rouvy – the katsura tree

Best bicycle mountpoint for action camera

Lockdown, maturing technologies and dropping prices have nudged me into an unexpected new diversion: filming bike routes. I’ve always found virtual trainers and on-line cycling simulators vaguely interesting. But not that interesting. I was into Zwift for a long time, then one morning I woke up and discovered I wasn’t.

Nowadays for the time being Rouvy is my virtual cycling world. It suits me fine. Rouvy is real life video and real life routes, some user submitted and the quality can be variable. Having said that, there is a certain quirky charm in some of the amateur videos that is missing from the professionally shot official routes. I fancied giving it a go for my tri club‘s local routes. With cameras coming down in price, it was worth a punt.

There’s loads of good information out there on how to film from your bike, but I couldn’t find much on the best way to attach the camera to your frame. I’ve tried filming from under the handlebars on a quadlock mount, and that works ok but the bike movement is noticeable, especially out the saddle.

So I decided to try an experiment. I chose to film a 5km out and back run along the Broom Park railway lines just outside Durham. Rouvy allows you to create run routes too and I thought this might make a good test (even though I don’t have a treadmill).

I have two Apeman action cameras, an A77 and an A100. The A77 was an impulse buy and is a bit cranky. I lost the battery cover while trying it out in the Tweed Valley Tunnel run in 2019 and haven’t done much with it since. The battery cover loss isn’t actually such a big deal after all since a bit of old cut-to-fit credit card does a surprisingly decent job.

I’ve just bought the A100 as it was cheap, and has image stabilisation. So here was my setup:

The A77 (top) is easy to fit on the quadlock mount. Fastening a camera to the headtube of a modern frame on a gravel bike with a chunky headtube was far trickier than I expected. The diameter of the headtube is close to 60mm and most brackets or clamps are not designed for this size. I wasn’t averse to putting the big chunky clamp for my DSLR on the headtube but it was just beyond its limits and didn’t really work. So in the end I chose the universal solution: A couple of plastic cable ties and a basic bolt. Seems ok.

Results were mixed. Exasperatingly, the A100 was tilted slightly upwards, and gave a nausea-inducing view of the just-above the path. The A77 view was much cleaner. The difference was less conclusive than I thought; I was expecting the headtube mounted camera to be much more stable, but in reality I’m finding it difficult to be objective about how much difference there really is. The A100 is a fraction of the price of a GoPro so I expect some differences, although the basic spec (4K and image stabilisation) is what I’m really interested in.

The main question I have about the A100 at the moment is what FOV to choose: Narrow, Medium or Wide. What’s best for a Rouvy film? I suspect Wide but I need to do a few tests to compare the differences. Annoyingly, the A100 seems to forget changes to its defaults between power cycles so everytime I switch it on, I have to change the FOV to whatever I want. I don’t know if the default setting can be changed.

The video below shows my handlebar/headset comparison. I have to say I’m pretty disappointed. The uptilt on the A100 makes the comparison very difficult. The video has about 4 or 5 clips of the same section of railway path. The brake leavers on the A100 are a frequent intrusion, and I’m going to scratch my head on getting the camera further out front. Attaching a broom handle to the top tube (cable-ties!) was one thought. Always interested in others.

The Saris M2 trainer – 6 months later

Six months down the line and my slightly scary Saris M2 seems to have settled down. A query on my August 2020 post where I noted the spectacular demise of my tacx trainer tyre prompted me to have a quick recap.

I replaced my tacx trainer tyre with a Zaffiro pro trainer tyre and have now done about 800km on it.

I did another ‘turn test’ – something I hadn’t done for a while. It now takes 4 or 5 full turns to get to the click, compared to the 6 to 8 back in July 2020.

The only other observation I might make is the importance of tyre pressure. I usually have the tyre at 110-120 psi, and any underinflation tends to noticeably increased noise. Wear seems very slightly off-centre on mine but it seems to be of no consequence.

the tacx tyre and the Saris M2

So there I was, plodding up some pass in the Czech republic, slogging steadily through the middle of a Rouvy server, minding my own business, when BANG! A mechanical! On a smart trainer. A blowout.

I hopped off my bike and moved to the side of the road. Fished out the tyre levers, which had conveniently appeared on a nearby workbench, and had a look at the problem.

I have my doubts about the Saris M2. I miss my Tacx Vortex. It’s a pity that it fizzed and sparkled when I tried to fix it. It’s pretty much unfixable now. It’s now a perfectly functioning not-smart trainer though. The Saris M2, with its scary clutch knob, makes me nervous. So much tyre squashing involved.

Moving to a dedicated trainer tyre was definitely a good move. It’s quieter and smoother. Not sure about the battery life though. Is it connected to the trainer? Is it a co-incidence that since changing to the Saris M2 that the cracks have begun to show in my trainer tyre?

On the surface of things I certainly seemed to have a smoking gun. But I’m a bit puzzled. Admittedly the tyre is not looking its best, but I was surprised it had resulted in such a rowdy blowout. I decided that it was just a fluke, and the tyre had a bit of life in it yet.

I wasn’t wrong. There was some life yet. Another 12 minutes to be precise. Then there was another loud bang, not dissimilar to the first one. Luckily I had another nearly new tacx trainer tyre, I just had to find it. I was nearly at the top of the pass too. Annoying.

The new tyre is on, and today I managed to complete the ride without a mechanical. But I’m really, really not sure about the Saris M2. Surely a tyre was never designed to be quite so squashy …

Adjusting the Clutch Knob tension on the Saris M2 Smart Trainer

Goodbye to the Tacx Vortex

My Tacx Vortex packed up. I was a bit surprised. Five years it’s been turning round and round with various degrees of resistance and I’ve always taken it a bit for granted. Then recently, when trying to pair it to Rouvy, I noticed nothing was happening. No lights, no connection.

I was in denial for a while. When something has worked fine for 5 years then doesn’t, it’s a bit of an adjustment. After changing the power and plug and digging around with a multimeter, it was clear it really was dead.

I lifted the hood, had a google, and it looks like it’s a thing. A thread on www.electronicspoint.com reveals a few people with tantalisingly dead Tacx trainers. I don’t fancy prodding the PCB with a soldering iron – my soldering skills were never that good – but I might one day revisit this.

The R64 resistor gets a few mentions and the one on my PCB looks like after 5 years it’s decided to call it a day.

I was discovering how much I’d taken the trainer for granted. The Tacx Vortex is a really simple affair, with a straightforward lever that presses the roller against your back wheel. I liked it a lot. Smart Trainers are hugely variable in price and the Vortex hit a nice entry-level sweet point when I bought it in 2015, and now it was dead.

Or, more precisely, dumb. Clearly I could use it as a dumb trainer, like in olden times. But I was surprised how quickly I’d adjusted to, and taken for granted, having real virtual resistance, or virtually real resistance, as I slogged up an alpine pass in the comfort of my garage. I needed another smart trainer.

Looking for a new Smart Trainer

Time for another shock or two, or three. The Vortex was no longer available. And Smart Trainers are pretty expensive. Even cheap ones are expensive. And as we pedal virtually through the post-Covid apocalypse, they are very hard to find.

I was quite keen on a wheel-on trainer. Mostly because they’re a lot cheaper, but also because you can use it to road-test, or pseudo-road-test, an old bike. During lockdown I’d rebuilt two old steel bikes and learned a lot on the way. Mostly about the joys of trying to adjust a cup and cone Campag bottom bracket. And I was discovering that the best place to discover that you hadn’t adjusted it very well, was not somewhere a long way from home.

I bought the Saris M2 Smart Trainer. I wanted to stick with Tacx, but price and availability were against me. Despite many of the alarming reviews for the M2 I decided to take a risk. It was the only one around my budget, in stock, and available to pick up.

There’s a good review, or non-review, of the M2 by Jeff Whitfield on the velonut.com website. I have to say my experiences almost exactly reflect his. He mentions adding the bolt action tube was difficult. My experience was identical. Initially I assumed that I was doing something wrong and I kept checking and re-checking the documentation until I was convinced I was right, and eventually it jostled in. It was in the right place but the engineering felt a bit rough.

Same story for the resistance unit. After many “this can’t be right” moments I tentatively tapped the bolt through the frame, having constantly checked the alignment and kept everything lined up. There isn’t a huge amount of thread to be engaged on the bolt and I spent a few minutes fruitlessly using a socket wrench turning a small amount of empty space mistakenly assuming I’d finally managed to get the bolt far enough along to engage the thread.

But I got there in the end. And then, the adventure really begins. The Clutch Knob.

The Saris M2 Smart Trainer Clutch Knob

It’s hard to find a review of the M2 that doesn’t mention the Clutch Knob. It is, as they say, a fine idea. In principle, you might add. But it scares me. So much tension in such a delicate looking piece of metal. So much so, that on my first attempt, I lost my nerve. I’d read so many reviews mentioning so many bent frames and so many clutch knobs not clicking, that I really didn’t want to end up with a bent 2-day old trainer. The last one the shop had in stock.

It was very frustrating. After spending a lot more time assembling the frame than I expected, and thinking I was finally getting somewhere, I wasn’t sure I wanted to risk turning, and turning, a yellow piece of plastic hoping, waiting and praying for it to click. What I hadn’t been able to find out from my searches was a ball-park figure for how many turns, how much pressure, would be required to get to the click. How much should the tyre be squashed?

One article suggested 2.5 to 3 turns, but that was for a black knob. The M2 was yellow. Did this matter? Another support article suggested the tyre should be squashed in about a 1/3 to get the required resistance, with the tyre inflated to just a little of maximum. And a thread on www.trainerroad.com shows that I’m not the only one to be uneasy about the level of deformation or turning required.

Having lost my nerve and resigned myself to never knowing, and then discovering the joys of software calibration, I was feeling pretty dejected. Compared to the simple engagement mechanism used by the Vortex to apply pressure to the wheel, I was finding the M2 pretty inelegant. I visited the Saris website, registered the product, and put in a support ticket explaining my sad story.

Waiting for the click

To set things up I was using a 1983 Alan bike with a wheel and tubular of similar age. The tubular is pretty frayed but it would fine to test things out.

Next after the Alan was to try it with my road bike with a dedicated Tacx trainer tyre inflated to the recommended 120psi.

After my first failed attempts I decided to man-up and try again. But this time I decided to video my attempts so that I had something concrete to show to Saris.

The twenty minute adventure is available for viewing but you might have some paint you need to watch drying somewhere, so here’s my key discoveries.

TL;DR

Old tubular at approx 120psi: Start tightening clutch-knob (06:47).
Old tubular at approx 120psi: Clicks after 8 and a bit turns (08:27).
Deformation in tubular (09:34).
Old tubular at approx 130+psi: Start tightening clutch-knob (12:50).
Old tubular at approx 130+psi. Clicks after 6 and a bit turns (13:25).
Tacx trainer tyre at approx 120psi: Start tightening clutch-knob (17:10).
Tacx trainer tyre at approx 120psi: Clicks after 6 and a half (18:00).

So that’s between 6-and-a-bit and 8-and-a-bit turns. To me that feels like a lot of turns, even if the tyres are slightly under-inflated. I get that you need decent thunk of pressure to get that potential 15% of slope, but even so, it doesn’t feel right. Perhaps I could inflate the tyres a smidgen more and it’d bring the number of turns down but I’m not so sure. There feels like a lot of tension in that L-bolt. I’ve had one training session on the M2 and the jury’s out. Time will tell.