RESET pin not always 12V tolerant?

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

I've been experimenting with HVSP on some ATtinies again, and this time I wondered what happens if 12V is applied to a running AVR.  It's obviously 12V tolerant for some 10s of us after startup in order for it to enter HVSP mode.

I used a t13 powered from a USBasp, and then manually connected 12V to the RESET pin.  The USBasp immediately disconnected, so I removed the 12V.  I removed then re-connected the USBasp, and the t13 looked dead.  It still responded to serial programming commands, properly reporting the signature.  When I switched the power from 5V to 3V3, the t13 started working again.  Repeated tests at 5V always failed, and 3V3 always worked.

 

I'm not sure why the t13 was damaged.  Even though there's no clamping diodes to Vcc, there must be some path from RESET to VCC.  I did some separate tests applying only 12V to RESET (VCC floating), and with the t85s I tested, I measured around 1V on VCC.  With a couple t13s, I measured 1.6-2V.  An equivalent schematic for the RESET pin would help explain things, but I have never been able to find one.

 

I have no special talents.  I am only passionately curious. - Albert Einstein

 

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

What does the Maximum Electrical Specs say for the T13? The spec sheet I have says Absolute Max for /RESET is 13V. Could you have fried USBasp?

 

Jim

 

Until Black Lives Matter, we do not have "All Lives Matter"!

 

 

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

If the USBASP was connected,  you have applied 12V to PB2 on the USBASP's ATmega8.

 

The Tiny13 should survive.   But the Mega8 will die.

 

I can understand the odd project with Tiny25/45/85.   I like the Tiny2313/4313 too.

However Tiny13 just appears to be a waste of space.

 

David.

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

Does it have a weak pull-up enabled? I wonder if a shunt, maybe 100k, could give some clues about how much current is going into the reset pin.

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

I think that you need to have Vcc then reset get's 12V. (and my guess is that it isn't written anywhere)

 

The main function for the 12V is to write to flash (if you write with vcc there is a internal charge pump at least on the old chips)

so there could easily be some leak current.

 

 

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

ka7ehk wrote:

What does the Maximum Electrical Specs say for the T13? The spec sheet I have says Absolute Max for /RESET is 13V. Could you have fried USBasp?

 

Yes, that's what I see in the datasheet, -0.5 to +13V.  The USBasp is fine.  As I mentioned, " It still responded to serial programming commands, properly reporting the signature. "

I also switched out the t13 for a t85 and it works fine at 5V and 3V3.

 

I have no special talents.  I am only passionately curious. - Albert Einstein

 

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

ron_sutherland wrote:
Does it have a weak pull-up enabled?
Yes (RPU)

ron_sutherland wrote:
100k
two order-of-magnitude less (typical to increase ESD tolerance when RESET is attached to an operator's switch)

 


DC Characteristics | ATtiny4 / ATtiny5 / ATtiny9 / ATtiny10

Connection of RESET Pin on AVR Devices | AVR® Microcontroller Hardware Design Considerations

 

"Dare to be naïve." - Buckminster Fuller

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

david.prentice wrote:

If the USBASP was connected,  you have applied 12V to PB2 on the USBASP's ATmega8.

 

The Tiny13 should survive.   But the Mega8 will die.

 

Good point.  I forgot to disconnect the reset line from the USBasp.  However the Mega8 seems to have survived.  I switched to a different USBasp, and I can still re-flash the t13, but it still doesn't run at 5V any more.

 

I have no special talents.  I am only passionately curious. - Albert Einstein

 

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

gchapman wrote:

ron_sutherland wrote:
Does it have a weak pull-up enabled?
Yes (RPU)

ron_sutherland wrote:
100k
two order-of-magnitude less (typical to increase ESD tolerance when RESET is attached to an operator's switch)

 


DC Characteristics | ATtiny4 / ATtiny5 / ATtiny9 / ATtiny10

Connection of RESET Pin on AVR Devices | AVR® Microcontroller Hardware Design Considerations

 

 

The t13 datasheet, s 18.2, says 30-80kOhm for Rpu on the reset pin.  I thought there is a diode between Vcc and the output driver and pullup pFET to keep 12V from feeding back to Vcc.  The characteristics of reset as IO in the datasheet show the max Voh for the Reset pin at about 0.8V below Vcc.

 

 

I have no special talents.  I am only passionately curious. - Albert Einstein

 

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

ralphd wrote:

Good point.  I forgot to disconnect the reset line from the USBasp.  However the Mega8 seems to have survived.  I switched to a different USBasp, and I can still re-flash the t13, but it still doesn't run at 5V any more.

There is no series resistor on PB2 line of USBASP.   So +12V would go directly to the substrate diode to VCC pin.

 

I would expect that diode to evaporate.

You can compare with other GPIO pins on the Mega8 by using a DMM on a disconnected USBASP.   e.g. resistance between GPIO pin and VCC pin

 

I would be amazed if the diode is still there.   Of course a series resistor on 12V might have limited the diode current.

 

David.

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

david.prentice wrote:

ralphd wrote:

Good point.  I forgot to disconnect the reset line from the USBasp.  However the Mega8 seems to have survived.  I switched to a different USBasp, and I can still re-flash the t13, but it still doesn't run at 5V any more.

I would expect that diode to evaporate.

You can compare with other GPIO pins on the Mega8 by using a DMM on a disconnected USBASP.   e.g. resistance between GPIO pin and VCC pin

 

I would be amazed if the diode is still there.   Of course a series resistor on 12V might have limited the diode current.

 

David.

 

Prepare to be amazed with how tough some of the AVR parts are.  I measured 0.73V for the diode between PB2 (target reset) and VCC.  Same measurement on the other pins I checked.

 

I'm not sure what went wrong with the t13 though.  I may end up throwing it in the trash, but for now I've set it aside in a baggie with a note to remind me that it won't work at 5V.

 

I have no special talents.  I am only passionately curious. - Albert Einstein

 

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

"Dare to be naïve." - Buckminster Fuller

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

I tested a t85, this time with the RST line to the USBasp disconnected, and applying 12V causes the t85 to stop running, and it resumes after removing the 12V.

With no power to VCC, there's a small leakage from RST to VCC.  With 12V on RST, I measured 0.3V on VCC.

I have no special talents.  I am only passionately curious. - Albert Einstein

 

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

 

The RESET pin has no VCC shunt.  Otherwise you couldn't use high-voltage programming.

 

 

I have no special talents.  I am only passionately curious. - Albert Einstein

 

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

VCC to GND

 

"Dare to be naïve." - Buckminster Fuller

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

Have you programmed the device incircuit? have you not accidentally killed the rest of your PCB by applying 12V to the reset line?

 

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

gchapman wrote:

VCC to GND

 

So is there something like a zener that goes from VCC to GND on all AVRs to protect them from overvoltage?

 

I have no special talents.  I am only passionately curious. - Albert Einstein

 

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

meslomp wrote:

Have you programmed the device incircuit? have you not accidentally killed the rest of your PCB by applying 12V to the reset line?

 

 

No it's just a small breadboard with the AVR, and a red LED with 1k resistor.  The USBasp is connected using my breadboard adapter cable.

http://nerdralph.blogspot.com/20...

 

I have no special talents.  I am only passionately curious. - Albert Einstein

 

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

Only CMOS's parasitic bipolar transistors (limit VCC and dVCC/dt, limit injection current)

AVR don't have a latch-up spec though are tested for latch-up during wafer fab qualification.

 

Application Note 339 Fairchild's Process Enhancements Eliminate the CMOS SCR Latch-Up Problem In 74HC Logic (ON Semiconductor)

[second sentence]

The latch-up mechanism, once triggered, turns on a parasitic SCR internal to CMOS circuits which essentially shorts VCC to ground.

KSRA-07HVQF409 | Product Change Notification | Microchip

CCB 4031 Final Notice: Qualification of Microchip Technology Tempe – Fab 2 (TMGR) as an additional fabrication site for selected Atmel products.

[bottom of right column]

Attachments

PCN_KSRA-07HVQF409_Qual Report.pdf

[page 10]

 


How to search for Microchip PCNs

AN1785 - ESD and EOS Causes, Differences and Prevention

[page 12, middle of right column]

Additionally, latch-up sensitivity is tested according to JESD-78, and is performed on sample units with a minimum stress current of ±100 mA and an overvoltage of 1.5 * VMAX.

TB3013 - Using the ESD Parasitic Diodes on Mixed Signal Microcontrollers (latch-up on page 3 middle of right column)

Advanced High-Speed CMOS (AHC) Logic Family (Rev. C) (Texas Instruments)

[page 7]

Latchup Protection

 

"Dare to be naïve." - Buckminster Fuller

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

ralphd wrote:
... and then manually connected 12V to the RESET pin.

...

I'm not sure why the t13 was damaged.

STK500 limits the 12V current; IIRC, 10V is enough (12V -> CCS -> RESET)

Atmel STK500 Schematic | Wayback Machine (signal V12P7, lower right of page 4, bottom center of page 6)

 

"Dare to be naïve." - Buckminster Fuller

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

@gchapman, this link you referred to shows a zener on Vcc to protect against Vcc spikes:

https://e2e.ti.com/cfs-file/__ke...

 

But you seem to be saying AVRs don't have that, so I still don't understand what you are referring to when you said I fried " tiny13's VCC shunt ".

 

Can you be more specific?  Are you saying there's a diode connecting RST to VCC, even though you can put 13V on RST?

 

I have no special talents.  I am only passionately curious. - Albert Einstein

 

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

 

My understanding is that all pins have the protection diodes except the reset pin, it only has the gnd side diode.

The DS shows only a pull up resistor to VCC on the reset pin.

 

 

In any case, I'm still unsure of what damage you did to your 13 that would let it run on 3v3 but not 5v0, unless the BOD circuit was damaged when you connected 12v to the reset pin.

Normally this is only a short duration pulse for HV programming, and not a steady state condition.

 

Jim

 

 

 

 

(Possum Lodge oath) Quando omni flunkus, moritati.

"I thought growing old would take longer"

 

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

ralphd wrote:
Can you be more specific?
There's an SCR from VCC to GND (CMOS parasitic bipolar transistors)

ralphd wrote:
Are you saying there's a diode connecting RST to VCC, ...
No

 

P.S.

ralphd wrote:
... you said I fried " tiny13's VCC shunt ".
Appears to be excessively leaky at 5V such that the 5V power supply current limits.

 

"Dare to be naïve." - Buckminster Fuller

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

Back when I set up test systems for an SMD line, the quality people would send these sorts of failures to an outfit for analysis, and the problem was always the same, ESD.