Ramping Function Has Steps

[SLOBrewer]

I'm running a temperature profile that is supposed to be smoothly ramping the temperature by 4 degrees over 4 days. In general it looks good but you can see some periodic corrections in the calculated setpoint. Interestingly they're happening at the same time each day during the ramp.

Repro Steps:
1. Create a temp profile that stays at 65 degrees for 1 day and then ramps to 69 degrees over 4 days.
2. Start a dual stage (heating/cooling) session.

Expected Results:
Temperature ramping should be a smooth function rising evenly over the 4 day rise period.

Actual Results:
Observe small step in setpoint at (24 + 10) hours into session and larger step in setpoint at (48 + 10) hours into session.

[hebnern]

It looks like this is probably caused by the device experiencing a watchdog reset. We save off the temp profile progress every 4 hours, so it will restart roughly where it left off, but may backtrack a few hours. It looks like that biggest jump is about a 3 hour backtrack.

[SLOBrewer]

Okay, this problem seems to have returned. I just did two sessions that ramped down at 1 degree per hour over 25 to 32 hours and both exhibited significant stepping on the way down. It also took significantly longer to hit the final temp than originally scheduled (roughly double the time). This is a big problem both because the ramping isn't achieved in the specified timeframe but also because it results in lots of ups and downs, which defeats the purpose of a gentle transition.

https://brewbit.com/dashboard/devices/588/sessions/10352?token=5ce610132a3ec464628eedf963f784a3

https://brewbit.com/dashboard/devices/589/sessions/10351?token=e95440ec1dba85bb9e4bf57b5747f3e4

[bmoney]

Can you share a screen shot of your profile that was used?

[SLOBrewer]

[bmoney]

What software version are you running?

[SLOBrewer]

1.6.0 on both.

[SLOBrewer]

One interesting note is that both batches had a primary fermentation that had a ramping up from one temp to another and there was no stepping observed. So it seems to be something either with more significant ramps (more than 5 °F), ramping down, or ramping from the start of the program.

[grantanddenise1]

I am also having this issue and appears to have got worse with the recent updates. It seems to only occur during temperature ramping. During this period the temperature was set to ramp from 20C to 28C over 4 days (2c per day). As you can see from the trace there are clear step downs in the ramp, such that the temperature only rises less than a third of what it is meant to in the given day ie. 0.6C per day. In the end I had to change to a static set point and manually increase it myself periodically. You will also note that the temperature readout and set-point is in degrees F whilst the unit is degrees C, and the graph is in degrees C. Really confusing.

[fos]

I'm also seeing this issue, and have been unsuccessful at getting the Brewbit to properly perform the temperature ramp function.

Using a temp profile intended to ramp the beer from 66º to 40º over 4 days and hold at 40º:

I am observing the following session:

[bmoney]

The unit saves of its profile execution progress every 4 hours.When the unit experiences a reset it resumes the profile at the previously saved off point.

[grantanddenise1]

So this is the cause, what is the fix? That is a lot of resets.

[SLOBrewer]

Can we get an update on this? It's not something so benign as minor stepping that ultimately resolves itself. I've had several temp profiles entirely fail to be effective as a result of this bug. Right now I have a beer that refuses to step up to 72 °F even though it should have done so long ago. It's stuck at it's pre-ramp temperature, starts to climb, and then drops back down. At this point I consider ramps to be entirely non-functional.

[SLOBrewer]

Here's an example where at least the ramp actually happened but, oh my goodness, that's not a happy transition. This should be an easy to track down as it's happening on 100% of my ramps across two different devices across several different temperature profiles. In other words, every time I do a ramp of any kind I see this behavior.

[hebnern]

There are two things I can think of to address this:

1. Figure out the cause of resets and fix them. This probably entails adding additional logging to the device so we can track down exactly where this is occurring.

2. The server always knows where the device is in its profile so it could send the device an offset when it connects to get it back on track.

I am working nights and weekends at work for the next few weeks, so I probably won't get around to working on this for a while. If anyone wants to take a crack at it, I'd be happy to provide pointers.