NEM pocket standardization

"Jane Sullivan" wrote

Isn't that the whole idea with NEM & NMRA standards? Of course Hornby always know better.

John.

Just another example of Hornby ignoring standards. Considering the recent efforts to reissue the Rivarossi line in the USA complete with RP25 wheels, you'd think they've learned their lessons.

Only the coupler fixture is standardized, see

I was just thinking that the pocket mount itself wasn't covered by the specs - pity, would be better if it was.

Ian J.

Not needed. NEM 362 provides the necessary coupler pocket dimensions and height above rail. However, AFAIK, MOROP does not specify a standard coupler shank to fit that coupler box. OTOH, any manufacturer that pays attention to the coupler pocket should be savvy enough to make a shank that will fit the NEM box. Hornby didn't.

NMRA's standard includes the position and diameter of the shank pin, which implies a minimum length for the coupler shank. That's why it's possible to replace old horn-hook couplers (eg on 309-year old Athearn cars) with modern knuckle couplers with no problems. Except one: many manufacturers make shank that are too thin, which results in coupler droop in a standard NMRA pocket.

You can't win...

No, because many wagons/coaches/locos have different amounts of space available behind the bufferbeam. Standardization of the mount for the pocket might result in strange under chassis lumps of plastic or coupler pockets extending way out where we don't want them. In fact MOROP do have an NEM standard pocket mount for odd situations, used by Fleischmann and Roco in several situations.

Greg.P.

With all due respect Wolf, do you know of many standards that Hornby ignore ?

Cheers,. Simon

Erm, I think your question is a wee bit disingenuous, eh? ;-)

Jane Sullivan noted that Hornby and Bachmann aren't using the same standard for coupler pockets. Thus either or both are ignoring some standard (I think it's the NEM standard, which specifies an interior width of 3.2mm. It's unclear where Jane measured the 4.0 and 4.6mm dimensions, I suspect it was over the exterior of the pocket.)

Until the early 2000s at least, Hornby didn't even have a standard for their own wheels. Wagon, carriage, and locomotive wheel profiles were all over the place. How do I know this? By comparing wheel measurements as given in reviews of their products. They still don't use the same wheel standards as other manufacturers, although there is now a welcome trend to adoption of the RP-25 profile. The variation in wheel profiles prompted Peco to develop its "universal" turnout, which was intended to accommodate the widely different wheel profiles in use in the UK in the

1950s and later. I bought a couple back in the late 70s. Never again. I had to fill the frog with epoxy and cut new flangeways, ditto for the check rails. I did this because my budget was rather limited back then. Nowadays, I would just throw them in the trash (which is where they ended up eventually anyhow, when I started building turnouts.)

There are two main standards bodies for model railways: MOROP in Europe and NMRA in North America. Their standards are mostly compatible, thank goodness. It's significant IMO that Peco has brought out a line of turnouts that conform to NMRA standards (well, as interpreted by Peco, which means they're little bit off, but Peco is not alone: NMRA had a number of one-way dimensions, which mfrs interpreted as +/- instead of

0/- or 0/+ as intended. This is being rectified.)

NMRA was formed to "promote interchange and inter-operability", and standards have been refined over the years to achieve this goal. They're not perfect, but IMO they're better than any others. One reason is that within the standards there are recommended practices, or RPs. RP-25 is the one that defines the most widely used wheel profile.

HTH

It's been stated earlier that that dimension is not part of the standard - is that incorrect? If not, why is it Hornby that is wrong, and not Bachmann?

You can hardly blame Hornby for that - people were running Triang, Hornby, Hornby-00 (Wrenn), Peco, Trix, various kits with various wheels etc etc - I was! And greatfult to Peco I was too, as I could buy and run anything.

^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ If the standard is not capable, you can't blame the makers!

Well, if it's only a "Recomended Practice" Hornby cannot be accused of breaking the standard, in fact the can rightly claim to be adhereing to it!

NMRA is not an ISO standard or similar, it's effectvely a de facto standard, much as "will it run on Peco track" is effectively a standard in the UK - to a large majority of UK 00 modellers, the Peco standard is the one that matters!

Cheers Richard

One way variation in a standard is quite clear. But frequently mis-read by people. The problem Wolf outlines is a standard which is being mis-read.

One can have a standard which specifies the minimum clearance of a flangeway, of say, 1.00mm, with a variation of up to 10% larger (this would be 1mm, -0, +0.1). Thus, 1.00mm is within specification, as is 1.05mm, and

1.10mm. But 0.99mm is out of specification. However, such as specification is frequently mis-read (or mis-copied) as +/-10%, and in use becomes 0.90mm to 1.10mm. The problem arises when one combines two mis-read numbers; for the flangeway of 1.00mm, one might also specify a flange thickness of 0.9mm +0, -10%, meaning that 0.81mm to 0.90 are within specification, but 0.91 are outside the specification. But if this is also mis-read, along with the flangeway above, one can have a flange of 0.990mm trying to go along a gap of 0.90mm - clearly it doesn't fit. But, had the specification been read correctly, there would have been at least 0.10mm of clearance.

Non-railway example; UK car speedometer specification is -0/+10%, meaning that a speedometer in a car when manufactured must never under-read, but can over-read by up to 10%. Car makers understand this, and therefore always set the manufacturing process to give some degree of over-reading. How much over-read they set during manufacture depends on their manufacturing tollerance; it is manufacturing tollerance which gives the variation in any manufacturing process.

The key word is "capable", as in capable process. Im engineering there is always +/-ve tolerance. If it is only one way, then the dimension needs re-specifying! If a standard doesn't do that, it's asking for trouble, i.e not capable.

UK speedo's also always designed to read 4% high by the way (mechanical ones, anyway. Source: Test Engineers from Smiths I worked with). I seem to recall elctronic are not skewed, and are 5% +/- tolerence.

Cheers Richard

No, you misunderstand (as have the makers which Wolf was refering to). The standard can say -0, +10%. It is a completely clear standard, and capable of being manufactured within specification. The engineer who is making the part (or designing its manufacture) reads the standard and knows there are two limits: upper and lower. The engineer knows the accuracy of the measuring equipment available, and therefore makes the part within specification.

One can re-draw a specification to mid-points. (I once got involved with re-drawing a track specification to mid-points, from one which quoted some absolute max (or min) with undershoot (or overshoot). It took longer than I thought it might.)

Doesn't alter the specification, which is in the C&U regulations. The specification is -0, +10%. The manufacturers correctly interpret this as "must never under-read", and "may overread to up to 10%". They make their production items to +5% (or +4%) of actual speed, with their manufacturing tollerance around it, ensuring that all speedometers are within the C&U specification. Those makers understand the specification.

- Nigel

You're confusing design specs and manufacturing tolerances. If the design specs specify a minimum, that minimum has to mesh with another design spec on another part. It's a warning to the toolmakers to ensure that the manufactured parts will never interfere with each other, ie, will always fit within specs. Where two parts must fit, it often happens that one must be specified as a 0/+ and the other -/0 (or min and max dimensions.) If the design spec were +/- in such a case, it could happen that parts at either extreme of the tolerance wouldn't fit. That's the point of Nigel's very nicely explained example.

[snip]

That is, they design the speedos to read [5% above actual speed +/- 4%], thus they will read as low as [actual speed + 1%] and as high as [actual speed + 9%].

See?

Well, I've just asked the queston of one of our suppliers (who makes all sorts of intesting stuff with CNC machines), and he just laughed when I asked if -0, +10% was ok for a tolerance - "Only when they've got the drawing wrong and can't be bothered to put it right". He'd reject he drawing, as it would almost certainly be misread, and anyway he'd have to recalculate the dimension himself to allow for machine tolernaces, thereby making the drawing invalid. The only exception he could come up with would be doing a taper from say Xmm to nothing (e.g. point blades), where the "nothing" end cannot have a -ve tolerance (but then the minimum length would take over anyway), and some CAD systems insist on having a -ve value, even if it zero.

I think you'll find these lads understood it too - they made the test equipment used for many millions of speedos - and don't forget, the C&U regulations are not the be-all and end-all in the automotive world, SSMMT, IEEE, etc etc. Some bits are/were "overuled" by international standards/EU standards - that's why Ford has an internaional "Director" of fuel caps (really), daft as it seems. I have acutually worked for "those makers" and had quite a lot of dealings with the various standards - and it 'aint simple or straightforward. No one standard covers everything completely, even when it apparently does.

Cheers Richard

[...] [...]

First, I don't care squat which is wrong. Second, in my experience Bachmann conforms to applicable standards (where it can find them), and Hornby (Triang or not) goes its own way. There are no standards for the UK, more's the pity. I'm astonished that so many of you UK modellers put up with the wheel profile mess that entailed Peco's valiant (but IMO failed) attempt at producing a "universal" turnout. I wanted to model GWR many years ago, but gave up when I realised that the UK manufacturers just refused to work to a common standard wheel profile (among other things).

NEM 362 specifies a coupling box for "austauschbare Kupplungskoepfe", ei, "exchangeable couplers." The interior dimension implies a shank size. Etc.

NEM does not specify a mounting for the coupler box, only height above rail, and position relative to end beam and buffers. There is no standard for mounting the box itself, as long as it's mounted at the correct position. I interpreted jane's original comment as referring to the couplers to be mounted in the pocket. If that's wrong, hey, jane, kindly supply correction. (If jane is referring to exchangeable coupler pockets, then there should be a standard mounting method IMO.)

FWIW, NMRA specifies both car bolster (thickness) on the car and bolster height of the truck (bogie), which guarantees a correct height for coupler pocket mounting. NEM could do the same for bogie vehicles. NEM

303 (buffers) implies a standard floor height, but could be extended to explicitly specify a standard floor or coupler mounting pad height so as to place coupler pockets at the correct height.

HTH

I measured it across the triangular-shaped dovetail at the other end of the pocket. This is the bit that allows you to attach the pocket to the vehicle.

AFAIK all car speedos are now electronic (in N. America anyway) - and that's not the same thing as digital. Presumably the only variable that affects the reading is tire/tyre diameter. Manufacturers warn against intalling different diameter tires beyond a small tolerance (even if wheel size is changed). Apart from affecting speedo/odo accuracy, it can throw the entire electronic engine control system out of whack.

[...]

Ah, yes, that bit is not standardised. But as your experience shows, it should be IMO.

HTH