AMFLOW PX L – First Rides Review (Real World Alpine Use)

First test rides on my PX L (powered by my Partners and Local Bike Dealer Bike Bow in Hall in Tirol) are done – and overall, it’s a seriously impressive bike. More on the Ride Characteristics, the Motor, Battery Range etc coming soon. This about the first Technical Learnings.

1. Fork Setup & Geometry

The OEM FOX 36 failed (air transfer issue) within the first two rides and is currently at FOX service (~5 weeks including shipping).
In the meantime, I tested a FOX 38 (170 mm) from my own stock, followed by another 36.

Interestingly, I actually prefer the 36.

It gives the bike a much more balanced geometry. The 38 lifts the front noticeably – even after compensating with +0.5° and a lower BB. You can clearly feel that a 160 mm fork simply suits the bike better overall. A 170 mm fork results in a slightly “stalky” ride characteristic.

The replacement 36 had a slightly too short cut steerer from back in the Days, which resulted in a lower cockpit than intended. Surprisingly, it works – especially in steep terrain and tight switchbacks, where the compact stack delivers very direct front-end control.

For fine-tuning until the original fork returns and to reduce strain on my neck and shoulders I’ve switched to a handlebar with 50 mm Rise.

2. Issues big and small

a) Upper 0° Standard Headset Cup Play

With correct headset preload, the original 0° upper headset cup (5) shows slight play under full load.
With the front brake applied, one can both feel and visually detect minimal forward movement of the standard 0° headset cup inside the head tube and see the Top Cover (1) move back and forth slightly. Hhm.

This definitely needs a proper solution.

Current workaround:
Using the +0.5° headset cup included in the accessory box solves the issue. Due to its integrated rectangular guiding block design at the Cup sleeve, it fits snugly into the frame and does not exhibit any movement.

I measured the cup insertion sleeve diameter on both the standard 0° headset cup and the alternative cups. The sleeve diameter is identical across all cups at least with the analog measure.

However, the completely round standard 0° headset cup tends to tilt slightly inside the carbon head tubes as it is not pressed in.

So obviously it does not provide sufficient friction or contact surface with the headtube, as it is not press-fit but simply placed (by Hand) into the head tube.

One can also clearly notice that the standard 0° cup can be easily placed into and removed from the head tube by hand.

In contrast, the alternative headset cups require significantly more force to install and can be removed by hand, but with a lot of effort and finger and nails twitching. Their rectangular guiding block extension increases surface contact and friction with the frame and ensures proper alignment along the frame’s central axis, effectively preventing any movement.

b) Power Extension Cable of the SRAM Derailleur

The Power Extension cable of the SRAM derailleur has relatively free movement where it exits the chainstay and is prone to getting caught by branches or similar obstacles.

A simple fix with a bit of Tesa duct tape works well.

A zip tie is not effective here, as the trapezoidal shape of the chainstay – especially where it meets the seatstay at the Horst link – causes the zip tie to slide toward the Horst link, where it doesn’t belong and get’s loose.

c) Dust and Dirt entering the Lower Headset / Crown Race Gap

The lower headset, bearing, and crown race are not very well protected against dust and dirt ingress. I really noticed this while changing the forks from the OEM 36, to 38 and to my old Aftermarket 36…

The Images show the Fork lowered a bit to show the Gap.

On many other bikes, there is typically a dust seal integrated at the crown race interface toward the frame and lower headset bearing. This is not the case on the PX (and also not on the PL).

d) Steerer Stop Block Replacement

For those who remove the standard steerer stop block which is required by Law in certain countries for accident prevention, a proper replacement part is currently missing. Since the Times of the early PL’s users are solving this with a 3D-printed part (Cults 3D Block Replacement). While this is a nice DIY workaround, a dedicated replacement part could ideally be included with the bike. As well as an AirTag Holder maybe 🙂 (Air Tag Holder FOX 36 AMFLOW PL)…

e) SRAM AXS Pod Controller – Ergonomics Limitation with Gustav Pro

The SRAM AXS Pod Controller itself is a well-designed unit, offering excellent ergonomics, intuitive “click and fire” operation, and full customization via the AXS app.

However, in combination with the Magura Gustav Pro brake, there is a notable limitation:

The clamp of the AXS Pod cannot be positioned far enough inward (to the left on the right-hand side of the bar), as it is physically blocked by both the shape of the Gustav Pro brake clamp and the reservoir of the radial brake lever.

As a result, the Pod sits slightly too far outward, causing the rider’s right thumb to repeatedly contact the controller during riding. This becomes noticeable and uncomfortable, especially on longer or more technical descents.

This is clearly a compatibility issue between the two components rather than a flaw of the Pod itself.

f) Chainstay Protector – Functional but Overbuilt Edge

The chainstay protector itself is well designed in terms of coverage and protection. However, due to its material thickness, it adds noticeable bulk to the chainstay and features a relatively high, pronounced outer edge.

When repositioning the rear wheel during tight switchback riding I give the chainstay a quick impulse with my left or right feet instead of rotating the bike with the hips – I noticed that it is quite easy to catch the left shoe on that edge when I du left hand switchbacks.

This can be slightly disruptive in technical terrain and tight maneuvers.

Additionally, I have some concerns that the protector could lose adhesion over time in this specific area, as the exposed edge is likely subjected to repeated mechanical contact and stress. So I will apply the best solution there is: the 3M Scotch Rubber Mastix Electric Tape 2228, 5,1 cm x 3 m, black, which is a self sealing tape for Isolation and Seals, which softens that sharp edge a bit.

g) Front Hub Seal – Loose Fit (Left Side, Direction of Travel)

The sealing ring of the front hub on the left side (in direction of travel), located above the hub bearing, tends to come loose repeatedly.

During wheel removal and installation, it can easily shift out of position, which raises concerns about it potentially being lost over time.

This should ideally have a more secure fit to ensure proper sealing and long-term reliability.

h) Rear Hub Freehub Standard – Shimano HG vs. SRAM XD Compatibility / 4-Pawl vs 6-Pawl long term durability

The rear hub is equipped with a Shimano HG-style freehub body (4-pawl system).

The bike comes with a SRAM XS-1270 Eagle Transmission cassette (10–52T), which is specifically designed to be compatible with traditional splined freehub bodies (Shimano HG-style freehub body Standard).

According to SRAM specifications, the XS-1270 cassette fits:

• directly on standard 8/9/10-speed HG freehub bodies
• on 11-speed HG bodies with an additional 1.85 mm spacer
• https://support.sram.com/hc/en-us/articles/26988415305243-What-driver-body-is-the-XS-1270-Eagle-Transmission-cassette-compatible-with

The stock wheelsets (XMA-30 mullet version) use the following configuration:

• 12 × 148 mm rear hub
• 32H
• 4 pawls / 36T engagement
• HG freehub body

Unfortunately, the XMA-30 only features a 4-pawl system, whereas the PX Pro’s XMC-30 version comes with a 6-pawl hub.

As a result, the stock setup on the Amflow PX and PR does not support SRAM XD-compatible Transmission cassettes. Switching to an XD-based cassette therefore requires replacing the freehub body (if available) or the complete rear hub/wheel. Guys, thanks for nothing!

More critically, the reduced pawl count results in less direct engagement and raises valid concerns regarding long-term durability. Given the torque characteristics of the M2S motor (up to 150 Nm), a more robust engagement system would be expected. A 4-pawl setup appears underbuilt for this application, especially under repeated high-load conditions typical for alpine riding.

This is not just theoretical – in technical terrain, the slower engagement vs the PL Pro Rear Wheel is noticeable, and from a mechanical perspective, higher load per pawl will accelerate wear over time, and we are talking about 150 NM max. + Rider input.

See: AMFLOW PX Tech Specs

i) Flip Chip Orientation – Reduced Clarity on PX vs. PL

On the previous PL model, the flip chip orientation for 27.5” vs. 29” was clearly visible and intuitive.

On the PX, this has become less obvious due to the new flip chip plate design. The geometry and shape of the plate make it difficult to immediately identify the correct orientation.

The small engraved pictograms provide some guidance, but in practice they are not sufficiently clear or intuitive.

As a result, determining the correct setting requires closer inspection than it should, especially in workshop or trail-side conditions.

Adding to the confusion, the accessory box appears to include replacement flip chips intended for the PR model, which are not relevant for the PX setup. Please correct me if I’m wrong…

j) Avinox Spider Lockring Tool – Lack of Axle Retention

The Avinox lockring tool for the spider lockring lacks any form of retention or guidance on the crank axle.

Unlike well-designed solutions such as the Cyclus tool for Bosch Gen4 lockrings, which is securely threaded onto the crank axle first and then bump proof Spring Locked, the Avinox tool is not fixed in place during use.

With the Cyclus system, the tool is screwed into the crank axle thread, creating a solid and centered interface. It is then guided onto the lockring via a snap-in mechanism. This ensures perfect alignment and allows safe one-handed operation without any risk of slipping.

In contrast, the Avinox tool relies purely on manual positioning and pressure probably by your Body leaning on it from the side. Especially when operated one-handed, there is a constant risk of the tool slipping off the lockring.

This is particularly critical given that:

• the required torque is relatively high (~20 Nm)
• the lockring is made of aluminum rather than steel

A slip under these conditions can easily damage the lockring.

From a mechanical and usability perspective, this is a clear design weakness. A guided, axle-supported tool interface – similar to the Cyclus Tools solution – would significantly improve safety, handling, and reliability during installation and removal.

k) Lower Headset O-Ring – Seal Damage and Insufficient Retention

The lower headset assembly consists of the bearing (9), the lower headset cup (8), and the lower bearing seal (7).

In practice, the seal (7) is not a properly retained sealing element but rather a very thin O-ring that is simply rolled onto or loosely placed over the lower headset cup (8).

The complete assembly (9 → 8 → 7) is then inserted into the head tube from below.

During installation, the O-ring frequently shifts out of position, rolls off, or can even be damaged. It lacks any defined seat or retention feature and does not stay securely in place.

The fit is noticeably loose, and the sealing concept appears under-engineered for a component that is directly exposed to water, dust, and contamination from the front wheel.

From a mechanical and serviceability standpoint, this solution is inadequate. A properly guided and tensioned seal – ideally integrated into the headset cup or designed with a defined groove and interference fit – would significantly improve reliability, sealing performance, and ease of installation.

We found a better retenting O-Ring which fits in a Bosch Gen4 Box in the Bike Bow Workshop, see Image 2, the smaller one on the left side of the broken original one.

l) Chainring Size vs. Chainstay Clearance

When running a 32T front chainring, the chain passes extremely close to the chainstay protector when in the smallest rear cog (10T on the 10–52 cassette).

With a 34T chainring, the clearance improves slightly but remains tight.

This should be taken into consideration, as the minimal clearance could lead to contact under load, flex, or contamination (mud/debris), especially in real-world riding conditions.

BTW at first with the 32T Chainring, the Transmission Chain seems to pass by the Spider super duper close. But it works without any Problems.

PS:  Why do we run 32T front on eMTBs in Tyrol? Because we ride the steepest, gnarliest, most technical stuff Uphills worldwide. from 30%-50%+ trails.

So basically, we run 32T front for a reason, no Joke. But I can understand you wondering if you come from the Netherlands 🙂

3. Drivetrain efficiency Adjustment

After the first steep uphill ride, I immediately swapped the front SRAM Avinox Direct Mount chainring from 38T to 34T, and more recently to 32T.

The 38T clearly hurts efficiency and range (approx. -30% in my use case). For alpine riding, a smaller chainring – such as 32T – is the only logical choice.

4. Gustav Pro Brake Performance Upgrade

The Magura Gustav Pro required an upgrade. The 203 mm MDR-S rotor was for me not powerful enough, so I moved to a 220 mm MDR-P setup, which is officially supported by Magura.

Current upgrade path:

• Pads: MDP 13 BLACK → SILVER (+15% bite) or YELLOW (more power)
• Rotor size: 203 → 220 mm
• Rotor type: MDR-P 220 mm (2.0 mm, plug & play)
• Future: MDR-S 220 mm (2.5 mm) for improved heat management and more precise braking while maintaining good modulation

5. Seatpost

Next change will be the seatpost – likely FOX Transfer 200 or OneUp. The Original Seatpost moves a bit slowly and also needs Lubrication.

I need maximum drop for very steep terrain with large root step-downs. The OEM post still has a bit too much stack while it is an improvement to the 2025 PL’s stock seatpost.

In our riding terrain, even a 10 mm lower saddle height makes a noticeable difference.

Also: Big thanks for the short seat tubes on the PL and PR – this is highly appreciated in the Alpine rider community!!

6. Chainstay Setting

I set the chainstay to maximum length (with 27.5″ rear), which noticeably improves climbing performance!

7. Battery & System Performance

With the 700 Wh battery:

• ~10–15% less total elevation capacity compared to the PL
• Battery runs ~5°C cooler
• Motor also slightly cooler and extremely quiet
• 32T front helps to have the M2S on optimum efficiency on steeper climbs

8. Wanted! Compact Range Extender – VoC!!

Avinox please listen: For the upcoming big mountain season (September/October), my biggest wish to Avinox is a 250–300 Wh bottle-style range extender, ideally backpack-compatible.

Form factor similar to:

• Bosch 250 Wh PowerMore
• Specialized range extender

An ideal PX Range Extender solution would include:

• Bottle-style RE (250–300 Wh)
• Simultaneous use of main battery + RE
• Hot-swap capability (replace depleted RE mid-ride)

The current RS 600 Wh strap-on solution is:

• Too bulky
• Awkward
• Not compatible with size L PX
• RS 600 will feel heavy in the Backpack

a) Real-World Reference Setup (Simplon Rapcon CX)

For the Bike Test see here: One Bike – Three different Motors – E-Mountainbike Magazine

Previously over 400.000 HM and 6500 km, I used a very flexible system on my Simplon Rapcon CX which is now my Touring bike 🙂

• Swappable main battery: 750 Wh or 500 Wh
• 250 Wh Bosch PowerMore on the frame
• plus 250 Wh Bosch PowerMore in the backpack for the super Gnarly +2500 HM ascent rides

The 250 Wh unit (~1.4 kg) was perfectly acceptable in my backpack, it is small and weighs around 1.45 kgs. Barely noticeable.

This allowed:

• On-the-fly swapping of the downtube RE with a fresh one from the backpack
• Ideal flexibility for long alpine rides

Typical configurations:

• 750 + 250 + 250 Wh
• 500 + 250 + 250 Wh

The 1250 Wh setup worked perfectly for:

• 2500–3000 m elevation days (Dolomites, Ticino)
• No charging required during the day

b) Light eMTB Reference (TQ System)

On my Simplon Rapcon TQ:

• 320 Wh internal
• 160 Wh range extender
• Total bike weight: 19.3 kg

The RE was integrated via a Fidlock interface (developed together with Fidlock Germany) and worked extremely well.

9. Conclusion

Overall: The AMFLOW PX is an outstanding bike and a clear step forward compared to the already strong PL Series.

Congratulations to the AMFLOW and Avinox Team, they built and designed an outstanding Alpine Bike!

This Bike is another Game Changer, like the PL Series was. And we are not even talking about the PR Series!

The only one Topic where I think DJI has not yet understood what we want is the Range Extender solution. Volabike etc show what we want.

Since: Seriously? Routing the Cable BEHIND the Seattube to the Tire????

Why would I want to ride around with such a DIY solution? Come on Avinox, that can be done professionally – see my PowerMore or TQ examples above.

More Details & Impressions coming soon…

Cheers

Christoph Malin