Jump to content
 

Testing Bachmann 36-567 with ESU 53900 DCC Decoder Tester


Recommended Posts

Hi,

 

I've just bought two Bachmann 36-567 Next 18 DCC decoders for evaluation as it has Zimo features and has four function outputs and four logic level outputs.

 

I bought an ESU 53900 DCC decoder tester which includes a Next 18 Socket but the online manual says the 21MTC FO3,FO4,FO5,FO6 are at logic level.

 

The 36-567 also has FO5 and FO6 on the Speaker outputs of the Next18 socket - which shouldn't cause damage as the speaker on board the tester can be switched out of circuit.

 

post-29876-0-73589600-1508946717_thumb.png

 

 

My question is - will I cause damage to decoder or tester if I plug the two together (the 36-567 has Logic level outputs on FO3 to FO6 and the tester paperwork appears to imply Next18 socket FO3 to F04 LEDs wired by resistors to decoder positive)?.

 

 

Regards

 

Nick

Link to post
Share on other sites

There are two Next-18 socket specifications, the sound Next-18S and the non-sound Next-18. It would appear that the ESU decoder tester has the Next-18S version of the socket.

 

This all demonstrates what an ill thought out socket format the Next-18 is! You cannot plug a non sound decoder in to a socket with a speaker fitted without risk of damage to the decoder from the speaker connecting two outputs together which may or may not be in the same state. Zimo may have pre-empted the possibility and fitted protection to the AUX5 and AUX6 outputs, but who knows.

 

If the decoder tester has a high enough value resistor in series with the LEDs connected to the logic outputs such that they restrict the current to below 5mA damage should not occur to the decoder, but they will not show the output state. Probably best to measure the LED current before plugging in the decoder. The AUX3 and AUX4 outputs are re-use of the SUSI bus outputs so these pins will always be logic level. They are the equivalent of pins 3 and 4 on both PluX and MTC-21 which should have the indicator LEDs connected with their cathodes to 0V so there is a bit of odd wiring on the decoder tester.

Link to post
Share on other sites

There are two Next-18 socket specifications, the sound Next-18S and the non-sound Next-18. It would appear that the ESU decoder tester has the Next-18S version of the socket.

 

This all demonstrates what an ill thought out socket format the Next-18 is! You cannot plug a non sound decoder in to a socket with a speaker fitted without risk of damage to the decoder from the speaker connecting two outputs together which may or may not be in the same state. Zimo may have pre-empted the possibility and fitted protection to the AUX5 and AUX6 outputs, but who knows.

 

If the decoder tester has a high enough value resistor in series with the LEDs connected to the logic outputs such that they restrict the current to below 5mA damage should not occur to the decoder, but they will not show the output state. Probably best to measure the LED current before plugging in the decoder. The AUX3 and AUX4 outputs are re-use of the SUSI bus outputs so these pins will always be logic level. They are the equivalent of pins 3 and 4 on both PluX and MTC-21 which should have the indicator LEDs connected with their cathodes to 0V so there is a bit of odd wiring on the decoder tester.

 

Hi,

 

I've done a simulation of an idealised 5 volt logic output driving an LED through a 1k resistor supplied with 14V DC and the worst case current was 12ma.

 

I've looked at the 36-567 under an illuminated magnifier and the MCU is a PIC18F46K22 whose I/O pins are rated 25ma source and sink.

 

I might try a range of decoders on the ESU tester in the morning although I might wait for some of the new Bachmann 36-557 21MTC 4 function/2 logic function decoders that I've got on order.

 

 

Many thanks

 

Nick

Link to post
Share on other sites

Putting 25mA in to an I/O pin at high from a 12V supply could be way more than the decoder is consuming resulting in the 5V rail going above 5V which would not be good, as people with MTC-21 decoders have found out when mixing up Marklin and non-Marklin versions (Zimo label them with C and D suffix to differentiate them, others may not be so clear). I am sure that Zimo will not be expecting to see high voltage on either the SUSI bus or speaker connections of a Next-18 decoder.

Link to post
Share on other sites

Hi,

I've taken the plunge and plugged the 36-567 decoder into the ESU 53900 tester.

 

post-29876-0-69465200-1509034750_thumb.jpg
After changing the appropriate CV to make the Zimo made decoder change the SUSI outputs to logic level I found I could control F03 and FO4 LEDs on the tester (driven by transistor buffers on the tester).

As expected I couldn't control FO5 and FO6 presumably as they use the speaker pins (I switched the speaker out of circuit and I guess the tester does not have tracks between those pins and the FO5 and FO6 buffers (they must be driven by the FO5 and FO6 pins on the 21MTC connector - not sure about the Plux22 connector).

I also put FO3 and FO4 into servo mode as part of my playing.

I may get some fine solder plus enamelled wire and connect the speaker pins on the Next18 socket to the bases of the FO5 and FO6 buffer transistors (with an on-off switch for when testing sound decoders).

Regards

Nick

Link to post
Share on other sites

  • RMweb Premium

There are two Next-18 socket specifications, the sound Next-18S and the non-sound Next-18. It would appear that the ESU decoder tester has the Next-18S version of the socket.

 

This all demonstrates what an ill thought out socket format the Next-18 is! You cannot plug a non sound decoder in to a socket with a speaker fitted without risk of damage to the decoder from the speaker connecting two outputs together which may or may not be in the same state. Zimo may have pre-empted the possibility and fitted protection to the AUX5 and AUX6 outputs, but who knows.

 

Does this mean that anybody buying the new sound ready next18 socket Farish class 40 (with a speaker fitted as default so all that has to happen for sound is to plug in a next18 sound decoder), can't run it under DCC unless they fit an expensive sound decoder they might not want/need/afford.

 

Oh dear is all I can say if that is the case, and wonder what the reaction will be when this becomes common knowledge.

 

Izzy

Link to post
Share on other sites

Does this mean that anybody buying the new sound ready next18 socket Farish class 40 (with a speaker fitted as default so all that has to happen for sound is to plug in a next18 sound decoder), can't run it under DCC unless they fit an expensive sound decoder they might not want/need/afford.

 

Oh dear is all I can say if that is the case, and wonder what the reaction will be when this becomes common knowledge.

 

Izzy

 

Hi,

 

Good point but maybe a piece of sellotape/adhesive Kapton over the loudspeaker contacts on the Class 40 pcb will do the job. Could be the Class 40 PCB has a switch/link to switch out the speaker.

 

Regards

 

Nick

Link to post
Share on other sites

I think the problem is the confusion (and lack of clear labelling) which separates Next18 and Next18S.   The former has the eight function outputs, the latter has six function outputs and two speaker outputs.    Both of these standards mandate a space for the decoder, the "S" version mandates more space to accommodate the sound decoder.  See the MOROP website for the standards document NEM622 (German and French).  I've not read enough of these to see if the standard requires that a NEXT18 decoder is fully safe if attached to a NEXT18S socket; if it does, that clears up the standard. 

 

 

If there is a speaker fitted, it appears that the Farish 40 is a NEXT18S socket at the wiring level.   If there is sufficient space for the "S" sound decoders, then the label should be NEXT18S.  If insufficient space, then its a "bastard we made up" connection, complying to no standard  (such approach to standards is not unknown in UK model railways). 

 

 

Now, moving to compatibility.  Bachmann's own website describes the class 40 as having a Next18 socket, and says that is suitable for their 36-567 decoder.  So, that's black and white according to the manufacturer: a 36-567 works in the class 40, or if it doesn't Bachmann have no option but to repair/recompense anyone whose stuff goes pop.      It also sort-of implies that the logic level outputs won't be harmed by attempting to drive the loud speaker against each other (because as a consumer I might buy a 36-567 (or any other NEXT18 decoder), try it in a 40, then move it to something else and expect all eight outputs to still work).   The speaker won't make noises on the logic outputs (logic is 5v DC).   I've been pondering the use of the logic levels on the speaker, and am not yet convinced about the sense of connecting the unprotected low current outputs to a speaker, but perhaps something else will come out to explain why its harmless to the decoder.  

Link to post
Share on other sites

I think the problem is the confusion (and lack of clear labelling) which separates Next18 and Next18S.   The former has the eight function outputs, the latter has six function outputs and two speaker outputs.    Both of these standards mandate a space for the decoder, the "S" version mandates more space to accommodate the sound decoder.  See the MOROP website for the standards document NEM622 (German and French).  I've not read enough of these to see if the standard requires that a NEXT18 decoder is fully safe if attached to a NEXT18S socket; if it does, that clears up the standard. 

 

 

If there is a speaker fitted, it appears that the Farish 40 is a NEXT18S socket at the wiring level.   If there is sufficient space for the "S" sound decoders, then the label should be NEXT18S.  If insufficient space, then its a "bastard we made up" connection, complying to no standard  (such approach to standards is not unknown in UK model railways). 

 

 

Now, moving to compatibility.  Bachmann's own website describes the class 40 as having a Next18 socket, and says that is suitable for their 36-567 decoder.  So, that's black and white according to the manufacturer: a 36-567 works in the class 40, or if it doesn't Bachmann have no option but to repair/recompense anyone whose stuff goes pop.      It also sort-of implies that the logic level outputs won't be harmed by attempting to drive the loud speaker against each other (because as a consumer I might buy a 36-567 (or any other NEXT18 decoder), try it in a 40, then move it to something else and expect all eight outputs to still work).   The speaker won't make noises on the logic outputs (logic is 5v DC).   I've been pondering the use of the logic levels on the speaker, and am not yet convinced about the sense of connecting the unprotected low current outputs to a speaker, but perhaps something else will come out to explain why its harmless to the decoder.  

 

Hi,

 

I didn't know about the Next18S.

 

Regarding the 36-567 I can see that if both logic level outputs FO5 and FO6 were in the same logic state that little current might flow through the speaker however I've found testing the 36-567 in the ESU decoder tester its quite easy to press the wrong function button on my DCC system and enable a function output I didn't mean to.

 

The MCU is a PIC18F46K22 with 25ma source and sink from its I/O pins when in Output mode. I don't know what happens if the 25ma is exceeded. Maybe Zimo have put series resistors on FO5 and 6 - there are lots of resistors on the MCU side of the 36-567.

 

post-29876-0-41033000-1509094179_thumb.png

 

Regards

 

Nick

Link to post
Share on other sites

  • RMweb Premium

Hi,

 

Good point but maybe a piece of sellotape/adhesive Kapton over the loudspeaker contacts on the Class 40 pcb will do the job. Could be the Class 40 PCB has a switch/link to switch out the speaker.

 

Regards

 

Nick

 

I posted about this issue of the next18 v's next 18S sockets on the Farish class 40 thread and pauliebanger replied that isolating the speaker contacts would be required/work. I take it from this that there is nothing specific in the standards which cover next18 in 18S sockets.

 

Izzy

Link to post
Share on other sites

Hi,

 

I didn't know about the Next18S.

 

Regarding the 36-567 I can see that if both logic level outputs FO5 and FO6 were in the same logic state that little current might flow through the speaker however I've found testing the 36-567 in the ESU decoder tester its quite easy to press the wrong function button on my DCC system and enable a function output I didn't mean to.

 

The MCU is a PIC18F46K22 with 25ma source and sink from its I/O pins when in Output mode. I don't know what happens if the 25ma is exceeded. Maybe Zimo have put series resistors on FO5 and 6 - there are lots of resistors on the MCU side of the 36-567.

 

 

My concern is that a speaker is a nominal 8ohm resistance between FO5 and FO6.  Which is fine if both are at 0v, or both at 5v, because no current can flow.  But, if one is at 5v and the other 0v, then there is potential for current flow depending on the exact schematics of the decoder. 

 

I can think of a few tests, even non-risky ones, but may be simpler to drop a line to Zimo support and ask them what happens to those two outputs if the decoder is used in a NEXT18S socket, with speaker attached, and if one of the two outputs is accidentally activated by the user.

 

 

- Nigel

Link to post
Share on other sites

I'm finding this to be a very interesting conversation, however the technical (electronics) aspects of this are beyond me.

I don't know if this is of any help  (a section taken from NEM662)......

 

 

 

 

The contacts may be loaded with 0.5 A each.

The track connections as well as U + and GND (decoder plus and minus to rectifier) ​​are designed with two contacts each.

The load capacity of these connections is therefore 1.0 A.

 

The circuit boards of the locomotives as well as the locomotive decoders must be designed in such a way that a different use of the pins AUX5 and AUX6 does not lead to damage to the locomotive or decoder.

 

 

Unless I'm mistaken, it appears to say that decoder manufacturers must design the Next18 decoders with protection against such damage.

 

 

.

.

Link to post
Share on other sites

I posted about this issue of the next18 v's next 18S sockets on the Farish class 40 thread and pauliebanger replied that isolating the speaker contacts would be required/work. I take it from this that there is nothing specific in the standards which cover next18 in 18S sockets.

 

Izzy

 

Izzy,

 

My response was to address the specific point you raised re the GF Class 40. Strictly, there should be no problem anywway, but insulating the potential connections is a ssafety first' strategy which will work, cause no damage and be easy to achieve.

 

The issue of the differing functions at the pin positions shared by the speakers and Aux 5 and Aux 6 is covered in MOROP, NEM662. It puts the onus on the model maker to provide the correct electrical connections.

 

Function Outputs Aux 3 - 6 are logic level only.

 

In fact, Aux 3 and Aux 4 also  share functions with SUSI Clock and Data outputs.

 

2.2 Electrical properties

The contacts may be loaded with 0.5 A each. The track connections as well as U + and GND (decoder

Plus and minus to rectifier) ​​are designed with two contacts each. The resilience of these Connections is therefore at 1.0 A. The circuit boards of the locomotives as well as the locomotive decoders must be so be designed to ensure that a different use of pins AUX5 and AUX6 will not cause damage

 

 

Kind regards,

 

Paul

Link to post
Share on other sites

Looks to not be a problem then. Decoders should be happy with a speaker connected to the logic outputs 5 and 6, while locos should be OK having logic levels applied to the speaker - it is in the standard. The other way round generally won't be a problem because the decoder will not fit due to insufficient space, but even so if a small sound decoder is fitted to a non sound loco it should be happy having sound blasted into its function drivers while there is head scratching as to why no noise is coming out!

Link to post
Share on other sites

....and on the decoder manufacturer, if I read it correctly.

 

Ah yes, you are correct. My mistake.

 

I don't have a non sound Next 18 decoder to test, but a ZIMO MX658N18 sound decoder on my ZIMO decoder tester, MXTAPV, immediately connects the speaker, so it seems ZIMO know the correct specification and their responsibilities.

 

The Next 18 socket might be spec'd for Next18S only. There is a jumper on the MXTAPV to change from open collector to Logc Level FOs, so when I have chance I'll see if this makes any difference.

 

The Bachmann Next 18 are re-badged ZIMOs so probably all's well.

 

Can't throw any light on what an ESU test board will do, but I'm sure they will have thought of this, after all, the ESU website says '....just plug in any decoder.....' and if a non-21MTC decoder was fitted to their 21 pin socket, no harm was done.

 

Kind regards,

 

Paul

Link to post
Share on other sites

Slightly off topic but does anyone know if using the ESU test rig you can install two decoders at the same time, say to simulate a piggy back situation of setting up one decoder for driving and the other for sound as has recently been seen with I think Lenz or Zimo and TTS.

Rob

Link to post
Share on other sites

Looks to not be a problem then. Decoders should be happy with a speaker connected to the logic outputs 5 and 6, while locos should be OK having logic levels applied to the speaker - it is in the standard. The other way round generally won't be a problem because the decoder will not fit due to insufficient space, but even so if a small sound decoder is fitted to a non sound loco it should be happy having sound blasted into its function drivers while there is head scratching as to why no noise is coming out!

 

Sorry Suzie,

 

My earlier response was based upon a misreading of the NEM specifications.

 

I'm not confident that all parties will in every case adhere to or even know about their obligations regarding these dual role connections. One only has to look at the 21MTC fiasco to see the dangers on the horizon.

 

I've just written a feature about DCC interface types for Hornby Magazine, which is probably why I'm so animated. LOL

 

My personal preference is for the PluX interface which does seem to address all the shortcomings of other types whilst adding useful features, particularly the ability to safely use different 'size' PluX decoders. It has to be recogised, though that Next 18 gives the thinnest package yet possible, and that feature is very likely a priority for N gauge designers.

 

Kind regards,

 

Paul

Link to post
Share on other sites

If there is only space for a Next-18 then that will be the only choice - but I am sure it can't be beyond the possibility of manufacturers to find that extra 0.7mm of thickness in 4mm scale models! It is not like the Next-18 is small enough for Z or anything where PluX has the 8-pin which should be small enough (should anyone get round to making one!)

 

It is worth noting that Zimo already produce the MX648p16 sound decoder that is 0.1mm thinner and 5mm shorter than the Next-18S size spec (but 0.5mm wider alas) therefore is actually smaller than the functionally equivelent Next-18S MX658n18.

 

At the end of the day the key functionality differentiator between Next-18S and PluX-16 is the provision of a connection for stay alive which is completely missing from Next-18. It will be good to put the facts out there so that people will be able to lobby the manufacturers to fit sensible sockets.

Link to post
Share on other sites

If there is only space for a Next-18 then that will be the only choice - but I am sure it can't be beyond the possibility of manufacturers to find that extra 0.7mm of thickness in 4mm scale models! It is not like the Next-18 is small enough for Z or anything where PluX has the 8-pin which should be small enough (should anyone get round to making one!)

 

It is worth noting that Zimo already produce the MX648p16 sound decoder that is 0.1mm thinner and 5mm shorter than the Next-18S size spec (but 0.5mm wider alas) therefore is actually smaller than the functionally equivelent Next-18S MX658n18.

 

At the end of the day the key functionality differentiator between Next-18S and PluX-16 is the provision of a connection for stay alive which is completely missing from Next-18. It will be good to put the facts out there so that people will be able to lobby the manufacturers to fit sensible sockets.

 

Thank you for pointing out the MX648P16. Nevertheless, I stand by my claim that the Next 18 is the thinnest plug-in interface currently available.

 

I have both of the decoder types you mention in front of me. And a set of calipers. There are some discrepancies between the dimensional differences you state and the measured reality.

 

There is a much greater depth required to fit an MX648P16 than MX658N18 for two reasons.

 

1 ) the PluX decoder is thicker.

2 ) the PluX interface is much thicker.

 

The MX658N18 has a maximum thickness of 3.21mm, including the Next 18 plug. The MX648P16 has a maximum thickness of 3.59mm excluding the PluX connector which adds 4.28mm to the overall thickness.

 

You used the term 'functionally equivalent'. I think we can probably agree that neither decoder will function unless it is fitted to its mating counterpart on the model's PCB. That being the case, overall measurement of depth required to accomodate the working combination is the key factor which manufacturers will need to take into account when selecting an interface type and designing the model's internal spaces.

 

Whether the differences have a material impact on the performance of the model (weight, for example) or the way mechanical integrity may be affected is something the manufacturers will have to take into account.

 

If the only consideration was the logic of the electrical connections, PluX would be the obvious choice.

 

For all 'plug-in' interfaces, the thing which excercises me most of all is the generally poor access to all of the decoders' features provided on the models' PCBs.

 

I'm not a fan of Next 18 for a number of reasons, but I can see the attractions for manufacturers, especially for N gauge models. Space is less critical an issue in 00 gauge and above, so there is a stronger case for PluX decoders, a view which I have never refrained from expressing.

 

My concerns about the effectiveness of the Next 18 interface, (differences between Next 18 and Next 18-S aside) is the relatively low power handling, and probably as a result of this, the use of Logic Level FOs from Aux 3 and above. Let's hope manufacturers are more aware of this than they were in the case of 21MTC decoders.

 

Non of this helps the OP to resolve the question of whether it is safe to fit a Next 18 decoder to an ESU test board, so maybe a new thread would be the best place to continue this discussion?

 

Kind regards,

 

Paul

Edited by pauliebanger
Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share

×
×
  • Create New...