Duncan Werner

Thanks, but right back at you: this forum has been great. I really appreciate all the feedback and it's been really helpful so far. --- It's probably karmic retribution (I was a bit dismissive about this concern the other day), but this morning we had a two-hour power outage in my neighborhood. I wasn't cooking anything at the time, but it definitely brought home how much of a problem this would be if I were. After the fact I found out that our local power company (PG&E) maintains a map of power outages on their website, and they're a lot more common than I would have expected. So food for thought. And regarding coolers: I picked up a 9 quart cooler on Amazon the other day for about $20, it works like a champ. I wasn't expecting all the interest in coolers, but now that I have one I can see why people like them.

I've been thinking about that, but I was worried about the case where you just unplug it when you're done cooking (without stopping it). I do that from time to time, and I don't want it to come on unexpectedly the next time. We could make it explicit - a "power lock" setting, that you have to turn on, which then will restart it on a power failure. I think that might be safe and we could indicate it with some icon on the screen. But I have to ask, is this all that common? We used to get a lot of power outages when I lived in Nantucket, but other than that it's been pretty rare for me. I've heard that it's common in Seattle as well. (Of course it might happen if there's a serious problem, like a tornado, but then cooking would the last thing you need to worry about). In any event, that's just software so it shouldn't be too hard at least to test. [Just incidentally, we actually don't have any EEPROM but we can write to the flash memory on the micro - program space - and that's where we store settings. If you're curious, it's this series]. I forgot to mention that there's another change, in response to something you posted - we changed the countdown timer to explicitly set hours and minutes (up to 95:59). We used to do a kind of lumpy adjustment, so it was 1 minute from 0 to 30, then increments of 5 minutes up to 2 hours, then half-hours, and so on. With this update you can explicitly set 0:31, or 0:36, and so on.

Here are some of the basic dimensions, in case it's helpful: sidekic_dimensions.pdf We're working on getting something on the website that will make this a little clearer, but it's taking longer than expected. And there is a beeper in the current hardware, but it's not terribly loud. It should be noticeable but it won't wake you up if you fall asleep. It beeps a couple of times when it hits the target temperature, and it beeps continuously if the countdown timer expires.

Sorry about that, veering off topic a little bit. It's a varistor; specifically a metal-oxide varistor. Basically, it does nothing as long as the voltage is below a certain level. If the voltage gets too high, it starts to conduct. The idea is that if there's a voltage spike, current goes through the varistor before reaching any parts which could be damaged by the high voltage.

It's an interesting thought. If we get it back in the lab we can check that out. There aren't many components directly on the line, and there's a MOV on there which should shunt surges, but the best way to know would be to look at the parts individually.

Let's talk about it off-list, I'll send you a note.

If it just rose 2 degrees, that's probably coincidental, I don't think it's working. As you say it's almost certainly the controller. If you power cycle it, there's no state preserved for the PID so that would suggest it's hardware - either a component failure or faulty board connection.

That's terrible, I'm sorry. Honestly I don't know what might have gone wrong - although I don't think there's anything you could have done which would make this happen. It was probably defective hardware. Can I replace it? We have the rev 2 machines ready starting next week. I'd like to have a look at and see why it stopped working.

So here's where we are with this: I think the machine is safe, but it's likely that our power cord is underspecified. That is, it should be better insulated. Rather than fix that, however, I think we'll follow dcarch's line of thought and fully ground it. To that end, we're going to put the machine on hold for a week while we can refit some inventory. We'll figure out how to handle the existing machines over the next couple of days. Let me just reiterate that I think it's safe, but this is the right thing to do so we might as well do it now. Particularly for people who don't have GFCI in their kitchen, it's better that we err on the side of being extra cautious. Thanks to everyone for the kind words here, I'll update when I have some more information.

Well this is very serious. I believe we are fully conformant, but I'll check with our consulting engineer to get an official position. Actually these are intended for hot water/coffee machines, so they're rated 105C. (Edit: that's not to say that they will live up to the rating, but that's what they're rated).

So I'm pretty sure we've figured this out. It's basically what I described above, except for the dodgy part about environmental temperature. Here is what is happening: (1) As I said above, it's a small amount of water boiling at the top of the heater. This happens when the heater can't transfer heat into the water fast enough. There's some steam generated as well, but this can be hard to see. (2) Heat transfer is more efficient when the water is at lower temperatures. It's easier to heat up water at 25C than 63C. That's also why it takes longer to heat from 40C - 60C than from 20C - 40C. And it takes longer when there's more water. So the reason that it was happening is that the water temperature was already pretty high, and you had a lot of water. I'm guessing that that cooler is 16 quarts. When it was trying to heat up to 63C (or maintain that temp) and it had to run full power, heat transfer was less than efficient so there was some residual heat, boiling the water at the top. When the power level dropped down, it was able to transfer more efficiently - hence no boiling and quieter. We normally don't notice this because while it may happen briefly at high temperatures, it's only temporary while the heater is running at 100%. In your case apparently there was more power required to heat the water (and probably the cooler insulation as well) so it ran making noise for some time. I can't quite account for why your stopping and starting made it go away - that might have been a coincidence, but we're still tinkering. However while it might be annoying (sorry about that), it's not going to cause problems except for additional evaporation. One resolution might be to impose a limit on the heater of say 75%, but that would slow down heating so I'd rather not do it if we can avoid it. We will keep investigating. --- Thanks again for the thorough testing and analysis!

Well it definitely makes sense that it would be related to power. I said before that I thought it was boiling water on the exposed heater element and I still think that's the case (I could be wrong, but working hypothesis). [incidentally, there's a small icon on the screen, at the top-left, which shows the power output as a bar graph. It's not precise - the bar graph has 7 levels, while the actual power output is a 16-bit value - but it's a quick way to see what it's doing]. So you have a heat source - the heater - and you have a heat sink, the water. In power terms, a supply and a load. The heater is generating heat and that's being absorbed by the water. In most cases the heat will transfer efficiently - there's so much capacity in the water that all the heat is absorbed, and the heater (including any portion which is exposed) will be the same temperature as the water. If the water line is low, relative to the heater, then all the heat generated won't be absorbed, and the heater will be hotter relative to the water. If the difference is large enough, just above the water line it will boil the water. Surface tension will pull a small film of water up the heater, to where it's hot, and will boil it. This is not that unusual - in fact you can make it happen, if the heater is running, by slowly lifting it out until it's less and less in the water (if it's at full power, and about 1/2 of the "coil" is exposed, this will always happen). But ordinarily it shouldn't happen if the heater is submerged, because (1) the heat transfer is efficient, it's designed for that, and (2) again the water should have a pretty large heat capacity, which we're actually increasing by circulating it. In the ordinary case, you can resolve this by adding more water so that the heater is better submerged. The design is such that as long as the water is above the middle of the bottom window, roughly, it will be sufficient. There may be slight variations, but the top of the bottom window is the ideal point. Now if the heater is running at 1/2 or 1/4 power, even if it's still slightly exposed, you won't get this effect because there's less power to transfer. In your cooler setup I would guess that you could make this happen by adjusting the set point while it's running - set it to a few degrees above or below the actual water temperature - because it will adjust the heater to either 100% or 0%. I don't mind telling you that we've been running around today trying to figure out what's happening to you. Generally, we're working on the theory that there's capacitance. So going back to the power metaphor, we have a supply and a load. We assume that the load is resistive - that is, (loosely), it will absorb power. But if it's capacitive, then it will absorb some amount of power and then stop. At that point, it won't absorb any more power and you get the heat feedback loop I described above. Why would that be the case? You're using a pretty big cooler, so there's a lot of water which would suggest just the opposite - that water should be able to absorb a lot of heat. However if you were running outside, in freezing weather, then it would be slightly different - the environmental temperature would "hold back" the transfer of heat, (because of the relative heat capacities of water and air). But that's not the case here. If there were a big difference between the environment and the set point that might still happen, but I don't think that's it either, unless your house is extremely cold. But it could be the cooler. As it happens, my 16 quart cooler just came in so I can do some experimenting. Assume for the moment that the cooler itself is keeping the water cold. Then we try to heat the water, but it won't absorb the heat, and we get some feedback. There's a limit to the cooler's ability, so it doesn't happen when the heater is at 50%, but it does happen at 100%. And then, over time, the cooler's own temperature raises to 63C. At this point it won't resist the heat anymore, so we don't have that effect. Sorry for the long post. I may know more (for example, I may know that I'm totally off-base) when I get a chance to run my cooler.

It's proportional.

Sorry for leading you down the wrong path, then. I still think that's what it is, although I have to get somewhere where I can hear that video a bit better. If there's a big difference between the environment and the water temp then it might happen at the surface regardless. I hope it's not too annoying.

The water should be below the top of the top screen. I'm reading our manual and see that we are somewhat confusing about that, apologies. I generally have it just around the bottom of the top screen. Around the middle of the top screen would probably give you the best level to allow for some evaporation. Note to self: mark water line on housing.