bill@uunet.UU.NET (Bill Vermillion) (12/06/90)
Bias is basically used to linearize the response of the tape.
To adjust it you should either know the mfrs specs on the tape, or use a
rule of thumb type bias.
Two heads make it harder. Let me just explain the procedure on a
three-head to understand it, and then back that procedure two a two-head
machine.
Now I don't know the specs for cassettes, but let me give you 15ips reel to
reel. For 7.5 ips you can halve the frequency. Put the machine in
record, put a 10kHz tone, and adjust the bias from minimum. As you
increase the bias you will see the output increase. Continue increasing
bias until the output peaks. Continue increasing bias until the output is
3db less than the what is was at the peak output.
This works well for most pro tapes, eg Scotch, Ampex, Agfa.
Another way that is typically done is use a 1kHz tone. Do the same
procedure as above but only overbias by 1db.
To explain further, one of the specifications you see on tape is a
"hysterisys loop". This shows the output of the tape as relating to the
input signal. At the first application of the signal you get almost no
output. As you increase the output starts to increase also. Plotting this
you will find an area where the output increases linearly with the input.
Then the output increase will taper off and finally no ouput increase will
be seen for any input increase. THis is the saturation point.
Now when you have a tape that is magnetized and you apply the opposite
phase of signal to it, the output will not start falling until you have
gone negative for quite a way on the graph.
If you apply a bias signal of an amplitude sufficient to be at the point
where the original signal started to become linear, then you apply your
signal to the top of that, you have moved the signal into the linear
portion of the curve.
If you are familiar with film curves they are similar.
Now the output of a tape head fall about 6db/octave on either side of a
maximum output. THis peak is usually somewhere between 500Hz and 2000Hz
depending on design but is usually in the 70-Hz to 1000Hz area. To make
the response flat we boost the record response by increasing the high end
record signal in reverse of the output fall-off of an unequalized head.
The boost is pretty severe. I forget the exact specs but it is in the
neighborhood of over 10db at 10kHz. We boost the lowend on playback.
Because of the non-linearity of the head and the tape this is why an
underbiased tape will have an exaggerated top end.
Increasing bias lowers the distortion. An overbiased tape will have
diminished high-frequency response but better distotion specs (to a point).
Now, if you have a two-head machine, erase and rec/pb, the job gets a bit
more difficult. You will have to time (or mark) you tape. You also need
some way to measure the bias current. If it is not meterable, then there
is probably a testpoint in the circuit somewhere.
You start recording. Make a note of the current. Keep recording, wait a
bit, increase the bias, and note it (and the tape-counter or time), do this
for several instances.
Then you play back the tape and watch the output. If you are doing the 3db
overbias at 10kHz, not the time of the tape where this occured, look up
your notes on the bias level, and then using the meter reset the bias to
that point.
I haven't had to align a tape machine like that since the late 60's using
broadcast cart machines.
One other trick, which is my favorite way to align a reel-to-reel machine
is to use the minimum modulation noise method, which is a must if you use
something like dbx noise reduction.
To use this method requires only a tone generator and your ears. (Won't
work on a 2-head machine - at least not easily).
You set the tone generator at a suitably low frequency. You want something
much lower than your system can record or playback. We found at BeeJay
that a good frequency was 7Hz. (That's right SEVEN cycles/second).
With bias at minimum, and the tape in record with the 7Hz tone, you listen
to the noise output. (You may have to high-pass filter if you get the pure
7Hz through your machine). You will hear noise pulsed at a 7Hz frequency.
If you go too high you can't hear the pulsing and too low you may not get
enough signal to the tape to cause modulation noise.
As you increase the bias the character of the noise will change. But there
is a point between the different sounding noise images when the noise
output is at it's lowest. This is the minimum modulation noise point. It
is very close to the 3db over point for tapes such as Ampex 456. We found
that if you try this with Agfa you require so much bias that we overloaded
the machine and it shutdown. This make for a clean sounding tape though.
Many many months ago someone asked me about bias, and I refered them to the
Camras book. They said they were disappointed by my answer and had hoped
for a better explanatio than pointing to some expert.
I will try to go back and brush up, and post an article (though it may take
two) on bias and how it is done. (Don't hold me to that, though I will
try).
bill
wa
recording
--
Bill Vermillion - UUCP: uunet!tarpit!bilver!bill
: bill@bilver.UUCP