Hirose Mini-Velo


The Autumn 2016 Bicycle Quarterly includes a photo feature about riding in the Hirose Owners’ Meeting. I really enjoy these events, because they combine amazing bikes with wonderful rides. Many Japanese custom bikes are incredibly elaborate and beautifully crafted, yet they are intended to be ridden.

What makes “Hirose watching” so much fun is that each bike is completely different. Some have Hirose’s custom rear derailleurs, which are based on the classic Cyclo “pizza cutter” derailleur – except they are 10-speed compatible. I have seen at least four completely different front derailleurs Hirose has made. One bike shown in the BQ article has Mafac cantilevers, with an extra pulley to double the mechanical advantage of each brake.

Whereas most builders will turn you away if your ideas are too crazy, Hirose-san will look at you for a while and say: “That is an interesting question. Let me see how I can solve it.” The amazing thing is that he really does solve it: All his bikes work great. They have nothing of the “not-quite-there prototype” quality that you often get with one-offs.


At the last Owners’ Meeting, I was especially fascinated by this Mini-Velo. Mini-Velos use smaller wheels to make them easier to portage on narrow mountain trails. They also are popular for city riding because they are especially nimble. This one looked simple at first, but a closer look showed that it was anything but. Click on the images for higher resolution.


On this Mini-Velo, all cables run inside the frame tubes (except the front brake, which would require a cable run too convoluted to work well).


In the photo above, you can see the cables for front and rear derailleur, as well as the rear brake, enter the frame.


Here you see the crossed-over seatstays, and the exit for the brake cable, which then runs through the seat tube to the rear brake. Clever – but there is another reason why Hirose used the crossed-over seatstays.

The rear derailleur cable also runs through the top tube. This avoids having to route it around the bottom bracket – the straighter cable run makes for better shifting. The crossed-over seatstays allow the cable to enter the stay without having to get around the seatpost. If you didn’t know the cable was in there, you would never guess. All cables run inside small tubes that connect entry and exit points, so replacing a cable is easy. But just imagine assembling it all as you braze the frame!


The shifter cable exits the seatstay – also with a straighter cable run than if it used the usual path along the chainstay. The shifting is superb, which isn’t always the case with internally routed cables.


With the crossed-over seatstays and the elegant brake cable routing through the seat tube, the rear brake must be on the front of the seatstays. Hirose-san prefers centerpull brakes, and for this bike with narrow tires, he used an old set of Mafac Competitions. But with the small wheels, the brake sits much lower than usual, and the angled stays are too far apart for the brake bosses.

The solution? A curved bridge that provides the mounting points for the brake pivots with the right spacing. The brake pad holders are custom-made, too – Hirose-san does not like the riveted Mafac originals (which can loosen – this is not a problem with the one-piece Compass brake shoe holders). So he machines his own posts that screw onto modern pad holders, so he can use them with classic centerpull and cantilever brakes.


The decaleur also is a fabrication tour de force. It attaches both to the front and the rear of the (custom-made) stem! This is necessary to make it stronger and more stable, since there is no rack to support the bag. There top part of the part that attaches to the bag doubles as a handle.


The elegant bag support doesn’t need triangulation, since the weight of the bag is suspended from the saddle.


Chrome-plated lugs and fork crown add beauty, but the bike doesn’t take itself too serious – how about the custom-made holder for a whimsical front light?


The reflector attaches to the pump, making it easy to remove if you don’t think you’ll need it.


A custom bottle cage…


… and a beautiful taillight provide the finishing touches to this amazing machine. And having seen it on the road, it appears that it rides as well as it looks. It’s truly a show-case of Hirose’s genius.

Further reading:

About Jan Heine, Editor, Bicycle Quarterly

Spirited rides that zig-zag across mountain ranges. Bicycle Quarterly magazine and its sister company, Compass Cycles, that turns our research into high-performance components for real-world riders.
This entry was posted in Rides. Bookmark the permalink.

28 Responses to Hirose Mini-Velo

  1. gasconha says:

    I love Hirose’s work that I discovered via his youtube channel. There is so much information and “secrets” given in those videos, including the manufacturing of he 10 speed cyclo type derailleur, which is both fascinating and scary to watch (who can afford a custom derailleur that is not even the best!?).
    BQ’s issues on framebuilder are trully inspiratinal. I am a mechanical geek and I do love the creativity that are but into framebuilding by people like Hirose-san.
    Jan: did you had chance to get the geometry of this mini-vélo? In issue 53 there is not information about it? Do you think Hirose-san had to increase the trail because of the smaller wheel?
    Thanks for this article.

    • I don’t really like to divulge what could be trade secrets – the Hirose Mini-Velo I rode for a day in the mountains truly surprised me, because it didn’t handle like all the other small-wheeled bikes I had ridden (Moulton, Bike Friday). It rode like a good low-trail 650B randonneur bike, and that is high praise indeed. I probably don’t give too much away when I say that the geometry most definitely is low-trail. Increasing trail won’t compensate for the lower rotational inertia of the smaller wheels, just increase wheel flop.

      • vectorcircle says:

        Refining the handling seems to be extra critical for bikes using 20″ wheels. I built my own version of a mini-velo (based around an older BMX frame) 6 or 7 years ago in large part because I didn’t like the design and handling of the small-wheeled bikes available in the U.S. I had to try several different forks until I hit on the right formula, but I was surprised that the bike turned out to be one of the best handling bikes I’ve ever owned.

        If more small-wheeled bikes were as well designed as the Hirose, I think they’d become more accepted, although I do think this particular one could use bigger tires…

      • The Hirose I rode had huge 44 mm-wide tires, and it rode very nicely. I’d be interested in finding out what geometry you settled on for your Mini-Velo.

      • vectorcircle says:

        My frame has a factory spec head angle of 74 degrees (not unusual in the BMX world), and I’ve ended up pairing it with an 18mm offset fork. I wouldn’t say the bike has an especially forgiving ride quality, but it is extremely nimble and sticks to curves like crazy (in other words, it’s fun).

        I do wish my frame had a longer rear triangle like the mini-velo above, as I suspect that design choice helps improve the overall stability of a small bicycle.

  2. Cory says:

    The taillight is made by Kimura who also does some front torch models as well as rear reflectors. Very nicely machined pieces !

  3. Stephen Bamford says:


  4. John Duval says:

    I see the front derailleur cable headed down into the head tube, but how does it get past the bottom bracket and around the pump? This bike does not look rinko friendly. It seems like the bigger frame (due to smaller wheels) would not be good for that either.

  5. Niels Hansen says:

    Is that crossed seatstay not a possible trademark/design infringement of Gary Turners “Triple Triangle” design?

    I don’t know much about copyright etc, just wondering.

    Br Niels

    • Crossed-over seatstays have been around for many decades. It’s a common way of attaching the rear triangle to the front. Any trademarks on this design date from at least 80 years ago and would long have expired.

      The “Hellenic” stays of Hetchins date from before the war. The attachment of the Hirose is very similar to many Jo Routens bikes from the 1940s onward, which also ran the brake cable through the seat tube. You can see several Routens bikes in our book The Competition Bicycle.

  6. Michael says:

    Does housing run inside the frame on those cables that run inside the frame?
    How do you change the housing/cable inside the frame when one breaks?
    Does the performance suffer with internal housing? I have heard mixed things.

    • Running housing inside the frame is always a bad idea. The Hirose has small-diameter tubes brazed into the frame for the cable. Changing the cable is simple – you just thread the cable into the entry until it comes out of the exit. Making a frame that way is a lot of work, but it works flawlessly.

      Some older Italian frames simply punched a hole for the entry and another for the exit. Either the cable housing ran all the way, usually with some sharp bends, or the bare cable had to find its way past the bottom bracket. Those did perform poorly…

      • Larry Leveen says:

        I’ve worked on several bikes with internal routing of just a bare cable. Not fun. I found that the type of stainless cables I stocked in my shop were ferromagnetic and the large “cow magnet” I kept on my bench really saved the day by guiding the cable out the tiny hole provided by the manufacturer. Ugh!

        As someone who has never built a frame or been around frame building operations, it would be interesting to see step-by-step pictures of how the small diameter tubes are brazed into place on/in the frame tubes. I don’t believe you have covered that in detail yet, Jan.

        Thanks for making me appreciate a mini-velo, a type of bike I honestly had disdain for until now. Broadening readers’ frontiers and helping people consider alternate approaches is a strength of your publication (and blog).

  7. Michael Arciero says:

    The decaleur is really interesting! I imagine that to release the bag you would pull up on the handle part to slide it out of the lower part, then slide the bag down to unhook the loop from the lower part. But I cant see any joint where tubing of the handle part slides trombone-style into the lower part, if this is how it works. Perhaps it is obscured by the bolt there?

    Am very curious about decaleur design. This is one area I would address with a custom constructeur build. Am interested in viable alternatives to the standard after-market options.

    • The bag part of the decaleur does slide into the tubes of the part that is mounted to the stem. There is a spring-loaded locking pin that prevents it from releasing accidentally. Many different decaleurs have been tried, but we keep coming back to the Herse/Singer design (which Hirose uses with his own modifications), because it works best.

  8. Dante Casali says:

    Any idea of the overall length? I’d love to find a bike that could fit in the back of a minivan upright but transversely, which is 4 feet wide. Production 20″ road bikes include Soma’s discontinued minivelo, but I don’t know how it rides.

    • Sorry, I didn’t measure the bike. I suspect the wheelbase is similar to big-wheeled bikes (just over 1 m), so add the wheels, and you are there.

    • Bill Gobie says:

      Any Moulton with a separable frame will fit. http://www.moultonbicycles.co.uk/models.html

      • I think he meant a bike that doesn’t need disassembly… Otherwise, any bike short of a recumbent will fit – just take out the wheels and make the rear fender a Rinko ones. In fact, most of the Rinko nuts we sell are for that purpose rather than for train/air travel.

    • Jimmy says:

      The Mercier Nano sise L is 56″ overall length. That’s dead straight, I suppose if you turned the bars to the side and had a bit of space diagonally it might fit in the back of a 48″ wide mini van.

      It’s interesting because in some ways, they are really not that much smaller than a conventional bicycle -I owned a Nano and it really didn’t make it that much easier to fit in the back of a car than a standard bike -still had to pull a wheel at least.

      However, in other situations, the smaller size was very noticeable. For example, dismounted walking the bike through a farmers market was MUCH easier and the bike less obtrusive. You wouldn’t think so just from the numbers, but in practice it made a big difference. Also easier to find a spot in a crowded house to park it.

      By the numbers given the Nano is low-trail. I didn’t check actual geometry but it sure felt that way riding it. Front end is pretty light (depending on how high you have the seat post) and it handled better with a small front load, IMO.

  9. Bill Gobie says:

    That is a well thought out bike.

    This bike has plenty of room above the front wheel and below the bars for a frame mounted front rack. Carrying the load on the frame instead of the steering assembly decouples the load from the steering. The load has little effect then on the steering. Certain Moultons use a frame mounted front carrier very successfully. See this picture and this one.

    20″ tires to try on this bike, in the Compass spirit of high volume and light construction, are Schwalbe Shreddas. I think those are the 20″ tires Compass would sell if Compass made 20″ tires.

    • Carrying the load on the frame instead of the steering assembly decouples the load from the steering.

      Decoupling the load from the steering isn’t always desirable. Any load at the front will affect the handling, because it affects the bike’s wheel flop. If you carry the load attached to the fork with a low-trail geometry, it stabilizes the steering due to its inertia. However, at very low speeds, you get a tendency for the handlebars to turn due to the weight and the inclination of the steering axis.

      With the load attached to the frame, you have better low-speed stability and less high-speed stability. That is why the bikes of the postal carriers in Europe had the load attached to the frame, but the newspaper carriers – the famous porteurs – attached the racks to the fork.

      • Bill Gobie says:

        I think we can agree that there is no thorough theoretical understanding of bicycle dynamics. (“Everybody knows how to ride a bike, but nobody knows how we ride bikes,” in this misleadingly titled article: http://www.nature.com/news/the-bicycle-problem-that-nearly-broke-mathematics-1.20281.) Rules of thumb have been worked out experimentally by groping around in the parameter space. With the AM series Dr Moulton came up with a design that is not highly sensitive to front loads up to 20-25 lbs. I’ve gone camping with a significant front load on my AM in the Cascades and Olympics, descending at up to 40 mph, with no qualms about the bike’s handling.

  10. Russ says:

    Would be interested to know how he changed the angles on the lugged bottom bracket shell

Comments are closed.