> I'm trying to tune my A-110s to very precise frequencies and I'm

assuming I'd need some

> (probably digital ) device to be able to get one at say 587 Hz and

the other at 845Hz. What

> devices do people use to get exact tunings with analog oscillators
>
> I'm sure there is something very simple that performs this

function, I just don't know what

> it's called or where to find it

A 'frequency counter' or a 'universal counter timer' *might* do the
job for you, but you need to be careful - they are frequently
specified to operate over a large range, say 5Hz to 5MHz, so getting
an accurate reading (i.e. less than a hertz) at low (audio)
frequencies can be quite a challenge! I have a second-hand (and
rarely used) Racal-Dana 9902: don't know where you are based, but
there are loads out there, see for example:

http://www.telford-electronics.com/the_graveyard.htm

- these are the people in the UK where I got mine from.

Most have two modes of operation: for higher frequencies they count
the number of cycles of the applied signal over a known (fairly
small) time interval - for low frequencies this will not be very
many, hence the inaccuracy; so for lower frequencies it works the
other way round - it counts the number of cycles of an internal clock
during one period of the applied signal, and then displays the
_period_ of the signal, and hence you have to invert to get the
frequency (fiddly, but you do get a more accurate measurement of low
frequencies this way).

Another alternative would be to try and use a 'chromatic tuner': I
have a Seiko ST-747, which has a range from A0 (27.5Hz) to C8
(4186Hz), and which displays the deviation +/- (up to) 50 cents from
the nearest semitone. Thus again, in theory, with a little maths you
can work out the exact frequencies. Claimed accuracy is +/-1 cent.

With the counter reading a frequency of 848.2Hz, a period measurement
gave 1179us = 848.18Hz, the tuner was saying G#+35 cents =
(nominally) 830.61 * 2^(35/1200) = 847.57Hz (and I say 'nominally',
as this figure is from a book, and not what the tuner might be
calibrated to!).

My conclusion: just about do-able, but if you want readings around
the 'tenth of a hertz' region, you will probably have to work at the
maths a little bit!

Tim