[GreenKeys] Isochronous mode

Brooke Clarke brooke at pacific.net
Mon Jan 4 15:06:29 EST 2010


Hi John:

Do the KG signals get XORed with the TTY output in the SB-2244?  If not 
then why are the KG signals in the SB-2244?

Thanks for the links & Have Fun,

Brooke Clarke
http://www.PRC68.com


John Nagle wrote:
>> Date: Sun, 03 Jan 2010 12:44:18 -0500
>> From: Bill Horne<ehorne at speakeasy.net>
>> Subject: Re: [GreenKeys] SB-2244 TTY patch panel
>> To: greenkeys at mailman.qth.net
>>
>> I think it's just a throttling mechanism to slow the machines down while
>> the rotors align in the crypto gear.
>>
>> Bill, W1AC
>>
>> On 1/3/2010 11:52 AM, Brooke Clarke wrote:
>>      
>>> Hi:
>>>
>>> It may be related to TEMPEST.  A mechanical TTY machine generates RF
>>> signals at each on or off transition.  If a bunch of machines are
>>> synchronized so they all have a transition at the same time then their
>>> TEMPEST emanations may get mixed up.
>>>
>>> Have Fun,
>>>
>>> Brooke Clarke
>>> http://www.PRC68.com
>>>
>>>
>>> Jim Haynes wrote:
>>>
>>>        
>>>> Some implementations of Teletype with crypto equipment required that the
>>>> stepping the Teletype keyboard or tape reader at a fixed rate.  In the
>>>> Model 28 parts books you'll see mod kits to adapt things for "synchronous
>>>> pulsed operation".  A large station might have a TD stepper facility
>>>> consisting of a baud rate generator and then a bunch of boxes that
>>>> send out TD stepping pulses to all the TDs.
>>>>
>>>> I don't know why it was this way.  I don't have any first-hand knowledge
>>>> of crypto equipment, but I did once buy some TD stepping equipment at
>>>> a surplus sale.
>>>>
>>>> Where this is done each TD or keyboard has to have its usual loop and
>>>> then a loop for the stepping pulses.  So I'm guessing what you have
>>>> provides for switching both the signal and the stepping pulses.
>>>>
>>>>
>>>>
>>>> jhhaynes at earthlink dot net
>>>>          
>       That's called "isochronous mode".  The same bits are sent as if in normal
> async mode, but the equipment runs in sync with a constant clock.  This is
> normal cryptographic practice for fixed links.   The crypto gear (what NSA calls
> a "key generator", with a type beginning with "KG" generates psuedorandom bits
> at a constant clock rate, and those are XORed with the outgoing data.  When no
> data is being sent, the process continues, so the encrypted data stream doesn't
> indicate whether anything is being sent or not.  Bits go out even when the
> sender is idle.  This prevents traffic analysis.
>
>      For info on Teletype-era crypto, see
>
> http://www.nsa.gov/public_info/_files/cryptologic_histories/history_comsec.pdf
>
> Page 82 starts a clear discussion of TEMPEST problems in the Teletype era.
> Teletypes radiated enough to be readable 50-200 feet away.  Friden
> Flexowriters (especially the Justowriter, the first typewriter that
> could right-justify automatically) could be read from over 600 feet away.
> Teletypes could even be read by listening to the audio from a telephone.
> The details of what was done about this are in that document.
>
>      For the later generation of Teletype crypto (electronic, as opposed
> to older rotor systems) see
>
> 	http://www.jproc.ca/crypto/kw7.html
>
> The corresponding 60WPM Baudot rotor machine is the KW-2, but I can't
> find any pictures of that on line.  Electromechanical on-line rotor
> machines never worked very well - too many sliding contacts running
> at too high a speed.
>
> 					John Nagle
>
> ______________________________________________________________
> GreenKeys mailing list
> Home: http://mailman.qth.net/mailman/listinfo/greenkeys
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:GreenKeys at mailman.qth.net
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
>
>
>    



More information about the GreenKeys mailing list