Unconnected bottom pad = unresponsive device?

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

I've just soldered an Attiny85 MLF-20 onto my PCB wich breaks all the pins out for me.. Great... But i can't read the device signature with my STK500.
I've tried with HVSP and ISP.. Nothing.... I am completely lost as to why i can't read the signature from the chip..

I have noticed a flaw in my PCB though. I forgot to connect the bottom pad of the MLF-20 package to ground.

Could this cause the chip to not function even if i connected all the other pins of the chip has been successfully connected to the PCB and the STK500?

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

While bottom pad may well be Gnd it probably only really comes into play if you have a heavy current consumption and perhaps even if you need heat dissipation.

 

I doubt it's the reason for no response. More likely the package is rotated on the pads like pin1 is rotated 90 deg to the pin 16 position or something like that. You'd have to get something seriously wrong for neither ISP nor HVSP to work!
 

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

How did you solder it to the board?  Hot Air or soldering iron? 

Most likely you either have a missing connection or a solder short somewhere, do some continuity testing with a meter.

Post a photo of your board for more suggestions.

 

Jim

 

 

FF = PI > S.E.T

 

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

clawson wrote:
While bottom pad may well be Gnd it probably only really comes into play if you have a heavy current consumption and perhaps even if you need heat dissipation.

There have been discussions on this.  IIRC some models >>might<< use it as Gnd.

 

We don't do a lot of MLF work, but when we do we have a number of plated through holes on the board "pad" area, to allow soldering (and desoldering).

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.

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

Thanks for the quick response!
I've checked again and again that I've got the pinout and the rotation of the IC correct.

Here are some basic pics of the schematic, board and the chip soldered to the PCB. If you could take a look and spot and error that would be great!

Only the Attiny85 has been soldered to the board. The project is made in Eagle CAD. If you would be willing to take a look at the EAGLE project as well please let me know.

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

Ok, that looks hand soldered using an iron, so neatness counts, use some liquid flux and some small solder braid and run the braid + iron around the chip to clean up the area.

Then use some flux remover.   Most likely you have a short someplace between pins. 

I prefer to use solder with a water soluble flux, after soldering I wash the board under hot water and scrub with a toothbrush, then blow it dry with compress air.

Leaves the board clean and easy to inspect, looks like it was done on a pic-n-place/reflow oven.

 

Jim

 

 

FF = PI > S.E.T

 

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

 

Oops !

 

1. There is a difference between should be (a recommendation) and shall be / must be (a requirement). I've always connected the central pad to GND so don't know whether that applies here.

2. It is possible to apply too much solder to the central pad and thereby lift up the whole chip so that the other pins don't solder underneath.

3. The soldering on the MOSI MISO & SCK pins doesn't look that good.

4. I see no decoupling capacitors for VCC.

 

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

Which of the 2 headers is your ISP connector? Anyway 510R resistors are far too low for the ISP pins even if in series with LEDs, the programmer is not likely to be able to drive the pins correctly.

John Samperi

Ampertronics Pty. Ltd.

https://www.ampertronics.com.au

* Electronic Design * Custom Products * Contract Assembly

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

JS,

I have a product that uses a T25 with 68 ohm resistors on the LEDs and it programs just fine using either a dragon or USBASP programmer, so what he has should work.

 

Jim

 

 

Edit:  I just noticed he has the ISP connector (6 pin) between the resistors and the LEDs, that will not work! 

 

FF = PI > S.E.T

 

Last Edited: Mon. Nov 28, 2016 - 10:09 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Nothing but the Attiny85 is soldered onto the PCB right now. The soldering looks way worse on the pic than in real life. I will take the PCB under i microscope tomorrow and check my horrible solder job :)
I applied VERY little solderpaste to the bottom pad to avoid the issue that N.Winterbottom is talking about. I have no capacitor on VCC right now. This shouldn't be required, should it?

Lets say that the chip is soldered in correctly.. Where is the next place you would recommend troubleshooting?

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

OP said: I have no capacitor on VCC right now. This shouldn't be required, should it?

 

What does the datasheet say?

 

 

 

FF = PI > S.E.T

 

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

Use a multimeter on diode test to check that the pin is connected. You can measure the internal substrate diode ror each pin from gnd. Also check for shorted pins. Also your power tracks are too thin and no bypass capacitor.
Remember - it is easier to not solder parts to a pcb as opposed to trying to add them later.

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

... and your /reset pin is floating. I would tie it to Vcc via a 10K resistor.

 

Ross McKenzie, Melbourne Australia

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

Thanks Kartman!
I would never have thought about using the protection diodes to check the pin connections. Perfect!

 

For anyone wondering why this works, see Figure 10-1, page 53, I/O Ports.

http://www.atmel.com/images/atme...
 

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

You'll find most devices have the diodes on the pins. I did some qfn work recently where i reminded myself of the technique.

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

I did the test. I checked the following:

- No pins are shorted to any other pin

- On all the pins i could measure a diode voltage of 600-700mV to ground.

- I could also see a diode voltage of 600-700mV between Vcc and ground.

 

So i guess that this means that all the pins are wired correctly.

I'm using the 5v supply from the STK500 board to power the IC.

I also added 30uF bypass capacitance to the PCB.

 

Ill double check the setup of the STK500 when i get home from work.

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

valusoft wrote:
... and your /reset pin is floating. I would tie it to Vcc via a 10K resistor.

Hmm! For this AVR I disagree.

 

1. Whilst the PB5 pin is configured as external reset an internal pullup is always connected.

Port B, Bit 5 – RESET/dW/ADC0/PCINT5
• RESET: External Reset input is active low and enabled by unprogramming (“1”) the RSTDISBL Fuse. Pullup is
activated and output driver and digital input are deactivated when the pin is used as the RESET pin.

2. With a regular ISP programmer connected RESET will be driven high or low. A pullup won't contribute much here.

3. The OP has used High Voltage Serial Programming. A pullup resistor could cause mayhem under these 12V conditions:

 

 

 

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

Yes I am aware of how this works. However I read the OP's first post as saying that only in desperation (my interpretation) to determine the T85's signature had he used HVSP and ISP.

Ross McKenzie, Melbourne Australia

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

Just to provide some info as to how i connected the PCB/Attiny to the STK500 without using the 8 pin DIP socket i followed this guide:
www.robotroom.com/Atmel-ATtiny-S... (For ISP and HVSP).

 

Insted of poking jumpers into the socket i just used the PORTB male headers and some jumpers.

 

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

I also added 30uF bypass capacitance to the PCB.

But you MUST HAVE ceramic bypass caps 100nF (some use 10nF or both), 30uF will do nothing much for fast transients.

John Samperi

Ampertronics Pty. Ltd.

https://www.ampertronics.com.au

* Electronic Design * Custom Products * Contract Assembly

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

So.. I tried to add a more stable power supply. That did the trick. I was able to read the device signature, read the fuses and program the fuses, but..................
I set the disable reset fuse using the STK500 HVSP because i need the pin as an input in my project. Now i can't communicate with the device again ;(