Matho's controller

[MaxN68]

An increase by 0.25 is the standard. If we lower the resolution to 0.50 we have a faster response of the probe (minimum sampling time to reading) but the controller would not have reason to be a control to 0.25, if we raise the resolution we will have sampling times more lenses and an increased delay compared to real time.
The 0.25 resolution is perfect for our use.

 

[megabyte]

Wow, sorry to open up a can of worms. I thought it might have been smoothed out in the UI to save us pedantic users from sleepless nights trying to tune out 0.1c overshoots with our PID settings :lol: I had no idea it was done to keep the code size down. A very good reason.

Do people really need more than 0.5 accuracy for brewing?

I don't think the accuracy is all that important so long as it's consistent between your brews. The precision of 0.0625 degree steps could be helpful for the pid holding a steady temperature though.

I confirmed last night it is the probes. One temp probe I stuck in went up by .25, the other up by .5 (different styles of probe).

Wow I'll have to try this. The DS1820 (without the B ) can only do 0.5C so it sounds like it's smart enough to scale the precision based on the probe type.

Edit: B and ) was turning into a smiley so I added a space betwen them.

 

[lael]

I did some playing with this this morning. It looks like the temp probes that I'm providing with the kits are set to a resolution of 9bits. I'll post a video showing how to change it to 10 tonight ( will show .25 degree increments)

Essentially for the impatient or more savvy, download the libraries from the software update video and replace the old ones with the new ones. There will be an additional DallasTemperature library. Open up the DallasTemperature library folder, one up the basic example. Change the one wire pin to 11, and the resolution to 10. Upload the sketch, open tools>serial monitor. Check that it increments by .25.

Re-upload ardbir and go through set up. Watch the glorious .25 temp changes for hours on end

 

[dca]

OK, voice of reason here you guys need to pull this into perspective. I have been an EE for ~20 yrs, working on precision temp control systems for 14 yrs.

This DS1825 sensor is only accurate to 0.5*C / 0.85*F in the range of -10 to +85*C.....at boiling in the range of -55 to 125*C, it is only accurate to +/- 2*C (3.4*F).

So, having the extra resolution w/o accuracy does not really buy you anything.
Yes, you can calibrate, but it is still subject to accuracy of your source and the curve of the DS1825.

ave

 

[lael]

Yep, I agree, and am happy with 0.5 accuracy. It is nice to see the temp change faster though

 

[SBOB]

I did some playing with this this morning. It looks like the temp probes that I'm providing with the kits are set to a resolution of 9bits. I'll post a video showing how to change it to 10 tonight ( will show .25 degree increments)

I ran out of enthusiasm before working out what needed changing, so thanks for saving me more thinking

 

[SBOB]

OK, voice of reason here you guys need to pull this into perspective. I have been an EE for ~20 yrs, working on precision temp control systems for 14 yrs.

ave

Yeah, but pretend we are operators...Voices of reason dont work.. We just want to see little numbers moving around
(just had a discussion with one today wanting 4 decimal places of accuracy shown in his SCADA for a water turbidity meter... )

 

[MaxN68]

I'm happy with 0.125 accuracy on my ArdFerm

 

[lael]

I did some playing with this this morning. It looks like the temp probes that I'm providing with the kits are set to a resolution of 9bits. I'll post a video showing how to change it to 10 tonight ( will show .25 degree increments)

Essentially for the impatient or more savvy, download the libraries from the software update video and replace the old ones with the new ones. There will be an additional DallasTemperature library. Open up the DallasTemperature library folder, one up the basic example. Change the one wire pin to 11, and the resolution to 10. Upload the sketch, open tools>serial monitor. Check that it increments by .25.

Re-upload ardbir and go through set up. Watch the glorious .25 temp changes for hours on end

this is proving a little more elusive than I first thought. You might need to enjoy watching 0.5 temp changes for hours on end for a little while longer

 

[SBOB]

this is proving a little more elusive than I first thought. You might need to enjoy watching 0.5 temp changes for hours on end for a little while longer

why?...worked for me based on your one line explanation, I did it this evening....
or you mean elusive as you dont have time to do the video for people?

 

[lael]

why?...worked for me based on your one line explanation, I did it this evening....
or you mean elusive as you dont have time to do the video for people?

Have you turned it off and on again?

 

[SBOB]

Have you turned it off and on again?

damn you dashing my 0.25c hopes and dreams...
I had not

 

[lael]

Lol, yeah, that's what my reaction was too

 

[SBOB]

Lol, yeah, that's what my reaction was too

strange..
even when i enable the Dallas define as 'True' within the Ardbir code (which then is suppose to/appears to perform the same sensors.setResolution command) I still only get 0.5c resolution

Did have a small panic when I re-started a new Ardbir project and forgot to configure the correct board and language type.. had a screen full of solid lines 1 and 3

 

[zwitter]

Hey guys
While you are all knee deep in t he code, I have a request.

We have some LCD scrambling issues due to the use of a relay to switch the pump. If the wiring is neat it really is not an issue but for some it can be a big problem.

So the request is can the LCD be refreshed / re-drawn after the pump is turned on or off each time? Or maybe once a minute instead?

I Guess that is either easy or near impossible depending on how the code is written. Any future boards will use SSRs but as we have Mega boards space is not the limitation.

Thanks
Zwitter
James

Sent from my iPad using Tapatalk HD

 

[SBOB]

So the request is can the LCD be refreshed / re-drawn after the pump is turned on or off each time? Or maybe once a minute instead?

I Guess that is either easy or near impossible depending on how the code is written. Any future boards will use SSRs but as we have Mega boards space is not the limitation.

while others know the code much better than me, I'll say it would be quite difficult to implement.
As the screen sections/words are only written/changed when they are changed in the code, its not like there are 4 strings within the code holding what should be displayed at any time which you could simply re-write to the LCD. For instance the top line is written once on startup and never again.

I'm pretty sure its a memory reason, but if memory wasnt a concern (like in the mega) a simpler way would be to have 4 strings (one for each line) and you re-write them as required then output them to the screen. This would allow for the screen to be 'refreshed' whenever you wanted, whereas currently you would need to know what mode/display you currently have so you could re-create what may be scrambled on the screen

 

[tateg]

Hey guys
While you are all knee deep in t he code, I have a request.

We have some LCD scrambling issues due to the use of a relay to switch the pump. If the wiring is neat it really is not an issue but for some it can be a big problem.

So the request is can the LCD be refreshed / re-drawn after the pump is turned on or off each time? Or maybe once a minute instead?

I Guess that is either easy or near impossible depending on how the code is written. Any future boards will use SSRs but as we have Mega boards space is not the limitation.

Thanks
Zwitter
James

Sent from my iPad using Tapatalk HD

Hey zwitter,
Is it possible to use the ssr for the second heating element for the pump if I have no intention of a second element.

 

[lael]

Hey Tate,
The issue is when people don't organise their wiring well. There's been two reports from almost a hundred kits sent out(unsure how many have been built at this point). I wouldn't worry about it.

 

[real_beer]

Hey Tate,
The issue is when people don't organise their wiring well. There's been two reports from almost a hundred kits sent out(unsure how many have been built at this point). I wouldn't worry about it.

The beauty of your latest design lael is that the two halves of the box can be unplugged to work on allowing much less wiring that has to be squashed inside. I had scrambling issues with my original Matho's Controller but eventually sorted it out using the same method you advocate.

I'm also waiting on a slow boat from China for a USB panel mount to fit outside the box software updates can be done without having to keep opening the box and moving the wires around unnecessarily all the time. It's not such an issue with the new design but I think keep opening them up repeatedly is always best avoided if possible. My controller is mounted high and well away from the system so getting liquid and steam on it isn't a problem for me, although I'll still put a cover over the USB port. http://www.ebay.com.au/itm/USB-2-0-type-B-male-to-A-Female-host-case-panel-mount-screw-short-cable-cord-/310790581002?

Looks like your next design upgrade will be a wiring loom kit :lol:

Edit: You could probably charge $250 for it by itself :lol:

 

[tateg]

Hey Tate,
The issue is when people don't organise their wiring well. There's been two reports from almost a hundred kits sent out(unsure how many have been built at this point). I wouldn't worry about it.

Thanks Lael,
I had a few issues with the old controller but the box was pretty tight. Just thought that as a safe guard I could use the second ssr as I do not have any plans on adding a second element.
Cheers