Interference on input into Axe110 datalogger using an 18M2

[martinn]

I am using an Axe110 datalogger with a 18M2.

I have a simple normally open reed switch connected between power and input 2, the reed switch is connected via 300mm of insulated wire (not shielded cable). The Picaxe registers a changed input when I touch the reed switch wires, initially it was setup to use an interrupt, but the same applies even when using a simple “if pin2 = 1 then flash a led”.

When I replace the 18M2 with an 18X the problem disappears, I am guessing this has to do with the ‘touch’ capabilities of the 18M2.

Any suggestions to reduce the interference?

thanks

Martin

 

[hippy]

It shouldn't be a result of the 'touch' features and sounds more like electrical pick-up. I'm not sure why that should have more affect on the 18M2 than the 18X as the pin in both cases are TTL inputs.

Do you have a scope which will allow you to look at the signal where it enters the AXE110 ?

Ideally you should change the "If pin2=1" to "If pinC.2=1" for the 18M2 however they are both equivalent, will read the input from the same PICAXE leg.

 

[Minsik]

As the base devices get lower in power needs and higher in input impedance they are becoming more and more sensitive and so the surrounding cct is more and more important. Not so much as sh'eel be right but knowing good technical details.

I had similar problems getting an 18X current as low as it should be when it was supposed to be low, and in fact It did work fine in a breadboard but in a socket and veroboard was hopeless.

1) earth all unused inputs. These are of course more now than there ever was with newer chips with configurable inputs/outputs. I think most are inputs by default. Only when out toggled do they become an output. (perhaps a clue here!)
2) reed sw input does it have a pull down R? or just reed sw to the b+ rail? Even a 1-5meg ohm R will work wonders here instead of an open cct.

As said my troubles were all basic and rulez all ways are to terminate inputs with what you need and the cct requires.

My input and design was for low current and so I used a 6M8 R and this worked fine even though thats very high. Something is better than left to the chip input.


Good luck.

Minsik.

 

[hippy]

reed sw input does it have a pull down R?

Good point and that could well be it. I'd mis-remembered and thought the AXE110 had a pull-down on input 2 but looking in the AXE110.PDF it does not appear to.

Connect 10K between input 2 and 0V and that will hopefully fix things.

 

[martinn]

Thanks Hippy and Minsik
The pull-downs did not help, but the clue about electrical pickup made me check something I should have checked first, the problem only occurs when the download cable is connected.

The 18M2 works fine with no download cable connected (the cable does not have to be plugged into the usb port, but the problem is far worse connected).

I have tried several cables, the problem remains with them all, so I guess it is, as suggested, some sort of electrical pickup, and that the 18M2 is more sensitive to it than the 18 X.

Martin

 

[BeanieBots]

I'm shocked that the pull-down did not help. (assuming you tried 10k).
Are you sure you put it in the correct place.
Try 1k.
If that doesn't solve the problem, then you will need to use coax cable and/or remove whatever is generating the noise.

 

[eclectic]

Thanks Hippy and Minsik
The pull-downs did not help, but the clue about electrical pickup made me check something I should have checked first, the problem only occurs when the download cable is connected.

The 18M2 works fine with no download cable connected (the cable does not have to be plugged into the usb port, but the problem is far worse connected).

I have tried several cables, the problem remains with them all, so I guess it is, as suggested, some sort of electrical pickup, and that the 18M2 is more sensitive to it than the 18 X.

Martin

Where is the interference coming from?

Can it be masked/shielded/screened...?

e

 

[martinn]

I have tried 10k and 1k pullups (resistor between 0v and the input) and the setup on both input 1 and 2. The problem remains the same. If I remove the programming cable, I have no problems. If I unplugged the programming cable from the computer but leave it connected to the Axe110 the problem continues.

The unit is running on 3 AA batteries. If leave the cable in and walk around the house, whenever I come close to a TV or compact florescent lamp the interference returns. With the 18X in the same unit, running the same program no interference is evident. Even a 100mm insulated wire placed in input 1 or 2 creates the same problem.

It isn’t a major problem as when in situ the programming cable will not be fitted.

Martin

 

[hippy]

The unit is running on 3 AA batteries. If leave the cable in and walk around the house, whenever I come close to a TV or compact florescent lamp the interference returns. With the 18X in the same unit, running the same program no interference is evident.

Full marks for investigation and reporting. Out of interest what download cable are you using ?

 

[BeanieBots]

I have tried 10k and 1k pullups (resistor between 0v and the input) and the setup on both input 1 and 2. The problem remains the same....

Not that it will solve your issue, but a pullup goes to 5v. A pulldown goes to 0v.
Your house must glow in the dark if there is enough airborne noise to overcome a 1k resistor
Out of curiosity, can you try putting a cap between the input and 0v.
Make it a large one, say 1uF or maybe even 10uF (rather the 'normal' 100nF) and let us know what happens.

 

[martinn]

@Hippy I have tried a few download cables, including the serial one that I bought with an early datalogger and several homemade ones, it is worse (well more sensitive) with a usb convertor plugged in, but given I am making the pickup even longer that is not surprising.

Out of interest, it does the same whether plugged into the download or datalink sockets.

@Beanie I tried both 10uF and 1uF capacitors to no avail. If I ground the 0v anywhere on the board with my finger it does stop the problem.

 

[BeanieBots]

I tried both 10uF and 1uF capacitors to no avail. If I ground the 0v anywhere on the board with my finger it does stop the problem.

If that didn't cure the problem, then I very strongy believe that you actually have a circuit fault. Probably an open somewhere.
1k and 10uF needs real current to produce enough voltage to trigger an input. My guess is that are not actually in circuit for some reason. If they really are in circuit and the line really is being toggled and the noise really is airborne, then you really are standing inside a health hazard

 

[eclectic]

If that didn't cure the problem, then I very strongy believe that you actually have a circuit fault. Probably an open somewhere.
1k and 10uF needs real current to produce enough voltage to trigger an input. My guess is that are not actually in circuit for some reason. If they really are in circuit and the line really is being toggled and the noise really is airborne, then you really are standing inside a health hazard

@martinn

Could you please indicate
where and how
you connected the resistors / capacitors?

It may help to use the attached scan
draw on it,
then re-submit.

e

 

[martinn]

Thank you everyone

I went and got another Axe110 board and set it up the same and it works fine (even with no pull downs), I am guessing BeanieBots was right about a faulty circuit, I probably should have tested for that earlier. Nevertheless an interesting lesson, I assumed because it did not happen with an 18X the problem must have been the 18m2, it just proves never assume anything.

I guess it is safe to take the tin foil hat off

Martinn

 

[BeanieBots]

At least you had the foresight to wear a tin foil hat until the fault was found
Glad you got there in the end.

 

[premelec]

@ martinn I would be interested to know specifically what the problem was if you ever figure it out - perhaps by re-using the original board - broken trace? Solder bridge? Gremlins? It's good to know the specifics so we can avoid 'em