new Atmega8 not reponding

Go To Last Post
9 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi, Im a bit concernd.

 

I bought some bundles of m8s a while ago They have all worked well. Ive even taken to make a PCB.

 

Now I have avrdude giving "not responding " rc -1 , "please check.." etc.  The thing is that I have checked the programmer, and it works find with other

setups (same conctact layout, actually the identical PCB ), i have measured the paths for power supply, and for the programmer from connectors to the chip,

all seem fine.

 

Im hessitant, but tend to lean against the m8 being faulty...., I rather not though because I dont think the PCB will stand a shift of chip (throughhole DIP), the pads will go.

 

Any good thinking that could save the situation ?

 

 

This topic has a solution.

Last Edited: Wed. Dec 27, 2017 - 12:15 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

gechxx wrote:

Any good thinking that could save the situation ?

Not enough information given about the board, do you have a schematic?  Did you program the fuses wrong?

"void transmigratus(void) {transmigratus();} // recursio infinitus" - larryvc

"It's much more practical to rely on the processing powers of the real debugger, i.e. the one between the keyboard and chair." - JW wek3

"When you arise in the morning think of what a privilege it is to be alive: to breathe, to think, to enjoy, to love." -  Marcus Aurelius

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi, no I have not had the m8 connected at all before inserting at board. (out of delivery folio)  (as a follow; Im thus using internal clock 1Mhz for starter,will reprogram, well if I get it going) )

 

Well the board is well working in another copy. So the scematics/layout should not be of concern, and since I have measured the connectivity (buzzer on instrument) between 

programming PINs and the chip, and also PINs inbetween (for short-circuits), and both seem ok, i.e. contact where there should be, and no cross contact. the VCC pins have

the 5V visavi ground, OK I havnt actually measured contacting of the ground PINs to ground !.... thnx for question (makes you describe, and think, if that now is a possibility ..) there is still hope !

So Im a bit exhauseted as to what could be other than the chip itself.... 

Last Edited: Mon. Dec 11, 2017 - 08:57 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

As you seem to have problems with new/"virgin" chips, have you slowed down the ISP bit rate enough?

gechxx wrote:
So Im a bit exhauseted as to what could be other than the chip itself....

If you have not checked the integrity of the signals right at the chip e.g.

gechxx wrote:
OK I havnt actually measured contacting of the ground PINs to ground

...how could you have "exhausted" all possibilities?

 

-- Connect all power and ground pairs.  Have proper decoupling caps.

-- Check all signal levels >>right at the AVR pins<<.

-- Does /RESET drop when you do a Read Signature?

 

More complete checklists are in http://www.avrfreaks.net/comment... and links therein.  Those threads should be pertinent to you to "exhaust" your possibilities.

-- Does your AVR have the proper supply voltage on ALL Vcc and AVcc pins? (Don't just say "yes"; check each.)
-- Does your AVR have ALL Gnd pins hooked up? (Don't just say "yes"; check each.)
-- When you do a serial ISP Read Signature, does /RESET drop? Nice clean signal, not overly rounded? If not, find out why.
-- Trigger 'scope on that falling edge of /RESET, and put the second probe on SCK. Do you see activity with nice square edges during the Read Signature?
-- Repeat with PDI (which is NOT MOSI or MISO). If OK,
-- Repeat with PDO (which is NOT MOSI or MISO).

 

If all of the above is good, and still no results, either your AVR has no clock source, or is fried. Was this a new AVR? Do you think you've mucked with the fuses? If so, inject an appropriate clock signal of about 1MHz (you can get it right on your STK500) into XTAL1 and try again.

Given the '128 target, what devices are hanging on the ISP pins? In particular, a comms transceiver (MAX232, MAX485, etc.) connected on USART0 can cause problems.

The above is for a '128 target, so some differences...

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

Last Edited: Mon. Dec 11, 2017 - 09:10 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

OK, sorry , Im a bit rookie, thats why I ask for help, in a  hope I havnt exhauseted. And actually, no I havnt checked signals directly at chip, that is beyond my budget (dont have a scope with mem function, only steady signals, an old tube, green light if  you know). I have to cope with whats possible with an AVO more or less. (have also an old one, with mechanical hand, it can make some shiver for bits sometimes)

Do I need de-coupling capacitors on the m8, that one is not CMOS is it ?  (in that case Iv just lucky my previous board(s) works so well)

(that about ground was pure hope, measuring again, I realized I actually had done)

 

As to bit-rate for programmer: no I havnt slowed (it has always worked before, on this setup).  Was that -B wasnt it.

 

Actually, this chip can be the first from a new purchase-lot. It actually looks different from the previous. The "half-moon" end marker is much bigger on this, the surface looks more lead-greyish than blackis as the previous.

(could there be a CMOS variant of the m8 out there ? (in china, yes they are most certainly from there, aliexpress...) )

 

Last Edited: Mon. Dec 11, 2017 - 09:28 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 1

The plastic color of the chip doesn't matter, as long as it is a true Atmel AVR chip and not a knock-off, or a total fake.

 

You need by-pass caps for reliable operation.

You were lucky if things worked well before, without them.

Place a 0.1 uF cap across each Vcc/Ground and AVcc / Ground pair of pins.

Optionally place a 0.1 uF cap from ARef to Ground, do not connect ARef to Vcc.

Be sure ALL Vcc, Avcc, and Ground pins are connected to V+ and Ground.

 

Post a good photo of your PCB.

Post a schematic of your design.

Use the "Mountain range" icon, two to the left of the smilie face, to post pictures.

 

Does your PCB use an external crystal?

Does it have pads for one?

If it has pads for one, (and the two associated caps, typ 12 pF, +/- some), go ahead and install them.

 

Be sure to turn the ISP speed way down for your initial testing.

 

Be sure you oriented the chip correctly when you placed it on the PCB.

 

JC

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

well guess its a brick, or Im thick as one, or both. 

 

( I do have room for an Xtal and caps, but not mounted, wont waist on this guy, how difficult could it be for it to kvack an answer at say -B 10 (or -b 1200)  )

 

(would a cap on Aref ! make a diff and replace other "decouplings"?)

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

If the chip happens to be fused for an external clock, no speed will work, because the chip has no clock. You don't need the clock to be right to program the chip, but you may need it to be present.

This reply has been marked as the solution. 
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 1

HI.

For the sake of it I, and since I have gotten so much of your support, I feel I should conclude this, even though it is greately shameful.

It is a happy ending, but possibly not for my state.

 

It turns out, it was me that hade the fuse gone.

 

This was my first PCB, and most things turned out ok, but there were some twirks, among which that I had swithced the connecting of pin 2 and 4 of my 2x5

connector for the programmer........ That I HAD corrected on my first PCB, which was easy since pin 2 wasnt used anyway, a little jumper added.

(measuring contact to chip from programmer connector easily comes out positive for a mirror connectivity...)

Now being a bit intermittent AVR digger, there goes months between my 'attacks' all earlier twirks gone, and not reading my loggs for next round until now....

 

Well as you see; no shadow over Atmega8, it is still a nice small piece that functions most of time, regardless of being misshandle for things like good clock, and decoupling.

Will consider your points if & when going for another PCB round. (or getting into new  trouble)

 

With Whishes for a Happy New Year

Georg

Last Edited: Wed. Dec 27, 2017 - 12:18 PM