I’ve got it mostly solved but it’s a little funky. I use the mobile link app and connect my trainer via Bluetooth instead of ANT+ now. As long as my network connection is solid it works perfect. But if I drop the wifi signal it disconnects completely and takes a few seconds to reconnect. The second person I talked to over there was awesome(Jamie). It’s a bulky measure but I’m happy to wait for the new update.
I had not experienced the problem for 5-6 days until this morning during which I dropped to zero watts over a dozen times, which caused me to keep getting dropped from my group, very frustrating. I hope they solve it and get the update out soon (I too got the emailed response that they were aware of the problem and working on it).
I have something to add to this topic. I have a relatively new Hammer (bought last Saturday) which has firmware 31.31. I have been using it pretty much exclusively with my Quarq or SRM reporting power, except for one session on ERG where I used Saris FEC to report power.
That ERG session worked quite well, and the tracked power averages for the intervals worked out to within 3w of the SRM.
Today, however, I rode a different bike which has no crank-based PM. This required me to use, at first, Saris PWR for power (that worked badly) and then Saris FEC for power (that worked a bit better.)
I went over the London underground stairway three times. All three had the same dropout to zero watts and slamming the brakes of the bike (it goes from around 17mph to 0 instantly according to Strava.)
It did not matter how hard I pushed or whether I was pedaling or coasting going in to that section. It ALWAYS happened at the same spot halfway up the ramp.
My speculation: This always happens at the precise point where the simulation of momentum drops out and the simulation of grade becomes dominant. It has little to do with what the rider is actively doing with the legs.
It almost acts like a little kernel panic followed by a reset a second or two later.
There are a few things that add to the perfect storm of silliness:
The Zwift “coaster brake” that is applied when you stop pedaling. (by the way, I HATE this feature)
Momentum is simulated in a choppy and unrealistic way on Zwift. I have observed this for over a year.
Grade simulations seem to be also more extreme than on the Cycleops Virtual Training. I was riding at 100% sim on it and it was smoother on transitions that riding Zwift at 50%, similar grade types.
I figure Cycleops will address this issue with a firmware update, but I do hope that Zwift will also be in the loop. When people talk about how great the Kickr sim is on Zwift, I think that this is primarily due to it being the “standard” smart trainer that the Zwift developers have been working with from the start, and their simulations are designed around the way it responds. Clearly the Hammer responds differently, and hence should be treated differently by the app.
I won’t race with the Hammer except perhaps on 0% sim (maybe up to 30%?), using the Quarq or SRM. It just seems like a handicap.
My response from Zwift support was essentially passing the buck to Cycleops. Hence I have contacted Cycleops. I assume they will just add me to the list of users reporting this bugfeature.
I’m on rest day now, and tomorrow might be an outside day, but I am thinking of a few test protocol for this issue:
Try climbing the subway stairs at various % of trainer realism, and see if there is a threshold where the bug does not occur (50% definitely DOES trigger the bug). I assume that 0% would not trigger the bug.
See if it is possible to find a Cycleops Virtual Training course that has a similar terrain setup (i.e. a fast runup to a severely high grade climb, changing almost instantly.) See if the same problem occurs.
If anyone has too much time on their hands I would suggest they try these and report back. I may do it eventually.
The zero watts issue is no longer confined to the esculator or steep gradients for me, I will drop to zero watts while riding along on near flat gradients. I also have developed what I think everyone is referring to as the snapping sound. It sounds like my chain is slapping the side of the trainer but that’s not it, it is something internal. It is very intermittent and only happens once or twice over a 90 minute period.
Right now I have serious regrets about having bought this trainer, my third CycleOps. I hope word gets out so other people steer away from them until the issue is resolved.
The Zwift technical support folks looked at my log files and saw the my ANT+ connection was experiencing multiple dropouts each session. At their recommendation I have been using a Bluetooth connection for a while. I haven’t gotten back to the escalator to try it, however, I not only haven’t been experiencing dropouts but the riding feels much more natural with no apparent lag. I am using my Android Tablet running the Zwift app to make the connection. It sounds clumsy but once set up it connects automatically.
ANT+ is a 2.4GHz signal that apparently doesn’t do well in a noisy radio environment like my living room. This is also why it is recommended that you use a USB extender (an active cable would be better but more costly.) It is not clear whether ANT+ is so bad on the road but I suspect not. I don’t care what the connection method is as long as it works. We’re lucky that the CycleOps Hammer supports both options.
If anyone else tries this method out please update us. If I experience the problem again I will report back.
Yesterday I rode the Volcano circuit and after 5 minutes at 0% grade I dropped to zero watts again. I stopped and switched the trainer and power to bluetooth, leaving cadence and heart rate on ANT+. The trainer didn’t seem as responsive (that’s was just my perception, I don’t have any data to confirm that) so I switched the trainer back to ANT+ and left the power readings on bluetooth.
It worked perfectly for the next 30-45 minutes until it lost the bluetooth signal which I could not get restored (my laptop was only 2 feet away). I switched back to ANT+ and the bike took off at 174W with out me even pedaling. I unplugged the trainer and reset everything back on ANT+ even though bluetooth reappeared also and finished the ride without incidence.
Next time in London I’ll set it back on bluetooth and see what happens. As far as the snapping noise, I removed some of the grease from the pawls as instructed, it made the noise on the first small climb but stopped and it never occurred again.
I’ve finally gotten to try the escalator on the London Loop and sadly, even using the Bluetooth connection, I stall out the same way on the gradient. In some ways it seemed even harder to get started. I’m quite disappointed.
I tried Zwift racing on the Hammer yesterday. My Quarq had to be sent to the factory so I mounted my beater bike (no crank PM). Not sure whether it would get stuck on steep climbs I chose somewhere around 40% trainer effect. No actual incidents of dead-stopping, but the Hammer certainly wasn’t making it easy for me. It seemed like I was having an even tougher time than normal in the A field (it’s been rough since the speeds have gotten unrealistically high).
At first I wrote it off to crappy legs (which was true) but then decided to have a closer look at the data files once I got over the initial depression. A close inspection of the ride file on trainingpeaks using zero smoothing revealed that there are definite speed drops that correlate to the power troughs that the hammer produces. These can be quite pronounced, and I think have a cumulative effect of a handicap on the rider in Zwift races.
After giving up on the race, I mounted my TT bike (which has an SRM) and did a few minutes so I could do a spreadsheet of the values. Sure enough, there were definite power troughs (down to zero in some cases) where the SRM was showing higher numbers. Zwift seems to really respond badly to zero watt readings. It doesn’t respond well to slight dips in power during races either.
Has anyone had any strange experiences while using the Hammer in ERG mode during a structured workout? Specifically, resistance buildup over the course of the entire workout?
I am currently using my cycleops fluid trainer and a PT hub. It works fine but the idea of not wearing out my rear tires, reduced sound, and a smart trainer has appeal. These power drop off issues sound terrible from a racing standpoint, particularly because if you fall off for a few seconds in a tight race (or even a competitive group ride), you have basically wasted the time; and I ride a trainer rather than the road because time is a premium.
I’d like to pull the trigger on a Hammer because the trainer seems otherwise sound and the best of the whole group of smart direct drive trainers, maybe train on the Hammer and race on the fluid trainer with PT until the bugs are worked out?
There is no guarantee they will ever get the bugs worked out and anyone buying a Hammer today will be very disappointed if they want to us it on Zwift.
Saturday my trainer dropped to Zero watts and my Avator stopped twice while climbing Box hill and then the usual Escalator issue (on the escalator my Avitar got of the bike and unclipped) yet when reviewing the power graphs afterwards I could only see one of the sharp power dips on Box Hill and the corresponding speed drop. The escalator drop was not present on any of the power readings. Except for the one drop the post ride measurements equaled what I was recording on my Garmin from my Stages PM
I will go back to using the Stages PM for the power reading on Zwift (Saturday I was seeing how the Hammer and Stages compared in their measurements), buying the Stages became an expensive work around to the Hammer problem.
Lastly, the snapping noise reoccurred while on Box hill also, another disappointment with the Hammer.
I feel compelled to give my input as well. I am also starting to doubt that Cycleops will figure things out. I think that maybe the mechanical design of the Hammer’s internal power meter is not able to generate consistent and accurate measurements. They’ve had a bunch of time since 2/12/17 to at least get another beta out and all we are getting is radio silence. Makes me think they are really struggling to find a firmware solution for the zero dropouts (and lag too). My solution is to not do races in Zwift for now. I am going to buy a PM for my race bike this spring and I guess I will have to use that bike for Zwift racing events. That’s a $1000 solution… I suggest doing some research on a Kickr. Seems like people are having more luck with that smart trainer.
Wow, it is just amazing to see even having all these big problems of the Hammer revealed here, people still would want to buy this POS with a wait for a fix. Man, what in the world are we living here? It is a $1,100 boat anchor if it does not work properly. If anything, I would wait until the fall to see if they have Hammer 2 out with some of these fundamental problems addressed. I personally would wait for KICKR 3. If they would address the noise and thru-axle issues with a KICKR 3, I would buy one immediately and sell you my Hammer for half the price I paid for it.
I was having similar problems with the power dropping out and with slow adjustment to changes in slope angle until I stopped using my Garmin heart rate sensor. I am now only using a Garmin cadence sensor and the Cycleops Hammer via ANT+ and I have had no drop outs after about 10 rides. Personally, I don’t feel I absolutely need a HR sensor so it hasn’t been big deal for me. I am using a Microsoft Surface 3 (not the Pro) and so it doesn’t have a whole lot of processing horsepower. I wonder if the Surface 3 can’t handle input from 3 devices or is there a different problem. I would be interested to know if others have seen an improvement by not using a HR sensor.
I tried no Garmin HR monitor today for a while and still had the same old lag and zero watt dropouts. Boy, I can’t wait for the next Firmware update - almost 3 months now but I’m sure it’s still a top priority over there at Cycleops…
Just wanted to let you know that, because I have a friend who works in engineering at Saris, I was able to let them know about the palpable displeasure being expressed here on the forums. I even linked them to this thread so they are now very much aware of the problems you all are experiencing, and more importantly, the potential problem of Zwift users not wanting to buy their products in the future.
Whether they can/will solve these problems w/ a firmware update is anyone’s guess. But for now, at least, they know that it needs to be a big priority for them if they want these products to continue to be in demand.
I’m changing out the groupset on my bike in about a week which means I’ll have a nearly new Stages left crank arm PM for sell if anyone is interested and wants an immediate fix.
It’s a Shimano Ultegra arm and I’ll sell for half what I paid for it.
I’m interested, how do you want to proceed?