Huret Duopar; flawed genius at work?

For discussions about bikes and equipment.
Brucey
Posts: 37292
Joined: 4 Jan 2012, 6:25pm

Huret Duopar; flawed genius at work?

Postby Brucey » 14 Feb 2020, 11:01am

My first derailleur geared bike was fitted with a Huret mech; a 'Svelto'. It never shifted terribly well, (having a parallelogram geometry that appeared to be optimal should the larger sprockets happen to be on the outside of the freewheel..... :roll: ) but it did have some nice touches like ball-bearing jockey pulleys. It (of course) soon wore out and became incredibly floppy in the main parallelogram, so I retired it before it 'retired itself' by spontaneously plunging into the rear wheel.

I was surprised to discover that the Huret Allvit model -which had a much more sensible parallelogram geometry if you were using a 14-24 or 14-28 5s freewheel- actually predated the Svelto model. However that too used to wear out or seize up if it saw much weather.

I concluded that the inept parallelogram geometry in the Svelto was born of expediency; making derailleurs like this was not expensive and that was a big selling point in mass produced derailleurs. This conclusion was reinforced by the observation that the Huret 'Eco' model (which had equally poor geometry, but at least lasted a bit longer before it wore out) was the most usual to be found on hordes of Raleighs in the late 1970s and early 1980s.

I was surprised to discover that the manufacturer of these pressed-steel horrors also made one of the most expensive and complicated rear mechs that was ever produced; the Huret Duopar.

Image

This was launched in 1975 and boasted a stupendous capacity, exotic materials such as Titanium, and because of its clever double parallelogram design, the guide pulley would track any freewheel perfectly, even on the widest range touring ratios. For many years, when tested rigorously, it was found to be the best performing derailleur available. Even the (later, cheaper, heavier) Duopar Eco model performed well on test. This provoked both SunTour (Mountech and Trimec models) and Shimano (Superplate models) into a response which used similar ideas including a second parallelogram or pivot, all of which were meant to improve the way the guide pulley tracked the freewheel.

However all these mechs had the same problems; they were complicated (and therefore a bit heavy) and they were prone to wearing out, or just bending too easily. The SunTour mechs had a second tension spring set coaxial with the guide pulley, on a third pivot. The spring and the guide pulley were competing for the same space, and the net result was that the pulley was made in such a way as it would wear out, jam or break, and replacing it (should you be able to obtain the correct part) was fiendishly complicated.

The shimano superplate model effectively used a second parallelogram too, but they exploited the fact that one side of the second parallelogram was always loaded in tension so used a cable instead of a rigid link, #20 here; https://si.shimano.com/pdfs/ev/EV-RD-M700-SP-0615B.pdf

The guide pulley was a special one, and less easy to replace than a standard one, but it wasn't so heavily compromised as the SunTour ones.

The Duopar just used to wear out; the second parallelogram was engineered using parts which looked as if they had come out of an Allvit (and/or a Svelto), were exposed to all the crud thrown off the wheels and were usually coated in oil from the chain too, just to make sure the dirt would stick well and turn into really good grinding paste. It is also worth noting that every bump in the road would tend to cause the second parallelogram to articulate, which must have caused much of the wear. No surprise then that these mechs can last fairly well on (say) road tandems, where conditions were usually less bumpy and cleaner, with the RD is well away from the plume of crud coming off the front wheel. The duopar was also a hefty beast, all those extra parts weighed plenty, even if made in Titanium; the Titanium Duopar weighed almost exactly double the weight of a SunTour Cyclone II.

There is more to read about Duopars here

http://www.disraeligears.co.uk/Site/Huret_derailleurs.html (including many Huret documents, patents etc) and here

https://bikeretrogrouch.blogspot.com/2014/11/classic-equipment-huret-duopar.html

and here

http://www.disraeligears.co.uk/Site/SunTour_derailleurs_-_Frank_Berto_and_the_curse_of_Duopar.html

you can read how the Duopar tested well and this spurred other manufacturers to emulate it.

The thing is, the principle behind the Duopar is basically correct. Arguably despite the cost of the things, the main problems were, in a curious paralell with their cheaper mechs, ones of execution rather than conception.

cheers
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~

bgnukem
Posts: 264
Joined: 20 Dec 2010, 5:21pm

Re: Huret Duopar; flawed genius at work?

Postby bgnukem » 14 Feb 2020, 11:24am

Interesting, and a far cry from the Huret Eco gear I had on the back of my Raleigh Winner 10-speed as a teenager!

Isn't this just an over-complex way of tracking the freewheel contour that Shimano achieved using an inclined single parallelogram?

Brucey
Posts: 37292
Joined: 4 Jan 2012, 6:25pm

Re: Huret Duopar; flawed genius at work?

Postby Brucey » 14 Feb 2020, 11:36am

bgnukem wrote:Interesting, and a far cry from the Huret Eco gear I had on the back of my Raleigh Winner 10-speed as a teenager!

Isn't this just an over-complex way of tracking the freewheel contour that Shimano achieved using an inclined single parallelogram?


Shimano were not allowed to use a slant paralellogram before 1984, because it was a SunTour patent dating from the early 1960s. The 'modern' Shimano RDs with a slant paralellogram as well as sprung upper and lower knuckles do indeed provide reasonable tracking of the guide pulley over a range of freewheel sizes. And that range (of different sizes) is indeed slightly greater than that allowed by single-sprung slant mechs (older SunTour, most SRAM, and many current/recent shimano models).

However the double spring arrangement doesn't work perfectly in practice for several reasons;

1) the pivots have some sticking friction so don't allow perfect tracking
2) the springs are opposed to one another, and of course have both a rising force and a rising rate; this means the force balance (and therefore the guide pulley tracking) is never going to be perfect; for example using a different chainring causes the force in the B pivot to vary , and this change is at the expense of accurate tracking.
3) like the Duopar, riding on bumpy roads causes the mech to articulate and therefore the pivots wear faster than normal.

The Duopar tracks the sprockets more perfectly than shimano mechs and others similar, particularly because point 2) above does not apply. Shame it has other drawbacks though....

cheers
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~

ChrisButch
Posts: 960
Joined: 24 Feb 2009, 12:10pm

Re: Huret Duopar; flawed genius at work?

Postby ChrisButch » 14 Feb 2020, 4:20pm

I still have one, and yes, it did spend its working life (actually quite long - about 10 years) on a tandem.

iandusud
Posts: 244
Joined: 26 Mar 2018, 1:35pm

Re: Huret Duopar; flawed genius at work?

Postby iandusud » 14 Feb 2020, 6:01pm

I wonder how many Eco RDs Huret made, it must run into millions as they were fitted as OE to so many bikes in the 70s and 80s. As for the Duopar I had one on my one and only bike in the 80s. It would certainly cope with a huge range of gears and worked well but they were fragile, so not for the ham-fisted.

Ian

Bonefishblues
Posts: 7365
Joined: 7 Jul 2014, 9:45pm
Location: Near Bicester Oxon

Re: Huret Duopar; flawed genius at work?

Postby Bonefishblues » 14 Feb 2020, 7:18pm

I had the titanium flavour one one my 80s touring bike. It was a revelation.

photobike
Posts: 30
Joined: 7 Nov 2014, 5:51pm

Re: Huret Duopar; flawed genius at work?

Postby photobike » 14 Feb 2020, 7:32pm

I had one on an 80's Raleigh Classic 15 speed that I bought off my brother in about 1989. It ate itself on a tour I was doing about 1991 from West Wales to London. Somewhere in the Brecon Beacons I think. It was replaced by the only cheap and nasty RD available in the only bike shop for miles around.

I remember loving the tech but cursing the terminal event. They were extremely expensive if I remember so that bike never got a like for like replacement.

Peter

Bonefishblues
Posts: 7365
Joined: 7 Jul 2014, 9:45pm
Location: Near Bicester Oxon

Re: Huret Duopar; flawed genius at work?

Postby Bonefishblues » 14 Feb 2020, 7:44pm

They were - IIRC I bought mine after they were delisted? Or not, but I do remember it was very inexpensive.

PT1029
Posts: 726
Joined: 16 Apr 2012, 9:20pm

Re: Huret Duopar; flawed genius at work?

Postby PT1029 » 15 Feb 2020, 7:18am

As I recall, Duopar Eco £20, Titanium £50, but might have been £70? (which was a lot for bike parts back then).
As they gained a reputation for being a tad fragile, I replaced my Eco (with a Shimano Superplate) before a world tour.
That said, although I now run indexed, the Duopar the best shifting rear gear I have ever used.

Bonefishblues
Posts: 7365
Joined: 7 Jul 2014, 9:45pm
Location: Near Bicester Oxon

Re: Huret Duopar; flawed genius at work?

Postby Bonefishblues » 15 Feb 2020, 8:48am

£70 list rings a bell. Mine was, again IIRC about a third of that.

User avatar
speedsixdave
Posts: 692
Joined: 19 Apr 2007, 1:48pm
Location: Nottingham, UK

Re: Huret Duopar; flawed genius at work?

Postby speedsixdave » 15 Feb 2020, 10:00am

I had a Duopar Eco on my Raleigh Weekender in about 1986. It was a lovely bit of kit that I used to bang on proudly about (certainly compared to the standard Ecos on most of my friends' 1980s Raleighs), and worked well (inc LEJoG) until I lent the bike to a friend for some sort of charity ride. He put the mech into the spokes within ten miles and that was the end of that. The bike shop replaced it with a long-cage cheapish slant-parallelgram Shimano mech that handled 15 gears with quiet aplomb.

I sort of await Brucey's considered reappraisal of the very average Weinmann 730 single-pivot brakes that adorned my Weekender and most other cheap Raleighs of the era. The ubiquitous suicide levers certainly did not improve matters! We are, despite the general moaning, very fortunate with kit on cheap bikes nowadays.
Big wheels good, small wheels better.
Two saddles best!

Brucey
Posts: 37292
Joined: 4 Jan 2012, 6:25pm

Re: Huret Duopar; flawed genius at work?

Postby Brucey » 15 Feb 2020, 10:57am

speedsixdave wrote:...I sort of await Brucey's considered reappraisal of the very average Weinmann 730 single-pivot brakes that adorned my Weekender and most other cheap Raleighs of the era. The ubiquitous suicide levers certainly did not improve matters! We are, despite the general moaning, very fortunate with kit on cheap bikes nowadays.


They were always flexy, rather feeble brakes! Often the frame would be so 'gappy' that once better brakes and/or 700C rims were contemplated, a trip to the framebuilders seemed like the best option, to get some canti bosses affixed. Weinmann 750 centre pulls were an easier upgrade, but they too were not very powerful when used at full stretch. By comparison if clearances were smaller and 610 centre pulls could be fitted with the brake blocks mid slot, they were reasonably powerful.

The suicide levers can work and be slightly useful, but they more often do neither, and just rattle about uselessly.

cheers
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~

Brucey
Posts: 37292
Joined: 4 Jan 2012, 6:25pm

Re: Huret Duopar; flawed genius at work?

Postby Brucey » 15 Feb 2020, 4:21pm

Allvit derailleur adverts

Image
Image
from here https://restoringvintagebicycles.com/2020/02/02/huret-allvit-rear-derailleurs/

The early Allvit had two round pulleys; later examples had a toothed upper pulley and presumably worked better on chains with low side-plates (nearly all 'modern' chains have low side plates, starting with the Sedisport). Ball bearing pulleys were standard for most Allvits, and the clever parallelogram geometry gave the most controlled guide pulley gap then available, provided you used the right freewheel, that is.

One frequent criticism of Allvits is that the required cable tension to get the thing to move can be very high; this is most obviously the case when downshifting from top gear; there isn't much mechanical advantage to be had when the pinch bolt is in that position. However with modern cables and a few tweaks, the required tension can be acceptably low.

Another issue is that the pinch bolt is not free to articulate, or strain relieved in any way, and the angle change through the gear range Is quite large. This is likely to lead to cable failure near the pinch bolt, especially if the cable tension is high because the mech is seizing up. Even the most modest radius on a strain relief piece would probably have given the cable a much longer life, and would be easy enough to implement if you wanted to.

One unusual feature of an Allvit is that it is one of the few derailleurs where there is a simple (and strong) integral part of the mech that is likely to bear the brunt of the impact should the bike fall over. This means that the mech is likely to survive many forms of casual treatment, and if it does get bent, it can be straightened again.

The Allvit was manufactured for about twenty years. In that time it went from a top-of-the-line mech to a design which looked rather archaic, and was probably being made less well, to serve in a different part of the bicycle market.

Although it is (and was) heavy, quirky, flawed in execution, and can just as easily wear out as jam solid, for many years it was a better choice -for the money- than many other derailleurs, providing consistent shifting (on a typical 14-28 5s freewheel) with smooth-running jockey pulleys.

In many respects it works like modern rear mechs, whereby the bracket part of the mech stays in one position (although it will swing backwards when the wheel needs to come out), and the guide pulley tracks the block profile by virtue of the parallelogram geometry. In point of fact this approach might be inherently superior for many cassette/freewheel profiles, because it results in a curved path for the guide pulley rather than a straight one. This could be a better match for a 'normal' wide range cluster with sprocket sizes which increase geometrically.

I wonder if a 'modern Allvit' with (say) a carbon-fibre bash-guard/bracket, a better engineered parallelogram etc might be a good idea?

cheers
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~

drossall
Posts: 4700
Joined: 5 Jan 2007, 10:01pm
Location: North Hertfordshire

Re: Huret Duopar; flawed genius at work?

Postby drossall » 15 Feb 2020, 11:22pm

I started with a Simplex plastic gear. When it failed, I don't recall how, I got a Svelto instead. I don't remember much problem with it, but then at that point I had no experience of anything better. Then I upgraded to Shimano 600s when those were new on the scene.

User avatar
gaz
Posts: 13896
Joined: 9 Mar 2007, 12:09pm
Location: Kent, car park of England

Re: Huret Duopar; flawed genius at work?

Postby gaz » 16 Feb 2020, 1:04am

I had a couple of XT Super Plates BITD. Seemed pricey but probably cost me less than a M730 that was current at the time would have done.

Largely bought because they were there on the LBS shelf rather than my understanding (or even being aware of) its alleged advantages. First one ended up in the spokes on a stretch of roughstuff, second was still on the bike when it left my care.

Don't recall any issues for shifting on a 14-28T freewheel, don't recall it being any better than the long arm 105 'gold arrow' it replaced.
2020 : To redundancy ... and beyond!