Re: Potential hardware problem with RamFast

[DO11@calvacom.fr (Dominique OTTELLO)]

Dans <1994Apr08.175655.18409@calvacom.fr> DO11@calvacom.fr (Dominique OTTELLO) 
 ecrit:


>Hello from France.

>I apologize for my bad English. I hope you'll understand me.

>It seems it can have hardware design problems with RamFast. At least with
>mine. It is a short card Rev D.

>On this card there is three kinds of logical circuits : LS, HC and HCT.

>With 5 volts power supply:
>The minimum output logical one level for TTL LS is 2.4 volts
>The minimum input  logical one level for MOS HC is 3.5 volts
>The inputs of HCT circuits are fully compatible with outputs of TLL LS.

>If there is connection between TTL LS outpout and MOS HC input, the logical
>level one of the TTL LS can be not seeing as one by the MOS HC.

>It is very difficult to lay hands on this problem because I don't have
>schematics of the card.

>I have had this problem with electronic computer on Aircraft and the
>manufacturer wrote a Service Bulletin that ask airlines to replace MOS HC
>circuits by MOS HCT circuits.

>May be there is no potential problem with the RamFast if there is no
>connection between LS outputs and HC inputs.

>Thank you for your attention.


>* Dominique Ottello   -   do11�calvacom.fr    ! Sysop of Apple IIGS city on *
>* Paris - France - Apple II's user since 1980 ! Calvacom-RCI network        *

>> DO11�calvacom.fr (Dominique OTTELLO) writes:
>> May be there is no potential problem with the RamFast if there is no
>> connection between LS outputs and HC inputs.

> bazyar�netcom.com (Jawaid Bazyar) writes:
>  I believe this is the case. Also, note that most logic levels in most
> computers run very close to 0v and +5V; i.e., they almost never run
> near the limits of tolerance.

I agree with you. In most cases output logic level from LS is greater than
3.6V. Nevertheless, designers must calculate circuits with worst cases. Murphy
law is always true.

A great avionics manufacturer send a service bulletin to all airlines to
replace HC devices by HCT devices in a Fuel Quantity Indicating Computer
because sometimes there was failures with automatic refueling caused by LS
outputs connected to HC inputs.

  ==========================================================================
  = Dominique Ottello   -   do11�calvacom.fr    = Sysop of Apple IIGS city =
  = Paris - France - Apple II's user since 1980 = Calvacom-RCI network     =
  ==========================================================================