Homebuild Calibration

[Brian Garner]

Hi

I just looked at the picaxe, its quite cheap. I did not know about these when I started and I used a Stamp BS2 for version 1. I now am using a 16f877 pic. These have 8 or ten ( cant remember offhand) A/D inputs ( 10 bit). loads of ipins for I/O and 8k mem with 256 eprom.

I have a 4 line lcd display, showing
3 sensors, setpoint, ( its a kiss at the moment but I am thinking of ECCR in the future)
Time. Meant for a scrubber timer but displayed so if I want to i could use it for timing deco stops in a pinch.
Battery voltages 2 off.
And a line which dislays a bargraph or warning messages depending on the state of the sensors.

This is my main display. For a secondaty display I have 2 lcd dpm which connect directly to cells 1 and 2 and just display Po2.

These are calibrated by pots and are easy to adjust in air for 20.9 %

When I calibrate the master I can compare these to the master and see that the calibration ( done in 100%) is somewhere near.

My unit has been pool dived a few times and is almost ready for open water tests. Hopefully pool this thurs and if all is well I may have a do at Dotty this weekend.

Brian Garner

[Andy_Corkill]

Hi Brian

What sort of Op Amp have you used, I don't know if mine is that accurate. I've used a Single Supply Quad Amp (LM324N) with a +5V input, connected it up as a Non-Inverting amp.

I used pots to calibrate to 20.9 % O2 at a MV output of 10 MV's. When I raise the output of each sensor simulator up to 50 MV's I'm getting a descrepancy of up to 6% on each sensor reading (I know calibrating for 100% O2 will help to reduce the errors when I come to use it in anger).

I think this may improve when I move everything off the breadboard. But I was just a little surprised to get such a diffence in readings & haven't even connected the O2 cells yet!

Anyone had similar problems ?

Thanks

Andy

[NotDeadYet]

A question for those using the Picaxe controllers... (bearing in mind that electronics to me is the same as magic)...

What do you need to make it work? Is it simply a case of hooking up the three sensors to the three inputs and writing a bit of code to read the voltages? Or do you need to connect some, er, extra gubbins.

And how easy is it to hook up an alphanumeric display to it? Or some LEDs? I'm thinking a HUD system for my sidemount rebreather first then maybe work on a solenoid system.

The code programming I am comfortable with but I know nothing about electronics.

Cheers,

Stuart

[Brian Garner]

Hi Andy

Like you I used lm324 's. With no O2 sensor fitted I get readings of about 1% O2 instead of 0. This must be from the offset errors in the op amps. I did try several and found some were better than the others. With the O2 sensors in I calibrate and get a nice steady 21% reading in air rising to between 98 - 102 in O2. I use the opamps with a gain of about 40. Using cells which output 10mv @ 21% this gives me a 0 to 5V output from the op amp.
A useable range to about 2.5 bar Po2. ( Hope I got the figures right from memory.


To calibrate the master handset I flush the cells with 100% and get the readings. I use the max reading as 100% This loops untill the maximum reading is not increased. This is compared with the analog secondary visually. When the max o2 value is obtained I then use this as the calibration reference. If the analog readings are not 100 then I do it again.

From the calibration reference I do the calculations to get setpoint and warning limits. There are some innacuracies introduced in the conversion process but the overall readings are steady and sensible. I dont know what the accuracy of the inspiration is but I will not be far away I am sure. Another variable is also the tempertature which I have ignored for the moment. This my also produce errors. I dont expect them to be large though.

I know that I have at least 1% errors at 100% possibly more but I think it was Nigel who said his body wasnt calibrated to 1% accuracy. I dont think mine is either.

Hope that makes sense.


Brian

[Andy_Corkill]

Getting answers about about Picaxe controllers.

Hmmm. You must realise that my knowledge of electronics is limited to a few lights for my push bike, and this Pic contoller project for my KISS. So I know only a little more than nothing about electronics.

But on my experience so far I have found the PIC relatively easy to interface to. With good examples on their website
(http://www.rev-ed.co.uk/picaxe/) and studying how this guy on the
therebreathersite did his, I’ve managed to make some progress.

But I know it’s going to get a little more difficult from now on. Trying to maintain a constant supply voltage for my circuit is one thing, every time I light up an LED or fire my solenoid I get a voltage drop which in turn causes my O2 readings to go high ? Discrepancies between readings is another, and there will be more.

So I’d say as an electronics incompetent, you may be able to get something up and running with the PIC on a breadboard. But to overcome the other problems you must have some degree of electronics competence (or know someone how has).

Regarding the coding of the PICAXE. The language is very limited on program size To make the controller seem as if it is doing many things at once, you may find yourself coding in a most nasty way to get the most from the controller. Nasty coding like this is also prone to bugs, so you have got to be pretty dam sure of your coding.

Andy

[beanie]

Originally Posted by Andy_Corkill

But I know it’s going to get a little more difficult from now on. Trying to maintain a constant supply voltage for my circuit is one thing, every time I light up an LED or fire my solenoid I get a voltage drop which in turn causes my O2 readings to go high ? Discrepancies between readings is another, and there will be more.

so some kind of voltage regulator required then have you tried using a packaged voltage regulator?

[Andy_Corkill]

Originally Posted by beanie

so some kind of voltage regulator required then have you tried using a packaged voltage regulator?

I tried using one, but couldn't get it to maintain a consistant 5 volts to my PIC & OP Amp circuit when firing the solenoid. I've since decided to use two separate battery supplies, one for the LCD backlight & solenoid the other to drive everything else. When the solenoid fires the LCD backlight goes dim, which isn't such a bad thing. At least I have some sort of visual reference as to the state of the solenoid batteries.

But I've still got to spend some time with a voltage regulator for the PIC circuit, as when I light up my HUD Leds I'm still getting a small drop. I'm still on a learning curve, any suggestions on voltage regulator choice will always be welcome.

Thanks

Andy

[beanie]

Originally Posted by Andy_Corkill

I tried using one, but couldn't get it to maintain a consistant 5 volts to my PIC & OP Amp circuit when firing the solenoid. I've since decided to use two separate battery supplies, one for the LCD backlight & solenoid the other to drive everything else. When the solenoid fires the LCD backlight goes dim, which isn't such a bad thing. At least I have some sort of visual reference as to the state of the solenoid batteries.

But I've still got to spend some time with a voltage regulator for the PIC circuit, as when I light up my HUD Leds I'm still getting a small drop. I'm still on a learning curve, any suggestions on voltage regulator choice will always be welcome.

Thanks

Andy

um not sure but I would have isolated power supplies for the control and LEDs / solinoid then the operation wouldn't interfere with your regulated PIC circuit, You can use the same battery for both bits still.

well at least you should be able to

[Andy_Corkill]

Originally Posted by beanie

um not sure but I would have isolated power supplies for the control and LEDs / solinoid then the operation wouldn't interfere with your regulated PIC circuit, You can use the same battery for both bits still.

well at least you should be able to

I'm trying to Beanie mate. I must be doing something wrong, still reading manuals trying to understand how to regulate the circuit properly.
Any examples/components you can direct me to on the internet would always be welcome.

Thanks

Andy

[beanie]

Originally Posted by Andy_Corkill

I'm trying to Beanie mate. I must be doing something wrong, still reading manuals trying to understand how to regulate the circuit properly.
Any examples/components you can direct me to on the internet would always be welcome.

Thanks

Andy

what voltage do you need?