Do these customs and handling fees apply to shipments originating from within the Euro-zone?

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!

(I note that the PiDP-8 person drives to the Netherlands for affordable postal services.)
Edit: oops, not the swissmicros person.

Hi again Peter. Your remark about the '816 not violating its own requirement has jogged my thinking -- thanks! Certain info in the datasheet is implied, and for me in this case it fills in the blank -- the WDC doc does contain all the necessary detail for managing the break-before-make business. I thought the WDC datasheet had an error or omission, but we all know THAT never happens!

Agreed -- only one peripheral device at a time. What I meant to get at is, any device that wants to talk to the '816 without contention must play nicely -- its outputs need to walk the walk regarding the min & max times required to enable, and the min & max times required to disable. With '816, avoiding contention is a bothersome business, and I'm not saying every situation demands it -- it's not a judgment. Anyway, as I see it the options are:

• Using a transceiver you need to...
  - Determine when the transceiver's OE cue will arrive, then select a '245/equivalent ('HC245, 'AC245, CPLD?) that can respond in a timely fashion. ("Timely" meaning those four specs are met). Memory & I/O also have specs to meet, but those specs are far more forgiving because Ø2-low is all dead time on the non-multiplexed part of the bus. (The 'C02 data bus is easy for the same reason.)
  
  
• Omitting the transceiver you need to...
  - For the ROM, determine when its OE will arrive then select a product that can respond in a timely fashion
  - For the RAM, determine when its OE will arrive then select a product that can respond in a timely fashion
  - For I/O device A, determine when its OE will arrive then select a product that can respond in a timely fashion
  - For I/O device B, ...

Despite all the rigor, omitting the transceiver is still appealing, especially for a ROMless sytem with minimal IO.

I agree the point re low output-impedance may be quite important. Contention is only as bad as its strongest participants! So when deciding whether contention can be tolerated it's worth considering that an 'HC245 or 'AHC245 will result in less Vcc noise than an 'AC245. (OTOH if you've done the homework and satisfied the timing specs then there'll be no contention, and low output impedance is a non-issue.)

I feel the same way. The 65c265, which includes an '816 core, has a non multiplexed address bus, but the speed rating on the '265 is only 8 MHz.

cheers,
Jeff

_________________
In 1988 my 65C02 got six new registers and 44 new full-speed instructions!
https://laughtonelectronics.com/Arcana/ ... mmary.html

Yes they apply as well, Switzerland is still not in the Euro-Zone and is not even part of EU. Worse even many EU companies do not deliver goods to private persons in Switzerland. There are only some treaties regarding cooperation and even those, by many citizens and the strongest party, are under heavy criticism. The penalty (handling by swiss customs) to order something from most companies in the Euro-Zone is at least 28USD. Sometimes it is even worse, especially when something is sent by DPD. Luckily for me we there is a free trade agreement with China and so ALIEXPRESS is my best friend. If only the delivery times would be faster.


Actually in my designs there is only one OE and devices have individual CE/CS (chip enable/select). I already thought about creating a OE (in my design I call it MRD=master read) which is delayed when asserting. OE read times are normally much shorter than CS read times so this would not have any impact on speed but would avoid contention.