MTH G scale speed matching: too fast?

I just had another PS2 G scale -8 change it's speeds. Again, it's too fast. It's like something on the board changes (weakens), and then it needs a tighter flywheel gap? This engine ran fine up until now. I have searched for this solution before and now I will just make the gap smaller. I had ordered spare tach readers that didn't help any other time I tried.

The results are it rips apart large consists and breaks the electric couplers on me.

 This time I had warning. I tried to build a consist with matching the engine's speeds first while not connected. I saw the lead engine was faster. I just under-estimated how much faster it was. When I put them together, they didn't make it around and behind my garage without disaster. 50'?

 I'm starting to wonder if I should just tighten the gap on all of these? What changes to make the speed nearly double on five of these engines so far? Yet I have even older ones that are fine. I have first release G scale -8s that are stock. Seems like it's the time from maybe 05'-09'??

https://mthtrains.com/70-2073-1

I'm wondering if running these at higher track voltages is causing me grief? I try to keep it at 21 volts using the engine's on board track voltage reading.

I'm not clear how G diesels are put together but can you power it up and manually turn the flywheel?  IMO if you want to get to the bottom of it, you need to measure the actual tach signal voltages.  The tach sensor is missing reflected stripes thereby instructing the motor control to spin faster.  It can be a major nuisance to instrument the tach circuit to a voltmeter but it is sort of described in this post.

The idea would be to tabulate the white and black stripe voltages one at a time as you manually turn the flywheel (again, this assumes you have access with the circuit powered).  Of course this would be made easier with an oscilloscope but let's ignore that for now.  Ideally, you want the maximum difference between the white voltage and the black voltage.  If this is something you see yourself trying, I can elaborate.

Otherwise, I don't think 21V AC track voltage is a factor here.  The tach circuit runs off of regulated DC - 5V or less in this case.  OTOH I can speculate that the striping paint might have changed its reflectance (both the black and white) due to fading, accumulated grime including on the sensor window(s), or maybe even a change in the paint formulation itself.

stan2004 posted:

 OTOH I can speculate that the striping paint might have changed its reflectance (both the black and white) due to fading, accumulated grime including on the sensor window(s), or maybe even a change in the paint formulation itself.

 

 

I can only post that I also looked at that several times. Once, the flywheel did have gearbox grease on it. Yet after being cleaned, and in the rest of the cases it didn't seem to matter? I tried blackening the stripes with a sharpie. I cleaned with alcohol. I seem to remember even trying a printed tape. Because of the G scale truck design, I would think that grease could be a factor. I find it interesting that no other G scale model I have, has had this issue. Just the Dash 8s so far.

I run them more than any other though.

I had considered that heat or other factors warped the plastic casing changing the distance. This engine, and 2 others that I bought brand new don't show any signs of change.

 As soon as the tach sensor is brought closer, the engine runs correctly. This case in particular, I only brought it in a hair. I was surprised that it then worked.

The brand new PS3, I had to bring in around half the distance! I didn't measure the change but it took 2 or 3 grinding attempts to get the tach close enough to correct the speed. I thought I had it at first and yet it seemed to revert back to running fast. I ended up fully grinding thru the truck casing if I remember correctly.

 I was shocked that MTH did not make a tach with a shim for this PS3 model. I can't believe I got the only one that does this? I did not buy any more yet. 

I match my MTH G scale diesels to all their other G scale products. I expect them to run together without issue.

BTW, I don't have a scope. I don't know how to test for that.

have you ever tried switching the ps3 boards and see if the problem control issue motor run too fast  goes away ?part of the tach parts are on the ps3 stacker boards. quickest way would be to switch board and see if problem goes away!

I only had the 1 PS3 G scale engine at that time. I did have PS3 boards in other conversions that I had done. I don't risk good boards in troubled engines.

I chased this problem with PS2 engines. I tried several steps and always ended up moving the tach closer to solve the issue. If you searched for a solution and got one that worked, you would go with that solution from then on.

I did not ever figure out "the why" it was happening. I found that moving the tach closer solved it, and those exact engines ran perfectly ever since.

I am amazed that other MTH G scale -8 owners don't see this. I can't figure out how I got five different -8 MTH engines with speed issues and no one else did. I have the very first releases of the -8s and they run fine. 

If it were one or two, I'd guess it was an anomaly. I'm at 5 now and I can only guess that there'll be more?

and why haven't the other diesels like the F series, VO1000s or the PAs I own given me this issue yet?

Uh Oh! did I state that out loud?

Stan, the other thing is that they run an exact speed. I have paired them together and they will keep the same gap between them at this faster rate.

They are connected in this video I believe but they don't struggle. This was so long ago that it caught me by surprise when it just happened again.

The black 2-window sensor chip on the tach board has a published spec of optimal distance from the target (the tach stripes).  Too close and it doesn't work, too far and it doesn't work.  I have never seen a PS 2 or 3 tach board with a mechanical adjustment (screwdriver or otherwise).  So one can only assume MTH designed the nominal mounting distance to be at the "ideal" distance to the flywheel stripes.  Likewise, I have never seen any published troubleshooting techniques that are prescriptive with respect to aligning or maintaining the tach sensor.

To dig deeper I think you'd need to take meticulous tach sensor voltage measurements to confirm the alignment is as good as it can get.  I'd say this kind of DIY analysis would go 10 times (or more) faster with an oscilloscope than with "just" a voltmeter.  That said, if you don't have a scope I'm thinking it is what it is...

The 2 engines in the video above both came brand new and were too fast. The rest were all used engines that behaved most of their life and then all of a sudden, each one went faster. It happened to me and broke stuff so that I got used to speed matching every season before running large consists. This last one popped up mid to late season and even though I looked for it, it still bit me. I underestimated how far the speed was off.

 I contacted MTH directly after receiving the second brand new unit and was told I should talk with a tech about how to fix my issue. That was after them telling me I must be wrong about it. They said I probably was comparing it to some engine I built or upgraded myself. I explained I had 100 total ( O and G scale) of their engines and knew the speed was not correct. I think I have more than a dozen -8s alone?

 When it occurs, the speed is something like 30 (34?) % greater. I forget now. I have to assume that something gets weaker?

BTW The PA version in G has vertical can motors.

If you are missing stripes, the locomotive will run faster.  It's wanting to read a specific number of stripes for a set speed.  If you miss half of them, it'll run the motor twice as fast to achieve the count it's expecting.

The Omron EE-SY124 sensor used for PS/2 has an optimum sensing distance of .8mm, it starts to fall off rather quickly soon after the peak.  Note that you don't want it any closer than .8mm as the fall off in sensitivity is even faster as you get closer.

Engineer-Joe posted:

...

 When it occurs, the speed is something like 30 (34?) % greater. I forget now. I have to assume that something gets weaker?

...

Engineer-Joe posted:

...

Stan, the other thing is that they run an exact speed. I have paired them together and they will keep the same gap between them at this faster rate.

They are connected in this video I believe but they don't struggle. This was so long ago that it caught me by surprise when it just happened again.

 

Whoa!  So are you saying that the engine run at either

(1) the correct speed, or 

(2) exactly 30% (or whatever) faster than the correct speed? 

In other words when a -8 runs fast, it runs fast at the SAME faster speed?  Now that's what I call a head-scratcher!