AVR-Chat

----- Original Message -----

From: Valerij Avrealij

To: AVR-Chat@yahoogroups.com ; codevisionavr@yahoogroups.com

Sent: Sunday, February 13, 2005 11:13 AM

Subject: [AVR-Chat] ISP problem: very high current when ISP switched ON (more 200mA)

Hello 2all.

When I in system program my tiny15 the current more than 200mA!!!
The chip and 7805 are hot. I use Altera ByteBlaster programmer with special PC software.
Very interesting: when I connect swiched ON ISP the current is normal,
when I OFF it and then ON the current more than 200mA
How this problem can be solved?
P.S. Using this ISP cable I successfull program mega8 chip.

----- Original Message ----- 

From: Valerij Avrealij
To: Leon Heller
Sent: Sunday, February 13, 2005 11:53 AM
Subject: Re: [AVR-Chat] ISP problem: very high current when ISP switched ON 
(more 200mA)


Hello, Leon.

You wrote:



>> Hello 2all.

>> When I in system program my tiny15 the current    more than 200mA!!!
>> The chip and 7805 are hot. I use Altera ByteBlaster    programmer with 
>> special PC software.
>> Very interesting: when I connect    swiched ON ISP the current is normal,
>> when I OFF it and then ON the current    more than 200mA
>> How this problem can be solved?
>> P.S. Using this ISP    cable I successfull program mega8 chip.

> It sounds like the chip is going into  latch-up.

What you mean? What is  "latch-up"? Early this situation repeated with
90s2343. What it can be?
For info: Just after few seconds after new (just buyed) chip was
connected to the programmer the current ups to >200mA.


With CMOS devices, it is possible in some situations for the substrate to 
become biased so that the device behaves like an SCR, think of a 
forward-biased diode. Lots of current is drawn and the device can 
self-destruct. If power is removed quickly, they often recover.

It can be caused by applying a voltage to an I/O pin before the chip is 
supplied with power, and there are several other causes:

http://www.analog.com/library/analogDialogue/archives/35-05/latchup/

Leon 



-- 
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.300 / Virus Database: 265.8.7 - Release Date: 10/02/2005

Hello 2all.

When I in system program my tiny15 the current more than 200mA!!!
The chip and 7805 are hot. I use Altera ByteBlaster programmer with special PC software.
Very interesting: when I connect swiched ON ISP the current is normal,
when I OFF it and then ON the current more than 200mA
How this problem can be solved?
P.S. Using this ISP cable I successfull program mega8 chip.

Thanks in advance.
-- 

  Valerij Avrealij

On Feb 13, 2005, at 5:13 AM, Valerij Avrealij wrote:

> When I in system program my tiny15 the current more than 200mA!!!
> The chip and 7805 are hot. I use Altera ByteBlaster programmer with 
> special PC software.
> Very interesting: when I connect swiched ON ISP the current is normal,
> when I OFF it and then ON the current more than 200mA
> How this problem can be solved?
> P.S. Using this ISP cable I successfull program mega8 chip.

The problem is that your 7805 isn't protected against "back feeding." 
Its trying to regulate its output which is being driven higher than its 
unpowered input by your ISP cable.

You need a different model of regulator, else insert a diode between 
the existing regulator and the ISP connector.

Or you could power your circuit by normal means when using the ISP. 
This will keep your 7805 happy.

--
David Kelly N4HHE, dkelly@HiWAAY.net
========================================================================
Whom computers would destroy, they must first drive mad.

Hello, Leon.

You wrote:

  
  
>> Hello 2all.

>> When I in system program my tiny15 the current    more than 200mA!!!
>> The chip and 7805 are hot. I use Altera ByteBlaster    programmer with special PC software.
>> Very interesting: when I connect    swiched ON ISP the current is normal,
>> when I OFF it and then ON the current    more than 200mA
>> How this problem can be solved?
>> P.S. Using this ISP    cable I successfull program mega8 chip.

> It sounds like the chip is going into  latch-up. 

What you mean? What is  "latch-up"? Early this situation repeated with
90s2343. What it can be?
For info: Just after few seconds after new (just buyed) chip was
connected to the programmer the current ups to >200mA.


-- 

  Valerij Avrealij

Hello, David.

You wrote:

  
>  On Feb 13, 2005, at 5:13 AM, Valerij Avrealij wrote:

 >> When I in system program my tiny15 the current more than 200mA!!!
 >> The chip and 7805 are hot. I use Altera ByteBlaster programmer with
 >> special PC software.
 >> Very interesting: when I connect swiched ON ISP the current is normal,
 >> when I OFF it and then ON the current more than 200mA
 >> How this problem can be solved?
 >> P.S. Using this ISP cable I successfull program mega8 chip.

>  The problem is that your 7805 isn't protected against "back feeding."
>  Its trying to regulate its output which is being driven higher than its
>  unpowered input by your ISP cable.

>  You need a different model of regulator, else insert a diode between
>  the existing regulator and the ISP connector.

>  Or you could power your circuit by normal means when using the ISP.
>  This will keep your 7805 happy.


Thanks for your help. How you think, the chip was damaged? Or it can
be used?


-- 

  Valerij Avrealij

----- Original Message -----

From: Valerij Avrealij

To: Leon Heller

Sent: Sunday, February 13, 2005 12:56 PM

Subject: Re: [AVR-Chat] ISP problem: very high current when ISP switched ON (more 200mA)

Hello, David.

You wrote:

  
>  On Feb 13, 2005, at 5:13 AM, Valerij Avrealij wrote:

 >> When I in system program my tiny15 the current more than 200mA!!!
 >> The chip and 7805 are hot. I use Altera ByteBlaster programmer with
 >> special PC software.
 >> Very interesting: when I connect swiched ON ISP the current is normal,
 >> when I OFF it and then ON the current more than 200mA
 >> How this problem can be solved?
 >> P.S. Using this ISP cable I successfull program mega8 chip.

>  You need a different model of regulator, else insert a diode between
>  the existing regulator and the ISP connector.
Which model of regulator you recommend? Where I must insert diodes?
There are VCC, GND, MISO, MOSI, CLK,~RESET.

>  Or you could power your circuit by normal means when using the ISP.
>  This will keep your 7805 happy.
Do you mean power through battery?



-- 

  Valerij Avrealij

Hello, Leon.

You wrote:

>  With CMOS devices, it is possible in some situations for the substrate to
>  become biased so that the device behaves like an SCR, think of a 
>  forward-biased diode. Lots of current is drawn and the device can 
>  self-destruct. If power is removed quickly, they often recover.

>  It can be caused by applying a voltage to an I/O pin before the chip is
>  supplied with power, and there are several other causes:

> http://www.analog.com/library/analogDialogue/archives/35-05/latchup/

Very interesting effect. How it can be overcomed?


-- 

  Valerij Avrealij

On Sun, 13 Feb 2005, David Kelly wrote:

> 
> The problem is that your 7805 isn't protected against "back feeding."
> Its trying to regulate its output which is being driven higher than its
> unpowered input by your ISP cable.
> 

I've been using the Altera Byteblaster to program Altera CPLD's.  The 
Byteblaster I use has no power source in itself and gets its power from 
the board it's programming.

It could very well be latch-up.

Zack

>>
 >> The problem is that your 7805 isn't protected against "back feeding."
 >> Its trying to regulate its output which is being driven higher than its
 >> unpowered input by your ISP cable.
 >> 

>  I've been using the Altera Byteblaster to program Altera CPLD's.  The
>  Byteblaster I use has no power source in itself and gets its power from
>  the board it's programming.

>  It could very well be latch-up.

Thanks to all for quick help!
I think it is time to buy new ISP-programmer like PonyProg or
somethink like this.

-- 

  Valerij Avrealij

On Feb 13, 2005, at 5:57 AM, Valerij Avrealij wrote:
>
>>  The problem is that your 7805 isn't protected against "back feeding."
>>  Its trying to regulate its output which is being driven higher than 
>> its
>>  unpowered input by your ISP cable.
>
>>  You need a different model of regulator, else insert a diode between
>>  the existing regulator and the ISP connector.
>
>>  Or you could power your circuit by normal means when using the ISP.
>>  This will keep your 7805 happy.
>
>
> Thanks for your help. How you think, the chip was damaged? Or it can
> be used?

Its hard to hurt a 7805. The problem is that your ISP was sourcing 
current as hard as it was able, only 200 mA. Should be thermal and 
current protection in the regulators and all should be fine once they 
cool down.

If my diagnosis is correct the AVR was not a factor in your problem.

Some regulators are safe in this situation, others are not. Some 
specifically state that current will not flow backwards thru the 
device. Others say something very similar with "protected against 
reverse battery connection" which is slightly different.

Curious as to whether the ISP was able to hold 5 volts on the AVR under 
these conditions? If so then you might "get away" with it as you should 
only need the ISP for short bursts.

--
David Kelly N4HHE, dkelly@HiWAAY.net
========================================================================
Whom computers would destroy, they must first drive mad.

On Feb 13, 2005, at 8:30 AM, Zack Widup wrote:

> I've been using the Altera Byteblaster to program Altera CPLD's.  The
> Byteblaster I use has no power source in itself and gets its power from
> the board it's programming.

Valerij is using an AVR and the Atmel AVR-ICE provides power to the 
circuit. Don't know what ICE is actually being used.

> It could very well be latch-up.

It could but a 7805 is an ancient part and IIRC will suck its output 
side down rather than let it float higher than its input.

--
David Kelly N4HHE, dkelly@HiWAAY.net
========================================================================
Whom computers would destroy, they must first drive mad.

On Feb 13, 2005, at 6:53 AM, Valerij Avrealij wrote:

>>  You need a different model of regulator, else insert a diode between
>>  the existing regulator and the ISP connector.
>
> Which model of regulator you recommend? Where I must insert diodes?
> There are VCC, GND, MISO, MOSI, CLK,~RESET.

One diode in series between regulator output and the AVR Vcc. The AVR 
Vcc should connect directly to the ISP header's Vcc.

>>  Or you could power your circuit by normal means when using the ISP.
>>  This will keep your 7805 happy.
>
> Do you mean power through battery?

I don't know your circuit. If your device is normally powered with a 
battery thru the 7805, then, "yes." If this works then my diagnosis is 
probably correct and it would be worth your while to insert the diode 
above. Then you could program using the ISP power only.

--
David Kelly N4HHE, dkelly@HiWAAY.net
========================================================================
Whom computers would destroy, they must first drive mad.

Hello.

�Some�words�about�7805,ISP�and�my�AVR.

�First�of�all�my�programmer�works�fine�with�mega8�and�7805,�but�when�I

�connect�my�ISP�to�tiny15�or�90s2343�also�powered�from�7805�or�from

�another�source�(for�example,�power�supply�(AC�220V�->�DC�5V))�the�current

�is�very�high.

�Also�very�interest�situation�is�that,�when�I�connected�swiched�ON�ISP

�to�the�board�all�are�OK,�but�if�I�connected�ISP�with�breaked�VCC�to

�the�board�(but�board�VCC�is�not�breaked�(it�is�swiched�ON)),�and�than

�turn�on�the�ISP�I�have�very�high�current.

�

�Sorry�for�tangled�note.

�This�is�realy�interest�situation.

--�

��Valerij�Avrealij

On Feb 13, 2005, at 1:16 PM, Valerij Avrealij wrote:

> Hello.
>
>  Some words about 7805,ISP and my AVR.
>  First of all my programmer works fine with mega8 and 7805,

Thats nice but we still don't know much about "your" circuit. The 
reason you are having problems is that something is different. Is this 
board of your design? Is everything identical, same hardware, but for 
the exchange of Atmega8 and tiny15/90s2343?

>  but when I
>  connect my ISP to tiny15 or 90s2343 also powered from 7805 or from
>  another source (for example, power supply (AC 220V -> DC 5V)) the 
>  current is very high.

Once again is it the exact same board or just same regulator P/N? If 
two 5 volt regulators are connected in parallel (the one on your board 
and the one in your ISP) then its very unlikely they both agree on the 
same value of "5.0 volts". With luck one will simply source while the 
other sinks (both get hot). Without luck the two will oscillate (both 
get hot).

>  Also very interest situation is that, when I connected swiched ON ISP
>  to the board all are OK, but if I connected ISP with breaked VCC to
>  the board (but board VCC is not breaked (it is swiched ON)), and than
>  turn on the ISP I have very high current.

I don't understand the term "breaked VCC" in the above context. Am lost 
as to the difference between "breaked" and "switched." Am guessing you 
powered the circuit, then powered the ISP, and found high current. Also 
guessing that previously you did NOT power your circuit but the 
circuit's 7805 got hot when the ISP was powered, which is what lead to 
my previous suggestion to put a diode immediately after the 7805 
regulator before 5 volts reaches either the AVR or ISP header.

If the 7805 does not get hot with the board power off and the ISP power 
on, and the device programs, then that is the way things are supposed 
to work. Is what you were supposed to be doing in the first place. But 
if in this case the unpowered regulator gets hot you need to disconnect 
it, or use a diode rectifier in the circuit to automatically disconnect 
it, or select a different model 3-terminal 5 volt regulator.

>  Sorry for tangled note.

Would help a bunch if you didn't send HTML formatted email. I had to 
reformat your text to form a proper reply.

--
David Kelly N4HHE, dkelly@HiWAAY.net
========================================================================
Whom computers would destroy, they must first drive mad.

Hello, David.

_ANSWER_EMBEDDED_

You wrote:
>  On Feb 13, 2005, at 1:16 PM, Valerij Avrealij wrote:

 >> Hello.
 >>
 >>  Some words about 7805,ISP and my AVR.
 >>  First of all my programmer works fine with mega8 and 7805,

>  Thats nice but we still don't know much about "your" circuit. The 
>  reason you are having problems is that something is different. Is this
>  board of your design?
> Is everything identical, same hardware, but for the exchange of Atmega8 and tiny15/90s2343?
The problem with tiny15/90s2343 exists when I use any hardware logic.
There are NO problems with mega8, even if I use 7805. So, it is not
a 7805-problem.

 >>  but when I
 >>  connect my ISP to tiny15 or 90s2343 also powered from 7805 or from
 >>  another source (for example, power supply (AC 220V -> DC 5V)) the 
 >>  current is very high.

>  Once again is it the exact same board or just same regulator P/N?
Yes.
> If two 5 volt regulators are connected in parallel (the one on your board
>  and the one in your ISP) then its very unlikely they both agree on the
>  same value of "5.0 volts".
No, I have just ONE power regulator on my board. The ISP is powered
from target (I using Altera ByteBlaster).

 >>  Also very interest situation is that, when I connected swiched ON ISP
 >>  to the board all are OK, but if I connected ISP with breaked VCC to
 >>  the board (but board VCC is not breaked (it is swiched ON)), and than
 >>  turn on the ISP I have very high current.

>  I don't understand the term "breaked VCC" in the above context. Am lost
>  as to the difference between "breaked" and "switched." Am guessing you
>  powered the circuit, then powered the ISP, and found high current.
No, I realy mean that VCC line is "breaked" or "disconnected". The ISP
is not self-powered, it powered from the board. When I connect VCC
line I have very high current.
But, if I connect all wires (VCC, GND, MISO,...) at one time the
current is normal.

>  my previous suggestion to put a diode immediately after the 7805
>  regulator before 5 volts reaches either the AVR or ISP header.
Even I use another power source (i.e. AC-DC convertor from 220V to 5V)
I have some problem.

By taking all described above into consideration I think that it isn't
7805-problem.

Thanks for suggestions!

-- 

  Valerij Avrealij