Re: Extensive article on Rivendell and Grant Petersen

Am Wed, 25 Sep 2024 18:00:21 -0400 schrieb Zen Cycle
<funkmaster@hotmail.com>:

On 9/25/2024 3:26 PM, Wolfgang Strobl wrote:
>
<BIG snip>
>

 But I'm quite sure that a simple switch using  a low
power radio signal to communicate with a derallieur that is essentially
reduced to a sealed microcontroller operating a single actuator has a
lot less failure points than a mechanical Rube Goldberg device that has
to fit into a brake lever and has to communicate by a degrading wire
rope running over several corners, merging both control and power into
that single, unreliable channel.

>
It depends on how deep you want to go with your root cause failure
analysis, or preemptively, your FMEDA.
>
(this comparison is ignoring the parallelogram/jockey wheel cage
assembly; e.g. the basics mechanics of the derailleur)

I ignored that assembly because both systems have these basic mechanics.
But ok, let's compare them in detail.  There is a relevant detail, where
both are quite different: The delicate mechanical clockwork inside the
brifters have to work against the varying losses from the Bowden cable
and a spring inside that derallieur assembly, in addition. That spring
has to be strong enough to move  it reliably into the other direction,
over the full range.

The actuator, on the other hand, just sits there, without having to
overcome force from a spring in either direction, so it has to apply
only enough force to move the derailleur against its own friction and
mass, now and then.

>
But at the top level they have about the same number of failure points:
the points you note above VS a switch, battery, and ECM/derailleur.

Have you ever disassembled a modern mechanical brifter like for example
those Shimano sold in 2010, in order to remove a broken cable? I have.
Wasn't fun.  Replacing a CR2032 after two years or changing an battery
for recharging every 500-1000 km is a breeze, on comparison. The former
action just needs a coin, the other one needs no tool at all.

>
A down-tube system is even simpler - no ratchet mechanism

But still not failsafe. My Peugeot PR60/L from 1978 had only friction
shifters (indexing didn't exist then), but my Peugeot PR3000 from 1995
can be switched between friction and indexing.  Quite a hassle to find a
replacement, after one of both broke, years later.  Found a compatible
replacement from SunTour by asking the owner of a somewhat obscure bike
shop during a business trip to Berlin. 

.

>
What can go wrong? How easy is it to repair?

That's easy. Just do without changing gears, as our ancestors did,
problem solved.

I'm comparing reasonly recent systems, only.  In addition, my current
focus isn't on repairability when being stranded somewhere in the
Sahara, but on convenience under our current conditions, meaning single
day trips around where we live, or somewhere in Europe, do single day
trips during our vacation. 

>
The downtube:
- not much can go wrong, even if it's indexed. You might break a cable.
Easy to diagnose, easy and extremely cheap to fix.

Sure. Did it often enough, even during a bike tour on vacation, once.
 

A failed shift lever
isn't likely, at least, in my 40 years of riding, I've never seen a
failed downtube shifter that wasn't from abuse.

I did, see above. No abuse, just heavy use, mostly while commuting,
bordering to abuse.  

>
The integrated mechanical system - A bit more than the down tube, but
still extremely easy to diagnose.

Or so you think. But perhaps "Doesn't shift in one direction, anymore,
but no way to find out, without disassembling a fragile mechanism that
is broken enough to prohibit disassembly" is a diagnose, IYO.

If it happens to be the shifter, it's
usually a replacement, but sometimes repairable for the Fore
mechanically inclined (If the mfr sells part: ratchet, bearing,
spring....). The complete assembly might be expensive depending on the
model. If you can get parts, relatively cheap.

That's a big if.  Well, the broken part was a tiny spring deep inside
the ratchet mechanism, as I found out much later, after I had partially
disassembled the broken brifter after a shop sold and mounted a new one.
Found out that, while shimano sold most parts of that specific brifter
via their distributor, they didn't have a part number for that spring.
After looking into the disassembly instructions later, just out of
curiosity, I was quite sure about the reason for. Mounting that tiny
spring not only needs complete disassembly, there must be a specific
tool used in production, to mount it inside. In short, it's not
repairable, when broken.
   

>
The wireless:
- the switch contains a battery, physical switch, microcontroller, and
transceiver (latter two likely integrated).

All sealed and only one simple moving part, a switch.

   - The derailleur contains a battery, transceiver, microcontroller,
stepper motor, worm gear.

The battery is an easily replaceable part, a microcontroller isn't
something wearing out within a few years.   Actuators are typically sold
with a MTBF around 20.000 hours*)

>
Lots to go wrong there, none of it repairable.

A broken brake lever assembly can easily be found by looking into the
spare parts catalog from SRAM and then using a search machine for
looking up "ED BRAKE LEVER ASSEMBLY (PADDLE AND ELEC POD) RIVAL ETAP AXS
DISC LEFT".

After replacing batteries fails, It's still not clear. Is it the switch
or the derailleur?

Try the other lever. If that one can operate the derailleur, it's most
probably a problem with the lever.  Look at the LED on the inner side of
the lever. Does it light up, when you try to shift?  If yes, check
whether the derailleur is connected to the failing lever or not, If not,
try another battery and check again. If this fails too, inspect the
battery contacts, carefully. If there isn't an obvious problem, try all
three batteries in the other, working lever. If all three work, you
probably have a problem with the electronics, not much different from a
broken part inside a mechanical brifter - no way to fix that on the
road. 

Considering they're a matched pair, it's irrelevant.
You're fucked either way, and need to replace the entire shifter and
switch set (unless you're lucky enough to have a firmware bug that can
be fixed with a new download).

If you are paranoid, you can buy a "wireless blip"
<https://www.sram.com/en/sram/models/ec-blip-b1> and carry that
somewhere on the bike. It is quite simple to reconfigure these switches
using the SRAM app on a mobile phone, and better than not being able to
shift, anymore. Just pushing a button on the handlebar is certainly less
difficult than operating a lever on the downtube.

>
If you want to dig deeper into the failure analysis- Is it a mechanical
failure (corrosion/breakage)? Firmware bug (how would you tell except to
try a new download)? or did the silicon just decide to quit? Lot's more
points failure there than a mechanical system.

That's all just idle speculation and arbitrary counting imaginary
failure points. You could also count each contact point of the inner
wire of a brake cable as a single point of failure and would be less
wrong.

>
So which would you rather have? A system that has easily diagnosable
exposed mechanical parts with the likely possibility of a cheap repair,
or an expensive system with no replaceable electronic parts?

For our purposes I want a system where some fragile mechanical parts
that quickly wear out are replaced by electronics that easily outlive
those parts. So far, I'm quite optimistic.

>
There's no right, or wrong answer.

Actually, it's the wrong question.



*)
<https://www.linengineering.com/news/selecting-the-right-stepper-motor-for-medical-applications>

"Stepper motors typically have an MTBF of over 20,000 hr of continuous
operation. When stepper motors operate at their bearings’ rated axial
and radial loads or less with temperatures kept to less than 50°C,
stepper motors usually last 20 years, assuming a 50% duty cycle."

 

--
Bicycle helmets are the Bach flower remedies of traffic