4 block of info down. Taken

4 block of info down. Taken from their instruction manual. 

Check the maths

Dear Ed,

According to the Switch-8 Mk2 webpage, the max current per Physical Output is 40milliAmps

https://ncedcc.zendesk.com/hc/en-us/articles/200526429-Switch-8-Mk2

​Now, comparing that to the known "operating" and "stall current" specs of a Circuitron Tortoise

http://www.circuitron.com/index_files/tortoise.htm

Operating : 4milliAmp
Stall : 15 - 16milliAmp

it would appear that even in worst-case "both motors stalled" condition, the cumulative 32milliAmp load of "2x Tortoises connected to a single output" is within the 40milliAmp output capability of the Switch-8 Mk2.

That said, I have to ask, how are you actually entering the command to "throw the pair-of-turnouts"?
If you are using any system which allows triggering of "macros"
(Say, an NCE command system, a MiniPanel, or JMRI),

then you _could_ well hang each Tortoise off a single Physical Output (and associated discrete address),
let the Human User Interface "trigger a macro",
where the macro (series of instructions) sends the two-required-address-throw commands
(appearing to the End User "as one self-same single-command --> both turnouts throw" cause-effect event).

The bonus of "solving it with a macro" (IE use the "control input" part of the signal-path to achieve the desired result)

is that keeping each Tortoise on it's own physically-addressible output maintains flexibility and control
(even if the context of a "paired-turnout crossover" means "having one turnout throw without the other" feels kinda illogical... I have a current layout plan where being able to "split-throw half-a-crossover" is actually critical to the S.O.P. ). 

I hope this helps...

Happy Modelling,
Aim to Improve,
Prof Klyzlr

PS just read the Switch-8 MK2.DOCX 
https://ncedcc.zendesk.com/hc/en-us/article_attachments/201221809/Detailed_Instructions_for_Switch-8_Mk2.docx

and it occurs to me...
the basic process is
- select the Switch8 Output to be programmed (say "Output X")
- press the "Program Output" button
(this effectively tells the Switch8 to "listen-for and remember the next Accessory DCC throw command you hear")
- using the DCC system, issue the desired ACCY THROW command (address) in question (say, ACCY 1)
- and thus Output X is now programmed to "listen and respond" to ACCY 1 command

Assuming the above to be true, 
what's to stop the modeller:
- Programming "Output X" as above to "ACCY 1"
- Then doing the same programming-process for "Output Y", but throwing "ACCY 1" _again_?
(NB that if this works, and both "Outputs" are now listening/responding to a common "ACCY 1" DCC command)
then the Switch-8 can have each output "reversed" after-the-fact if required,
using CVs 561 - 568 (Outputs 1 thru 8 respectively).

Different instructions - found it

@Neil, Thanks for the pointer to the instructions.  I didn't spot it because the paper version I have is pretty fine print and I didn't catch that pargraph in the "Other Technical Stuff" section.

@Prof, the plan is to be abl to control either through the NCE handset, or through JMRI, or via push buttons via the button board.  For now it will be the handset until my fascia is mounted to and then I will add the push button control.

Thanks,

-Ed

You could also...

...use a set of contacts on one machine to power the second. That way there is no possibility of overloading the MK 8.

The Old DIODE MATRIX - Graeme's solution replaces it!

[disclaimer:  I am about to make the DCC conversion, and have not yet made these potential mistakes]

As I understand your query, the machines when driven simultaneously can overload your circuit.  As such, Graeme seems to point to a work-around.   One that I find particularly interesting.

Consider the old DC setup with a diode matrix route-selector.  One switch would drive several machines.  Graeme seems to have replaced this  once simple setup with an elegant and almost as simple solution.  With the circuit arrangement Graeme supplied, one could have a set of machines operate in a cascade, almost like handing off a baton at a relay.   And potentially, one could add a minor delay in the chain by including a small capacitor... - on the wire between tortoise 1 and tortoise 2... etc ...but that would be only for the truly daring.

Working!

Thanks to the reassurances in terms of current ratings and the pointer to the instructions, I now have my paired crossover turnouts working off a single accessory address as well as two single tortoise switch machines.  The Switch8 Markj II works great and the button board is attached but not wired up to push buttons yet.  But I love the Switch8 and the tortoise switch machines.  This was my first time ever on a ayout using switch machines and I am 100% thrilled with them.  Due to cost I won't be using them in yards, but for all my mainline turnouts they will get used.

Thanks for the pointers everyone!

-Ed