serial pin on a 40x1 internally connected directly to Vcc?

[tiscando]

Hi all,
I have a picaxe 40x1 and it can connect to the picaxe programmer, but it woudn't download the first byte of the program.
I have correctly insalled the download circuit and when i have investigated, I have found that the serial in (rxd) pin seems to be almost directly internally connected to it's vcc power rail. I tried pulling it down to gnd (0v) with a resistor as low as 82 ohms (it warmed up but the picaxe chip didn't) and it's voltage only went as low as 4.7v (vcc=4.9v regulated). The serial out (txd) works normally (as rxd is high, the txd flashes like how i have investigated on a working picaxe chip with it's rxd high)
Does this mean the chip is damaged or can something be done?

 

[hippy]

There's something very wrong if a resistor connected to Serial In gets hot. Unless there's a wiring fault, circuit error or you've mis-identified the pin it sounds to me that the PICAXE has been damaged.

 

[Dippy]

It would be a good idea to post your schematic EXACTLY as you have it right now.

Also tell us if you are using breadboard and whether you've inadvertently mis-wired it previously in other experiments/testing (even if briefly). Have you ever accidentally connected a power supply over 5.5V to the PICAXE?

Have you got a proper Rev-Ed proto/project board to try it out in/on/with?

IF the chip is damaged then a decent service and burial is in order

Generally, it is always a good idea to buy the cheapest project board for the desired chip to get up and running. I know this will bring on the Wrath Of Stan, but when it's all new and you want it working in 10 seconds flat then breadboard is not ideal for the inexperienced - regardless of how well you paint it with felt-tipped pen.

It may also seem a waste of money to buy a board you'll only use a few times, but these issues crop up regularly (usu in conjuntion with breadboard+newbie) and maybe a chip has been popped so another one has to be bought and the carriage costs and time. So, maybe buying a proto board is NOT a waste of money after alll...?

 

[stocky]

I'm with you dippy (sorry stan!)
I have Rev-Ed boards for all but the 14/20 (on order!)
I then have header pins to connect to my various breadboards where I put the ancillary stuff.

Stocky

 

[tiscando]

I have earlier, and now again, tried the 40x1 on a different breadboard and the same happens. the former try involved connecting the 82r resistor between the serial in and the ground. I didn't mean the resistor getting hotter than 40'c!

See the attachment: the picture is a bit bity but i should have cleared it up a bit. The transistor Q1 in the regulator was a kt850b npn on the first try with the main breadboard and then a bfy51 on the test breadboard. before i found out the fault, i misconnected the power supply onto the base of the kt transistor rather than the collector. the highest voltage for the supply is 7.5v.

 

[hippy]

The resistor shouldn't be getting hot or warm at all, it shouldn't change from ambient. If this is the same 40X1 as you used before you could have damaged it, burnt out its innards and it may always now behave in a faulty manner.

I'm not clear what you have as the regulator here, but if you've put 7.5V through the chip it could well have been destroyed, anything above 6.5V and you can say goodbye to it unless you've been very lucky.

I would recommend getting a new 40X1 ( either unused or known to be working ) use three AA batteries as a 4.5V supply and try to get that working before trying more complicated configurations.

Also, having zoomed in your image it appears that there's a PICAXE on the left. Is what's on the right the 40X1 ? What exactly do you have and what are you trying to do here ?

In your first post you said, "have a picaxe 40x1 and it can connect to the picaxe programmer"; what do you mean by PICAXE programmer ?

 

[tiscando]

1. I mean the picaxe chip does connect to the (picaxe programming editor, via download cable, downloader, the window with the progress bar that pops up after i have pressed the 'run' button), but it can't download because the serial in pin is faulty.

2. many 0.25watt resistors between 100r and 47r heat up a bit if their voltage drop is more than 4.5volts. under 0.25w, over 9v, or under 47r, they heat up considerably more.

3. anyway, my 40x1 chip either has been damaged by the misconection of the regulator, in an unusual way, or if not, it might be from a bad batch and was still in Rapidonline's boxes. so, i will just try again with another 40x1 chip.

4. what i have at the moment are:
- 2 18x chips, one as a bingo game, 2 28x1 chips, one as a brushless motor driver oscillator, and some 08m's, one as a fan controller.
I sometimes have the problem when the program in the picaxe chip runs inconsistently, but this is usually solved if i add some capacitors around 0.1uf on the power rails.

 

[tiscando]

with the 40x1, i am experimenting with a vcr loader (wired up to my large breadboard)

 

[Dippy]

I assume the thing marked 'PIC/AXE' to the left of your 40X1 is the Picaxe serial connection to your PC?

When you said "I have found that the serial in (rxd) pin seems to be almost directly internally connected to it's vcc power rail" - is that WITH the download cable connected AND connected to your PC?
If so, what are the measurements of the serial-in pin WITHOUT the download cable connected?

Your original incorrect power connection may have damaged the PICAXE. I just can't tell from this distance.

And did you check/test your home-brew regulator BEFORE you connected it to PICAXE?

Once again, this is an excellent reason to buy the cheapest project board when you are new to PICAXE/Electronics (sorry again Stan, I've sinned twice now in one thread) . And also an excellent reason to buy a proper regulator. Fewer variables means fewer cock-ups. But it's up to you, it's your money and time.

 

[moxhamj]

Dippy is giving some good advice. My first picaxe was a 40X. But my second picaxe was an 08. I'd also recommend the cheapest project board to start. And I like to use sockets for picaxes, and maybe I'm paranoid but before I put the chip in I check the 5V and 0V are coming out of the sockets as they should.

I've never heard of a shop supplying a damaged chip. But I do recall a conversation with sales staff where they told me that many customers think they have been given a dodgy chip. Then later the customer has to admit that yes, they might have put 240VAC on the power supply pins, but it was only briefly...

Re the 0.1uF caps - I'd suggest putting them in next to every chip on the board.

 

[BeanieBots]

Input pins are "clamped" to Vcc via an internal diode. If excessive current has been injected into the pin, then it is possible that the clamp diode has been melted and causing the effect you see.
You can check with a simple meter set to ohms.
With the chip out of the socket and connected to nothing else, check the resistance between serin and Vcc. If it is the same value with the meter connected both ways and it is less than 10 ohms, then the chip is fried.
If it is not, then compare the readings (in both directions) with a known working chip. The exact value will depend on the type of meter you have so I cannot give an actual number for you to look for.

 

[BCJKiwi]

Just be careful with the meter used - some (typically older) multimeters have 9V batteries which can put 9V accross the circuit being tested in certain ranges. RTFM (for the meter of course)!

 

[Dippy]

Good idea BB. And even easier/better if your DMM has a diode check setting. I've just tried it.

Is that true BCJ?
I've just checked 3 DMMs. On Ohms they chuck out 0.3V and on diode check they use approx 3V.
All 9V PP3. One Fluke, two Maplins.
Maybe just really ancient/really cheap/analogue ones do that? (Actually one of the Maplins was real cheap).

RTFM - now that has given me my first laugh of the day... thanks.

 

[BeanieBots]

BCJkiwi is correct.
I have one (Sinclair, circa 1970) which goes up to 12v when OC.
It is actually a very useful feature for testing zeners. The drive current also has several values going up to 10mA but I'd be surprised if putting a typical (1mA) into any pin via any other pin on a PICAXE would do any harm.

 

[Dippy]

Blimey, do many people have things like that? When BCJ said 'old' I assumed he meant this century or the 90's.

And, yes a 'Zener' setting on a multimeter would be really handy.
A PICAXE project if ever there was one.

 

[BeanieBots]

As you get older, older becomes relatively older.
I went to Beaulieu Motor Museum a few years ago and was rather horrified to see a car which I used to have for regular comuting in!

As for old meters, I still have a wind-up Mega capable of 1kV.
Something else which has sadly gone with the last few decades is the ability to play office pranks without being sacked or sued. Many years ago we had a new 'technician' join the test department and thought it might be amusing to "measure his personal resistance" with the Mega

We hooked him up to the device and tentatively activated it at 250v.
Nothing...
Increased to 500v.
Nothing. Not even a twitch.
Cranked it all the way to 1000v. The needle wacked hard over.
Still no reaction!!!
We asked "are you OK?".
Through gritted teeth, we got the reply. "Yes, but could you please hurry up. This doesn't half hurt!"
One born everyday

 

[westaust55]

Age and multimeters

As you get older, older becomes relatively older.

my thoughts are when you are 1, a years is 100% of your life and is forever.
when you are 100, 1 year is 1% of your life . . . which is why they flash past.

I also have serveral meters, all in mint condition . A couple of digital units and

one old analogue (1972 era University MVA50 which has 34 ranges). Still have orginal box for the MVA50 which also still ls sees frequent use. Has outlasted a few digital meters where the soldered in battereis and solar charging schemes have failed or electronics failed - NOT thru misuse.

 

[BeanieBots]

Oy, I'm not 100 yet!

 

[westaust55]

apologies to BB

Nor am I but it is the goal to aim for (as long as it goes with reasonably good health)

in the meantime think of it as
when we are 50, a year WILL BE ( ) just 2% of our life and the years seem to flash by . . . .

 

[BCJKiwi]

Well I don't have the meter I refered to as it got cooked - don't ask, I can't remember why but I was still using it up to around 3 years ago. It had both 9V and 2x1.5V batteries in it, used for different ranges. The manual did cover which ranges were used by which batteries.

The meters I now use have only 2x1.5V batteries.

Is it still OK to apply 9V to say the PICAXE V+ or a port even if the current is limited?

Which is the critical factor for that magic smoke - voltage, current or power?

 

[westaust55]

Excessive current will cause over heating of the transistors within the PIC chip. If a transistor exceeds its maximum junction temperature then the transistor will usually fail then and there.

If one reads about the time duration permited to solder the leads of an IC it is related to keeping the transistors associated with that lead/pin from overheating.

By comparison, the Microchip datasheets indicate pin MCLR which is also used for programming as Vpp can handle 13.5 Volts.
Electrostatic dischange can be many times the rated voltage may not result in immediate failure but can stress the IC transistors so that failure might occur later.

However, using a voltage higher than the Max/Absolute rating given in the datasheet will allow higher current to flow leading to excessive temperatures and transitor junction failures.

Those in the PC world who try to run CPU’s and other chips at higher speeds often increase the core voltage of the CPU or graphics chips. The biggest issue is that they need to add some serious cooling systems to compensate for the greater heat and keep the chips from thermally “cooking”.

Power is Watts and seen as heat generated by the IC as a result of the current drawn. ie it is a byproduct of the current.

So IMHO, excessive current would be the greatest cause of IC failure.

 

[hippy]

Providing the current is limited enough you shoud be okay. You would have to check the particular datasheet to know what the maximum injection current can be.

I'd say current was the critical factor, but voltage, current and power are all inter-related. The way I see it is that, as V=IR, the voltage across any in-chip R is a result of the current through it so if that's limited so is the voltage across R, so is power. Thus explosion, destruction and escaping magic smoke depends on current not the voltage the current comes from. No idea how that applies to semi-conductor parts though.

Get it badly wrong and magic sparklies will accompany the magic smoke. At an extreme, teleportation of the entire chip can occur, often accompanied by a loud magic bang

 

[Dippy]

I just measured an out-of-circuit PIC with a DMM on 'Diode test' setting and it gave the 'correct' PN voltage on an I/O pin (when the DMM leads were correctly connected of course).
So, whilst everything everyone has said I'm sure is true this gets away from the fact that testing the clamp diodes is easy and feasible - as long as something else isn't muggered up of course.

Obv (generally) if the I/O configuration has anything different then you may get odd results e.g. GP3 of a 12F683. I'm sure someone has more spare time than me to test. So, do the test in conjunction with the PIC Data Sheet which also shows max clamp current.

 

[BeanieBots]

Just an extra word about all this volts/amps/watts thing.
The maximum voltage you apply to any pin depends on circumstances.
The absolute maximum voltage ratings apply to ANY circumstances.
A "normal" input (ie not MCLR pin) can only cope with Vcc + 0.6v.
So, for example, a PICAXE running at 5v could have 5v applied to its inputs with no problem. However, a PICAXE running at 3v would be destroyed INSTANTLY if you applied 5v to an input.
However, if the 5v was applied via a current limiting resistor, the internal clamp diode would clamp the VOLTAGE on the input pin to Vcc + diode drop thus avoiding instant destrucion.
The clamp diode can only cope with a certain amount of current. Often only around 10mA or so. Hence, the resistor should be quite a high value.

Another point to note is that the clamp current has to go somewhere. It is dumped into Vcc. Consider a PICAXE being run from 4X 1.5v batteries with a diode in series to drop the power rail down to around 5.5v.
If 12v is applied to an input via a current limit resistor, there is nowhere for the 'excess' current to go because the power rail diode blocks it. The PICAXE will now be powered by 12v (less whatever voltage is lost due to the resistor and current drawn by the PICAXE).

Just be careful and understand what is actually going on.

 

[tiscando]

What i mean by the title of this thread is that even with the download cable disconnected, the serial in pin is high even with the correct download circuit.
I have always been using 'home-brew' regulators for logic circuit and the picaxe chips - no problem. it is probably that i used a different transistor for the first time when i used the picaxe 40x1 when i got jumbled up with the transistor connections, once again damaging the 40x1 in that unusual way.

again, I'll try again with a new 40x1 with more caution.

 

[tiscando]

The zener diode in the regulator is 1.3 watts so it can withstand up to 100ma at 12 volts. The current through the 2.2k resistor is 6ma at 12v volt drop. what i mean by getting jumbled up is that - see my shematic in an earlier post, i have acidently touched the supply lead at the base of the transistor instead of the collector, destroying the zener diode. The nominal voltage for my unregulated psu is 6v at 1a, but i measured it up to 10v when no current is drawn.

 

[Dippy]

Jolly Good. Carry on testing.