40X2 power connection required for programming

wapo54001

Senior Member
#1
If the only power requirement is for programming which should be a pretty low power activity, is it still necessary to provide +5 to both pins 11 and 32, and ground to both pins 12 and 31?

I would like to power up only pins 31 and 32 while programming 40X2s.
 

erco

Senior Member
#2
Dunno, but it's easy enough to test. You can't hurt anything, either it works or it doesn't. By now you could have tried it and let us know. :)
 

stan74

Senior Member
#6
why couldn't the ground pins on a 28x2 be next to each other,or better still, joined together as 1 pin..but what about the OTHER PIN!!! lopsided
 

wapo54001

Senior Member
#7
A quick 'Google', gives the answer "YES" ... and 'er "NO" :confused: ...

... The datasheet doesn't seem to elaborate, or if it does, I can't find it.
I have looked at the datasheet and I could not find the answer. 'Course, pretty much I've yet to find any answer in that massive thing so no surprise. Without a chip to test I'm SOL, so I'll just order a chip and test later.

It appears that the serial In/Out pins and one set of +5/Gnd pins are identically placed on the 28X2 and 40X2, so I could use a 40 pin ZIF to program both a 40X2 and a 28X2 with identical wiring provided the 40X2 doesn't require the other set of power pins to be connected for programming.
 
Last edited:

stan74

Senior Member
#8
has anyone tried running 40x2 with only 1 ground or supply? I ain't messing with my chips...except putting them in the socket the wrong way round.
 

Buzby

Senior Member
#10
I've powered a 40X2 using just two pins, but that was only for a test with no external IO.

My belief is that if you power only one side of the chip, then all the current for the IO on the other side has to pass across the chip itself, and it may not have tracks big enough to handle all that current. The high current tracks will only be connected to the power pins on the same side.

I wouldn't risk it for anything using IO, in fact, I'd even be suspicious about programming, as that might use high current anywhere on the chip.

Cheers,

Buzby
 

premelec

Senior Member
#11
I suggest testing the resistance between the supposed same function pins - should give a clue if they ARE the same... ;-0 [note those tiny wires can be melted... don't use high test current...]
 
#13
I suggest testing the resistance between the supposed same function pins - should give a clue if they ARE the same... ;-0 [note those tiny wires can be melted... don't use high test current...]
The pins are connected within the chip BUT the path cannot tolerate much current. The resistance is relatively high: two tiny wires connect the external pins to the silicon and there are paths through the silicon. Capacitors should be mounted as close as possible to the pins anyway - remember that most of the electrical noise on those power pins is generated within the chip.
 

hippy

Technical Support
Staff member
#14
The bottom line would seem to be that if both power and both 0V are not connected no one can tell what might go wrong.

Like not using decoupling capacitors, not wearing a cycling helmet or seatbelt, there is no definitive answer as to what the consequences of doing that will be.
 
#15
The bottom line would seem to be that if both power and both 0V are not connected no one can tell what might go wrong.

Like not using decoupling capacitors, not wearing a cycling helmet or seatbelt, there is no definitive answer as to what the consequences of doing that will be.
Yup, nothing is certain until it's certain.

I'm going to do a pcb with a 40 pin ZIF with power to the two 40X2 power pins that are identical to the 28X2 power pins and give it a try. Maybe I'll throw in the traces for a DPST switch to power the other two pins "just in case."

I've looked at so many posts and publications that they've all blurred together. However, someone mentioned in a post that they measured the power pins on two sides of a 40X2 that in one direction it was a dead short, and when switching leads then the other direction there was a small resistance. In another place I read that for basic testing purposes powering only one set of pins was acceptable. Given those statements plus my assumption that surely programming current is not going to be significant at all, I'm going to assume that powering one side for programming purposes only is going to work.

PS I've been trying to upload a JPG of my latest pcb design for a week, using two computers (Win 7), three browsers, different copies of the image, clearing all caches, etc, and nothing works. I've uploaded lots of JPGs to the forum before without issue, now, it's been impossible for a while. It's difficult for me to believe the problem is at my end. I go through the process, click the upload button, a box appears "Waiting for Picaxe.co,uk" which lasts for maybe ten seconds and disappears, and the little "wait" symbol just goes around and around. Does anyone have a clue?
 

hippy

Technical Support
Staff member
#16
PS I've been trying to upload a JPG ... Does anyone have a clue?
Not really but it seemed to depend on how quickly the upload would complete so may in some way be file size related. When I tested it, small files uploaded easier than bigger files. How big is your JPG file ?
 
#17
Not really but it seemed to depend on how quickly the upload would complete so may in some way be file size related. When I tested it, small files uploaded easier than bigger files. How big is your JPG file ?
I just tried an upload from my Chromebook, 415kb JPG, which uploaded successfully. It appeared that I got a different program at the Picaxe end -- there was a graphic progress bar that went halfway and then stopped for the longest tine, then suddenly disappeared and the file was uploaded. My other computers do not show a progress bar. The file I struggled with is probably smaller. I will try my other machines again with this file that successfully uploaded from the Chromebook.
 

Circuit

Senior Member
#18
A definitive answer can be found in the Microchip presentation on PIC power "Advice on power supplies".

Multiple VSS and VDD pins
All VDD and VSS pins must be connected for proper operation.

"Many PICmicro MCUs have more than one VSS or VDD pin. This is common on
devices with higher pin counts. These pins are generally connected internally but
through a finite impedance. To insure proper operation, all VSS and VDD pins need
to be properly connected externally."

"Some PICmicro MCUs feature analog VSS and analog VDD as separate ground and
power connections dedicated to the analog sections of the device. Separating the
digital from the analog power and ground can reduce noise in the analog sections of
the microcontroller. Reducing the noise to the analog sections will typically allow
for more accurate A/D conversions. Analog VSS and analog VDD must be properly
connected and must typically be within 0.3 volts of VSS and VDD respectively, for
proper operation.
Because of the high frequency power demands of a microcontroller, it is
recommended that bypass capacitors be placed at the VDD and AVDD pins. These
bypass capacitors provide a low impedance path to ground for the high frequency
current demands of the microcontroller. This will also reduce the amount of noise
radiated by the power traces. Ideally, a ceramic bypass capacitor of about 0.1uF is
placed as close as practically possible to each power pin on the microcontroller."
 

geoff07

Senior Member
#20
has anyone tried running 40x2 with only 1 ground or supply? I ain't messing with my chips...except putting them in the socket the wrong way round.
I can confirm that in at least one case, not connecting all the pins leads to the chip not working.

Remarkably I just had an 18M2 survive being plugged in the wrong way. I don't suppose it helped prolong its life, but so far, no problem.
 

stan74

Senior Member
#21
I can confirm that in at least one case, not connecting all the pins leads to the chip not working.

Remarkably I just had an 18M2 survive being plugged in the wrong way. I don't suppose it helped prolong its life, but so far, no problem.
Connect the supply the wrong way around and it's fried...literally, and the 7805 reg I use was too hot to touch in seconds. There's benefit to using a proper dev board and not birds nest breadboard.
 

lbenson

Senior Member
#22
Connect the supply the wrong way around and it's fried...literally
Not necessarily with 08M2. I have several times made one briefly very hot by reversing the supply connections, and yet it still seemed to work (so far as I used it). Wouldn't be putting it into any DIY pacemakers, though.
 

AllyCat

Senior Member
#23
Hi,

I think it will largely depend on the Power Supply (or the current that it can deliver). With Alakline or Zinc-Carbon cells you'll probably get away with accidental reversal, but with Rechargeables or a regulated supply (e.g. a 7805) you might not.

AFAIK all standard ICs have an internal "Substrate Diode*" (normally reverse-biassed) between their Ground and Supply connections, so the reversed voltage across the chip should not exceed about 0.7 volt. Thus the chip may survive a (short-term) reversal if the available current is no more than an Amp or so. An (unreversed) 7805 may deliver up to a couple of Amps and if it is also reversed, probably even more (since it also contains a substrate diode). :(

But overheating is not the only hazard; a high current may "fuse" the Vdd bondwire (or however the pad is connected to the lead frame), or melt the (very thin) aluminium interconnect on the chip itself.

Adding a fuse or a PTC in series with the power source is generally the best solution, but a simple precaution is to include a "film" (MR/CR) resistor of perhaps 1 - 10 ohms, in series with the PICaxe Vdd pin, or with the battery supply itself. This resistor can also assist decoupling the PICaxe, and even 1 ohm may give an early warning by emitting some "magic smoke" if the supply is reversed. The resistance obviously can't be too large if the PICaxe is required to "work hard" (e.g. driving lots of LEDs active high), but that's another good reason to drive LEDs Active Low (i.e. with their anodes connected more directly to the power supply). ;)

*Google doesn't seem to find much related to substrate diodes as used in integrated circuits, but perhaps this explains a little.

Cheers, Alan.
 

stan74

Senior Member
#24
Only because you mentioned it Alan. A uk home linux system uses a poly fuse because people sometimes over load it. Says wait and recheck wiring somewhere in all the griff. I usually have a diode before the reg but I was using a low dropout 7805,regulates from 5.7V, with li ion batteries. I would recommend an in line fuse with these batteries.
 
Top