Message-ID: <5206ABF4.1020800@penguindevelopment.org>
Date: Sat, 10 Aug 2013 23:09:08 +0200
From: Peter <link AT penguindevelopment DOT org>
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:17.0) Gecko/20130707 Thunderbird/17.0.7
MIME-Version: 1.0
To: geda-user AT delorie DOT com
Subject: Re: [geda-user] T flip-flop
References: <201308101615 DOT 39595 DOT ad252 AT freeelectron DOT net>
In-Reply-To: <201308101615.39595.ad252@freeelectron.net>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Reply-To: geda-user AT delorie DOT com

On 10/08/13 22:15, al davis wrote:
> I have a need for a "T" flip-flop ......   to map a momentary
> pushbutton to on or off.
>
> The application needs 20 of them in a tight space, preferably
> arranged in 4 groups of 5.
>
> Looking for suggestions of a chip to do this, with a minimum
> package count and space requirement.
>
> Wishing for a "hex T flip-flop", but I can't find it.
>
> The best I can find seems to be a dual JK flipflop.  That's 10
> packages.
>
> I found a quad JK flipflop, but they share a common clock so
> that doesn't work for this application.
>
> I can do better in cost and space with discretes.
>
> ideas???
>
What are the timing requirements for this? Because if you can afford a 
few microseconds of delay, I'd imagine the best option in terms of space 
requirement would be to simulate a set of flip-flops using a tiny 
microcontroller. For example, an ATtiny20 could act as 4 independent T 
flip-flops while running from the internal 8 MHz oscillator (the 
necessary code would be pretty simple), and has tiny package options 
available (VFQFN, UFBGA...). I doubt you're going to get 4 independent 
*real* T flip-flops in that space for that cost.

-Peter