Bigdaddyale
Well-Known Member
is there any science behind blocking out the focal field of the sensor? seems to me there should be a better solution using the information to solve the problem. It's electronics
Brew Bubbles: Web-Enabled Airlock
Help Support Homebrew Talk:
ChrisThomas
Well-Known Member
I'm thinking the same thing, all I've proved with this "test" is the technology CAN work. I'm going to get the 'scope out and see if there's any change on the receiver that is easily measurable/detectable. That is, can a threshold be set to trigger the count.
My current fermentation is nearly finished so my real life useage is going to be down for a few days 'till I get my next batch on the go. Or I could just blow into lock I suppose?
Chris
Naesstrom
Well-Known Member
I just attached my first device today and it jus shows 0.0 in bpm so I'm going to try this. Thanks for the tip! Could you by any chance measure how far from the bottom the hole should be it would be fantastic!
OP
OP
Would a few of you please measure the width of the tube part of your airlocks? The part where the sensor goes. I'd like to get an idea of how wide it is on most airlocks to see if we can come up with a sexier solution.
ChrisThomas
Well-Known Member
So here's my kit of parts from yesterday;
Ordinary household scissors, map pin, cardboard and double sided tape.
I used the scissors to cut a strip of cardboard roughly the same width as the sensor housing. I then cut a piece off that strip that was just a bit shorter than sensor housing. I then put some double sided tape covering the bottom part of the cardboard, so the tape didn't cover the receiver only hold the cardboard in place, and could be repositioned easily. Then by eye I made a very small hole in the cardboard that would be right above the middle of the receiver. To make the hole I only pushed my pin so the end was just coming through and I then turned the cardboard over and did the same from the other side. This hole size seemed to work for me, as when I pushed the pin all the way through it didn't make any difference to the detection, so the hole size does matter. To big a hole and it doesn't work. I then positioned the cardboard in place and fine tuned the position until I was getting successful bubble counts.
This is what my working arrangement looks like;
This bit of tinkering is just to prove that this concept can work successfully, I don't necessarily see this as the final solution. I picked these things as everyone should have something similar lying around and get their own solution working for now.
A more permanent aperture could be fashioned. But as @Bigdaddyale said we should be able come with with an electronic solution and that's what I'm going to have a look at today. But my fermentation has stopped, time to get some tube into the bottom of a lock and start blowing?
Good luck everyone!
Chris
Ordinary household scissors, map pin, cardboard and double sided tape.
I used the scissors to cut a strip of cardboard roughly the same width as the sensor housing. I then cut a piece off that strip that was just a bit shorter than sensor housing. I then put some double sided tape covering the bottom part of the cardboard, so the tape didn't cover the receiver only hold the cardboard in place, and could be repositioned easily. Then by eye I made a very small hole in the cardboard that would be right above the middle of the receiver. To make the hole I only pushed my pin so the end was just coming through and I then turned the cardboard over and did the same from the other side. This hole size seemed to work for me, as when I pushed the pin all the way through it didn't make any difference to the detection, so the hole size does matter. To big a hole and it doesn't work. I then positioned the cardboard in place and fine tuned the position until I was getting successful bubble counts.
This is what my working arrangement looks like;
This bit of tinkering is just to prove that this concept can work successfully, I don't necessarily see this as the final solution. I picked these things as everyone should have something similar lying around and get their own solution working for now.
A more permanent aperture could be fashioned. But as @Bigdaddyale said we should be able come with with an electronic solution and that's what I'm going to have a look at today. But my fermentation has stopped, time to get some tube into the bottom of a lock and start blowing?
Good luck everyone!
Chris
ChrisThomas
Well-Known Member
The outside dimension of where my sensor is positioned is roughly 7mm, maybe a bit more but not as much as 8MM, measured by eye. I've checked on a couple of other locks and they're the same.
Chris
P.S. i think I've just blown up a sensor! I had two PCBs assembled and added two wires onto the sensor PCB pads to attach my 'scope, it doesn't count any more! Moved the wires onto the D1 mini pads with another PCB and all appears to be fine.
![Craft A Brew - Safale S-04 Dry Yeast - English Ale Dry Yeast - For English and American Ales and Hard Apple Ciders - Ingredients for Home Brewing - Beer Making Supplies - [3 Pack]](https://m.media-amazon.com/images/I/41vpOQ2OvTL._SL500_.jpg)
Naesstrom
Well-Known Member
Integration with brewfather seems to be working really well... you can never have to much data right! 
(green is bpm) Fun to see the correlation between fridge temp (black) and BPM (green) with that little "bump"
About the Airlock, I'm using this model and I found that I had to tilt it a bit forward for the sensor to be positioned in the right spot, was easy to see when the led when dark and the signal was blocked. Thinking that maybe a more elongated hole so that it can be adjusted would be an idea... will try and redesign your 3d model and test it out!
(green is bpm) Fun to see the correlation between fridge temp (black) and BPM (green) with that little "bump"
About the Airlock, I'm using this model and I found that I had to tilt it a bit forward for the sensor to be positioned in the right spot, was easy to see when the led when dark and the signal was blocked. Thinking that maybe a more elongated hole so that it can be adjusted would be an idea... will try and redesign your 3d model and test it out!
ChrisThomas
Well-Known Member
I've definitely destroyed a sensor! Built another PCB and all's well. Seems that the receiver is a bit sensitive to heat?
I've read the data sheet for the GP1A57HRJ00F and it implies that it's a detect/not detect output. From my quick test with another lock, filled with just water initially and with nothing in front of the receiver, there's no change in the signal that is detectable.
Placing my "filter" Infront of the receiver and using the same lock and water I can detect the meniscus. I've done a very short video of my quick test but can't upload it, it's 1.1 MB.
Chris
I've read the data sheet for the GP1A57HRJ00F and it implies that it's a detect/not detect output. From my quick test with another lock, filled with just water initially and with nothing in front of the receiver, there's no change in the signal that is detectable.
Placing my "filter" Infront of the receiver and using the same lock and water I can detect the meniscus. I've done a very short video of my quick test but can't upload it, it's 1.1 MB.
Chris
OP
OP
If Chris has a scope, the answer is "maybe." I'm more than happy to adjust the circuitry if we can prove there's a way to "trim" the sensor a bit. But, there is a science to this. As I said the aperture is creating a collimator which is like a small pinhole camera and it focuses the light. The sensor will happy detect any light, that focusing is probably letting it only see straight in front.
The reason I'm asking for some measurements is to see if we can maybe add a 3D printed collimator to the bracket. The width of the sensor is 10mm. If we can assure ourselves that most airlocks are 8mm or less, then a printed aperture is possible.
OP
OP
I thought I read a question here somewhere about the temperatures behaving strangely when switching from C to F or vice-versa. Assuming I did not dream that:
This behavior is a known thing and it's on my list of things to (maybe address). The reason is I use a circular buffer to filter the temperature and bubble readings. The buffer has a length of 5 and the sensors are polled every minute. The bubble counter itself is a ~60-second average. I count every pulse and when polled for the value, I divide it by the actual milliseconds passed since the last poll. The temperature is a point in time reading every minute.
Since I use a circular buffer the temps look like this:
| 70.2 | 69.8 | 69.9 | 70.1 | 71.2 | = 70.2
If we change to Celcius, it looks like this after the first minute:
| 21.2 | 70.2 | 69.8 | 69.9 | 70.1 | = 60.2
It's going to take five minutes to actually have five readings in Celcius and therefore a valid reading.
There are several ways for me to "fix" this but there are two reasons I don't see it as a critical issue needing to be fixed right away:
Oh, and be sure to keep the heat on @Thorrak to add this to Fermentrack. (Actually, he HAS been working really hard on something else, so be patient.)
This behavior is a known thing and it's on my list of things to (maybe address). The reason is I use a circular buffer to filter the temperature and bubble readings. The buffer has a length of 5 and the sensors are polled every minute. The bubble counter itself is a ~60-second average. I count every pulse and when polled for the value, I divide it by the actual milliseconds passed since the last poll. The temperature is a point in time reading every minute.
Since I use a circular buffer the temps look like this:
| 70.2 | 69.8 | 69.9 | 70.1 | 71.2 | = 70.2
If we change to Celcius, it looks like this after the first minute:
| 21.2 | 70.2 | 69.8 | 69.9 | 70.1 | = 60.2
It's going to take five minutes to actually have five readings in Celcius and therefore a valid reading.
There are several ways for me to "fix" this but there are two reasons I don't see it as a critical issue needing to be fixed right away:
- People are not going to switch back and forth a lot
- I am working on other projects (finally!) and doing another change right after a release is something I'm actively avoiding.
Oh, and be sure to keep the heat on @Thorrak to add this to Fermentrack.
(Actually, he HAS been working really hard on something else, so be patient.)ChrisThomas
Well-Known Member
I have looked at the signal output from the sensor with my 'scope and I've also had a read of the datasheet. There is nothing I can see on the output from the receiver that is usable. My interpretation of the datasheet is that is definitely a go/no go output, the trigger level for this is set internally. There's nothing I can get from the datasheet that says that this trigger level is user adjustable.
If there is something that can be mass produced and is more reliable and robust than a small offcut of cardboard and sticky tape then I for one would be interested. What I have found is that the relative positions of the lock and pin hole are sensitive. Now, it may be that a different size hole would not require "setting up", but I haven't tried that myself.
@wd16261's idea of shrink wrap is a good one, but not everyone has shrink wrap or the ability to shrink it safely. I've already destroyed one sensor whilst soldering on my test leads.
Not sure what the next step is, but happy to help/test... Mine is to get some fruit juice from the supermarket!
Chris
OP
OP
I think we can come up with something. Need some more measurements from folks to make sure I'm not barking up a light pole instead of a tree.
This is the location were I made a hole in my shrink tube
I was able to shrink it safely with a lighter but this gave me less result, I think its because the hole came closer to the emitter so that the focus changed. The best result for me is just by sliding the tube (with a max 1 mm hole at 10mm) over one site an leave it like that.
I just put a small bit of glue on it so that it doesn't move when I put it on the airlock.
The measurements of the tube of all my airlocks is 9 mm so there is not much space left. Actually with the shrink tube(wrap) almost none!
I know that it would be nice if these small changes were not necessary but I've already spend a lot of time to get mine working that I'm happen with this solution and that I finally can start using it!
@LBussy I completely agree with you that the wrong temperature after switching between °F and °C is a minor issue it was just in case you didn't knew.
Last edited:
OP
OP
I have a "draft" 3D model with a collimator. It's printing now so I'll see how it turns out. If any of you are experiencing the issue AND you have a 3D printer handy, let me know. I'll get you a Beta model when it's at least been proven a little.
It will get fixed, I'm just in no hurry on it.
Okay, I knew I read it somewhere.
It will get fixed, I'm just in no hurry on it.
ChrisThomas
Well-Known Member
I have a "draft" 3D model with a collimator. It's printing now so I'll see how it turns out. If any of you are experiencing the issue AND you have a 3D printer handy, let me know. I'll get you a Beta model when it's at least been proven a little.
Okay, I knew I read it somewhere.
It will get fixed, I'm just in no hurry on it.
@LBussy The temperature thing I only mentioned in case it was one of those things "overlooked", it's not an issue! Also, I don't have a 3D printer myself but I do have a friend with one and I get most of my "toys" built by him. More than happy to keep him busy!
Anything else you can think of I'm more than happy to try.
I was thinking if a slit on the transmitter side and a pin hole on the other, or maybe slits at 90degrees on opposite sides would help? Perhaps, instead of fitting collimator to the sensor, could something be fitted to the lock and provide the same result? Starting to overthink this now!
Chris
OP
OP
@gromitdj is helping me actually do this correctly, but here's what I cobbled together:
That's a 1mm hole and will come out a bit better once it's moved away from the edge a bit. It's a 1mm thick "card" leaving room for up to a 9mm wide airlock. I think the emitter is small enough not to need a pinhole, besides, we're out of room.
That's a 1mm hole and will come out a bit better once it's moved away from the edge a bit. It's a 1mm thick "card" leaving room for up to a 9mm wide airlock. I think the emitter is small enough not to need a pinhole, besides, we're out of room.
Although it looks very nice I think I will stay with my solution for now. I don't have a 3D printer and I don't know anyone in my neighborhood who can help me with this.@gromitdj is helping me actually do this correctly, but here's what I cobbled together:
View attachment 676829
View attachment 676828
That's a 1mm hole and will come out a bit better once it's moved away from the edge a bit. It's a 1mm thick "card" leaving room for up to a 9mm wide airlock. I think the emitter is small enough not to need a pinhole, besides, we're out of room.
Naesstrom
Well-Known Member
Looks awsome! I can print it and test it out to! A suggestion, instead of making the "tower" all the way to the bottom make a 45º piece and extend the tower from that? Will use a bit more plastic but reduces the risk of it breaking at the bottom due to momentum.
Oh and while I'm being rude and coming with suggestions ;D Might i suggest a change on the PCB. if the order of the pins for the temp sensors were the same as on the sensor (gnd,dat,vcc) instead of gnd,vcc,dat it would be possible to solder a ds18b20 directly to the pcb for the room temp. vessel/beer temp would still need to be on a wire tough.
OP
OP
I understand. Having access to one is very nice. Once I have this nailed, I was going to offer to ship some out, but I think shipping to Europe may be cost-prohibitive.
I would not print it yet - the hole does not line up.
I am only responsible for the idea - Donnie will make it make sense.
That "fin" is very secure, however. I have it supported along the full length at the back.Well, there are two reasons I did that:
- This is how we've sort of standardized on the sensor layout with BrewPi. I needed to pick something so that's what I went with.
- I tried soldering a DS18B20 directly on the board. It ran almost 10°F hotter due to the heat from the chip. I thought that would be cool too. The reason iSpindel can get away with it is that it's in deep sleep mode in between transmissions and therefore not heating. Since this is meant to be accessed all of the time, that was not an option.
Naesstrom
Well-Known Member
I tried soldering a DS18B20 directly on the board. It ran almost 10°F hotter due to the heat from the chip. I thought that would be cool too.
Ah yeah I had the same problem with some sensors I put around the house not that you mention it! Sorry to bother you with another question... I added a ambient sensor and it works great but even tough I don't have a vessel probe connected it shows a value:
OP
OP
That might be a product of me averaging values before determining if it's connected. I stumbled across something like that in another project I'm working on. I've opened an issue on this to keep tracking it.
Last edited:
I can provide you an shipping address in Florida.
I have friends that go there very often but not at this moment due to the Covid-19 virus situation
OP
OP
Yeah, the whole virus thing sucks. Of course, that's why I had the time to finish the last version.
Donnie did a new version of the bracket and it's printing now. Should be able to try it out this evening.
Donnie did a new version of the bracket and it's printing now. Should be able to try it out this evening.
@LBussy another small issue that I've noticed in Brewfather (I don't know if it is related to Brewbubbels or to Brewfather) but the description of the temperature is different and a bit confusing.
The Ambient (Room) temperature of the Brewbubbles is Temp in Brewfather and the Vessel Temp is Room Temp. See screenshots below:
The Ambient (Room) temperature of the Brewbubbles is Temp in Brewfather and the Vessel Temp is Room Temp. See screenshots below:
OP
OP
That looks correct to me:
- BB Vessel: 15.5, BF Temp: 14.5
- BB Ambient: 23.0, BF Room: 23.1
OK thanks for the info!
I'm also using the iSpindle so I've already got the temp of my beer. Maybe someday there will be an option to change the labels in Brewfather?
Or in Brewbubbles so we can chose between room or fridge?
OP
OP
Adding another configuration item is not something I can get excited about. I would be happy to switch the point to "Fridge Temp" if there's no dissent among users.
ChrisThomas
Well-Known Member
No problem for me. My integration is going to be with Fermentrack when my PSU finally arrives... By then maybe I can log BrewBubbles data?
Next wine just started, yeast just starting to yawn and stretch!
Chris
OP
OP
Tracking this as #24 in the Issues list. If there's no dissent when I get to doing more work, I'll make the switch.
OP
OP
Awesome!
ChrisThomas
Well-Known Member
@LBussy I'm in need of another BrewBubbles setup, I've got everything except the bracket, is there anything new on the 3D printing horizon? Or shall I get a print of the existing version?
I know I'm getting greedy having several fermentations on the go at once, but that's exactly how BrewBubbles can help the stressed home wine maker!
I know I'm getting greedy having several fermentations on the go at once, but that's exactly how BrewBubbles can help the stressed home wine maker!
OP
OP
Hi Chris - so glad you are getting good use out of these.
Well, if you are getting good bubble registration now, I don't think I have an improvement for you. I have a new design I've been playing with having a collimator but I am not certain it's "good enough" yet.
Too bad you're not closer, I'd send you one of what I have.
Well, if you are getting good bubble registration now, I don't think I have an improvement for you. I have a new design I've been playing with having a collimator but I am not certain it's "good enough" yet.
Too bad you're not closer, I'd send you one of what I have.
ChrisThomas
Well-Known Member
OP
OP
Sure, you can check out the new one in the devel branch:
https://github.com/lbussy/brew-bubbles/tree/devel/bracket
https://github.com/lbussy/brew-bubbles/tree/devel/bracket
- Joined
- May 24, 2020
- Messages
- 377
- Reaction score
- 115
I just put together one of these devices to test out. Initially I got very low BPM (less than 1) but when I tested the trick with focusing the diode and that made it better. Now I get 5-6 BPM. my fermentation has not yet take off so we will see in a few days.
I also measured the airlocks I bought here in sweden and they measure 8.8 mm where the sensor should be....
I also measured the airlocks I bought here in sweden and they measure 8.8 mm where the sensor should be....
ChrisThomas
Well-Known Member
So... I've got the 3D print back from my friend. The fixed colimator is really difficult for me to get working.
On the PCBs I made I can't get them to remain in the relative position to the hole in the support and I eventually lose the count. On the print I received I had to take off apoximately 1mm from the back of the support to get the sensor to fit behind the support, and the hole needed some "help" to clear, not sure if the printing is the best quality!?
I think I'll next experiment with the shrink wrap as @wd16261 has tried. My initial cardboard solution needs adjustment whenever I move it onto another FV.
Chris
On the PCBs I made I can't get them to remain in the relative position to the hole in the support and I eventually lose the count. On the print I received I had to take off apoximately 1mm from the back of the support to get the sensor to fit behind the support, and the hole needed some "help" to clear, not sure if the printing is the best quality!?
I think I'll next experiment with the shrink wrap as @wd16261 has tried. My initial cardboard solution needs adjustment whenever I move it onto another FV.
Chris
OP
OP
Yeah the alignment is the challenging part.
Ideally there would be a collimator on either side of the tube. With a 10mm gap, that only leaves 8mm assuming a printed 1mm "card".
Thinking through this, I wonder if one used something like electrical tape (black) and a pinhole if it would be easier?
Ideally there would be a collimator on either side of the tube. With a 10mm gap, that only leaves 8mm assuming a printed 1mm "card".
Thinking through this, I wonder if one used something like electrical tape (black) and a pinhole if it would be easier?
Similar threads
Latest posts
-
-
-
-
-
-
why does kegged commercial beer always taste better than bottled ?- Latest: Clint Yeastwood
-
-