[DSP-10] Fwd: HEY !! Bob !! Re: Corrupted Monitor on DSPx

Sat Dec 5 08:15:23 EST 2015

> Hi Bob:
> 
> Thanks for the in depth description, and sending the Flash IC's. Based on this information I think I have a much better understanding of how to proceed. With regards to R108, I am not sure at this moment. It was a couple of years ago when I built the DSP-10, that I would have to pull my notes from the build to determine if I made this change. Based on your description of what R108 achieves, I am thinking maybe I did not make this change. When I get home this evening, I will certainly check this out.
> 
> If I understand R108's function, and I did not make the change, this would certainly explain a lot of the trouble I have recently run into. Until recently, I did more monitoring or receiving with my DSP-10. It was not until I started transmitting that I noticed the problem. When monitoring, it is entirely possible I was just getting by, and the timing issue was not effecting me. 
> 
> 73
> 
> Rich, AJ3G
> 
>> On Dec 4, 2015, at 7:26 PM, Bob Larkin <boblark at proaxis.com> wrote:
>> 
>> Mike and I had more discussion about this this morning.  Then I went back and reviewed the DSPx software (that is available from the DSP-10 page).  A couple of thoughts evolved that might answer your question about the monitor use, Rich.
>> 
>> AS Mike pointed out, the DSP-10, running with the DSPx, always starts with Bank 0, the monitor (specifically the program XMON).  It first does some initialization, including the DSP-10 shift registers, examines JP1 and if present, it boots up another byte by byte program from Flash Bank 1.  This program for the DSP-10 is UHF3X.  Once this happens, XMON is not used again.  Therefore, once the UHF3X starts running the health of the monitor is not important.  But, for UHF3X to ever run at all, the XMON must be there.
>> 
>> Also, In the EZKit version of UHF3, there is a version of the monitor loaded into high program memory.  In the case of the DSPx, this is not needed because of Bank 0, and so it is not included.
>> 
>> One more thought.  Do you have the R108 change?  From the DSP-10 main page, under "Hardware Information, "R108 should be 4.7K Ohms (not 470). Some combinations of U107 and U104 have been found to have the data line at U104 switching too early. These are cascaded shift registers and the timing is quite close. The R108 resistor change slows down the data line going to U104, making the timing less critical."
>> 
>> And, also, a couple of the flash IC's went out in today's mail.
>> 
>> Good luck with getting it to work.
>> 
>> 73, Bob  W7PUA
>> 
>> 
>>> On 12/03/2015 01:49 PM, Rich Miller via DSP-10 wrote:
>>> Hi Mike:
>>> 
>>> What you have written certainly makes sense to me. You are correct, the monitor was present in some form until last night. This is when I noticed the FL1 had completely turned off. Upon recycle  it was clear something was amiss. When I attempted to reload UHF 3, it became clear the monitor was gone.
>>> 
>>> Based on your description, I am having a problem understanding how corrupted monitor could continue to allow certain operations to work, while not allowing other functions to operate correctly. I would think if the monitor was corrupted in anyway, all operations would cease to work (upon reboot). If this is true, then I think there may be a problem with the ADSP-2185 which was causing the original issue, and I somehow managed to corrupt the monitor working on the original symptoms.
>>> 
>>> 
>>> 73
>>> 
>>> Rich, AJ3G
>>> 
>>>>> On Dec 3, 2015, at 4:25 PM, kd7ts <kd7ts at comcast.net> wrote:
>>>>> 
>>>>> On Thu, 03 Dec 2015 10:30:28 -0800, Rich Miller via DSP-10 <dsp-10 at mailman.qth.net> wrote:
>>>>> 
>>>>> Just out of curiosity, is it possible in your opinion for a DSP-10 to partially run with a corrupted monitor? I really can't find any fault in any of the other parts of the RF Chain on the main board. When serial programming works and programs the shift registers, then the PLL's work great.
>>>> 
>>>> Hi Rich,
>>>> 
>>>> Opinions ? Sure, I wrote this up a few years ago, and will add some clarification ..
>>>> 
>>>> The ADSP2185 is a specailized processor optimized to handle FFT calculations, but also does many other jobs to make the DSP-10 operate. It does nothing on its own and requires instructions to be loaded into its memory. A Flash memory chip, 39VF040, is used to store the programs that the ADSP2185 uses.
>>>> 
>>>> When the DSP-10 is started the ADSP2185 looks for something to run.  In the case of the ADSP2185 used in the DSP-10x, this is BANK0 of the 39VF040 flash memory. The first few instructions are a "boot loader" that load the monitor stored in the flash to the low memory of the ADSP-2185 and start the monitor running from the ADSP-2185.  Once running, the monitor looks at the jumpers, JP1 and JP2 on the interface board to see what should happen next.
>>>> 
>>>> The monitor, still running from the ADSP2185, checks  JP1 for further instruction and if JP1  is closed,  looks in BANK1 of the 39VF040 and loads the UHF3 DSP program into the ADSP2185 memory and turns execution over to the DSP program. At this point, we have the monitor loaded in low memory and UHF3 loaded above it somewhere, but still in the ADSP-2185. The flash is not needed until the next time the DSP-10 is started.
>>>> 
>>>> If JP1 is open, the monitor, running in low memory of the ADSP-2185, waits for instructions from the outside. This is where the monitor is used to load FLASHUTB to the ADSP2185 and execution is turned over to the "FLASH UTILITY".
>>>> 
>>>> The flash utility knows how to talk to the flash memory, and we use it to load new insructions and/or programs to the appropriate flash memory bank. The monitor is still in the ADSP2185 low memory, but program execution is under control of the "FLASH UTILITY" , FLASHUTB.EXE.
>>>> 
>>>> As long as power is on, the ADSP2185 memory is retained, and we can use this feature to write a copy of the monitor back to the FLASH MEMORY. If you erase or corrupt the monitor in BANK0, it can be written back by the flash utility from the copy in ADSP2185 memory.
>>>> 
>>>> If you erase or corrupt the monitor in BANK0, AND reset the DSP-10 or cycle the power, then there is no longer a copy of the monitor in the ADSP2185 low memory and no copy in the 39VF040, so you can put this away until you get another 39VF040 flash chip with a monitor in BANK0.
>>>> 
>>>> The monitor is the first thing we use and the heart of these operations, and without it there is no reading, or writing of the flash memory.
>>>> 
>>>> So, the opinion .. if there is not a monitor in the flash chip then UHF3 will never be loaded to the ADSP-2185 and not much is going to happen. If the PLL gets programmed, then there must be a copy of the monitor in Bank0 of the 39VF040.
>>>> 
>>>> http://www.proaxis.com/~boblark/dspyour's10_dspx_flashbank1.htm
>>>> 
>>>> There are many hints available from the software used to update the UHF3 program from the link above.
>>>> Remember there are 2 jumpers, for the two banks, JP0 controls access to BANK0 where the monitor is stored, and JP1 controls use of BANK1 of the 39VF040 where UHF3 is stored.
>>>> 
>>>> I am guessing that the problem must lie in the BANK1 data, or a problem in moving the program data stored in BANK1 to the ADSP-2185 memory.
>>>> 
>>>> Mike KD7TS
>>>> 
>>>> How's that for an opinion ??
>>> ______________________________________________________________
>>> DSP-10 mailing list
>>> Home: http://mailman.qth.net/mailmanyour's/listinfo/dsp-10
>>> Help: http://mailman.qth.net/mmfaq.htm
>>> Post: mailto:DSP-10 at mailman.qth.net
>>> 
>>> This list hosted by: http://www.qsl.net
>>> Please help support this email list: http://www.qsl.net/donate.html
>>> 
>>>