issue Elite - R8247 - Tortoise point not changing polarity

[JohnUSA]

Newbie to DCC, Using new Elite 1.4.5 controller initially hooked up to new R8247 decoder via program feed I have checked address 0001 has CV33 set standard mode as 0 and CV3 I set to 0 enabling continuous feed as Tortoise specify this is required.  Hooked up channel 1 with one negative port to location 1 on the motor and the other negative port to location 8 on the motor.  As only testing the motor not hooked up anything to the "C" common port. 

Proven that there is feed from both negative ports using a meter.  When I first made the final connections the point motor worked once. Tried to throw the point back, didn't work. Tried again nothing.

Changed the feed on the elite from Prog to the track, point not working.

Hooked up the point motor to 9v battery its working fine.

So how and where is the polarity changed which enables the point to be thrown, why did it work once?

[JohnUSA]

I should add the point motor is the only item connected to the decoder which is the only item connected to the Elite controller as setting up on desk before adding to layout. 

[Chrissaf]

It is never going to work they way you have connected it. Both the terminals you have used on the decoder are both switched negative [but not at the same time]. When one of them is switched negative the other terminal is 'floating' and not a proper positive. It is the 'C' terminal that has the positive power voltage on it, but this terminal is permanently positive and can't be switched negative. This is also the terminal that you have not connected anything to.

 

The reason it might possibly have worked once and once only, might have been that the internal CDU [used to operate Solenoid point motors, for which this R8247 decoder is designed for] happened to be charged up and provided the 'use once' power. This possible explanation still seems somewhat unlikely to me as the integral CDU is between the switched negative and common C positive terminal that you have not made a connection to.

 

The Tortoise motor is a two wire stall point motor. You apply a DC voltage across the two wires to make the motor drive in one direction [one wire positive the other negative]. Then to make it drive in the other direction you have to reverse the polarity of the applied voltage to the same two wires.

 

The R8247 Accessory Decoder is a 3 wire decoder not a 2 wire output decoder and is incapable of providing the required voltage reversal output itself.

 

However, all is not lost. DCC Concepts make a 3 wire to 2 wire adaptor that can provide the necessary conversion from 3 wire to 2 wire 'reversal' operation.

 

These adaptors are very simple circuits made up of two power transistors, two diodes and two resistors. This following previous post publishes a circuit schematic for such a circuit found on the Internet [4th post up from the bottom on the page that opens]. So if you have the electronics construction skills, you could make your own 3 wire to 2 wire adaptors very cheaply.

 

Operating a 2 wire Stall Point Motor with a 3 wire R8247 Decoder

https://www.hornby.com/uk-en/forum/cobalt-omega-and-dcc-acc-decoder/?p=1

 

TIP: As a newbie poster on the forum, just be aware that the 'Blue Button with the White Arrow' is not a 'Reply to this post' button. If you want to reply to any of the posts, scroll down and write your reply in the reply text box at the bottom of the page and click the Green 'Reply' button.

.

See also – further TIPs on how to get the best user experience from this forum.

https://www.hornby.com/uk-en/forum/tips-on-using-the-forum/

.

[JohnUSA]

Hi Chrissaf,

 

Many thanks for your detailed response.  It confused me as to why it worked once so I tried to de-bug it and not having any electrical expertise wasted a lot of time.  I tried searching the forum and using google but got more confused therefore reached out.

 

The last time I worked on a layout was in the late 50's, recently decided to get involved again to have a hobby while surving Chicago winters.  A lot of changes since the 50's but looking forward to making progress thanks to you and this forum. 

[Chrissaf]

The 3 wire to 2 wire adaptor I have documented will solve your issue and would be 'Cents' to make yourself if you have the ability to solder electronic components together. If you don't feel confident to make your own, then you can buy the ready made 'DCC Concepts' ones for a few 'Dollars'.

[JohnUSA]

Hi Chris, took some time for me to get to test the DccConcept 3 wire to 2 wire adaptors.  It's still not working correctly.  It's permanently wanting to move the switch in only one direction.

 

I have established the best result I can get is when programming the CV settings as follows.

 

CV33 set to mode 2, i.e. 4 independent channels to drive 4 independent tortoise motors.  Then I have set for each channel CV3 -CV6 to zero for continuous operation as required by the tortoise motors.

 

Then setting it up using the main track outlet it results in the motor being driven to one side only.  i.e. just seems polarity is not being changed.  I took out the power to the Elite, manually changed the motor by gently sliding the arm across.  As soon as I hook up the elite again as the power is continuously on the motor works by the arm moving back.

 

So to me it looks like the adaptor isn't changing polarity when elite sends the message to change the motor arm direction.

 

Any ideas?  Have I correctly set the CV values?

 

Thanks

 

John  

[Chrissaf]

As you have mentioned CV33, this means that you have a Version 2 R8247 which supersedes the more common original R8247.

 

CV33 = 2 is not the correct mode to use.

Using Mode 2 results in both the + and - terminals on the port being switched together, this will totally confuse the 3 wire to 2 wire adaptor.

 

In order to use the DCC Concepts 3 wire to 2 wire adaptors, you need to set the following CVs

 

CV33 = 0 [Operation Mode Zero]

CV3 = 0 [Port 1 Always On]

CV4 = 0 [Port 2 Always On]

CV5 = 0 [Port 3 Always On]

CV6 = 0 [Port 4 Always On]

 

Then for the port addresses you set CV1 as follows:

 

CV1 = 1 for DCC port addresses 0001 to 0004

CV1 = 2 for DCC port addresses 0005 to 0008

CV1 = 3 for DCC port addresses 0009 to 0012

CV1 = 4 for DCC port addresses 0013 to 0016

etc

etc

CV1 = 63 for DCC port addresses 0249 to 0252

 

Plus for the above DCC Address table to be valid set CV9 = 0

 

I suggest that before checking / setting the above CVs that you write 8 to CV8 first to perform a factory reset to ensure that you starting from a clean slate.

 

 

[JohnUSA]

Hi Chris,

previous post should have read CV3=0

As noted in my initial post I had CV33 = 0 and the first port CV3=0, the setting I missed completely was the need for CV1 settings as you described above.

 

Thanks again.

[96RAF]

@John

Good to hear you are getting to grips with V2.0 with Chris’ help.

 

The updated R8247 has become a fairly complex beast with some confusing options, meaning that you have to get your head round what the device can do in each mode - standard or specials, then you have to understand and apply the correct values to the associated CVs to get it to achieve the required actions.

 

 

 

[Chrissaf]

As noted in my initial post I had CV33 = 0 and the first port CV3=0, the setting I missed completely was the need for CV1 settings as you described above.

 

Yes, but that was before you obtained the 3 to 2 wire adaptors. You then went on to state in your follow up post that you set CV33 = 2 when you added the adpators.

[JohnUSA]

Hi Chris and RAF96

It's working!  Thanks Chris.  RAF96 your not kidding re complex, but beginning to understand it.

Can't believe how basic the process is to program the elite.  Took me back nearly 50 years when I was involved in debugging financial software.  Be easier when I hook up the laptop.

Question to RAF96.  Did you ever get to RAF Halton Aylesbury Bucks.  I played rugby for Aylesbury back in the late 70's.   Loved the air shows the RAF held there.

[96RAF]

@John

 

RAF Halton - oh yes. My home for 3-years. I did my apprenticeship there as an airframe fitter. I was there 1960-1963 and it set me up for life.  I run our Entry old boys website - see here http://www.halton96th.co.uk/index.html

 

There are a few Ex-Brats on the forums.

 

I was more a hockey player than rugby and well remember being soundly thrashed (literally) by a girls school we played.

 

R8247 v2.0 - I was a beta tester for the update so I know a reasonable amount about it and how it works. Feel free to ask more questions.

 

[Chrissaf]

😎 Glad all now OK.

[JohnUSA]

RAF96 / Chris

 

I'm on the point of going crazy, it shouldn't be this complicated.  I got the first PAD driving 4 Tortoise motors address 1 thru 4.  Then posted here everything was OK.  Thinking the next PAD would be straight forward.

 

Well I can't get the next PAD working correctly, set it up and checked it reflects addresses 5 thru 8 and checked all cv settings I got control of ACC Adr:0005.  You can hear the motor briefly fire up when clicking control 1 then also hear it again when changing direction by clicking control 2 but the arm doesn't move.  The switch motor is working fine when I hook it directly to a 9v battery and switch the wires.

 

Now on setting up PAD 2 I didn't run an Elite reset to factory settings for obvious reasons, however I tested the 1st pad again and all four switch motors working fine and I don't want to have to set that all back up again.

 

I have gone thru the set up on PAD 2 and read back all the address and CV settings for it, I think 3 times to triple check all ok before I reach out to you.  The 2nd PAD address 0002 being 5 thru 8, CV1 is set to 2, CV33 is set to 0, CV3 thru CV6 set to 0 for 5 thru 8.  Also my understanding is as long as CV3 thru CV6 are set to 0 then CV38 thru CV45 will have no effect.  Chris you mentioned CV9 = 0 but I believe that's not relevant until I get to be above the address top of the range being 252 however its set to 0. 

 

One thing that does concern me though is unplugging the elite and on plugging back in you have to be very careful not to create a contact creating a few sparks.  I check PAD 1 working fine after this occurred but could that cause a system issue?

 

OT:  RAF96 I saw an article that Halton camp to close 2025, that's not good?

[Fishmanoz]

Sounds like it's set to pulsed output, not continuous?

[Chrissaf]

Do you have a third R8247 to try, there's always an unlikely but possible chance that R8247 number 2 is faulty and not accepting the 'programming' commands.

 

Does R8247 number 2 provide appropriate measurable 'always on' voltage outputs on the + & - terminals when the outputs are not connected to anything i.e no 2 wire adaptors fitted.

 

When taking voltage measurements, remember the C terminal is positive and the + & - terminals should be switched negative alternatively when the port is operated from the controller. A steady 'always on' voltage between 13 volts and 15 volts DC should be measured. But I strongly suspect that you are getting a 100ms [1/10 of a second] pulse [see text further below] and not an 'always on' output.

 

The be 'belt n braces' it would be prudent to temporarily wire a dummy resistor load across the terminals. The value of the resistors isn't critical but I suggest a value between 330 and 680 Ohms. Then measure the output voltage across the resistors, moving the meter from port to port as appropriate [hence the dotted lines in the drawing].

 

/media/tinymce_upload/c61d5e3138b41a4e85bd8d064c4a7365.jpg

 

Remember that the voltage will either be across the + & C terminals OR the - & C terminals, and then swap over when the appropriate DCC address is operated.

 

If you are not reading a 'steady always on voltage' across the resistors [ R ] with the Multimeter [ V ] that changes with the appropriate DCC address commands being sent from the controller, then I would put the decoder back onto the Elite 'PROG' output and perform the write 8 to CV8 factory reset anyway, then re-instate the CV configurations. Then test again on the Elite 'TRACK' output. If you now get the 'always on' voltages across the Resistors as expected, remove the resistors and replace with the 2 wire adaptors and test again.

 

When connected to the Elite 'PROG' output terminals, the R8247 being configured needs to be the ONLY device connected to the Elite 'PROG' output.

 

To test, the R8247 being tested needs to be removed from the 'PROG' output and reconnected back onto the 'TRACK' output of the Elite.

 

It is a lot of faffing about keep re-terminating the R8247, but unfortunately this needs to be done for all the documented tasks to be correct and valid.

 

You can hear the motor briefly fire up when clicking control 1 then also hear it again when changing direction by clicking control 2 but the arm doesn't move.

 

This is telling me that your writing zero to CV3, 4, 5 & 6 to set the output to 'Always On' has not been accepted in the R8247 configuration. The R8247 output is still sending out a 'pulse'. The default is likely to be a pulse of 100ms duration [1/10 of a second], hence the initial motor click but no noticeable movement.

 

This [pulsed output] is also Rob's [RAF96] diagnostic deduction too, as indicated in his reply above mine.

 

An 8 to CV8 reset would definitely be prudent to perform and follow that up with checking and setting all the CVs again.

 

Now on setting up PAD 2 I didn't run an Elite reset to factory settings for obvious reasons

 

I am really not sure what you mean by this statement as the reason for not resetting the decoder is not obvious to me. Writing 8 to CV8 resets the R8247 Decoder configuration to factory settings when connected to the Elite 'PROG' output, and has no impact on the Elite itself or any other R8247 that is not connected to the Elite 'PROG' output.

 

The second R8247 is totally isolated on the Elite 'PROG' output [or should be ... see 'Important' note below]. Therefore, whilst the second R8247 is being configured on the Elite 'PROG' output. The first working R8247 that is still on the Elite 'TRACK' output is not affected in any way.

 

IMPORTANT: Do not remove everything from your track except the R8247 being programmed and then just connect your track to the Elite 'PROG' output. The 'PROG' output is very low power and trying to programme a decoder via the physical track is not electrically efficient. The track can act as an aerial and introduce noise into the low power programming data. For optimum and best programming success results. Provide a temporary pair of direct wires between the Elite 'PROG' output and the R8247 DCC input of the decoder being programmed.

 

One thing that does concern me though is unplugging the elite and on plugging back in you have to be very careful not to create a contact creating a few sparks.

 

I suggest [recommend] powering down the Elite whilst you undertake physical wiring changes ... better safe than sorry. An inadvertent connection between the 'TRACK' and 'PROG' outputs has been known to fry an Elite programming output.

[96RAF]

It was Fishy talking about pulse Chris, but please allow me to chip in now.

 

Elite has 2 ways of programming an address to an accessory decoder.

1. Using the Menu - Direct - Acc - Address ...method you can enter the port address (e.g. 5, 6, 7, 8 or subsequent for PADs 3, etc) and the Elite will calculate the Group address and write that to CV1 for you.

2. Using Menu - Direct - Acc - CV ... method you enter the Group number into CV1.

 

Pulse length/continuous - write value 0 to each of CVs 3 to 6. As you say with this set  any value in CVs 38 to 45 will have no effect.

 

You can also write to these decoders in Operation Mode - i.e. PoM, bit there is no read-back possible, but handy for changing certain values on the fly once you have got addresses firmly established.

 

A couple of simple checks to ensure the Elite is programming correctly. Try reading back CV7 - it should give you 135 for version 2.0.

If you read-back CV1 it gives you the Group address not the port range. What we are looking for with PAD 2 is  value 2, Pad 3 value 3, etc.

 

A further check as you go method that I use is to connect an R406 signal lamp across any pair of ports. An led with a 1K ohm series resistor across each port will do just as well. A pulse will flash the led once, whereas continuous will obviously keep it lit. I found this lamp method was handy when I got to setting and checking the various pulse length and/or number of flashes on single or pair of ports during testing.

 

To avoid confusion I have relabelled my Hornby ports from ( + C - ) to ( < + > ) instead, where < and > are switched negatives.

Currently I have 4 x R8247 v2.0 on the layout and 2 x R8247 on the test rig. All of these programmed simply using the Elite Acc or CV methods and read-back to confirm all was OK before wiring into the layout.

 

[Chrissaf]

It was Fishy talking about pulse Chris

 

Ooops so it was. Funny how the brain works. Once I'd made the incorrect observation, Every time I looked at the thread I never spotted the mistake.

[JohnUSA]

Hi RAF96

 

Just checked PAD 2 CV 0007 its showing 134 not 135?.  I'm going to reset PAD2 back to factory settings and re-program it.  Bit busy today with election, will focus on PAD 2 tomorrow.

 

Did you see my note re Halton due to close 2025.  Major location to be closing.  

[96RAF]

Yes it has been on the cards to close Halton for years now. More RAF Heritage swapped for house building. It is hoped to buy St George’s church, which will secure the stained glass windows and then convert it to house the heritage museum, etc.

 

Appendix 1 Page 8 of the manual confirms 135 is the value of CV7 for v2.0. Try a PAD reset by writing value 8 to CV8 then check it operates on addresses 1 to 4 in pulse. I will check with Hornby about 134.

[Chrissaf]

Just a further post to add clarity.

 

In my reply timed at 18:50 on Page 1 of this thread I included a list for setting the value of CV1 to configure the DCC addresses using CV1 Group Addresses ... extract replicated below:

 

/media/tinymce_upload/2759388da9bfa9c02f606ab9f214729e.jpg

 

The CV1 Group Address method allows the R8247 to be used by any branded DCC controller as it is compliant with the NMRA Standard.

 

However, Hornby in their wisdom have included a special R8247 Accessory Decoder addressing mode in the Elite and Select controllers. Hornby included this special DCC Addressing mode, I assume because they thought it would simplify R8247 address configuration. It probably would have done if they had also explicitly documented it a little bit more adequately in the manuals.

 

Let me try and explain.

 

There are two ways in which you can configure DCC Addresses in an R8247 using the Elite [using 'Direct' Service Mode programming].

 

Method 1 - The 'special addressing mode' way as described on Page 58 of the Elite manual ... Page 58 extract below:

 

/media/tinymce_upload/81eebe862ff2b26eceef2ac4a13aa318.jpg

 

In this mode, one ignores the Group address numbering scheme I highlighted in my Page 1 post and enter the actual address directly for each Port 1 of each R8247 being configured with an address. Note that no mention of CV1 is made or required in this 'special mode' programming method.

 

So:

 

R8247 number 1 port 1 = DCC Address 1 gives/sets port addresses 1 to 4

R8247 number 2 port 1 = DCC Address 5 gives/sets port addresses 5 to 8

R8247 number 3 port 1 = DCC Address 9 gives/sets port addresses 9 to 12

etc

 

It is the number in bold that is actually used to input the address when using this method. The Elite 'special mode' takes the actual DCC address you enter and automatically converts it [using a routine in the Elite firmware] into a 'Group Number' then configures that calculated 'group Number' into CV1. Thus it is the 'Group Number' that is 'read back' in a CV1 Read action.

 

Method 2 - In this method, you ignore the R8247 Addressing method documented on page 58 of the Elite manual. Instead, you use the direct CV writing method documented on Page 60 of the Elite manual ... Page 60 extract below:

 

/media/tinymce_upload/dbb44e2d1c150adfe079771505ed3b23.jpg

 

Using the Page 60 CV writing method, means that you use the Group Address and write that value directly into CV1. In other words, using the DCC Group Number addresses listed in my reply on Page 1 of this thread.

 

I provide this additional reply to give additional clarity, because if you enter the Group Address value 2 [for addresses 5 to 8] using the Elite method documented on page 58 of the manual. You will not be entering address 5 to 8, but invalid addresses 2 to 5 instead [that is if they even take and get accepted].

 

I'm just wondering if the last paragraph above was what was actually performed on the second R8247. And if so, could that account for the errors in the second R8247 configurations that you are experiencing.

 

[96RAF]

All indications (CV7 = 134) are that that particular PAD is a v1.0 device and will not respond to anything you put into CV33, in fact it may even be confusing it. 

 

I would recommend a reset (CV8 =8) and try the simple Elite Menu Acc addressing (address 5, 6, 7 or 8) and see if that sticks.

 

Also read-back CV7 for your other PADs as confirmation of actual version.

[JohnUSA]

Hi RAF69, Some what surprised that these PADs could be v1.0 considering recently purchased direct from Hornby.

 

I will reset everything including the Elite and start all over.

 

Then I will validate version on each PAD and read up on v1.0 to see what other implications there maybe.

 

However having already reset on previous tests PAD2 had CV33 default as 0, on reset I checked and it's 0, I didn't attempt any changes as not required.   I also have been using the simple Elite Menu Acc addressing (address 5, 6, 7 or 8) for PAD2 

[Chrissaf]

I also have been using the simple Elite Menu Acc addressing (address 5, 6, 7 or 8) for PAD2

 

Remember that if using the 'Elite Menu Acc Address', you need to enter Address 5 and not Group Address 2 for the second R8247. See my earlier reply above for details.

[96RAF]

I await your PAD version listing.

The PADs can be updated and apart from the distance involved it could be worth contacting Hornby and telling them you bought new but ... some were, some weren’t.

 

Version 1.0 is essentially the same as Version 2 in CV33=0 mode, except you don’t get the points group firing facility E.g. set Point 1 to fire Points 2, 3 and/or 4 and ditto from other points. The only options you have are pulse length or always on.

PAD2 box looks like this with a sticker on it. Some in-stock units were done retro by Hornby but the rest were ex-production run.

/media/tinymce_upload/7c595ef8caba276e2cd6fbd99dd68430.png