Braumeister NEXTGEN Build

[tuckerle71]

I also wonder why they left so much space below the lip in the malt pipe?

On the 50L Braumeister, there are two elements. One inside the base of the maltpipe and one between the maltpipe and the outside wall. The temperature sensor is inside the base of the maltpipe, if I remember correctly.

 

[Edak]

On the 50L Braumeister, there are two elements. One inside the base of the maltpipe and one between the maltpipe and the outside wall. The temperature sensor is inside the base of the maltpipe, if I remember correctly.

Thanks Mr. Tucker but you were beaten to the answer.

 

[Malted]

I do also have a question. do you think the shape is important, I'm assuming the reason the braumeister is high with a small diameter is to allow the pump to push water through as much grain as possible, does this sound correct? If this is true, then from some looking around the 20L version seems to have more options for pre-made pots. A lot of the Pots around 85L seem to be shorter with larger diameters.

Yes I would suggest a cylindrical shape would be ideal.

Perhaps shape is important to optimise extraction efficiency?

"There are two ways to improve the uniformity of the flow: increase the depth of the grainbed or add more drains. Increasing the depth of the grainbed (see Figure 168) puts more of the grain higher up in the regions of flatter gradients. Adding more pipes and spacing them efficiently (see Figures 169-173) also flattens out the pressure gradients and make the flow through the grainbed more uniform."
Tun Geometry and Flow Potential. Palmer, How to Brew: http://www.howtobrew.com/appendices/appendixD-2.html

"A lower fluid velocity and a uniform distribution of collection points are desirable for achieving maximum extraction efficiency from the lautering process." http://morebeer.com/brewingtechniques/libr...3.4/palmer.html

Deeper grainbeds have more uniform rinsing, all else being equal but some say that a tall/skinny vessel will be more prone to stuck runoffs because of the greater weight/area pushing down on the grain. On the other hand a thinner grain bed from a shorter/wider MT takes longer to get a clear runoff because you're running through a thinner filter bed.

I have seen on the interwebs that a height to width ratio of 1 to 2 is ideal for a brew kettle to promote evaporation. However, the taller a pot is given a fixed volume, the taller the headspace will be and less chance of boilovers (eg preboil volume of 66L in the 85L vessel of the 50L BM).

I'd say Speidel have one foot in each camp because if my maths are correct, the 50L BM has a height to width ratio of 1.2 to 1 and if you're correct about the 20L BM dimensions, it would appear to be a height to width ratio of 1.15 to 1.

 

[tuckerle71]

Thanks Mr. Tucker but you were beaten to the answer.

That'll teach me not to refresh my browser when I get to work.

 

[angus_grant]

Hey gang,

I have finally got around to doing some temp maintaining runs on the sparge boiler. And I am coming up against some temp reading problems.

I have a 40mm thermowell which I put a sheathed SB18B20 that I purchased. There is a temp difference of 1.5 degrees Celsius between temp probe and digital thermometre I have inserted into the water.

Do you guys use some thermo-paste, a naked DB18B20 temp probe?? Or do you code for the difference in temp?

My PID settings are not far off. I think I need to adjust the integral variable a bit higher to actually get to my ramp temp (which in the tests was 40). It consistently sat on 39.1 as per probe and would not go higher. Which equated to 40.5 measured temp of water.

How are people handling temp reading?

Thanks,
Angus.

 

[Edak]

Hey gang,

I have finally got around to doing some temp maintaining runs on the sparge boiler. And I am coming up against some temp reading problems.

I have a 40mm thermowell which I put a sheathed SB18B20 that I purchased. There is a temp difference of 1.5 degrees Celsius between temp probe and digital thermometre I have inserted into the water.

Do you guys use some thermo-paste, a naked DB18B20 temp probe?? Or do you code for the difference in temp?

My PID settings are not far off. I think I need to adjust the integral variable a bit higher to actually get to my ramp temp (which in the tests was 40). It consistently sat on 39.1 as per probe and would not go higher. Which equated to 40.5 measured temp of water.

How are people handling temp reading?

Thanks,
Angus.

Hi Angus,
You should always use thermal paste to ensure that the sensor couples with the thermowell. My sensor was spot on so I didn't need to make any adjustments, but if yours has an offset then apply it after the measurement is taken in software.

If your temp is not reaching target then you increase the P factor slightly, the I factor should not be relied upon for this. I is intended to correct long term small errors.

My analogy is P is the accelerator and D is the brake. If you don't accelerate enough you won't get to your target but overdoing it will require the brakes. Once you reach your destination you don't use either pedal and if you are on a slope you might start rolling away from the target and need to apply accelerator again. The slope is related to the I factor.

edit: are you sure your digital thermometer is right?

 

[glenwal]

There is a temp difference of 1.5 degrees Celsius between temp probe and digital thermometre I have inserted into the water.

Are you sure there is a difference, or is there just a lag because you're not using thermal paste? If it is actually a difference, there should be no problem with correcting it in software.

If its lag, then you don't want to be adjusting or you're measurements will then be out once the probe has come up to temp. To reduce the lag, you want to make sure the heat can transfer as eaisly as possible from the liquid to the temp probe (eg. by using thermal paste)

 

[angus_grant]

Thanks Edak. I may end up having to strip off the little stainless shield that is placed over the sensor to get good contact with the thermowell.

Off to Jaycar tomorrow night to get some thermal paste.

Thanks for the analogy. Certainly when testing my PID code with the little kitchen kettle it was all about "applying the brakes" so maybe why I got stuck on making adjustments with integral adjustments. I had no problems ramping up temps. So it is all a bit different when heating up 10L of water in a big pot.

I will check my digital thermometer against another digital, and also against my voltmeter which can also measure temperature. From previous checking the other digital one was out of whack a bit though. Maybe I should pony up for an proper industrial glass thermometer as well. You can't have too many thermometers, can you?

 

[angus_grant]

Glen: Something else to test.

I was running the tests last night last thing before going to bed so didn't have time to do some decent investigation. I have a feeling it may be that the temp probe is not "attached" fully to the system, and not really lag. Something the thermal paste should help equalise.

Maybe eventually the thermowell will reach the same temp as the water, but the problem is that will result in temp overshoot as it would take the thermowell a long time to match internal temps to external temps

 

[Edak]

So it is all a bit different when heating up 10L of water in a big pot.

And it is even harder when you have over 25L of water and grain...

To further complicate my above analogy; unlike a car, the accelerator in our PID can only be fully on or off and you are always going up hill (because you apply heat to something that naturally cools). Adding insulation reduced the incline of the slope. Also, when I say "brakes", given that we don't actively cool, it would be more like the car has some intelligence which detects the distance and speed and cuts off the accelerator rather than applying brakes. This means D overrides P.
The I factor is hard to explain, but it could be like a back-seat driver (kids) going "are we there yet?". After a while you get frustrated with the little *******s so you plant the foot down a bit more. Actually that is a terrible description so I will end it there but note that it is an accumulative value that can get out of hand if left to accumulate over a long period of time, which in this case would be during ramping.

In my case:
* I have reduced the incline of the slope dramatically by adding a jacket and a lid (fully covered during mash and partially covered with a domed lid during boil).
* I have a 1-second window, which means I can apply the accelerator once per second (The duration of which is dependent on my accelerator and the car overriding it)
* To combat the "I" value getting out of hand during ramp times, a trick I use is to only start accumulating the "I" when the measured temperature is near the set point within a tolerance of about 0.5 degrees. This is great for taking it out of the original equation when ramping because it will not really contribute to overshoot but will help the system to hold the temperature a bit better.

 

[angus_grant]

yeh, when I was working out the PID variables on my kettle testing I knew I'd have to change them fairly radically going full size, and major changes again when I add grain.

I'm going to get some spent grain from someone to do some testing with the grain mass in there as well. Or I may do a BIAB with the grain I currently have in the sparge boiler and then use that for testing the full vessel.

I'm going to add insulation in an effort to reduce ramping times and increase boiling rate, and it will provide some thermal protection during the mash process.

To combat PID kookiness, my code fires the element full-tilt until the current temp is 85% of target temp and then switches over to PID control. I will tweak this percentage when working with full water and grain, but it will probably increase towards 100%

I look forward to the next 4 pages as we all try and develop an analogy for the PID algorithm. he he...

 

[angus_grant]

Added some thermal paste last night and the temps are now closer. Just inside 1 degree difference

I used a digital thermometer I grabbed from CraftBrewer and had another digital thermometer I grabbed from another homebrew store and none of them agree. The two digital temp probes are touching each other in the pot so theoretically should read almost the same, but just over 1 degree difference.

I am now thinking about getting one of these and inserting my DS18B20 which is sheathed inside a stainless shield (from manufacturer). This should get me fast results at least.

So how do I test the accuracy of the DS18B20? Which thermometer do I trust as being correct? Or do I just assume the DS18B20 is accurate (or more to the point, it should be consistently inaccurate and I just adjust my systems) and move on with my project. :huh:

 

[Edak]

Added some thermal paste last night and the temps are now closer. Just inside 1 degree difference

I used a digital thermometer I grabbed from CraftBrewer and had another digital thermometer I grabbed from another homebrew store and none of them agree. The two digital temp probes are touching each other in the pot so theoretically should read almost the same, but just over 1 degree difference.

I am now thinking about getting one of these and inserting my DS18B20 which is sheathed inside a stainless shield (from manufacturer). This should get me fast results at least.

So how do I test the accuracy of the DS18B20? Which thermometer do I trust as being correct? Or do I just assume the DS18B20 is accurate (or more to the point, it should be consistently inaccurate and I just adjust my systems) and move on with my project. :huh:

Looks like a good thermowell idea!

I used this sensor in my system, which is a DS18b20 already in a thermowell.

For calibration, although I could use a multitude of calibration tools which I have at work, I usually trust a good old thermometer, not a mercury one though. I am lucky enough that all of my thermometers, analog and digital agree on what the temperature is.

 

[bonj]

I am now thinking about getting one of these and inserting my DS18B20 which is sheathed inside a stainless shield (from manufacturer). This should get me fast results at least.

That's exactly what I have here. I will be using it in a t-piece on the mash return to the tun.

As for accuracy, my perspective is that every system is different and needs to be learnt anyway, so why worry about inconsistencies between sensors, and just start brewing with it. You will soon figure out if you need to adjust your mash regimes.

 

[matho]

hey angus the ds18b20 is accurate to +- 0.5 deg c over a range of -10 to +85 deg, I have used it along side of my brewing thermometer and they both say the same thing. I have found that my digital thermometer is different by about 1.5 deg at mash temp but then agrees with the lab thermometer at boil which the ds 18b20 doesn't. The actual temperature isn't all too important just repeatability is.

cheers steve

 

[woodwormm]

I used this sensor in my system, which is a DS18b20 already in a thermowell.

can anyone explain the wiring on these probes? the site says it's single wire interface, yet shows 3 wires and the thermometer they link is 12Volt, . Is there any chance these would work on an STC1000?

 

[bonj]

can anyone explain the wiring on these probes? the site says it's single wire interface, yet shows 3 wires and the thermometer they link is 12Volt, . Is there any chance these would work on an STC1000?

Unlikely they'd work on an STC1000. These are digital sensors.

The 3 wires will be +5V, DATA, and GND

I can't remember the colours, but DATA will likely be YELLOW, and GND either BLUE or BLACK, +5V should be RED

 

[angus_grant]

:lol:, nice to hear that no-ones readings match across different thermometers.
How on earth do we make decent beer when we can't get the mash temp accurate to 0.1 deg?
OMG, I'm freaking out!!!!!! DMS, stuck ferments, no bubbling in airlocks!!! waaaahhhhh!!!!!!!!!!! <<SLAP!!!!!!!>>


As I alluded to before, as long as the temp readings are consistently incorrect I can adjust the software. And Matho's comment about repeatability. And confirmed by Bonj's comment about brewing on the damn thing and adjust later..

Edak's temp sensor is a cheaper option ($US10) than what I linked to. $US13 for compression fitting + $US20 for temp probe + postage from US. What I currently have is the DS18B20 sensor in a stainless tube but will require some form of compression fitting to the pot. Looks like I'll be ordering four of Edaks sensors. One for the sparge boiler, one for the main system, and two for my fermentors as I am going to convert them over to software controlled instead of the FridgeMates.

 

[woodwormm]

Unlikely they'd work on an STC1000. These are digital sensors.

The 3 wires will be +5V, DATA, and GND

I can't remember the colours, but DATA will likely be YELLOW, and GND either BLUE or BLACK, +5V should be RED

aaah, i'm pretty well educated on 240 and 12 volt systems... now gotta learn digital, but that makes sense. so essentially they've got a power supply (+5V), information supply back to equipment/thermometer (DATA) and a Ground.

looks like a pretty cheap, super useful thermometer to couple the probe and the readout together.

 

[bonj]

aaah, i'm pretty well educated on 240 and 12 volt systems... now gotta learn digital, but that makes sense. so essentially they've got a power supply (+5V), information supply back to equipment/thermometer (DATA) and a Ground.

looks like a pretty cheap, super useful thermometer to couple the probe and the readout together.

Yep, the DATA wire is bi-directional using the "OneWire" protocol. Basically, the sensor has a unique 64 bit address and is polled individually. When it receives a request addressed to it, it will send the data back on the same wire. You can put a number of them (I think 15) on a single data wire and poll them using their unique addresses.