[GreenKeys] Western Union Ticker-5A Stock Quotation Machine

Wed Apr 18 13:01:17 EDT 2007

Hi Bob:

Now we're cooking.

The 5-A coil is 0.15 H, 20 Ohms and let's say 60 ma operation.
The printing rate is 500 characters per minute with 6 bits per character (no 
start, no stop bits) or 3,000 bits per minute, or 50 bits per second, or 20 ms 
per bit.  i.e. half that baud rate you mentioned.  So the time interval might 
be T = 1/500 or 0.002 seconds.

Now:  V = L * dI / dT = 0.15 * 0.06 / 0.002 = 4.5 volts which is close enough 
to a standard 5 volt logic supply.

Thanks for an explanation I could get my teeth into!

How do the coil specs for the 5-A compare to the ASR-33 that I used to use with 
my first computer?

Have Fun,

Brooke Clarke, N6GCE
w/Java http://www.PRC68.com
w/o Java http://www.pacificsites.com/~brooke/PRC68COM.shtml
http://www.precisionclock.com

Bob Camp wrote:
> Hi
> 
> Yes, some have done the experiment you describe. In my case it was  
> sometime in the late 1960's (gulp!) . Results:
> 
> 1) With the diode on the coil the system simply does not work at all.  
> The magnetic field does not collapse fast enough to release the  
> armature in time. It's not a matter of distortion in this case, it's  a 
> matter of not having any copy at all.
> 
> 2) The high voltage on "break" allows the energy to be dumped faster.  
> When the volts X amps is higher the coil dumps it's stored energy in  a 
> shorter period of time.
> 
> 3) In order for the mechanical part of the system to work correctly  the 
> electrical signals need to be pretty accurate. Any significant  
> distortion in the electronics mucks up the mechanical side pretty fast.
> 
> Here's some math ....
> 
> The basic formula for current in a coil is:
> 
> voltage = inductance X (delta current) / (delta time).
> 
> At  100 baud and 10% distortion T =  1/1000 of a second.
> Current is about 60 ma.
> 
> so:
> 
> V = L x 0.060 / 0.001
> V = 60 * L
> 
> That's for a constant current loop ...
> 
> If L is in the roughly couple of Henrys range then you get voltages  at 
> or above 60 for any reasonable level of distortion.
> 
> Bob
> 
> 
> On Apr 18, 2007, at 12:34 AM, Brooke Clarke wrote:
> 
>> Hi Peter:
>>
>> Did you try using a current source instead of a voltage source?   Did 
>> you use a protection diode across the coil?
>>
>> I've been studying this for most of today and I've found two key  
>> factors.
>>
>> 1) the loop time constant is L / R.  It may take 3 to 5 time  
>> constants for a signal to settle so the time constant needs to be a  
>> very small part of a bit time or you're going to have problems.   But 
>> using a current source brings the time constant near zero, i.e.  R is 
>> near infinity.
>>
>> 2) the coil in the electromagnet does not like to change currents.   
>> If there is already current in the coil and you turn it off the  coil 
>> tries to generate a voltage that's high enough to maintain the  old 
>> current.  This is how the spark coil in a car works.  This  voltage 
>> pulse can easily destroy the driver transistor.  The common  way of 
>> fixing that is to put a diode across the coil with it's  polarity such 
>> that when the coil is forward biased the diode is  back biased, but 
>> (forgetting about the bi polar drive for a minute)  that will not work 
>> and at the same time allow the coil to operate  quickly, instead the 
>> diode slows the coil way down since now  there's a current circulating 
>> in the diode and coil with it's own  long L/R time constant.
>>
>> I think the fix for (2) is to use a bi polar Transil.  It's sort of  
>> like back to back zener diodes (which probably would also work)  with 
>> a breakdown voltage greater than the DC drop across the coil  when 
>> saturated but much much faster and rated to absorb spikes.
>>
>> I'd like to hear from someone who has done this type of  experiment.  
>> My gut feeling is that a properly designed driver can  work with a low 
>> voltage supply.  My brain is telling me that  there's a lot of people 
>> using high voltage that would not be if  there was an easy way to 
>> avoid it.
>>
>> Have Fun,
>>
>> Brooke Clarke
>> w/Java http://www.PRC68.com
>> w/o Java http://www.pacificsites.com/~brooke/PRC68COM.shtml
>> http://www.precisionclock.com
>>
>>
>>
>> Peter Gottlieb wrote:
>>
>>> Well obviously a power op-amp (Burr-Brown) or homebrew  
>>> (complimentary pair, see National app notes) will take RS-232 and  
>>> make bipolar.
>>> I am not too familiar with the dynamics of polar relays but TTY  
>>> selector magnets need a high voltage to get current into the high  
>>> inductance rapidly.  I couldn't get my 2B to run at 28 volts even  
>>> though I needed a series resistor to drop the current, but go to  100 
>>> and it ran like a champ.
>>> Peter
>>> Brooke Clarke wrote:
>>>
>>>> Hi:
>>>>
>>>> I'm trying to bring a Western Union Ticker-5A Stock Quotation  
>>>> Machine back to life.  This is the machine that replaced the  older 
>>>> stock ticker machines that were under glass domes to keep  down the 
>>>> noise.  I'd say that it's father was the Edison  Universal stock 
>>>> ticker that changed Edison from someone who was  broke to a multi 
>>>> millionaire.  It's mother was the model 26  Teletype machine since 
>>>> they both use a similar, but not  interchangeable, selector cage.
>>>>
>>>> It's my understanding that the 26 was good for 100 CPM yet the 5A  
>>>> was good for 500 CPM.
>>>>
>>>> The 5A uses a permutation code that consists of 5 bits but  instead 
>>>> of using a character for shift to numbers and another  character for 
>>>> shift to letters the 5A uses another code bit for  this so the total 
>>>> number of bits in the code is 6.  This solves  the stuck in figures 
>>>> problem.
>>>>
>>>> But unlike teletype machines the 5A does NOT use start and stop  
>>>> bits.  This is one of the things that makes it fast.
>>>>
>>>> The thing I'm working on now is how to drive it.  The input goes  to 
>>>> a polar relay that expects to see current polarity reversals,  not 
>>>> the on off keying that a normal teletype machine uses.   Testing has 
>>>> revealed that it wants to see 50 to 100 ma and being  a 20 ohm coil 
>>>> a little over 1 volt is all that's required.
>>>>
>>>> Are that circuits already in existence that will take in RS-232  and 
>>>> supply the bi polar drive current?  I think that might be  more 
>>>> common in the UK or Australia than in the US.
>>>>
>>>> Why do teletype machines typically have loop voltages in the 100  to 
>>>> 200 volt area?  Could these be replaced by current sources  that 
>>>> have a voltage compliance a litter higher than the voltage  drop 
>>>> across the input relay coil?
>>>>
>>>> For more on the 5A see my web page:
>>>>
>>>> http://www.pacificsites.com/~brooke/WU5A.shtml
>>>>
>>>> Have Fun,
>>>>
>>>> Brooke Clarke, N6GCE
>>
>> _______________________________________________
>> GreenKeys mailing list
>> GreenKeys at mailman.qth.net
>> http://mailman.qth.net/mailman/listinfo/greenkeys
>>
> 