welshhighland

They appear to be like the Talyllyn machines, but without the state-of-the-line dial... what say you Marquis?

Martin Coombs wrote:Not adding anything to the technical discussion, but I mentioned a while back about having seen similar instruments in use in Argentina.

Well, just for interest, here is one such machine spotted last month on display in the station at San Carlos de Bariloche, at the end of the former FC del Estado route from San Antonio Oeste. Click on the image to call up an enlargement.



The machines were built under licence in Argentina, though I believe 'Talleres (workshops) Esquel' almost certainly refers to a factory name rather than to the town of Esquel in Patagonia. 'FNGR' is the abbreviation for the former nationalized Ferrocarril Nacional General Roca, which ran the Bariloche line.

Others more erudite than I may be able to comment on any differences from the FR MTS machines, or indeed the IE machines destined for the RhE.

Martin

H.E.Bogge wrote:They appear to be like the Talyllyn machines, but without the state-of-the-line dial... what say you Marquis?

..Er.. depends what you mean by 'Talyllyn' machines.....

If you are referring to the current single-line machines worked by the TR, then it is the same shade of red for the Pendre-Brynglas section - but the the visual similarity ends.

If, OTOH, you mean the ex-Talyllyn machines that are now in the Port - Minffordd section on the FR then, yes they are nearly identical ..but..

I fear to get the complete answer, I might need to post the photograph elsewhere - if I can call upon your good nature Mr. Coombs - at the present moment, I think the following:

• undoubtedly, it is built by a contractor out in Argentina. Exactly what this means, I am not utterly sure in the case of the Railway Signalling Company (RSCo)[1] - I suspect that unfinished castings and 'bits' were sent overseas and fettled/assembled in licensed workshops - rather than shipping over complete instruments.
• the design of the instrument is very, very peculiar. On the surface it looks a bit like an 'M' pattern carcass, but has more in common with the 'S' pattern. I say this on several counts - the galvo is the rather enormous type favoured by the 'M' pattern [2] (and evidenced on some British examples - LINK for the F configuration, 'M' pattern Boston Lodge intermediate, which I think is converted from a F config terminal instrument, formerly Wharf - Pendre [unfortunately, it shews what looks like an early 'S' pattern unpolarised galvo, fitted as part of the conversion to an intermediate])
• For comparison:
  
  
  These are my own instruments (over in the country library), both 'S' pattern and set up for magneto working (sorry about the duff picture, off a mobile phone) - you can see the similarities with the Argentinian instrument.

I think on reflection the number 'S. 32' [3] and the lack of indicator in the top left of the instrument proves that this is an 'S' pattern instrument - I can't quite tell from the guideways in the picture - if so it is exceptionally interesting, as I think it is the lowest numbered 'S' pattern instrument I have ever seen. Coo.


[1] Fazakerley - the RSCo 'built' all the METS in the world - just in case later readers didn't realise..
[2] This is a potential pitfall - on some British and Irish 'M' pattern instruments the large galvo seemed to be favoured - but not on 'colonial' instruments - it all depends on when the instrument was constructed.
[3] As part of the licencing the numerical series was I think worldwide.

Marquis,

By all means distribute the photo wherever you think fit.

Just to confuse matters, here is another Argentine pic, though a slightly out of focus one, this time of an FC General Urquiza instrument in the museum at Retiro station in Buenos Aires.



I look forward to hearing of your deliberations in due course.

I hope by the way that the moderators will excuse my provoking international and therefore off-topic discussion, in the hope that this will build up our understanding of how such machines - soon to be in use on RhE - actually work.

Martin

Fantastic!

That picture shows exactly the pitfalls in trying to eliminate which type the instrument is from the galvo alone!

I guess that is an 'S' pattern instrument, built for magneto working (no bell tapper) fitted with an 'M' pattern galvo mounted at the wrong angle - the stop pins are all a-flunter for the deflection.

Doesn't look as if this was provided with a serial number but does have S...No [blank]. Of course, my theory could go to pot if a picture of a staff from these instruments appears and they are 'M' pattern and the S stands for Serial..

EDIT: On second thoughts, your first picture is a modified instrument anyway - the train going to/from twiddly knob (used as a cut-out switch to drop the galvo and tell the bloke at the far end to stop winding the mag) has been replaced with a push-button type.

Marquis DeCarabas wrote:
If, OTOH, you mean the ex-Talyllyn machines that are now in the Port - Minffordd section on the FR then, yes they are nearly identical ..but..

I presume that these machines were sourced via the TR then? IIRC the W&L acquired a number of MES machines from Sierra Leone in c1975 (along with the Hunslet loco and the coaches) which were sold on to the FR. Where are they now? One MES "token" from these Sierra Leone machines is on display with the Hunslet in the NRM - I don't know if any more exist. What would happen if one was inserted into the token machine now - would it be accepted?

Nick

Nick Griffiths wrote: One MES "token" from these Sierra Leone machines is on display with the Hunslet in the NRM - I don't know if any more exist. What would happen if one was inserted into the token machine now - would it be accepted?

If the configuration matches - A,B,C,D,E,F and the instrument pattern 'M' or 'S' (ISTR that SLGR was 'S' pattern - perhaps the origin for the use of S pattern above Tan-y-Bwlch) then there is nothing to stop the staff going into the machine - getting it out again is a different matter......

Marquis DeCarabas wrote:

Nick Griffiths wrote: One MES "token" from these Sierra Leone machines is on display with the Hunslet in the NRM - I don't know if any more exist. What would happen if one was inserted into the token machine now - would it be accepted?

If the configuration matches - A,B,C,D,E,F and the instrument pattern 'M' or 'S' (ISTR that SLGR was 'S' pattern - perhaps the origin for the use of S pattern above Tan-y-Bwlch) then there is nothing to stop the staff going into the machine - getting it out again is a different matter......

The first METS instruments we had from Tan y Bwlch to Dduallt in 1968 were the former LNWR ones from Verney Junction to Buckingham. From memory, the pair cost £30. I may have the stamps for relabelling them somewhere. I have - or had - staff No.1 from that set; another is in the County Museum and the other 28 went to the FR with the instruments.I remember Norman Pearce saying there was something unusual about the configuration. It was one we did not already have on the FR - was it perhaps an F? I forget. I may be wrong after so long; only A3 or A4 intelligence.

Pedr

pnjarvis wrote:I remember Norman Pearce saying there was something unusual about the configuration. it was one we did not already have on the FR - was it perhaps an F? I forget. I may be wrong after so long; only A3 or A4 intelligence.

I suspect that it might be because they were 'S' pattern instruments, rather than 'M' pattern which was the FR standard.

Delving further into the Sierra Leone connection there would have been at least 36 pairs of instruments - they appear to be 'S' pattern and set up for magneto working.

I find this thread fascinating, and can't wait to see the blending of early 20th. century cast iron and copper coils with 21st. century digital technology.

Uncle John wrote:I find this thread fascinating, and can't wait to see the blending of early 20th. century cast iron and copper coils with 21st. century digital technology.

But you never will, unless you volunteer with S&T, as it will all be hidden in the machines or adjacent grey boxes!!

JGF

Adrian

Adriangrayfr wrote:

Uncle John wrote:I find this thread fascinating, and can't wait to see the blending of early 20th. century cast iron and copper coils with 21st. century digital technology.

But you never will, unless you volunteer with S&T, as it will all be hidden in the machines or adjacent grey boxes!!

JGF

Adrian

Me too, really. I await with hopeful (and educated) interest that it works well, and that all possible failure modes are covered. [1] Sadly I'm now too far away and busy moving/renovating a house to be of much practical use. However, come the time and the opportunity I would very much like to help with reconditioning the IR instruments - the little grey boxes are another matter...

[1] NB. that isn't to be interpreted as a negative - although I understand the elegant simplicity (yes!) of the METS [2] whether worked on DC or AC; how it will all tie in with RFC1122 and RFC1123 and their updates will be fascinating.

[2] polarized locks, now that's *very* clever Click - if those who think I've come up with a complicated set of suggestions - read the Patent for the Permissive Unidirectional Staff Instrument..

German and French translations available.

"Quite simple" to describe:
The interface between the very simple token machines and an "internet" backbone would have to turn the amps/volts/phase into data at the transmitting end, add a whole pile of addressing, encryption and integrity data to that , then do the reverse at the receiving end. But as ever the devil is in the detail....
Replace (simple) machines and some wire with the same (simple) machines, some bits of wire, interface boxes, more wire, a couple of servers, more boxes, more wire.....

I'm sure The Marquis would agree - on a simple railway simple signalling is the best (most economic) solution. Token machines have been around for many years, have had most (if not all) the problems identified and "solved". There are few failure modes, and few of those are "beyond understanding". Modern communications systems such as "the internet", or "internet technology" are many orders of complexity removed, have many more failure modes, many of which are hard to understand.
Expanding on what I said in my previous post.
Using a modern network backbone (Internet TECHNOLOGY, not "The Internet") for communicating the signalling data and commands adds several layers of equipment - you need to convert your input into something that can be transmitted as "0s and 1s", you then need to add some addressing (so it knows where it is going and where it is coming from), you then add encryption (to make it hard for someone/something to "hack into" the signalling control network, you then add data integrity data (to make sure that what you sent is what you receive). The conversion is, in computing terms "fairly simple" to do, as are the encryption, addressing etc. So what starts out as a simple switch state (on/off) "message" carried on a pair of wires becomes a long message of several tens of bytes. But its still not left the source! Now you pass it through a box (an Network Interface Card - NIC) that puts it onto the network. This NIC adds a little more addressing data and a little more checking data, and off it goes. Along the way it may pass through a number of other NICs, some of which my just be there to boost the signal, some of which may be to route the message, each adds a bit of addressing data (so you know how it got to the end). Finally it arrives at the NIC at the fat end, and on to the receiver's processor. Here all the unwanted addressing is stripped away leaving the data that was sent from the transmitter remains. This is checked to make sure its all there (remember that integrity stuff I talked about?), that it has come from a valid source (the source address is correct), that it is correctly encrypted. All this data is removed, and we are left with the data that represents the initial message (the state of our switch).
Now that's a lot of effort to get the state of a single simple switch from point a to point b. Looking at a token instrument there are only a few contacts, and by clever use of phase/voltage/frequency the state of these can be be transmitted quite easily on a pair of pairs (or a single pair) using traditional techniques. However think how many wires are needed to take the state of a thousand switches. It makes more sense as each "pair" of wires can securely carry the status of many thousand switches, compared with a point to point (traditional) system requiring many thousands of wires (or pairs of wires...) to do the same.

Adriangrayfr wrote:

Uncle John wrote:I find this thread fascinating, and can't wait to see the blending of early 20th. century cast iron and copper coils with 21st. century digital technology.

But you never will, unless you volunteer with S&T, as it will all be hidden in the machines or adjacent grey boxes!!

If I (a) was younger, (b) didn't already have commitments to local charities and my family and (c) lived nearer and/or had a driving licence, I'd be there like a shot.

bobsmith wrote: *interesting layman's guide snipped*

Now that's a lot of effort to get the state of a single simple switch from point a to point b. Looking at a token instrument there are only a few contacts, and by clever use of phase/voltage/frequency the state of these can be be transmitted quite easily on a pair of pairs (or a single pair) using traditional techniques. However think how many wires are needed to take the state of a thousand switches. It makes more sense as each "pair" of wires can securely carry the status of many thousand switches, compared with a point to point (traditional) system requiring many thousands of wires (or pairs of wires...) to do the same.

I remember seeing the modern/old interface in graphic detail when BT was modernising its telephone exchanges in the late 1970s and early 80s. By far the biggest part physically was the interface between all the thousands of pairs of copper wires with traditional relephones on the end and the all-singing all-dancing modern switch. It was quite a challenge to get across to the suppliers the concept that you can't at a stroke get rid of (a) the need for 75V ac ringing current (b) several tens of milliamps of dc microphone current and (c) a forest of rotary dial tepephones.