Jim Lesurf wrote in message ...
In article , steven
robinson wrote:
Jim Lesurf wrote in message
...

[snip loads]


Do you have impedance/frequency plots of the individual units? Any
idea how these values were altered by the cabinets and/or combined
useage? What amplifier did you use, and is it unconditionally stable?


Hi Jim Thanks for the reply, a lot of what you are saying is foreign to
me, but this is what I know: The spec of the 8Ohms 50watt speakers in
the parallel leg a Nominal power 30watts Peak power 50watts SPL @
1W/1m 82db Frequency range 100Hz to 20kHz Resonant Frequency 130Hz Max
cone displacement 5mm Voice coil diameter 20mm Magnetic inductance 0.8
Tesla Magnetic flux 200u Weber

Alas, the above does not tell you what you'd need to know to be able to
start the process of trying to design an 'in series' design. TBH such
designs, using dissimilar speakers, would be a nightmare to get right
without a lot of test gear, etc. Best avoided in my view as trying this (as
you have found) can be more grief that you'd expect.

Withount using cross overs how can I block damaging frequencies lower
than 100Hz and higher than 20kHz from the drivers? or isn't that a
concern?

I can't tell what may or may not be the real cause of the problem you had.
The problem is that there is too much we don't know that might have an
effect. However the resonance frequency (which I assume is 'free air') of
130 Hz seems quite high for a 'full range' speaker. I may well be wrong as
I lack data. Someone who knows more about speaker building may well be able
to give a better comment on this. However I'd suspect that below 100 Hz the
cone movements might become too great unless you use the unit in a cabinet
that acts as a closed box to limit this.

Otherwise, to control movements, you will need a cross-over - although this
could be an 'active' type applied before using distinct power-amps to drive
the individual units. This may all be to complicated for you to wish to do,
though.

For the above reasons it is possible that the speaker was damaged by being
forced to move too far to provide LF output. i.e. by exceeding its
mechanical limits, not by overheating by excessive power.

It is unlikely that signals above 20kHz will do any damage. Chances are
that there is only a little power at such frequencies, and the units will
essentially ignore it. :-) (This assumption may not be correct if the
amplifier went into oscillations and was producing very high powers at an
ultrasonic frequency. However normal music and speech does not usually have
much in the way of sustained high powers above 20kHz.)

Unfortunately I don't know anything about the 4Ohms speaker, except its
Peak power is 80watts.

Likewise the amp is a bit of a mistery, it is an amp called a v-series
manufactured by an amusement company in leeds. Its output is supposed to
be 70watts into 4Ohms.

Unless you can establish that it is an unconditionally stable design, then
it is possible that it went into oscillation and this caused problems. This
is probably less likely than other causes, but I can't rule it out at this
point.

I'm getting the impresion that putting dis-similat drivers in series is
a bad idea.

It is, unless you know a great deal about the units, and can carry out the
relevant design/measurement process. Afraid it isn't as simple as assuming
the power is divided up as if the units were plain resistors. In general,
'hi fi' designs avoid speakers in series without any cross-over or
adjustment networks.

What if I put two 8Ohms in series twice and parallel them
together, would this be a better idea? I could then use the same
speakers with the same properties.

This is more likely to work as the inherent properties of the units are the
same. However it may still not work as you expect for various reasons. For
example: Putting the units into the same cabinet mean they couple
acoustically. This alters their properties, so even though you have
connected them in 'series/parallel' the resulting impedances, etc may not
be the same as for a single driver in the same cabinet.

In this configuration you would expect the output to be spread evenly
across the four speakers, wouldn't you? What do you think?

The output would probably be distributed reasonably evenly, but it is hard
to be certain about this as the cabinet has an effect. It would be better
than using dissimilar units, but it is not something I'd normally recommend
as the results may still not be what you anticipate. For example: there
will be acoustic interference between the units that will alter the
resulting total response, etc.

To be honest, trying to do anything like this is risky without having a
fair amount of data, and/or doing the right measurements, or proceeding
with care and taking a long time experimenting.

FWIW I've happily designed and build various amplifiers, and tuners, and
other 'electronic' items over the years. I also cheerfully modify
commercial electronics to alter the performance to suit. However I have
always avoided trying to self-design speakers. Simply too difficult and
time-consuming to get right. :-) That said, many people do develop their
own designs, and produce good results. However this can involve a lot of
effort and patience and understanding. Some of the 'kits' and published
designs can work well if made with care, but these are obviously based upon
someone else having already put in the time and effort to sort out the
necessary details!

Slainte,

Jim


Hi Jim
I seem to be going round in circles with this one. I am starting to
understand the phrase "lot of effort and patience and understanding"
you quoted.

I am considerring having to use a crossover in my design, do you think
this would solve some of my power distribytion issues? I would still
use the two paralleled 50W 8Ohms for mid to high fq and the 80W 4Ohms
for low fq, providing the frequency range of the speakers are
suitable. I'm either going this way or the two paralled/series setup
with the same drivers, but without crossovers, what do you think?

Thanks and regards Steve