[time-nuts] FE-5680A ("New" version) fine frequency adjust

Peter Bell bell.peter at gmail.com
Fri Nov 18 19:45:46 UTC 2011 

The Xilinx chip (which is strictly speaking a CPLD) is programed via
JTAG - all the required pins are on the test connector on the edge of
the board.  In theory, you can also read back the fuse maps using the
JTAG port if the chip hasn't been secured. I haven't actually tried
it, because I expect the chip IS secured - and a raw fuse dump is not
that useful anyway for anything except copying the device.

In any case, I'm pretty sure the servo loop in this unit is analog -
once the board initializes, the CPU spends most of it's time in halt
and only wakes up on a 1Hz interrupt to do some sort of housekeeping
functions then halts again, while the servo is operating continuously
The serial data lines that program the DDS are also connected directly
to the CPU rather than the XIlinx chip - and are only active when the
board is starting up (and, I presume, when you send it the trim
commands - but I haven't actually verified this).

The code that the CPU runs (which is held in that Waferscale/ST PSD
chip) should be accessible, since it's just addressed like a flash
chip by the CPU - the only possible difficulty would be figuring out
how the bank mapping and decode PLD are set up - these are also
programmable via a (different) JTAG port, bit are also presumably
secured - there has to be some mapping going on because the PSD has
1mbit (128KB) of flash and the MCU only has a 64KB address space -
having said this, it's likely that just blindly dumping the first 64K
of the address space would turn up at least something interesting -
this is actually on my list of "things I will try when I have some
spare time" :)

Regards,

Pete


On Sat, Nov 19, 2011 at 1:03 AM, Don Latham <djl at montana.com> wrote:
> Let me add my thanks. My knowledge of FPGA's is about the same as my
> knowledge of the female of our species. Is it possible to read the
> program that's been entered into the chip? I'm not clear as to the data
> flow in this beast. If the control loop is indeed digital, then there's
> a PID controller in software? It's possible then to control the
> frequency as some want to do with the beast as it is, using an Arduino
> or Picaxe or some such to adjust the frequency with an external loop and
> digital corrections via the rs232 rather than all this crude analog
> resistor stuff :-)
> Don
>
>
> EWKehren at aol.com
>> Pete
>> Thank you, I will look at its some time in the future, right now I let
>> it
>> cook for a couple of weeks undisturbed and than I may cut the trace that
>> seems  not to do any thing. If you have a 100 MHz scope it will trigger
>> on any
>> pulse in  this unit unless your scope is defective. I think there are
>> features hardware  wise in this unit that can be activated at
>> manufacturing
>> depending on price and  demand. I am sure you are aware of instruments
>> that a
>> hardware wise the same and  cost any where from $15 000 to $40 000 the
>> only
>> difference is the software  imbedded and many times you can buy an
>> upgrade.
>> Bert
>>
>>  message dated 11/18/2011 10:26:02 A.M. Eastern Standard Time,
>> bell.peter at gmail.com writes:
>>
>> All the  pins have things connected to them - but pin 6 seems a little
>> strange -  it's connected to the 'ACT240 (in fact, it goes through 3
>> sections of the  chip in series!) and finally to the xilinx chip - but
>> none of these signals  ever appear to do anything and the output is
>> just constant high.  It's  possible that this the pulse is just
>> extremely short - the fastest scope I  have at home is only 100Mhz -
>> but I would have expected to be able to  tirgger on it at least.
>>
>> Regards,
>>
>> Pete
>>
>>
>> On Fri, Nov 18,  2011 at 9:09 PM,  <EWKehren at aol.com> wrote:
>>> Pete
>>>  Thank you for doing all the work. This is great info and will help
>>> all
>> of
>>> us that want to discipline the beast. One more question, since you
>>> have
>> done
>>> so  much testing is there one pin on the connector that  has nothing
>>> connected to  it?
>>> Thanks again   Bert  Kehren
>>>
>>>
>>> In a message dated 11/17/2011 9:59:14 P.M.  Eastern Standard Time,
>>> bell.peter at gmail.com writes:
>>>
>>>  OK, I  just managed to get some spare time to have a look at this,
>>> and
>>> here are  some conclusions:
>>>
>>> 1) The fine  frequency adjust pot is not connected to  the c-field
>>> circuit.  All the connections to it go across the board to the
>>> MAX1246
>>> ADC  chip.  The track is wired between ground and the vref  output  on
>>> this chip and the wiper goes to a little pi filter (ceramic cap,
>>> 100K,
>>> 1uF tant) that feeds channel 1 (pin 3) on the ADC.  As far as I  can
>>> see, it goes nowhere else.
>>> 2) The  c-field current is fixed.   The circuit is very simple - there
>>> is  a 5.1V reference generated using a  zener and a 681R resistor to
>>> the
>>> 8.8V and this is fed into a potential  divider consisting of  a pair
>>> of
>>> 120K resistors.  The centerpoint of  this divider  is connected to pin
>>> 5
>>> (+ve input) of a TLC27M4BI - the output  of  this amp (pin 7) is
>>> connected via a 270R resistor to pin 2 of the FPC  connector and hence
>>> to the c-field coil.  The return side  of the coil  (pin 1 on the FPC
>>> connector) connects to pin 6 (-ve)  of the opamp and via a  620R
>>> resistor to the 5.1V  reference.
>>>
>>> If there is any other input to  this circuit,  I have not been able to
>>> find it, despite probing all over the  board.
>>>
>>> Conclusion:
>>>
>>> This version carried  out the fine frequency  adjust digitally by
>>> changing the reload  values on the DDS chip - the reason  the pot does
>>> does nothing in  this specific model is that the DDS trim is  now
>>> being
>>> carried  out over the RS-232 port and as a result the pot value  (if
>>> it's  being read at all) is not being used for  anything.
>>>
>>>  Adjusting the c-field still works, though - if you connect a  DC
>>> supply
>>> set at about 2.55V via a high value resistor (I used 100K) to  pin  5
>>> on
>>> the TLC27M4BI next to the flex cable connector then you  can trim  the
>>> frequency by adjusting  it.
>>>
>>>  Regards,
>>>
>>> Pete
>>>
>>>  _______________________________________________
>>> time-nuts  mailing list -- time-nuts at febo.com
>>> To unsubscribe, go to
>>>  https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>>> and follow the  instructions there.
>>>
>>>  _______________________________________________
>>> time-nuts mailing list  -- time-nuts at febo.com
>>> To unsubscribe, go to
>> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>>> and follow the  instructions  there.
>>>
>>
>> _______________________________________________
>> time-nuts  mailing list -- time-nuts at febo.com
>> To unsubscribe, go to
>> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>> and follow the  instructions there.
>> _______________________________________________
>> time-nuts mailing list -- time-nuts at febo.com
>> To unsubscribe, go to
>> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>> and follow the instructions there.
>>
>
>
> --
> "Neither the voice of authority nor the weight of reason and argument
> are as significant as experiment, for thence comes quiet to the mind."
> R. Bacon
> "If you don't know what it is, don't poke it."
> Ghost in the Shell
>
>
> Dr. Don Latham AJ7LL
> Six Mile Systems LLP
> 17850 Six Mile Road
> POB 134
> Huson, MT, 59846
> VOX 406-626-4304
> www.lightningforensics.com
> www.sixmilesystems.com
>
>
>
> _______________________________________________
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.
>

More information about the time-nuts mailing list