somewhat apparent that this one did.  When the motor is installed, we'll see what other mysteries await.

I have never had to "break in" a can motor, is that really a factor?  I'm willin' to learn, anything's possible.

The go fast RC crowd swears by breaking in can motors..by running them under water for an hour or so... 

I don't think I'm going to run my motors under water, so that's out!

Did you happen to see the motor part number (such as RS-365SH)? Did you find an exploded parts diagram showing the MTH part #?  And how much do these things cost?

Stan, your point is well-taken and absolutely correct.

So how is this correct and how does it play out?  Mechanically coupling may damp the effect, but it is going to effect one motor or the other if they are not in sync to a set tolerance.  One motor is absorbing more load or current and will run hotter ultimately leading to further divergence and finally a failure.

If one motor starts significantly later than the other whether on blocks or free standing, you need to eliminate the mechanical effects, and if none pull the motor plug from the board and then apply direct current with an amp meter in series to each motor and evaluate.

Stan, You have changed the condition of his problem for your example.  It was not the rear motor with tach reader, it was the front motor without tach reader that was the problem.  Additionally,  in the very next post he discussed that not only did it start running significantly later, but that it ran jerky until higher voltage was applied.  In fact mechanical coupling would make it behave less jerky.  So explain to me why you need to keep experimenting again?

This is not a balance system with feedback.  The mechanical coupling (interesting term since it assumes perfect traction at each wheel) probably wasn't a design philosophy for system speed control.

If your proportioner valve on your brakes are out of wack, and the rear brakes carry 80% of the load,  the front and rear of the car will stop together, probably with a weird learch, but it ain't suppose to be that way.

Just as a clarifier,  the motor manufacturer part number PH 365 and 385 are very common numbers, but they are not the MTH part number.  That is the motor part number before Fly wheel, tach bracket and worm gear.

Even though the gear ratio is the same for all diesels, the length and position of the worm gear are different, which leads to different MTH motor part numbers.  Fly wheel size and tach reader bracket are other factors which can lead to different part numbers.  Motors MSRP normally around $25.  G

STan is correct when he says that if you have the loco blocked up, intermittent binding in the tach motor will result in greater speed fluctuations in the non-tach motor, as the PS@ unit increases and decreases the voltage to make up for the effect of the binding.  If one does not beleives this, try it---squeeze the flywheel of the tach motor and the other will speed up, proposrtionately more than the fluctuations of the tach motor.

GGG, if both motors are running at the same RPM, as should happen if traction tires are not slipping, then the amperage draw---and electrically-caused heating---of each should be the same, barring an internal short, broken winding, or bad brush.

RJR,  You guys are changing the conditions of the reported problem.  I don't disagree with what Stan and you are saying in general, but it is not how John described his problem.  Your assuming the tach truck was jammed.  Even if so, the speed control circuit would see that the rear motor (with tach) was not turning at the desired speed and longer pulse widths would be applied and the forward motor which was not jammed would not only start sooner, but also run faster.  So I agree with you on a bad or jammed Tach motor.  BUT,

That is not what John said.  He said the forward non tach motor started much later and ran jerky until higher voltage applied.

So if Stan is saying rolling block are bad, don't use them, or becareful rolling block can jam, Ok, but that is not how I interpret his response.

I stand by my statement that by the 7th or 8th post clear facts existed that said the forward (non tach) motor is bad.  Late start and jerky operations. 

As far as mechanical coupling, I understand the principle.  My point was as you state, the motors should operate the same. (Most do) The fact that the problem is less obvious when on the track doesn't mean it shouldn't be looked at.  If I had a dual motor unit where one motor ran significantly different then the other, I would fix it.  A dual motor unit that has out of tolerance motors, will result in one motor either pulling or pushing the other and it will draw more current, run hotter leading to failure sooner, and it is the good motor that get's damaged early.

This problem gets compounded with Make Ups and with AA units with slave boards, since the slave doesn't have a tach reader.  Now your banking on 4 motors matching.  Probably part of the reason you see a lot of Slave boards with shorted motor FETs.  G 

I found the same motor in a different locomotive, and they're $25.