Lattice MachXO2 breakout board - replacing FPGA with different one ?

Hi,

I have a couple of that neat boards with XO2-7000 and not much else.

I managed to burn the FPGA on one of them and since Farnell didn't have the exact 7000HE model that was on the board, I used 7000HC and bridged VCC so that chips gets 3.3V for core power that it wants.

But now I can't program it through FTDI. I've noticed several variations of the board. Some had XO2-1200 while mine have XO2-7000HE.

Shoould this with pin-compatible chip just work or do I have to have different EEPROM content for FTDI ?

BTW, trivial "blinkenled" example on _this_ board  fails with:

*************
Device#1 LCMXO2-7000HC: Failed to verify the ID 
(Expected: 0x012BD043 Read: 0xFFFFFFFE).

ERROR - Check configuration setup: Unsuccessful.

ERROR: pgr_program failed.

ERROR - Programming failed.
**********

If it doesn't work through FTDI, I'll use ISP cable on it, but it's nuisance...

On 7/8/2016 12:26 PM, Brane2 wrote:
> Hi,
>
> I have a couple of that neat boards with XO2-7000 and not much else.
>
> I managed to burn the FPGA on one of them and since Farnell didn't have the exact 7000HE model that was on the board, I used 7000HC and bridged VCC so that chips gets 3.3V for core power that it wants.
>
> But now I can't program it through FTDI. I've noticed several variations of the board. Some had XO2-1200 while mine have XO2-7000HE.
>
> Shoould this with pin-compatible chip just work or do I have to have different EEPROM content for FTDI ?
>
> BTW, trivial "blinkenled" example on _this_ board  fails with:
>
> *************
> Device#1 LCMXO2-7000HC: Failed to verify the ID
> (Expected: 0x012BD043 Read: 0xFFFFFFFE).
>
> ERROR - Check configuration setup: Unsuccessful.
>
> ERROR: pgr_program failed.
>
> ERROR - Programming failed.
> **********
>
> If it doesn't work through FTDI, I'll use ISP cable on it, but it's nuisance...

I'm not sure what's wrong.  If the ID were all 1's, I'd say it was a bad 
solder joint, but with the lone 0 in the returned ID that doesn't seem 
to fit.  Still, that ID is just not right either and it would seem 
something is wrong with the JTAG.  Is the "expected" ID for a 7000HC or 
a 7000HE?

-- 

Rick C

Dne petek, 08. julij 2016 20.57.18 UTC+2 je oseba rickman napisala:

> I'm not sure what's wrong.  If the ID were all 1's, I'd say it was a bad 
> solder joint, but with the lone 0 in the returned ID that doesn't seem 
> to fit.

I sorted it out. It works now. YEAH ! ;o)
BOard was damaged in very weid way. I seems that I managed to stab the board with something like Xacto knife in exactly the "right" spot - TDI line at ISP connector. It was so small it was practically invisible. 

After patching that track, XO2 Breakout Board works fine with HC chip now ;o)






  Still, that ID is just not right either and it would seem 
> something is wrong with the JTAG.  Is the "expected" ID for a 7000HC or 
> a 7000HE?
> 
> -- 
> 
> Rick C

On 7/12/2016 9:28 PM, Brane2 wrote:
> Dne petek, 08. julij 2016 20.57.18 UTC+2 je oseba rickman napisala:
>
>> I'm not sure what's wrong.  If the ID were all 1's, I'd say it was a bad
>> solder joint, but with the lone 0 in the returned ID that doesn't seem
>> to fit.
>
> I sorted it out. It works now. YEAH ! ;o)
> BOard was damaged in very weid way. I seems that I managed to stab the board with something like Xacto knife in exactly the "right" spot - TDI line at ISP connector. It was so small it was practically invisible.
>
> After patching that track, XO2 Breakout Board works fine with HC chip now ;o)

Yes, that makes more sense.  A floating TDI line could result in nearly 
anything being returned.  I thought the tool had a diagnostic mode that 
would tell you that sort of thing.  But maybe you have to select that.

I build a board with an XP device and used to do all the testing myself. 
  Every once in awhile I would get a flaky JTAG connection and it would 
report what it thought was wrong.  That was using the programming 
software with the old classic tools, but that should be the same with 
the diamond tools.

-- 

Rick C