TL866A and trouble reading ATMEGA328P

[fitch]

First, I apologize if this is not the right place for this question.

I built a $20 LCR tester purchased from eBay, and all seems well with the tester.  I also just purchased a TL866A in case I needed to recover from incorrect fuse programming of the ATMEGA328P used in the LCR tester.

So, I just received my TL866A, and I downloaded the latest version of SW (6.5) from the web.  It installed fine (Windows 10) and did not offer to update the firmware (3.2.72), so that must be current as well.

Weird thing is if I try to read a programmed ATMEGA328P it reads as blank in code and data sections, but it can read the ID and it can read the fuses.  I am also able to reprogram at least some of the fuses using the TL866A, and I have been able to read an old 2732 eprom.  All the self tests pass.

I read reply #714 in this thread https://www.eevblog.com/forum/testgear/$20-lcr-esr-transistor-checker-project/msg517065/#msg517065 and it looks like he has the same problem, but no one came forth with any solutions.

Does anyone else have a TL866 that behaves this way?  Maybe it is just buggy software, but I would like to know in case I have to return it for refund.

Any info is really appreciated!

Thanks,
Bill

[Bicurico]

If the Atmel has the protection fuse on then you obviously cannot read out the memory contents. All you get is &H00 bytes.

Try your device with a new Atmel and flash it with some data without the protection fuse and read it out. It will certainly work.
Be careful with messing around your preprogrammed Atmel as you might end up erasing the memory contents and then you can buy a whole new device because the manufacturer certainly will not give you the software for you to flash.

Regards

[fitch]

I should say Doh!  Of course, the part has its lock bits programmed.  Another reads fine.  Thanks for the help!

Bill

[Bicurico]

There are tons of literature on how to circumvent the lock bit, the sum up with modern controllers is that you basically cannot. But very interesting reading, anyhow.

Regards

[SoundTech-LG]

I have a TL866, some other cheap ISP programmers, and a quite expensive Dataman 40 Pro. All the programmers are able to read Atmel AT89C51ED2 devices (the TL866 I do need to use SST 64K MCU to get contents). These Atmels reside in HDMI switcher/processor/scalers from Key Digital Systems. In the earlier version switchers I can also write the HEX files from the memory buffer contents. However...
on later switchers, even though I can look at the data contents on-screen, if I try to write a HEX file I get all FF in the files written !

[Bicurico]

I am by no means an expert on this matter but I think that the newer devices you want to modify might be using memories which are write once. You get them with all bits equal 0 and you can switch the bit only once to 1. It might actually be the other way around. This allows for cheaper chips and explains the behaviour you are experiencing.

Regards

[jazer]

Is there a simple way to pop an ATMEGA328P back to factory default state?

I had the misfortune of accidentally applying fuse settings from an ATMEGA32 onto the ATMEGA328P with predictable results: a bricked ATMEGA328P.  You'd think Atmel would make some effort at backward compatibility. Not even close.

It's no longer visible via the USB ISP I used to program it. I attached an external crystal in case the internal RC osc. got turned off but it didn't make any difference.

Any ideas, or do I just chalk it up to a learning experience?

[TheSteve]

You can google a high voltage programmer which will likely recover it. Probably not worth effort though as it is a real pain to do.

[jazer]

Thanks. I knew that can be done for the ATtiny versions but wasn't aware it's available for the larger footprint devices.

You're right that is a pain though.  The RESET has to be held to 12V and the programming has to be done in parallel, not SPI.  There are some resources to build a high voltage programmer such as:
http://denki.world3.net/avr_rescue.html

There's also at least one pre-built one available on ebay that is a standalone device that resets all the fuses so that programming can be done via the SPI interface again:
http://www.ebay.com/itm/ATMEL-AVR-ATMEGA-ATTINY-Fuse-Repair-Programmer-Fuse-Doctor-/222337203152?hash=item33c4547bd0:g:acgAAOSwo4pYRpzK
(The manufacturer also sells it directly on their web site: http://www.microcontrollerprog.com/fuseprog.html)

I'd have to have quite a few bricked chips to use either solution, so since I have plenty of good ones, my one dead bug will have to stay dead for now.

[SA]

If I reset the fuse of Atmega328p by using high voltage programmer, will I be able to read from read protected chip?

[ebastler]

If I reset the fuse of Atmega328p by using high voltage programmer, will I be able to read from read protected chip?

That would not be a very effective protection then, right? No, it does not work that way. "The Lock bits can only be cleared by executing Chip Erase", see section 28.7 of the chip's manual.

[coromonadalix]

unless as stated you use some glitch procedues   or even uncap the ic,   reading it is a NO  by normal means

resetting the fuse will kill the content,  as stated an high voltage programmer will do the same as erasing and resetting the bits, once done  the content is dead


you have fuses for spi access
fuses for read only  etc ....