I had the opportunity to use a Scace thermofilter hooked up to a Fluke yesterday on a friends GS3. The boiler temp was set at 200f and the Scace readings several times ramped up to 205f within 3-4 sec and stabilized . I did do a crude check of my Fluke by putting it in boiling water and it read 212f. Was the GS3 temp off or is the Scace temp reading not the same as the boiler temp? If so on double boiler machines with PIDs are you really reading the true brew temp by using the machine reading of the boiler temp. I would be interested to have Greg's comments as I suspect he has experience in this...

I take it that you did have the Fluke set for "T" Type Thermocouples, which are what the Scace Devices use?

I have the Scace II and also use it with my GS/3 and the readings are spot on.

Did you check GS/3 Coffee Boiler Temperature Adjustment?

jmp wrote:I had the opportunity to use a Scace thermofilter hooked up to a Fluke yesterday on a friends GS3. The boiler temp was set at 200f and the Scace readings several times ramped up to 205f within 3-4 sec and stabilized . I did do a crude check of my Fluke by putting it in boiling water and it read 212f. Was the GS3 temp off or is the Scace temp reading not the same as the boiler temp? If so on double boiler machines with PIDs are you really reading the true brew temp by using the machine reading of the boiler temp. I would be interested to have Greg's comments as I suspect he has experience in this...

Sounds like the boiler temperature offset value is incorrectly set in the machine. Water doesn't gain heat between the boiler and the group. It always loses heat, so the temp at the group is less than the temp within the boiler. LM uses an offset correction to account for this, so that you get your desired temperature at the group, not at the boiler.

Lemme know if you need more he'p.

-Greg

Interesting. I posted the same observation not long ago, but got somewhat different answers.

I think some of the responders are confused by the way you phrased it. When you say the boiler temperature is set to 200F, that's really the boiler temperature minus the offset (i.e., the target brew or group temperature.) If the Scace rises to 205F during the shot, it's likely that the offset is programmed at about 5-6 degrees.

This isn't about the offset being wrong or your Fluke not being calibrated. It's about the GS/3 temperature profile during a shot.

My Scace shows that the temperature at the group increases 3-4 degrees after starting the shot, then stabilizes, just like yours (and, yeah, I set the digital thermometer for a type T thermocouple and calibrated over a steam bath.) For example, if I target a brew temperature of 196.5F, then I must set the boiler temperature to 200.5F. That's because LM programmed my machine with a temperature offset of 4F. When I run the shot, the Scace reading starts out at 196.5F, but quickly ramps up to about 200F.

Greg is correct that the water temperature in the boiler is higher than at the group head, presumably because the group head sticks out in space and there's less thermal mass to keep the water at the same temperature (i.e., heat dissipates more easily from the group head than the boiler.) That's why the temperature offset exists: it's designed to more or less accurately show you the temperature of the water in the group head, taking the heat loss into account.

Here's what I think is the explanation: When you start brewing, the cooler water that's resting in the group flows through the gicleur into the 3-way valve and thence through the TL30 tube to the dispersion block and puck. That's why initially you see the target temperature. But at the same time, hotter water flows into the group head from the boiler, raising the temperature of the water in the group. The gicleur (intake) is positioned at the back of the group head, close to the opening into the boiler, so the water around it is heated quickly. I suspect that the cooling effect of the group head isn't instantaneous, especially back near the boiler, nor is the cooling that takes place during the trip through the gicleur, 3-way and TL30. I think this is why the Scace shows the exit water ramping up 3-4 degrees. I would speculate that with +4F water flowing continuously into the group head over a 30 second period, there isn't time for the excess heat to be lost, and that's why we see stabilization at higher temperatures. I suppose there might be some counteracting effect of cold water flowing into the brew boiler (the temperature there usually drops .5F or so during a shot), but remember that the intake water is pre-warmed via heat exchange in the steam boiler, and the PID is working hard to keep the brew boiler at the temperature you set.

Going back to what Greg said, while it's true that the water won't be higher in the group than it is in the boiler, it most certainly is possible for hotter water from the boiler to flow into the group and raise the temperature of the water that's already there.

[NB, it would be interesting to see a temperature profile on a Speedster. The gicleur is way out near the dispersion block, so maybe the water temperature doesn't rise as much.]

I would think LM knows all this. So why do they bother having an offset that shows only the resting temperature of the group and not the actual brew temperature? Perhaps it's because the actual brew temperature is an average of the temperature curve throughout the shot, so it's neither the starting nor maximum/stabilized temperature during the shot. Hard to say exactly what the temperature is, isn't it?

I have no explanation for EricC's claim that the Scace readings are spot-on, presumably meaning that the temperature stays at the target value (i.e., brew boiler temperature minus the offset) throughout the shot. Frankly, given the geometry of the group head and water flow, I don't see how this is possible. But I'm just an amateur at this.

Focusing on the fact that we both see a rise in temperature, what does this mean? If the roaster recommends a brew temperature of 200F, does that really mean the water hitting the puck has to be 200F, or are they using a similar LM machine (likely in many cases) and they're simply reporting the target temperature they set on the machine (i.e., they're actually brewing several degrees higher)?

The best advice I got was to treat the temperature as relative. Use the roaster's/reviewer's recommendation as a starting point and adjust the temperature for best taste. That's what people who don't have LM machines have to do anyway. I could lower my offset to 0-1 degree, to better report the temperature that will prevail at the puck during most of the brew cycle, but since it's unlikely roasters or other GS3/LM owners do this, the information wouldn't be particularly useful. In other words, regardless of what the offset means, it's likely that LM uses the same procedure to set it on all their machines, establishing a common reference that makes it possible for LM owners to share temperature information.

The good news is that no matter what approach you take, the saturated group makes the temperature curve predictable, stable and reproducible, which is the most important thing for finding the right temperature and using it to consistently produce the same shot profile.

Yes I did mean the temp on the gauge outside the machine which after reading represents the boiler temp plus the offset. As I said it was a friends machine and my only experience with it before was looking at it from afar and being jealous of someone having a double boiler PID machine. My Fluke is set on the type T thermocouple but I checked with Fluke on calibration and they recommended a crushed ice bath instead of boiling water as being more accurate. Indeed by this method my meter was almost one degree too high. This still means that on this GS3 there was around a four sec rise in temp above the reading on the gauge and it seems the temp profile isn't exactly flat as i thought and maybe the offset is not correct. I have a Simonelli Appia that after using the Scace as feedback I can dial in a brew temp that is very flat(less than a degree change). I have to use a a time consuming flush and recovery routine but maybe I shouldn't be so jealous of the GS3 if the temp during the pour varies that much. Greg I also got that improvement on your Scace and it does work as it held up to a lot of use during the past few months without breaking...thanks

Peppersass wrote:I have no explanation for EricC's claim that the Scace readings are spot-on, presumably meaning that the temperature stays at the target value (i.e., brew boiler temperature minus the offset) throughout the shot. Frankly, given the geometry of the group head and water flow, I don't see how this is possible. But I'm just an amateur at this.

Apologies for any confusion, yes, of course what I meant to say was that once the initial ramp up had taken place that the temperature for the shot was spot-on the set temperature (taking into account the offset).

Regards
Eric

EricC wrote:Apologies for any confusion, yes, of course what I meant to say was that once the initial ramp up had taken place that the temperature for the shot was spot-on the set temperature (taking into account the offset).

That's what I thought you meant, but it's not what I see here. After ramp up, the group temperature is 3-4 degrees hotter than the set temperature less the offset (i.e., pretty close to the set temperature.) It seems like the OP is seeing what I see, not what you see.

Given the geometry of the group head, it's hard to understand why there's a difference in what we see. From what I've seen posted here and on Gs3cafe, the typical programmed offset is -4F to -7F degrees. Sure, the offset programmed by LM into my machine (-4F) may be incorrect, but if that's the case the correct value should be 0F or -1F. This would indicate that the water temperature in my group head after the ramp up is barely less than the temperature in the boiler. This seems unlikely, given that it's sticking out in space. And why would my group head be several more degrees efficient at retaining heat than yours?

One possible explanation might be flow rate. Mine is fairly healthy, about 450 ml/sec. Well within spec and about what you would expect. If your flow rate is significantly lower, then maybe that would explain the difference. However, I'm skeptical of that explanation because, again, it implies that there's virtually no heat loss in my group head.

Another possible explanation is that the temperature sensor in my brew boiler is faulty, and the actual boiler temperature is higher than the temperature I'm setting. Temperature experts: Would this be a typical failure for the type of thermocouple used in the GS/3?

wrt failure mode of thermocouples - usual failure mode is open circuit. And ain't no way that heat gets added from the ambient, either. The offset's not correct. Actual boiler temp has to be several degrees higher than the group.

-Greg

gscace wrote:wrt failure mode of thermocouples - usual failure mode is open circuit. And ain't no way that heat gets added from the ambient, either. The offset's not correct. Actual boiler temp has to be several degrees higher than the group.

The last two sentences appear to be in conflict on my machine. The offset would have to be 0 or -1 degree to match the reading I get after ramp up, not minus several degrees.

Suppose the thermocouple is covered with scale? Would that potentially cause it to read several degrees lower than the actual boiler temperature? Something like that would explain what I'm seeing.

Otherwise, it doesn't seem likely that either the thermocouple in the Scace or the GS/3 could be off by 3-4 degrees. I got the same reading from the first Scace I had (the one I returned.) I calibrated the digital thermometer using a steam bath and adjusted for altitude (~700' ASL.) It was only off by about .7 degree as I recall. Doesn't seem likely the device is bad, but maybe I can figure out a way to verify that.

I never suggested that heat has been added from the ambient. What I suggested is that since the gicleur is positioned very close to the opening into the boiler, hotter water from the boiler is raising the average temperature of the group water that enters the gicleur. Water from the boiler only has to travel a short distance from the boiler opening into the gicleur, then through a 1" tube to the 3-way, then though a tube that runs the length of the group head into the dispersion block. Both tubes run through the hot water that's been sitting in the saturated group, so presumably the heat loss is minimized (though not zero.) The question is, how much heat is lost on that journey through the saturated group? I ran a 3-second warming flush before mounting the Scace and taking the readings, so the water in the group was somewhat hotter than it would be at idle. I can see where water sitting in the group at idle is several degrees colder than the water in the boiler, and this is the water that first enters the gicleur, but my theory is that within a few seconds hotter water has moved from the boiler into the vicinity of the gicleur.