I'm re-gluing a loose magnet inside a Shimano DH-3N20 dynamo hub
and having trouble removing magnetic debris. It's in the form of
fine powder, mostly rust particles. They're too fine to pick out,
but there's enough to potentially interfere with getting the
magnet seated so it clears the armature.

I've tried using masking tape to "tack up" the particles, but it's
not very effective and won't reach into crevices. I don't have a
source of compressed air but am unsure it'd do more that push the
rubbish around.

Has anybody found a solution to this sort of problem? One thought
is something like Plasti-Dip (tool handle coating), but it has the
potential to make an even bigger mess if it won't peel off clean.

Thanks for reading,

bob prohaska

On Sunday, July 31, 2022 at 8:35:23 AM UTC-7, bob prohaska wrote:
> I'm re-gluing a loose magnet inside a Shimano DH-3N20 dynamo hub
> and having trouble removing magnetic debris. It's in the form of
> fine powder, mostly rust particles. They're too fine to pick out,
> but there's enough to potentially interfere with getting the
> magnet seated so it clears the armature.
>
> I've tried using masking tape to "tack up" the particles, but it's
> not very effective and won't reach into crevices. I don't have a
> source of compressed air but am unsure it'd do more that push the
> rubbish around.
>
> Has anybody found a solution to this sort of problem? One thought
> is something like Plasti-Dip (tool handle coating), but it has the
> potential to make an even bigger mess if it won't peel off clean.
>
> Thanks for reading,
>
> bob prohaska
Try Plumbers Putty which is a clay-like material that won't stick to the metal but will push into all of the crevices.

On 7/31/2022 11:35 AM, bob prohaska wrote:
> I'm re-gluing a loose magnet inside a Shimano DH-3N20 dynamo hub
> and having trouble removing magnetic debris. It's in the form of
> fine powder, mostly rust particles. They're too fine to pick out,
> but there's enough to potentially interfere with getting the
> magnet seated so it clears the armature.
>
> I've tried using masking tape to "tack up" the particles, but it's
> not very effective and won't reach into crevices. I don't have a
> source of compressed air but am unsure it'd do more that push the
> rubbish around.
>
> Has anybody found a solution to this sort of problem? One thought
> is something like Plasti-Dip (tool handle coating), but it has the
> potential to make an even bigger mess if it won't peel off clean.
>
> Thanks for reading,
>
> bob prohaska

Have you tried cotton swabs, AKA Q-tips? If they don't do the job dry,
perhaps moisten them with acetone. That would also help remove any oils
that might interfere with adhesive bonding later.

You might also take a sharp steel scribe, or even a steel nail ground
down to a point. Attach the strongest magnet you have to the shank. The
pointy end will be magnetized and let you pick up particles in the
corners. Maybe that will be sufficient. Then wipe up with an
acetone-moist Q-tip.

I've got a small collection of small magnets. Some I've bought, but I've
also disassembled old hard drives and saved their magnets. They're very
strong, and affixed to steel mounts that can act as handles. Without
those, I'd have trouble getting them off the furnace's sheet metal where
I store them.

About compressed air: Before I had a compressor, I put an adapter on an
old propane tank. I could fill it at a compressor (say, in the school's
lab) and take the tank of air home with me. It doesn't last long, but
sometimes it was handy.

--
- Frank Krygowski

Frank Krygowski <frkrygow@sbcglobal.net> wrote:
> On 7/31/2022 11:35 AM, bob prohaska wrote:
>> I'm re-gluing a loose magnet inside a Shimano DH-3N20 dynamo hub
>> and having trouble removing magnetic debris. It's in the form of
>> fine powder, mostly rust particles. They're too fine to pick out,
>> but there's enough to potentially interfere with getting the
>> magnet seated so it clears the armature.
>>
>
> You might also take a sharp steel scribe, or even a steel nail ground
> down to a point. Attach the strongest magnet you have to the shank. The
> pointy end will be magnetized and let you pick up particles in the
> corners. Maybe that will be sufficient.

I really wanted that idea to work, but no luck. Even with a good
disk seek magnet the debris stuck in the dynamo. I ended up dry
wiping the surfaces to be glued and crossed my fingers. So far it
looks good. The clearance inside the hub is bigger than I thought,
more than ten mils. Maybe it was much ado about nothing.

The hub runs smoothly and generates power, so I'll declare victory
and retreat for now.

Next time the hub comes apart I'll give the sticky putty method
a try. The debris seems to accumulate on the magnet faces and
armature poles. A little oily residue will do no harm there.

As an aside, has anybody tried re-magneting a shimano hub? The
factory magnet appears to be ferrite. It's reasonably strong,
but NeBFe bar magnets should be a big improvement. The size
would be about 30 mm long, 4 mm wide and no more than 3 mm thick.
A quick check on Amazon found nothing that shape, let alone with
the right field orientation.

Thanks to all who replied!

bob prohaska

On Monday, August 1, 2022 at 11:30:37 AM UTC-7, bob prohaska wrote:
> Frank Krygowski <frkr...@sbcglobal.net> wrote:
> > On 7/31/2022 11:35 AM, bob prohaska wrote:
> >> I'm re-gluing a loose magnet inside a Shimano DH-3N20 dynamo hub
> >> and having trouble removing magnetic debris. It's in the form of
> >> fine powder, mostly rust particles. They're too fine to pick out,
> >> but there's enough to potentially interfere with getting the
> >> magnet seated so it clears the armature.
> >>
> >
> > You might also take a sharp steel scribe, or even a steel nail ground
> > down to a point. Attach the strongest magnet you have to the shank. The
> > pointy end will be magnetized and let you pick up particles in the
> > corners. Maybe that will be sufficient.
> I really wanted that idea to work, but no luck. Even with a good
> disk seek magnet the debris stuck in the dynamo. I ended up dry
> wiping the surfaces to be glued and crossed my fingers. So far it
> looks good. The clearance inside the hub is bigger than I thought,
> more than ten mils. Maybe it was much ado about nothing.
>
> The hub runs smoothly and generates power, so I'll declare victory
> and retreat for now.
>
> Next time the hub comes apart I'll give the sticky putty method
> a try. The debris seems to accumulate on the magnet faces and
> armature poles. A little oily residue will do no harm there.
>
> As an aside, has anybody tried re-magneting a shimano hub? The
> factory magnet appears to be ferrite. It's reasonably strong,
> but NeBFe bar magnets should be a big improvement. The size
> would be about 30 mm long, 4 mm wide and no more than 3 mm thick.
> A quick check on Amazon found nothing that shape, let alone with
> the right field orientation.
>
> Thanks to all who replied!
>
> bob prohaska
I would be a bit careful with that since it is just as easy to end the magnetism by reverse polarization.

On 8/1/2022 2:30 PM, bob prohaska wrote:
> Frank Krygowski <frkrygow@sbcglobal.net> wrote:
>> On 7/31/2022 11:35 AM, bob prohaska wrote:
>>> I'm re-gluing a loose magnet inside a Shimano DH-3N20 dynamo hub
>>> and having trouble removing magnetic debris. It's in the form of
>>> fine powder, mostly rust particles. They're too fine to pick out,
>>> but there's enough to potentially interfere with getting the
>>> magnet seated so it clears the armature.
>>>
>>
>> You might also take a sharp steel scribe, or even a steel nail ground
>> down to a point. Attach the strongest magnet you have to the shank. The
>> pointy end will be magnetized and let you pick up particles in the
>> corners. Maybe that will be sufficient.
>
> I really wanted that idea to work, but no luck. Even with a good
> disk seek magnet the debris stuck in the dynamo. I ended up dry
> wiping the surfaces to be glued and crossed my fingers. So far it
> looks good. The clearance inside the hub is bigger than I thought,
> more than ten mils. Maybe it was much ado about nothing.
>
> The hub runs smoothly and generates power, so I'll declare victory
> and retreat for now.
>
> Next time the hub comes apart I'll give the sticky putty method
> a try. The debris seems to accumulate on the magnet faces and
> armature poles. A little oily residue will do no harm there.
>
> As an aside, has anybody tried re-magneting a shimano hub? The
> factory magnet appears to be ferrite. It's reasonably strong,
> but NeBFe bar magnets should be a big improvement. The size
> would be about 30 mm long, 4 mm wide and no more than 3 mm thick.
> A quick check on Amazon found nothing that shape, let alone with
> the right field orientation.

Taking things in sort of reverse order: No, I've never heard of anyone
re-magnetizing a Shimano hub dynamo. I remember Jobst had built a
fixture to do that with the old Sturmey-Archer Dynohubs. Those would
supposedly de-magnetize immediately if disassembled without a keeper. I
made my own keeper to work on mine.

I'm curious about how a fancier magnetic material would work. I've
thought about that for the old Dynohub, but it works fine for me as is,
since this bike needs little beyond a "be seen" light.

But modern bike dynamos are usually engineered fairly carefully to
self-regulate pretty well, putting out ~0.5 Amp and ~6V over a wide
range of speeds. I wonder if a stronger magnet might ruin that self
regulation.

I'm also curious about how your hub lost its magnet in the first place!

--
- Frank Krygowski

In rec.crafts.metalworking Frank Krygowski <frkrygow@sbcglobal.net> wrote:
> On 8/1/2022 2:30 PM, bob prohaska wrote:
>> Frank Krygowski <frkrygow@sbcglobal.net> wrote:
>>> On 7/31/2022 11:35 AM, bob prohaska wrote:
>>
>> As an aside, has anybody tried re-magneting a shimano hub? The
>> factory magnet appears to be ferrite. It's reasonably strong,
>> but NeBFe bar magnets should be a big improvement. The size
>> would be about 30 mm long, 4 mm wide and no more than 3 mm thick.
>> A quick check on Amazon found nothing that shape, let alone with
>> the right field orientation.
>
>
> Taking things in sort of reverse order: No, I've never heard of anyone
> re-magnetizing a Shimano hub dynamo. I remember Jobst had built a
> fixture to do that with the old Sturmey-Archer Dynohubs. Those would
> supposedly de-magnetize immediately if disassembled without a keeper. I
> made my own keeper to work on mine.
>

Sorry, I was writing in the sense of replacing the original ferrite
magnets with neodymium-boron-iron. Trying to re-magnetize an existing
magnet is an ambitious project unless the geometry is very simple.

> I'm curious about how a fancier magnetic material would work. I've
> thought about that for the old Dynohub, but it works fine for me as is,
> since this bike needs little beyond a "be seen" light.

It'd probably have to be done with some combination of soft iron field
parts and very small NdBFe insert magnets. Maybe possible, but not
likely worth it. An original Dynohub is something of a treasure.
>
> But modern bike dynamos are usually engineered fairly carefully to
> self-regulate pretty well, putting out ~0.5 Amp and ~6V over a wide
> range of speeds. I wonder if a stronger magnet might ruin that self
> regulation.
>
Increasing magnet strength mostly increases voltage at low speed.
The regulation comes from leakage inductance, magnetic flux in the
armature that misses the field magnet. At high speed that leakage
inductance is what limits the current. Think of an AC power supply
with voltage equal to a constant times frequency. Driving an inductor,
the current that flows will be proportional to the ratio of that
constant to inductance. The ratio isn't dependent on frequency.

The regulating point will shift up. By how much is hard to estimate.
NdBFe is about double the strenth of ferrite. That'd be the upper
limit. It's no good for a matched load like a tungsten filament lamp.
It's more useful for charging a battery. Ideally an MPPT circuit
would make the best use of the added power. I think doubling the
power for a given speed is plausible if one doesn't try to use
the extra current.

> I'm also curious about how your hub lost its magnet in the first place!

To be clear, the magnet didn't lose its magnetization.

The magnet came loose from the field ring, apparently because
enough moisture got in the hub to cause rust. The swelling rust
broke the glue holding the magnet and at some point the magnet
starting moving and making noise.

Thanks for reading!

bob prohaska

On 8/2/2022 12:06 AM, bob prohaska wrote:
>
> The magnet came loose from the field ring, apparently because
> enough moisture got in the hub to cause rust. The swelling rust
> broke the glue holding the magnet and at some point the magnet
> starting moving and making noise.

Got it. That's a failure mode I wouldn't have guessed.

--
- Frank Krygowski

On 8/2/2022 12:04 PM, Frank Krygowski wrote:
> On 8/2/2022 12:06 AM, bob prohaska wrote:
>>
>> The magnet came loose from the field ring, apparently because
>> enough moisture got in the hub to cause rust. The swelling
>> rust
>> broke the glue holding the magnet and at some point the
>> magnet
>> starting moving and making noise.
>
> Got it. That's a failure mode I wouldn't have guessed.
>
>

Don't underestimate entropy:
http://www.yellowjersey.org/us70a1.jpg

--
Andrew Muzi
<www.yellowjersey.org/>
Open every day since 1 April, 1971

AMuzi <am@yellowjersey.org> wrote:
> On 8/2/2022 12:04 PM, Frank Krygowski wrote:
>> On 8/2/2022 12:06 AM, bob prohaska wrote:
>>>
>>> The magnet came loose from the field ring, apparently because
>>> enough moisture got in the hub to cause rust. The swelling
>>> rust
>>> broke the glue holding the magnet and at some point the
>>> magnet
>>> starting moving and making noise.
>>
>> Got it. That's a failure mode I wouldn't have guessed.
>>
>>
>
> Don't underestimate entropy:
> http://www.yellowjersey.org/us70a1.jpg
>

I'll admit to being in Frank's camp, at least initially.

Asked to check out the bike when it was for sale, I thought
the funny little noise couldn't _possibly_ be coming from
the dynohub. After purchase, the noise persisted and was
eventually traced to the hub. Opened up, rust was visible on
the armature. I thought the rust was rubbing the magnets.

Cleaned up, the hub was quiet for maybe two years. Then the
noise returned as a clicking sound. A loose magnet made the
most sense and sure enough, there was a loose magnet.

It's unclear how the hub got damp enough to rust as it did.
The shaft seals look reasonably good and the rest of the bike
looked well cared for. Certainly not left out in the rain.
(granted, it doesn't rain that much around here).

Much as I like dynamo hubs, this has been a cautionary tale.
Trouble with these things is either very expensive or very
tedious to fix.

I wonder if somebody could be convinced to start making
Soubitez roller dyamos in a sidewall-drive configuration.
I used one for years and it was magnificent. Still have it,
actually, but the plastic yoke that holds the roller has
cracked through and repair looks difficult.

Thanks to all for reading and guiding!

bob prohaska

On 8/2/2022 10:57 PM, bob prohaska wrote:
> AMuzi <am@yellowjersey.org> wrote:
>> On 8/2/2022 12:04 PM, Frank Krygowski wrote:
>>> On 8/2/2022 12:06 AM, bob prohaska wrote:
>>>>
>>>> The magnet came loose from the field ring, apparently because
>>>> enough moisture got in the hub to cause rust. The swelling
>>>> rust
>>>> broke the glue holding the magnet and at some point the
>>>> magnet
>>>> starting moving and making noise.
>>>
>>> Got it. That's a failure mode I wouldn't have guessed.
>>>
>>>
>>
>> Don't underestimate entropy:
>> http://www.yellowjersey.org/us70a1.jpg
>>
>
> I'll admit to being in Frank's camp, at least initially.
>
> Asked to check out the bike when it was for sale, I thought
> the funny little noise couldn't _possibly_ be coming from
> the dynohub. After purchase, the noise persisted and was
> eventually traced to the hub. Opened up, rust was visible on
> the armature. I thought the rust was rubbing the magnets.
>
> Cleaned up, the hub was quiet for maybe two years. Then the
> noise returned as a clicking sound. A loose magnet made the
> most sense and sure enough, there was a loose magnet.
>
> It's unclear how the hub got damp enough to rust as it did.
> The shaft seals look reasonably good and the rest of the bike
> looked well cared for. Certainly not left out in the rain.
> (granted, it doesn't rain that much around here).
>
> Much as I like dynamo hubs, this has been a cautionary tale.
> Trouble with these things is either very expensive or very
> tedious to fix.
>
> I wonder if somebody could be convinced to start making
> Soubitez roller dyamos in a sidewall-drive configuration.
> I used one for years and it was magnificent. Still have it,
> actually, but the plastic yoke that holds the roller has
> cracked through and repair looks difficult.
>
> Thanks to all for reading and guiding!
>
> bob prohaska
>
>
>

Panasonic made a very nice rim drive dynamo in the 1980s for
use with MTB where tire drive isn't reasonable. Large
diameter like the under-BB models. Can't find an image but
looked much like this:

https://i.pinimg.com/736x/07/9f/94/079f947e52bcaa3687b35a2baea5fc25.jpg

--
Andrew Muzi
<www.yellowjersey.org/>
Open every day since 1 April, 1971

On Tuesday, August 2, 2022 at 8:57:53 PM UTC-7, bob prohaska wrote:
> AMuzi <a...@yellowjersey.org> wrote:
> > On 8/2/2022 12:04 PM, Frank Krygowski wrote:
> >> On 8/2/2022 12:06 AM, bob prohaska wrote:
> >>>
> >>> The magnet came loose from the field ring, apparently because
> >>> enough moisture got in the hub to cause rust. The swelling
> >>> rust
> >>> broke the glue holding the magnet and at some point the
> >>> magnet
> >>> starting moving and making noise.
> >>
> >> Got it. That's a failure mode I wouldn't have guessed.
> >>
> >>
> >
> > Don't underestimate entropy:
> > http://www.yellowjersey.org/us70a1.jpg
> >
> I'll admit to being in Frank's camp, at least initially.
>
> Asked to check out the bike when it was for sale, I thought
> the funny little noise couldn't _possibly_ be coming from
> the dynohub. After purchase, the noise persisted and was
> eventually traced to the hub. Opened up, rust was visible on
> the armature. I thought the rust was rubbing the magnets.
>
> Cleaned up, the hub was quiet for maybe two years. Then the
> noise returned as a clicking sound. A loose magnet made the
> most sense and sure enough, there was a loose magnet.
>
> It's unclear how the hub got damp enough to rust as it did.
> The shaft seals look reasonably good and the rest of the bike
> looked well cared for. Certainly not left out in the rain.
> (granted, it doesn't rain that much around here).
>
> Much as I like dynamo hubs, this has been a cautionary tale.
> Trouble with these things is either very expensive or very
> tedious to fix.
>
> I wonder if somebody could be convinced to start making
> Soubitez roller dyamos in a sidewall-drive configuration.
> I used one for years and it was magnificent. Still have it,
> actually, but the plastic yoke that holds the roller has
> cracked through and repair looks difficult.
>
> Thanks to all for reading and guiding!
>
> bob prohaska
When the dynamo is engaged and generating power it heats up. When you stop and it cools down it pulls in air past any seals. If the air is humid it will condense inside the generator case. This heating and cooling also eventually destroys the attachments to the dynamo interior. This would be bypassed in newer dynamos by a number of methods such as making the capacity of the dynamo such that it never heats up under load as the smaller one do. Also we now have flexible attaching compounds that do not harden or become brittle over time. Technology marches on.

On Wednesday, August 3, 2022 at 6:57:09 AM UTC-7, AMuzi wrote:
> On 8/2/2022 10:57 PM, bob prohaska wrote:
> > AMuzi <a...@yellowjersey.org> wrote:
> >> On 8/2/2022 12:04 PM, Frank Krygowski wrote:
> >>> On 8/2/2022 12:06 AM, bob prohaska wrote:
> >>>>
> >>>> The magnet came loose from the field ring, apparently because
> >>>> enough moisture got in the hub to cause rust. The swelling
> >>>> rust
> >>>> broke the glue holding the magnet and at some point the
> >>>> magnet
> >>>> starting moving and making noise.
> >>>
> >>> Got it. That's a failure mode I wouldn't have guessed.
> >>>
> >>>
> >>
> >> Don't underestimate entropy:
> >> http://www.yellowjersey.org/us70a1.jpg
> >>
> >
> > I'll admit to being in Frank's camp, at least initially.
> >
> > Asked to check out the bike when it was for sale, I thought
> > the funny little noise couldn't _possibly_ be coming from
> > the dynohub. After purchase, the noise persisted and was
> > eventually traced to the hub. Opened up, rust was visible on
> > the armature. I thought the rust was rubbing the magnets.
> >
> > Cleaned up, the hub was quiet for maybe two years. Then the
> > noise returned as a clicking sound. A loose magnet made the
> > most sense and sure enough, there was a loose magnet.
> >
> > It's unclear how the hub got damp enough to rust as it did.
> > The shaft seals look reasonably good and the rest of the bike
> > looked well cared for. Certainly not left out in the rain.
> > (granted, it doesn't rain that much around here).
> >
> > Much as I like dynamo hubs, this has been a cautionary tale.
> > Trouble with these things is either very expensive or very
> > tedious to fix.
> >
> > I wonder if somebody could be convinced to start making
> > Soubitez roller dyamos in a sidewall-drive configuration.
> > I used one for years and it was magnificent. Still have it,
> > actually, but the plastic yoke that holds the roller has
> > cracked through and repair looks difficult.
> >
> > Thanks to all for reading and guiding!
> >
> > bob prohaska
> >
> >
> >
> Panasonic made a very nice rim drive dynamo in the 1980s for
> use with MTB where tire drive isn't reasonable. Large
> diameter like the under-BB models. Can't find an image but
> looked much like this:
>
> https://i.pinimg.com/736x/07/9f/94/079f947e52bcaa3687b35a2baea5fc25.jpg
> --
> Andrew Muzi
> <www.yellowjersey.org/>
> Open every day since 1 April, 1971
Since you can lace up a hub generator, don't you think that it would be a better idea?

On 8/3/2022 9:17 AM, Tom Kunich wrote:
> On Wednesday, August 3, 2022 at 6:57:09 AM UTC-7, AMuzi wrote:
>> On 8/2/2022 10:57 PM, bob prohaska wrote:
>>> AMuzi <a...@yellowjersey.org> wrote:
>>>> On 8/2/2022 12:04 PM, Frank Krygowski wrote:
>>>>> On 8/2/2022 12:06 AM, bob prohaska wrote:
>>>>>>
>>>>>> The magnet came loose from the field ring, apparently because
>>>>>> enough moisture got in the hub to cause rust. The swelling
>>>>>> rust
>>>>>> broke the glue holding the magnet and at some point the
>>>>>> magnet
>>>>>> starting moving and making noise.
>>>>>
>>>>> Got it. That's a failure mode I wouldn't have guessed.
>>>>>
>>>>>
>>>>
>>>> Don't underestimate entropy:
>>>> http://www.yellowjersey.org/us70a1.jpg
>>>>
>>>
>>> I'll admit to being in Frank's camp, at least initially.
>>>
>>> Asked to check out the bike when it was for sale, I thought
>>> the funny little noise couldn't _possibly_ be coming from
>>> the dynohub. After purchase, the noise persisted and was
>>> eventually traced to the hub. Opened up, rust was visible on
>>> the armature. I thought the rust was rubbing the magnets.
>>>
>>> Cleaned up, the hub was quiet for maybe two years. Then the
>>> noise returned as a clicking sound. A loose magnet made the
>>> most sense and sure enough, there was a loose magnet.
>>>
>>> It's unclear how the hub got damp enough to rust as it did.
>>> The shaft seals look reasonably good and the rest of the bike
>>> looked well cared for. Certainly not left out in the rain.
>>> (granted, it doesn't rain that much around here).
>>>
>>> Much as I like dynamo hubs, this has been a cautionary tale.
>>> Trouble with these things is either very expensive or very
>>> tedious to fix.
>>>
>>> I wonder if somebody could be convinced to start making
>>> Soubitez roller dyamos in a sidewall-drive configuration.
>>> I used one for years and it was magnificent. Still have it,
>>> actually, but the plastic yoke that holds the roller has
>>> cracked through and repair looks difficult.
>>>
>>> Thanks to all for reading and guiding!
>>>
>>> bob prohaska
>>>
>>>
>>>
>> Panasonic made a very nice rim drive dynamo in the 1980s for
>> use with MTB where tire drive isn't reasonable. Large
>> diameter like the under-BB models. Can't find an image but
>> looked much like this:
>>
>> https://i.pinimg.com/736x/07/9f/94/079f947e52bcaa3687b35a2baea5fc25.jpg

> Since you can lace up a hub generator, don't you think that it would be a better idea?
>

I own and use one, I sell them and build wheels around them.
Nice product series and modern units are much better than
classic ones in many ways.

That said, a tire or rim drive dynamo is less expensive,
lighter weight and fully disengaged when not in use. People
make decisions based on their own interpretation of various
criteria, which is fine.

--
Andrew Muzi
<www.yellowjersey.org/>
Open every day since 1 April, 1971

On 8/3/2022 10:28 AM, AMuzi wrote:
> On 8/3/2022 9:17 AM, Tom Kunich wrote:
>> On Wednesday, August 3, 2022 at 6:57:09 AM UTC-7, AMuzi wrote:
>>> On 8/2/2022 10:57 PM, bob prohaska wrote:
>>>> AMuzi <a...@yellowjersey.org> wrote:
>>>>> On 8/2/2022 12:04 PM, Frank Krygowski wrote:
>>>>>> On 8/2/2022 12:06 AM, bob prohaska wrote:
>>>>>>>
>>>>>>> The magnet came loose from the field ring, apparently because
>>>>>>> enough moisture got in the hub to cause rust. The swelling
>>>>>>> rust
>>>>>>> broke the glue holding the magnet and at some point the
>>>>>>> magnet
>>>>>>> starting moving and making noise.
>>>>>>
>>>>>> Got it. That's a failure mode I wouldn't have guessed.
>>>>>>
>>>>>>
>>>>>
>>>>> Don't underestimate entropy:
>>>>> http://www.yellowjersey.org/us70a1.jpg
>>>>>
>>>>
>>>> I'll admit to being in Frank's camp, at least initially.
>>>>
>>>> Asked to check out the bike when it was for sale, I thought
>>>> the funny little noise couldn't _possibly_ be coming from
>>>> the dynohub. After purchase, the noise persisted and was
>>>> eventually traced to the hub. Opened up, rust was visible on
>>>> the armature. I thought the rust was rubbing the magnets.
>>>>
>>>> Cleaned up, the hub was quiet for maybe two years. Then the
>>>> noise returned as a clicking sound. A loose magnet made the
>>>> most sense and sure enough, there was a loose magnet.
>>>>
>>>> It's unclear how the hub got damp enough to rust as it did.
>>>> The shaft seals look reasonably good and the rest of the bike
>>>> looked well cared for. Certainly not left out in the rain.
>>>> (granted, it doesn't rain that much around here).
>>>>
>>>> Much as I like dynamo hubs, this has been a cautionary tale.
>>>> Trouble with these things is either very expensive or very
>>>> tedious to fix.
>>>>
>>>> I wonder if somebody could be convinced to start making
>>>> Soubitez roller dyamos in a sidewall-drive configuration.
>>>> I used one for years and it was magnificent. Still have it,
>>>> actually, but the plastic yoke that holds the roller has
>>>> cracked through and repair looks difficult.
>>>>
>>>> Thanks to all for reading and guiding!
>>>>
>>>> bob prohaska
>>>>
>>>>
>>>>
>>> Panasonic made a very nice rim drive dynamo in the 1980s for
>>> use with MTB where tire drive isn't reasonable. Large
>>> diameter like the under-BB models. Can't find an image but
>>> looked much like this:
>>>
>>> https://i.pinimg.com/736x/07/9f/94/079f947e52bcaa3687b35a2baea5fc25.jpg
>
>
>> Since you can lace up a hub generator, don't you think that it would
>> be a better idea?
>>
>
> I own and use one, I sell them and build wheels around them. Nice
> product series and modern units are much better than classic ones in
> many ways.
>
> That said, a tire or rim drive dynamo is less expensive, lighter weight
> and fully disengaged when not in use. People make decisions based on
> their own interpretation of various criteria, which is fine.

Agreed, each style has its benefits and detriments. Our bikes have two
hub dynos, three roller dynos and three bottle dynos, IIRC. All work
well for our uses.

--
- Frank Krygowski

Am Wed, 3 Aug 2022 12:52:56 -0400 schrieb Frank Krygowski
<frkrygow@sbcglobal.net>:

>On 8/3/2022 10:28 AM, AMuzi wrote:

[about hub generators]

>> I own and use one, I sell them and build wheels around them. Nice
>> product series and modern units are much better than classic ones in
>> many ways.
>>
>> That said, a tire or rim drive dynamo is less expensive, lighter weight
>> and fully disengaged when not in use. People make decisions based on
>> their own interpretation of various criteria, which is fine.

There is still the erroneous opinion that hub generators consume a lot
of electricity even when they are switched off, because it takes effort
to turn the lifted wheel slowly, caused by what we call "pole
sensitivity" (Polfühligkeit) in German.

>Agreed, each style has its benefits and detriments. Our bikes have two
>hub dynos, three roller dynos and three bottle dynos, IIRC. All work
>well for our uses.

I've replaced all of these by hub generators, long ago, on our bicycles,
when these became available. Far to often I was stranded without
working lights on my commute, because of a failing generator. I've
ruined two roller generators and more bottle generators than I'm able to
remember, over the last ~45 years.

When tiny and cheap microcontrollers became available, I used one to
implement an automatic switch, which sensed pulses from a bottle
generator, switching the lamps to a battery when the generator failed,
mostly when snow or water made the tire wet and slippy.

<https://www.mystrobl.de/ws/pic/standlicht/standle.htm>

Four years earlier I had implemented a related design for my own bike (a
Peugeot PR 60/L bougth in 1978), which switched a commercial single lead
acid battery between two lamps (high beam and low beam, so to speak).
That battery sat on a socket with an integrated relay, which connected
the bottle dynamo to the low beam lamp, when removed. This construct was
required by law at that time.

All because the bottle generator was unrealiable as hell.

<https://www.mystrobl.de/Plone/radfahren/technik/komponenten/licht/lichtc1.jpg/image_preview>

Translated from a German language posting in the German group
de.rec.fahrrad in '95

Light computer with Basic stamp

I recently threw away the components of the craft described here, but am
documenting the idea and execution here so I might try again sometime
with current material. At the end of 2007, 12 years later, a product
with the desired performance characteristics is still not on the market,
although components are now available (hub dynamos, more powerful
microcontrollers, lithium-ion batteries, gold caps, better FETs) with
which cheaper and easier to build better solutions.

<snip>

Times have changed, both hub dynos, battery lights and led lamps became
reliable and cheap enough to replace halogen bulbs and bottle
generators.

I wouldn't mind retaining an old bottle dyno as a "Schutzmanntröster"
(policeman comforter?) on an oldtimer used mostly by daylight.
--
Bicycle helmets are the Bach flower remedies of traffic

On 8/4/2022 9:36 AM, Wolfgang Strobl wrote:
> Am Wed, 3 Aug 2022 12:52:56 -0400 schrieb Frank Krygowski
> <frkrygow@sbcglobal.net>:
>
>> On 8/3/2022 10:28 AM, AMuzi wrote:
>
> [about hub generators]
>
>>> I own and use one, I sell them and build wheels around them. Nice
>>> product series and modern units are much better than classic ones in
>>> many ways.
>>>
>>> That said, a tire or rim drive dynamo is less expensive, lighter weight
>>> and fully disengaged when not in use. People make decisions based on
>>> their own interpretation of various criteria, which is fine.
>
> There is still the erroneous opinion that hub generators consume a lot
> of electricity even when they are switched off, because it takes effort
> to turn the lifted wheel slowly, caused by what we call "pole
> sensitivity" (Polfühligkeit) in German.
>
>> Agreed, each style has its benefits and detriments. Our bikes have two
>> hub dynos, three roller dynos and three bottle dynos, IIRC. All work
>> well for our uses.
>
> I've replaced all of these by hub generators, long ago, on our bicycles,
> when these became available. Far to often I was stranded without
> working lights on my commute, because of a failing generator. I've
> ruined two roller generators and more bottle generators than I'm able to
> remember, over the last ~45 years.

I don't do nearly as much night riding as I used to. For one thing, I'm
retired, so never have to ride home from work at night. For another
thing, my best current night riding friend has contracted a serious
health problem and we haven't ridden together for many months.

But my roller and bottle dynamos have been very trouble free -
understanding that I only rarely rode at night in the rain. The problems
I've had with dyno systems have been contained in the headlamps: before
LEDs, I had a halogen bulb whose center contact got eroded by vibration,
giving intermittent contact. I had one very inexpensive B&M Lyt headlamp
whose switch needed work. I had another off-brand LED lamp whose LED
failed. (I replaced it, which was more trouble than it was worth.) A
couple years ago I had a weird intermittent problem with a B&M Eyc. I
never diagnosed it, because it cured itself.

I've worn out bearings on a couple dynos and had minor problems with
their mounts, but I can't recall other problems. But I don't doubt you
have much more experience riding at night than I do.

--
- Frank Krygowski

On 8/4/2022 8:36 AM, Wolfgang Strobl wrote:
> Am Wed, 3 Aug 2022 12:52:56 -0400 schrieb Frank Krygowski
> <frkrygow@sbcglobal.net>:
>
>> On 8/3/2022 10:28 AM, AMuzi wrote:
>
> [about hub generators]
>
>>> I own and use one, I sell them and build wheels around them. Nice
>>> product series and modern units are much better than classic ones in
>>> many ways.
>>>
>>> That said, a tire or rim drive dynamo is less expensive, lighter weight
>>> and fully disengaged when not in use. People make decisions based on
>>> their own interpretation of various criteria, which is fine.
>
> There is still the erroneous opinion that hub generators consume a lot
> of electricity even when they are switched off, because it takes effort
> to turn the lifted wheel slowly, caused by what we call "pole
> sensitivity" (Polfühligkeit) in German.
>
>> Agreed, each style has its benefits and detriments. Our bikes have two
>> hub dynos, three roller dynos and three bottle dynos, IIRC. All work
>> well for our uses.
>
> I've replaced all of these by hub generators, long ago, on our bicycles,
> when these became available. Far to often I was stranded without
> working lights on my commute, because of a failing generator. I've
> ruined two roller generators and more bottle generators than I'm able to
> remember, over the last ~45 years.
>
> When tiny and cheap microcontrollers became available, I used one to
> implement an automatic switch, which sensed pulses from a bottle
> generator, switching the lamps to a battery when the generator failed,
> mostly when snow or water made the tire wet and slippy.
>
> <https://www.mystrobl.de/ws/pic/standlicht/standle.htm>
>
> Four years earlier I had implemented a related design for my own bike (a
> Peugeot PR 60/L bougth in 1978), which switched a commercial single lead
> acid battery between two lamps (high beam and low beam, so to speak).
> That battery sat on a socket with an integrated relay, which connected
> the bottle dynamo to the low beam lamp, when removed. This construct was
> required by law at that time.
>
> All because the bottle generator was unrealiable as hell.
>
> <https://www.mystrobl.de/Plone/radfahren/technik/komponenten/licht/lichtc1.jpg/image_preview>
>
>
> Translated from a German language posting in the German group
> de.rec.fahrrad in '95
>
> Light computer with Basic stamp
>
> I recently threw away the components of the craft described here, but am
> documenting the idea and execution here so I might try again sometime
> with current material. At the end of 2007, 12 years later, a product
> with the desired performance characteristics is still not on the market,
> although components are now available (hub dynamos, more powerful
> microcontrollers, lithium-ion batteries, gold caps, better FETs) with
> which cheaper and easier to build better solutions.
>
> <snip>
>
> Times have changed, both hub dynos, battery lights and led lamps became
> reliable and cheap enough to replace halogen bulbs and bottle
> generators.
>
> I wouldn't mind retaining an old bottle dyno as a "Schutzmanntröster"
> (policeman comforter?) on an oldtimer used mostly by daylight.
>

Yes, that was once an actual product:
http://www.yellowjersey.org/soustan1.jpg
http://www.yellowjersey.org/soustan2.jpg

Despite various complaints about tire drive dynamos, these
never sold well.
--
Andrew Muzi
<www.yellowjersey.org/>
Open every day since 1 April, 1971

AMuzi <am@yellowjersey.org> wrote:
>
> Yes, that was once an actual product:
> http://www.yellowjersey.org/soustan1.jpg
> http://www.yellowjersey.org/soustan2.jpg
>

The pictures are completely baffling to me.
What does it do, and how?

Thanks for writing!

bob prohaska

On Thursday, August 4, 2022 at 2:51:25 PM UTC-7, bob prohaska wrote:
> AMuzi <a...@yellowjersey.org> wrote:
> >
> > Yes, that was once an actual product:
> > http://www.yellowjersey.org/soustan1.jpg
> > http://www.yellowjersey.org/soustan2.jpg
> >
> The pictures are completely baffling to me.
> What does it do, and how?
>
> Thanks for writing!
>
> bob prohaska
It appears to, be a weight to hold a lamp horizontal to the ground using a counterweight. The roads around here are simply too rough for it to work as designed.

On 8/4/2022 4:51 PM, bob prohaska wrote:
> AMuzi <am@yellowjersey.org> wrote:
>>
>> Yes, that was once an actual product:
>> http://www.yellowjersey.org/soustan1.jpg
>> http://www.yellowjersey.org/soustan2.jpg
>>
>
> The pictures are completely baffling to me.
> What does it do, and how?
>
> Thanks for writing!
>
> bob prohaska
>

Well, image #2 does have two popular languages.

It mounts between dynamo and lamp(s). If there's no AC
current, the lamps light by the battery inside the round
case. Otherwise known as a 'steady lamp' system. Modern
units use a capacitor to light an LED when stopped.

--
Andrew Muzi
<www.yellowjersey.org/>
Open every day since 1 April, 1971

On 8/4/2022 6:26 PM, Tom Kunich wrote:
> On Thursday, August 4, 2022 at 2:51:25 PM UTC-7, bob prohaska wrote:
>> AMuzi <a...@yellowjersey.org> wrote:
>>>
>>> Yes, that was once an actual product:
>>> http://www.yellowjersey.org/soustan1.jpg
>>> http://www.yellowjersey.org/soustan2.jpg
>>>
>> The pictures are completely baffling to me.
>> What does it do, and how?
>>
>> Thanks for writing!
>>
>> bob prohaska
> It appears to, be a weight to hold a lamp horizontal to the ground using a counterweight.

:-) So unintentionally funny!

--
- Frank Krygowski

AMuzi <am@yellowjersey.org> wrote:
> On 8/4/2022 4:51 PM, bob prohaska wrote:
>> AMuzi <am@yellowjersey.org> wrote:
>>>
>>> Yes, that was once an actual product:
>>> http://www.yellowjersey.org/soustan1.jpg
>>> http://www.yellowjersey.org/soustan2.jpg
>>>
>>
>> The pictures are completely baffling to me.
>> What does it do, and how?
>>
>> Thanks for writing!
>>
>> bob prohaska
>>
>
> Well, image #2 does have two popular languages.
>
> It mounts between dynamo and lamp(s). If there's no AC
> current, the lamps light by the battery inside the round
> case. Otherwise known as a 'steady lamp' system. Modern
> units use a capacitor to light an LED when stopped.
>
I'll admit up front that I missed the English fine print 8-)

Now I'm a little less confused. The pictures give no
sense of volume. But the descriptions imply batteries,
some electronics and a relay inside. When was it offered
for sale?

bob prohaska

Am Thu, 04 Aug 2022 11:54:56 -0500 schrieb AMuzi <am@yellowjersey.org>:

>On 8/4/2022 8:36 AM, Wolfgang Strobl wrote:

.....

>> Times have changed, both hub dynos, battery lights and led lamps became
>> reliable and cheap enough to replace halogen bulbs and bottle
>> generators.
>>
>> I wouldn't mind retaining an old bottle dyno as a "Schutzmanntroester"
>> (policeman comforter?) on an oldtimer used mostly by daylight.
>>

>
>Yes, that was once an actual product:
>http://www.yellowjersey.org/soustan1.jpg
>http://www.yellowjersey.org/soustan2.jpg

Actually not. Asista may have copied the design from Soubitez, though.
Thanks for the link!

I got both the idea and the case from a product made by Asista. Have a
look at this picture and my article again.

My text
<https://www.mystrobl.de/ws/pic/standlicht/standle.htm> mentions Asista

"Some years ago, I bought some steady-lights made by Asista, for my
childrens bicyles."

and the picture of the device shows it too.

<https://www.mystrobl.de/Plone/radfahren/technik/komponenten/licht/lichtc1.jpg/image_preview>

Actually, I originally wrote that on tripod, in 1998.

<https://web.archive.org/web/20000409002615/http://members.tripod.de/WStrobl/asista/standl.htm>

:-)

From that article:

<snip>

Some years ago, I bought some steady-lights made by Asista, for my
childrens bicyles. I paid about DM 15, (<$10), initially, which is
comparatively inexpensive, for such a device. It's built into simple
plastic box of about 9 cm diameter, with room enough for 4 AA sized
cells, and a 3x5cm printed circuit bord (PCB) located below the
batteries. The box has to be mounted on the generators socket, below the
generator. A wire coming out of the box has to be wired to the
generator, and the wire normally connected to the generator goes to a
simple knurled nut on the back of the box (not visible in the picture to
the right). The ground connection goes through the socket.

The device switches the lamp between the generator and the enclosed
batteries. As long as there is some input from the generator, it just
lets the generator feed the lamp. If the generator stops, it supplies
current from the batteries for about 10 to 15 seconds. If that's too
short (it usually is), the cycle can be restarted by just moving the
bicycle a little bit.

The advantages are: the device is cheap, very easily mountable, it uses
standard cells and has no switches to be operated. The disadvantage
(besides the relatively primitive design of the box and the connectors)
are: the timing isn't adjustable, and, more important: it doesn't give
any light when the bicycle is slowly moving, and the generator isn't
delivering enough current to light the lamps.

Lately, my children complained about the steady-light "eating"
batteries. In one case, it emptied a set of batteries in less than a
week. Instead of trying to buy a replacement, I decided to build my very
own version of the circuit for switching between generator and
batteries, this time. Given that I only barely managed to get a second
Asista steady-light for my other son, when he started to cycle to school
(Asista obviously doesn't build or sell these lights anymore, or perhaps
the local dealers don't carry them, anymore), I didn't have much of a
choice, anyway.

....
<snip>

Acutally, "some years ago" was in 1994, the original device was used for
about two years on my childrens bicycles, then I ripped the original
electronics out and installed my own microcontroller based design. Both
used their bicycles for riding to school, early in the morning, around
the year.

>
>Despite various complaints about tire drive dynamos, these
>never sold well.

Sure. Around here in Germany, most cyclists had put up with unreliable
lights that failed often, especially in snow and rain. Not me.

Fun fact, if you consider that funny: the helmet craze started about
that time, targeting children. "Think of the children!!" was the battle
cry. I rememember parents trying to talk me into forcing my children to
wear bicylce helmets, while I tried to get them aware about their
childrens broken head- and taillights and other safety hazards. Without
success, of course.

--
Wir danken für die Beachtung aller Sicherheitsbestimmungen

Am Thu, 4 Aug 2022 12:17:02 -0400 schrieb Frank Krygowski
<frkrygow@sbcglobal.net>:

>On 8/4/2022 9:36 AM, Wolfgang Strobl wrote:
>> Am Wed, 3 Aug 2022 12:52:56 -0400 schrieb Frank Krygowski
>> <frkrygow@sbcglobal.net>:
>>
>>> On 8/3/2022 10:28 AM, AMuzi wrote:
>>
>> [about hub generators]
>>
>>>> I own and use one, I sell them and build wheels around them. Nice
>>>> product series and modern units are much better than classic ones in
>>>> many ways.
>>>>
>>>> That said, a tire or rim drive dynamo is less expensive, lighter weight
>>>> and fully disengaged when not in use. People make decisions based on
>>>> their own interpretation of various criteria, which is fine.
>>
>> There is still the erroneous opinion that hub generators consume a lot
>> of electricity even when they are switched off, because it takes effort
>> to turn the lifted wheel slowly, caused by what we call "pole
>> sensitivity" (Polfühligkeit) in German.
>>
>>> Agreed, each style has its benefits and detriments. Our bikes have two
>>> hub dynos, three roller dynos and three bottle dynos, IIRC. All work
>>> well for our uses.
>>
>> I've replaced all of these by hub generators, long ago, on our bicycles,
>> when these became available. Far to often I was stranded without
>> working lights on my commute, because of a failing generator. I've
>> ruined two roller generators and more bottle generators than I'm able to
>> remember, over the last ~45 years.
>
>I don't do nearly as much night riding as I used to.

Same here, sort of. After I broke my bones in 2011 and couldn't ride for
a long time, I had to stop commuting by bike and still restrict myself
to riding at daylight and avoid rainy weather. As you probably know, I
rode around the year, independent of temperature and weather, the
decades before. I didn't, afterwards, because I couldn't trust my
strength and abilities like before, anymore.

I've partially recovered and regained some extra strength through muscle
training and especially gained endurance through long rides after
retirement, but that decision stands.

>For one thing, I'm
>retired, so never have to ride home from work at night.

I haven't to, either. I still try to ride about as much as I did when
doing 25 km each working day, around the year. I enjoy finally being
able to cycle far out into the local countryside and up the hills of the
Ahr Mountains on this side of the Rhine in friendly weather. For bad
weather I have a roller trainer.

>For another
>thing, my best current night riding friend has contracted a serious
>health problem and we haven't ridden together for many months.

That's a problem with depending on pair oder group rides: you sometimes
have to find new company.

For some reason, mostly because I dislike the sporty, fashion driven
style of biycle clubs, I always rode alone and still do, when doing long
rides.

>
>But my roller and bottle dynamos have been very trouble free -
>understanding that I only rarely rode at night in the rain.

That might be the reason. Freezing rain and sleet is even worse. But I
actually enjoyed riding through snow.

>The problems
>I've had with dyno systems have been contained in the headlamps: before
>LEDs, I had a halogen bulb whose center contact got eroded by vibration,
>giving intermittent contact.

Some lamps from B&M and other had an obvious design flaw, a simple screw
fixed both the contact to the bulb and the lamp in the holder on the
fork crown, which was the ground contact. The whole construct was able
to vibrate, producing contact problems en masse.

>I had one very inexpensive B&M Lyt headlamp
> whose switch needed work. I had another off-brand LED lamp whose LED
>failed. (I replaced it, which was more trouble than it was worth.) A
>couple years ago I had a weird intermittent problem with a B&M Eyc. I
>never diagnosed it, because it cured itself.

Well, I've consumed a specific B&M Lumotec headlamp like other people
use up inner tubes.

<https://www.mystrobl.de/Plone/radfahren/fahrten/urlaub/emsland1992.jpg.1>

1992

(This officially dangerous helmet was thrown away without replacement
shortly after this holiday)

<https://www.mystrobl.de/Plone/radfahren/technik/komponenten/licht/lichtc1.jpg/view.html>

Electronics 1995-1997, but the '78 Peugeot got its original lamp
replaced by a Lumotec much earlier.

<https://www.mystrobl.de/Plone/radfahren/fahrten/urlaub/wuerzburg2fuessen/25.jpg/download>

1996

<https://www.mystrobl.de/Plone/radfahren/IMG-2461.JPG/download>

Picture from 2007

In 2007 I switched to LED lamps and never looked back.

This is a short field report about the first Lumotec IQ fly, replacing
the Halogen Lumotec.

<https://www.mystrobl.de/Plone/radfahren/technik/komponenten/licht/lumotec-fly.html>

German language, but Google Translate gives a reasonable translation

Lumotec IQ fly field report

First experiences with two Busch+Müller Fly headlights (December 2007)

IQ Fly on my winter bike

Initial situation

After a battery system from the company Friwo (Friemann and Wolf,
primarily known for power supplies, chargers and small power packs of
all kinds) has served me well for almost 15 years on various bikes (at
least better than I had from a whole series of roller and side runners),
I had to take the crumbling of the last remaining battery case as an
opportunity to look around for an alternative at short notice.
Previously I had replaced the standard 1.2Ah sealed lead acid battery
used in this system with a new one every two or three years. This time,
however, the contacts were so worn out by corrosion and use that it was
foreseeable that the contacts would soon break.

What needed to be replaced?

It quickly became clear that replacing the battery was not enough. The
Friwo system has not been available for a long time and a comparable
system that can be looped in in a similar way as a dynamo replacement or
supplement is not available on the market. Since the newly purchased
winter tires (Continental TopContact Winter) were to be mounted on
another rim at the same time in order to enable a quick change back to
the slicks normally used (Schwalbe Marathon Slick), it made sense to
switch to hub dynamo power supply. This in turn made a new headlight
necessary, since the previously used switchless, round Lumotec, which
was only held together with a cable tie but upgraded with a 0.85A/5.2V
bulb, would not have got along with the hub dynamo.

Since the IQ fly had been announced at the time, but was not yet
available in practice, I first installed a halogen-based Lumotec fly N
plus, used it for a few weeks and then replaced it with the IQ fly senso
plus that had been delivered in the meantime and which I now have also
used for a few weeks.

Light distribution on the IQ fly

Projection onto the basement wall, shortly after switching to the
parking light mode. The IQ fly delivers a lot more light than the
standard fly, which is basically not bad either and is equipped with a
normal 2.4 W halogen bulb. However, the light has some unusual
properties that you have to get used to. The cold, pale color is
noticeable, but not really a problem. However, it is annoying,
especially when driving fast, that the light distribution has a hole in
the middle. The result is that when comparing the two headlights, you
see considerably more in the close-up range, but no more in the
long-distance range, perhaps even a little less.

Projected onto a basement wall, one has the impression of two
superimposed rectangles of slightly different colors (a central, more
yellowish central rectangle and a very broad, slightly deeper bluish
one, which, on closer inspection, appears as two circular structures
abutting). I tried to show the effect, which is not as clear in the
photo as it is with a direct view, by superimposing it with a drawing.
The initial image is contrast-enhanced compared to the original, it was
created by photographing the light spot on a cellar wall, distance a few
meters, standlight mode.

Based on a tip in news:de.rec.fahrrad I filed the circuit board slightly
and achieved a slightly better light distribution. Nevertheless, the
wish for a better high beam remains.

<snip>

LED lights have become much better in the meantime, but this one worked
well enough and AFAIK still does.

>
>I've worn out bearings on a couple dynos and had minor problems with
>their mounts, but I can't recall other problems. But I don't doubt you
>have much more experience riding at night than I do.

Most probably, and under worse conditions. Plus I've bought and used
most of the expensive ones, Union Turbo, Nordlicht, for example.

The friction roller is not a reliable method of power transmission, and
certainly not in the presence of wet and abrasive dirt. If you've ever
tried riding a Velo Solex, uphill through slush, you know what I mean.
:-)