Line following module problems

[a.a.mcmahon]

Hello!
I'm using the Microbot for a project designed to go around a factory/home/warehouse and stop infrequently to scan (via PIR) for intruders...

Line following has been great for ages with no problems. Now the red light I think seems to be dimmer and it's not following lines. After syntaxing and debugging I found the module to be reading from ~255 when raised in the air and when covered with my infer it is about 250ish. If I put my super bright LED iPhone flashlight against it I can get it to read ~10 even 0 on the debug. I have a feeling that some voltage has been lost to the LED on the module. Ie disconnected everything bar the module. Reattached it and replaced the power supply with very expensive fresh batteries even!
Anyone else having this problem or know what I could do about it?
Any help would be great fully appreciated!

 

[Technical]

There are no known issues with this module - has yours suffereed any damage?
Its basically just an LED and a light sensor, so quite hard to damage apart from physically.

Replacements are available in the BOT127 kit.
http://www.techsupplies.co.uk/epages/Store.sf/en_GB/?ObjectPath=/Shops/Store.TechSupplies/Products/BOT127

 

[a.a.mcmahon]

No nothing of any sorts, I make sure it is powered down when not in use and have never dropped it, I would be very surprised if it were to be damaged either physically or electrically.
It has now started to affect the motors too, the motors are acting as if they are not receiving a constant and sufficient voltage all the time, it barely has enough power to move itself at times, sometimes if a bumper is hit and it reverses the motion then I get a temporary increase in speed but it is as if the motors and line follower module are not receiving enough power. I have reattached all components, like I said, replaced batteries with decent quality ones... still its all starting to fail...
Could it be manufacturing defects during production? Is that a known problem?

Regards,

 

[Technical]

No, there are no known issues. IN the first post you stated your robot was working fine in the past, so that eliminates any manufacturing issue.

It sounds like you may have a short circuit somewhere on the robot e.g. a module connected the wrong way around - e.g. accidentally attaching the light follower upside down or backwards at the bottom of the support posts.

Is the motor issue only when the line follower is attached - if so that does imply that module has been damaged...
If all the time they we suspect you still have a faulty connection somewhere - remove all modules (including the battery pack), reconnect the battery pack with new batteries and then test the motors again before adding on any of the optional i/o devices.

Also make sure you have the 3xAAA batteries correctly inserted - it has been known to get one around the wrong way!

 

[a.a.mcmahon]

By manufacturing issues I meant gradual and deteriorating factors, such as internal motor bearing wearing out or something as benign as that!
I thought about short circuits, I have been using the multimetre on the device at various point and have found no obvious cross over. I've made sure to check with the line follower and its still suffering with the same problem.

I have once again, removed and reattached each individual part, testing them all individually as I go along. Problem still occurs. I have examined the main board for any signs of short-circuitness and nothing, even got a second opinion to see.

Do you have a list of places and ports I can test for resistance and voltage to see if they match/mismatch to try and locate the part of the board which may be faulty.

 

[Paix]

I have once again, removed and reattached each individual part, testing them all individually as I go along. Problem still occurs.

Problem still occurs, when/where in the testing?

Enumerate the individual steps of your testing and report the result after each of those steps please. Typing overall generalisations is unlikely to be productive.

I don't suppose that it has spent any extended periods outside in the rain or other adverse conditions? You seem overly concerned with manufacturing faults and motor bearings, but thin on specific detail.

 

[a.a.mcmahon]

That's one of the thoughts my team had when the lot of us were looking into it, we tested each individual part as we added it to the system, seeing if a particular part made it suffer but the motors start lagging as well as the line follower if they are the only thing attached!! We're certainly not making a hasty generalisation nor am I trying to get at the fact it is due to the manufacturers but we have eliminated many other causes.

No the product has not spend any time outside, it lives in a room were the air is always dry and in a cupboard away from sunlight even. We have ruled out any enivormental factors affecting the product, external factors are still a maybe, I always handle the product but the chances of someone getting the product out for a play and breaking it/ dropping it is very unlikely in our culture.

All in all, if its just the line follower module, not anything else, the brightness is still very dim...
if it is just the motor(s) in place then they are inconsistent/ sluggish and weak.

 

[Technical]

It does sound like a collapsing/weak power supply, does the LED change brightness when the motors are on/off - if so one of your gearboxes could be binding causing a stall current through the motor. Other than replacing the batteries we are out of ideas!

 

[a.a.mcmahon]

That's what I'm thinking, but I just can't find it nor find the reason for it! Also no, the LED brightness does not change with reference to the motors being on or off. Hmm, what would you do if you were in my shoes then?

Any ideas anyone?

 

[inglewoodpete]

I would be using a multimeter to measure voltages around the circuit: not always easy, of course. If proven OK, using debugging code to determine what values are being read from the inputs.

(I have never even seen one of these 'bots - I wish they were around when I was younger - but I have had a look at the manual that's available from the website.)

 

[a.a.mcmahon]

Yeah see I've been using the multimeter all over the circuit. Apart from the obvious I have nothing really to compare it to.
Reading all the adc's has been a little baffling. Inputs which arnt connected to anything are fluctuating quite a lot but it's not that unheard of.
Bots really are great. Jut when they work :-/

 

[hippy]

Hmm, what would you do if you were in my shoes then?

Disconnect all additions, download a 'do nothing' program and check all voltages were as expected. I'd use a bench supply so had a rock solid known voltage and also note the current draw.

Then download test programs to test individual lines and again check voltages and current draw.

For the motors, I'd write tests to control those one at a time, check the signals through the circuits, test with a resistive load and with motors in place.

I'd also test the motors, with and without gearboxes, and also the line sensor module, removed from the MicroBot and connected to the bench supply at appropriate voltage to see how they behave independently.

 

[a.a.mcmahon]

Do nothing?? Ah ok so, essentially a blank programme me thinks, measure all voltages with all components off... I don't know what they should all be with nothing attached but I think they should all be the same, test for resistivity too? I don't have a bench supply or anything of sorts available for the time being.
I suppose I could attach the line following module separate from everything... I think that is a good idea.

One thing is though, how do I measure the current of something? Call me a novice (and the only qualification I have is an GCSE electronics at A*) but I'm pretty sure I would have to connect in series wouldn't I? That would be hard to do with a BOT design were everything fits in so snugly...

Anyway thanks for your input! Really appreciate it, will get on it straight away!

 

[russbow]

Do nothing??

In program editor, go to Picaxe > clear hardware memory.

If it programs OK you have a nice clean Picaxe.

and the only qualification I have is an GCSE electronics at A*

Oh dear. Put your meter in series with the battery +ve pole and the +ve connection on your board.
Switch your meter to the highest Milli amps range.and switch on. If it's a moving coil meter you may have to reverse the leads. Switch the meter to a lower range until you get a sensible steady reading. Make a note of it and let us know.

 

[hippy]

Most good bench supplies will have meters for voltage or current so just a case of popping the batteries out and linking to the battery box with croc-clips.

If it doesn't show current then any meter which does can be used. That's put in series as you note so croc-clip from PSU to meter, another from meter to battery box.

If using batteries, I'd do the same but using a separate battery box instead of the bench supply.

 

[russbow]

Does this help?

 

[a.a.mcmahon]

One thing that may cause concern, an external PIC PCB has been made to control it as part of the project, three wires have had to be attached to the BOT because of it, one wire going into an input, another wire for an output as well as a 0V wire.

I have found that for sometime the wire that goes from the BOT and into the PIC's input has accidently been soldered into the 4.5V line instead of the data line.

Sometimes, when the BOT is powered off and the PIC PCB is powered on, the LED's around the BOT can flash very dimly but only for 5 seconds or so.

Could this be the cause of the problem? Some sort of damage to the BOT?

 

[BeanieBots]

.....I have found that for sometime the wire that goes from the BOT and into the PIC's input has accidently been soldered into the 4.5V line instead of the data line.......
Could this be the cause of the problem? Some sort of damage to the BOT?

At last, we are getting somewhere.
So, the BOT has been significantly modified and not even in the manner intended.
Yes, absolutely, it is very likely the cause of all your problems and has probably done irreversible damage.
The first step is to remove ALL modifications and then go back through all the suggestions in this thread and start again with fault diagnosis.
In particular, rebuild step by step and note what does and does not have an effect on supply voltage and opperation.

 

[a.a.mcmahon]

We I surely wouldn't say significatly! It is just those wires going into the inputs, I carefully scoured the manual and made sure everything was ok, the PIC as you know only provides what was it, 10mA, plus the wire that was on the 5V line was connected to an input, should have just drained the battery like any old module or component would have done.

 

[BeanieBots]

..., the PIC as you know only provides what was it, 10mA,....

No, No, No,
The current limit on a PIC output is NOT the maximum current that you force out of it, it is the current above which you will DESTROY it.
(that is why an LED MUST have a series resistor)
Also, if you connect supply volts to an input on a chip that is not powered, you will (attempt) to power it through the input protection diodes and probably destroy it.

From your symptoms (and late confession) it sounds like you have damaged one or more inputs/outputs on the PICAXE.
There is little point doing anything else until all the mods have been removed and the PICAXE has been replaced.

 

[a.a.mcmahon]

Ok I think I understand that, but does that not mean that the PIC that is located externally would be damaged and not the PIC on the BOT?

 

[Technical]

No. You said it yourself - when the microbot power supply was off the LEDs on the microbot still flashed. So it has to have been getting power via the wrong connection from the other board. So in normal use the two power supplies may also have been conflicting.

You have made a bad connection and probably damaged the ICs/circuits on the microbot. Simple as that.

 

[SAborn]

So i take it you have 2 power supplies on the bot, have you made both power supplies V0 common to both picaxe chips as they need to be.
If you have not got a common ground then this will be 1/2 your problem, the other 1/2 is more likely a dead chip from not having a common ground and wrong connections.

 

[a.a.mcmahon]

Resolved! At last, the motors were going incredibly slow because right at the beginning of the programming there was a command to halt the motors before getting them to go forward (as interrupt commands have been known to have certain things running without being reset).

Essentially what was happening is that it was being told to go forward and stop, thousands of times a second or whatever it is and so not unlike connecting the motors to a one way AC supply. Thanks for all your help though guys! The only problem that still remain is the line following module, I have bench tested it and all voltages and that are 'A' OK but still the light is dim, looks like I'll be making do with it.

 

[BeanieBots]

...Line following has been great for ages with no problems....

So it never occurred to you to go back to the version of software that worked or to let on to the fact that this problem started after you had changed the software?

 

[a.a.mcmahon]

I kept changing the software as I was testing for the line follower module and try and get it to work, during this time the bot was raised so I wouldn't of realised, I was also playing with the programming commands to change the speed of the motors so it didn't cause me for concern.

When I was trying to iron it all out I just copied the main part of the programming which would just push the bot forward when I was taking it back to basics, I only sorted it out when I forgot my USB stick and had to write new code, it worked fine then as I missed out the halt command parts.