PS2 vs PS3 Signal Strength - Update

Scott,

Some PS3 engines have a very small wiring "anomaly" referring the wire path of certain wires in regards to a toroid. The low signal strength is quite often corrected by simply slightly re-routing these wires.

Hopefully, GGG, Marty Fitzhenry, Gunrunnerjohn, or one of the other excellent MTH ASC's will respond to this thread with further details.

ScottM posted:

Do I need to tell my wife I need a Rev L?

 

 

Maybe,

My original Rev. G TIU's worked okay with my PS3 engines but I would have some minor intermittent issues once in a while.  After updating to a Rev. L, I never had a problem ever again.

In addition to the USB port, the Rev. L TIU also changed from ASIC to FPGA hardware. If my memory serves me correct, the PS3 engines also use FPGA vs the ASIC chip in a PS2.

Without getting to technical, FPGA chips are faster more responsive and also aided in reducing the size of the newer PS3 boards.

EDIT:

Credit to Barry as he did post a very detailed description about the differences between PS2 & PS3 boards:

https://ogrforum.com/...63#61683972398761363

H1000,

 If my memory serves me correct, the PS3 engines also use FPGA vs the ASIC chip in a PS2.

Your recollection is absolutely correct!

However, while I agree that a Rev. L TIU would be a worthwhile upgrade from his existing Rev. G, I don't believe that a Rev. L is likely to resolve his issue.

Based on this feedback, purchasing a Rev L is definitely planned. 

However, I look forward to learning more about the wiring anomaly and my subsequent testing. This is something totally unanticipated.

Thank you for the feedback thus far!

I did a search on the forum to understand the wiring/toroid issue. I then did a test were the wires on the unit moved away from the toroid to the greatest extent possible. This had a very minor effect, raising the signal from 3~4 to 5.

I also tested PS3 steamers and they registered a perfect 10.

A Rev L TIU is on order. Hopefully this has a positive effect. 

Both diesels are new models and purchased last year. I will test my other PS3 diesels, but is this common?

more good info

Scott,

The black coating is on the ground screw not the roller screw. 

There was also a batch of PS3 engines that had a very heavy coat of black on the wheel surface that would cause excessive sparking and poor signal strength.

Dave

If you get a lousy signal with a Rev G, don't expect a miracle from a Rev L.  I have 2 Rev Gs that give me same signal strength on my large layout and are just as reliable as the Rev L.  In fact, I have 2 original issue Rev Gs that are sound & reliable(one just went to a pair of grandkids for the layout I built them), and I'm on my 2nd Rev L, the first having been "trashed" due to a myriad of problems.

Follow Barry's point about the wire locations.  I hear it's helped many people.

I do believe there can be issues with Rev G when it comes to PS-3.  In some case a REV G can not load sound files successfully with PS-3.

There are case where a PS-3 engine with wiring interference can kill the signal for other engines.

But the first fixes should be to fold the capacitors on a diesel down horizontally and make sure the 40 pin wire harness is routed over those caps and away from the toroid near the 8 pin connector. Second is checking the screws mentioned by Dave. Third is making sure wheels and pickups are clean.  Spraying an electronic cleaner between roller pin and roller and adding a drop of oil for lubrication and to prevent corrosion between pin and roller.

See it that resolves your issues.  G

PS3 board wire rerouting.

https://ogrforum.com/...ce-needed-with-photo

I looked at the caps, they came down from the factory.

The ground wire to truck screw is silver and does not have a black oxide coating.

Wheels & roller were previously cleaned with solvent (yes-very dirty for a new engine).

These items have not improved signal strength.

Tonight's tests:

To bypass any potential issues with track, wheel sets, etc...I attached a test lead from an outer rail & connected to the silver screw for the ground wire. This had zero effect.

This same engine will deliver '10s on other area of the layout.  This particular zone (multiple 3' sections) delivers 3 & 4s.  Although one engine, this issue is common with PS3 units, diesels much more than steam. Solid 10s with PS2.

I previously reported the TIUs were Rev Gs - my error. The main TIU is Rev I, the second Rev H (they are a hassle to dismount & remount).  A Rev L is on order, hopefully the purchase wasn't a complete waste of money.

Are there other factors that can affect signal strength?

I noticed the what i call "gen 2 ps3" or those that will start up if you roll the thumb wheel in throttle mode have eliminated the black coated screws.

That is interesting.  So engine can deliver 10 on other sections of loop, but not a certain section, despite other PS-2 engines doing fine on that section.

So, what accessories and or switches in that loop.  Maybe something attached that is hurting PS-3 engines trying to communicate.

When you looked at diesel wiring did you still make sure harness wires routed away from the toroid?

Lastly, does a specific TIU channel feed that section?  If so, move the wires feeding that section to a different channel.  Does the low signal follow the TIU channel switch?  If so maybe some level of degradation on the channel that hurts PS-3 signal strength.

These are just some things to test and try and I admit I do not fully understand why PS-3 can have issues where PS-2 do not.  But there seems to be cases were it happens.  G

Thanks for the feedback G.

These problems can definitely be a challenge to solve! My hope is to ultimately have more robust layout operations and to eliminate the gremlins who only seem to come out when hosting visitors.

A new development: Prior to going to work this morning I decided to verify there was a gap in the center rail between track/feed sections. I found that I could not slip a sheet of paper into the gap. I then used a razor blade to force open the gap. Next, I retested the signal strength using the same engine as used last night and found it had a perfect 10!

Unfortunately I did not perform an initial test but last night this area delivered a 3.

I then tried to recreate a zero gap condition by connecting the center rails with a test lead but the signal did not drop below 10.

This will be reconfirmed tonight 

Scott,

When blocks touch, via center rail, due to expansion, your signal will easily go from a 10 to a 2.  Sometimes the location of the lowest signal will be directly over a feed.

Dave

David Minarik posted:

Scott,

When blocks touch, via center rail, due to expansion, your signal will easily go from a 10 to a 2.  Sometimes the location of the lowest signal will be directly over a feed.

Dave

this is good info to have.

Dave, 

Please explain what you mean by "Sometimes the location of the lowest signal will be directly over a feed"?

Also, I tried to recreate the low signal by re-closing the rail gap with a screw driver blade  but still got a perfect 10.   I was hoping to turn the problem on and off.  Any thoughts?

Thanks!

David Minarik posted:

When blocks touch, via center rail, due to expansion, your signal will easily go from a 10 to a 2.

I'm not sure that this is the case. I did some testing on this as well on my portable layout and couldn't reproduce any significant reduction of the signal level by closing the center rail gaps with PS2 or PS3 engines.

Also, when an engine or a lighted rolling stock rolls over these gaps, the rollers should also close the gap, but I have never noticed a drop in signal level on my home layout when this happens.

Every layout is electronically different.  There's no telling what the effect will be of doing most anything.

ScottM posted:

Dave, 

Please explain what you mean by "Sometimes the location of the lowest signal will be directly over a feed"?

Also, I tried to recreate the low signal by re-closing the rail gap with a screw driver blade  but still got a perfect 10.   I was hoping to turn the problem on and off.  Any thoughts?

Thanks!

 

Scott,

The feed is right where the wires are soldered on.  As RJR said, every layout is different.

I could always tell when my track had expanded by my signal reading on a certain area of my layout.  I didn't use fibre pins.  I have been going back and placing plastic in between blocks.  This seems to solve the expansion/signal problem.

Hope this helps,

Dave

Connecting two TIU channel outputs together does seriously degrade the signal, and the result is variable and unpredictable.

gunrunnerjohn posted:

Connecting two TIU channel outputs together does seriously degrade the signal, and the result is variable and unpredictable.

John,

Not sure I understand why anybody would do that.  Can you explain?

Dave

Dave, I was responding to the comment that connecting blocks together didn't have an effect.

I just replaced the primary TIU with a Rev L and now have 10s in areas that were 3~4 for PS3 diesels. Problem solved!

This is a relief given that I had run multiple tests trying to determine why one zone  was a 10 while the other was a 3 (always a perfect 10 with PS2 engines). I assume the chip set in the TIU is the key.

Is the practice of drilling a hole in the TIU cover and then routing the antenna wire thru it still recommended? I know that was once the practice to improve reception.

just mount the TIU vertically and you've done the same thing.

Exactly how they are mounted. Glad to hear the modification isn't necessary, too much of a pain to dismount.

gunrunnerjohn posted:

Connecting two TIU channel outputs together does seriously degrade the signal, and the result is variable and unpredictable.

gunrunnerjohn posted:

Dave, I was responding to the comment that connecting blocks together didn't have an effect.

 Maybe so, but closing the gap on two blocks that use the same TIU channel shouldn't have any effect on signal.

H1000 posted:

 

 Maybe so, but closing the gap on two blocks that use the same TIU channel shouldn't have any effect on signal.

You sure?  I've run across several instances helping a friend with his track signal... Tooo many short blocks can be a problem.... Here's an Example..... 30 feet of track divided into 3 ten foot blocks might yield a track signal of 6 & 7 in each block..... the same  30 feet as one block yields a perfect 10. The problem is .... You never really know  what's up until you try it and every layout is different.

The antenna extension really helps extending the remote  range . Our layout is 78 feet long with tiu at each end. The down side is...  It's very easy to break off  the antenna if having to move the tiu every so often. Once in  place  no problem.

Gregg, on my layout every turnout is the end of 3 blocks.  So I have many very short blocks, and signal strength is not an issue.  I have so many clocks because layout was built before DCS and designed for max control of multiple conventional locos.

H1000 posted:

gunrunnerjohn posted:

Connecting two TIU channel outputs together does seriously degrade the signal, and the result is variable and unpredictable.

gunrunnerjohn posted:

Dave, I was responding to the comment that connecting blocks together didn't have an effect.

 Maybe so, but closing the gap on two blocks that use the same TIU channel shouldn't have any effect on signal.

Not true.  It will also degrade the signal.

H1000 posted:

gunrunnerjohn posted:

Connecting two TIU channel outputs together does seriously degrade the signal, and the result is variable and unpredictable.

gunrunnerjohn posted:

Dave, I was responding to the comment that connecting blocks together didn't have an effect.

 Maybe so, but closing the gap on two blocks that use the same TIU channel shouldn't have any effect on signal.

Not true, the point of the recommended wiring for DCS is NOT to have any signal loops.  If you have two drops star wired from the main block and you connect the tracks together at the gap, you have created a loop for the RF signal.  This is what will cause missed commands and repeated commands.  They do get briefly connected together in normal operation with rollers on the different segments, but if they're constantly connected, it'll certainly affect the signal adversely.

The idea of closing the gap between blocks was implied by the short moments of time when engines and lighted rolling stock move through these sections. Should have stated that from the get go.

The old club layout was 42 x 20 and we would regularly run long lighted passenger trains that would close up to 3 and sometimes 4 blocks at a time and saw no loss in signal whatsoever.

When it comes to the DCS signal I don't think there's a perfect answer,,,, Go with what works for you.

H1000,

The idea of closing the gap between blocks was implied by the short moments of time when engines and lighted rolling stock move through these sections. Should have stated that from the get go.

John is correct. Connecting two adjacent blocks can and often will cause DCS signal degradation.

However, in the case of pickup rollers bringing a gap between two blocks, the degradation is transitory and quickly gone, and the the effect is typically not noticeable.

I don't disagree at all Barry. If I am going to wire a new layout, I will strongly recommend using blocks. But I won't tell somebody to completely start rewiring an existing layout just to accommodate DCS. Plug it in and see what happens first. And the case of the OP, a new TIU solved all of his problem without addressing any of the recommended track & wiring troubleshooting that was previously mentioned.

In the case of our old layout, 3 or 4 blocks were continuously connected via a 24 foot lighted passenger train. In this case the trasition was never quickly gone.

Blocks can be totally isolated from the effects of wheels closing the gap by using a short piece of 3" dead center rail track between blocks. 

I have never noticed any problems when my long trains of lighted passenger cars cross between blocks or between TIU channels. 

I would note, Dave, that 3" wouldn't be enough to break the circuit on long passenger cars, but could be enough to stop locos with less than 3" between rollers.

I do have one block that has two hot feeds.  There is an old UCS track in the block and I found I was losing power.  Not having time or inclination to unscrew track at fix it, I temporarily ran a second feed to the other side of the UCS.  That was about 10 years ago, and since I have had no signal problems, I never did do the permanent fix.  Sometimes the electronics just works out right.  Nothing, no nothing, beats being dumb & lucky.

H1000,

If I am going to wire a new layout, I will strongly recommend using blocks. But I won't tell somebody to completely start rewiring an existing layout just to accommodate DCS. Plug it in and see what happens first.

Agreed!  

That's what I did when DCS came out.  Worked almost perfectly, and a few lights on a few sidings made the whole thing perfect.  No rewiring, BUT, layout had all 14 gauge wire.