In this article, we use so-called affiliate links. With every purchase through these links, we receive a commission from the merchant. All relevant referral links are marked with . Learn more.
Following a redesign, the former Cycleops Hammer now trades under a new manufacturer's name. The H3 turns a very generous flywheel mass, which rolls out almost endlessly at idle. The brake is very powerful, simulates up to 20 per cent gradients if required and measures power with acceptable accuracy. The running noise is much more muted than before and very quiet up to medium speeds.
In Zwift mode, however, delays of several seconds occasionally crept in on undulating terrain, which is irritating when the Avatar is already rolling downhill but the brakes are still demanding uphill power. Saris does not (yet) have its own app; the manufacturer recommends the Rouvy app for updates, which we repeatedly found difficult to use. The folding mechanism is exemplary. The feet can be easily folded out of the trainer's chassis, providing solid and wide support for the device. The Smarttrainer is folded in seconds and can be transported and stowed compactly thanks to the carrying handle. A front wheel support is included in the scope of delivery.
Technical data of the Saris H3
- Price: 999 Euro >> e.g. available at Amazon
- Weight 22.3 kilogrammes
- Footprint W x L: 80 x 47 cm
- Storage dimensions W x L x H: 22 x 47 x 50 cm
- Max. power: 2,000 watts
- max. gradient: 20 per cent
- Volume at 27/40/66 km/h: 32/45/68 decibels
- Average deviation: 6 watts
Test summary of the Saris H3 Smarttrainer
- PLUS: outstanding pedalling feel, quick and compact folding
- MINUS: Delayed response in Zwift, no dedicated app
How TOUR tests roller/smart trainers
We used all the devices in a normally furnished living room measuring just under 20 square metres. Firstly, the manufacturer-specific app was set up on a Samsung S7 running Android, if available, and the trainer's firmware was updated to the latest version. A calibration ride ("spindown") was then carried out using the app after warming up. We used the app in ergometer mode with Bluetooth to complete an interval programme. We also checked the accuracy of the power display.
An SRM system was used as a reference, which in turn was synchronised with other power meters. We limited ourselves to endurance performances in the range of 150 to 350 watts and also tested the control speed of the trainers during power jumps in interval mode.
Finally, we completed various courses on Zwift and tested the response behaviour of the brakes - the reaction time to changes in terrain - and evaluated the ride feel. We used ANT+ as the connection to the PC. The sound level was measured at a distance of one metre from the device at three speeds: 27, 40 and 66 km/h; at pedalling frequencies between 90 and 120 rpm. It should be noted that the sound level has a logarithmic scale. An increase of 10 decibels corresponds to a doubling of the perceived volume. Low frequencies are not adequately recorded by the level measurement, which is why we also describe these in the text.
