Testing TS with driver vertical

[mashaffer]

In measuring TS parameters does it matter whether the driver is vertical or horizontal? The reason I ask is that it is not convenient for me to test vertically unless I just suspend it from the ceiling via some string and I presume that letting it hang free like that would be worse.

 

[djarchow]

No matter how you measure, the driver should be rigidly mounted.

If I am doing free air measurements, I use a large wooden clamp and clamp the driver magnet to the table edge with the drver oriented vertically. Also if the driver has a vented pole peice, you need to be sure and keep the vent well away from any obstructions.

Regards,

Dennis

 

[richie00boy]

I usually test horizontally and it's not really a problem unless the driver has very weak suspension.

Good tip on clamping the magnet to the edge of a table. I might add that to my measuring guide

 

[planet10]

If the driver is facing up or down, the weight of the cone will throw off your measurements. The bigger the driver (n general) the less accurate your results will be.

Ideally the driver is clamped rigidly and in a "facing forward" position, free from any close objects and in a perfectly quiet room (if the background tunes are turned up too high you can see that in your measures). In parctise -- with the small drivers i am most often measuring -- holding the driver with my elbow resting on something solid (my chairs arm rest) is sufficient for me to get repeatable results.

John Janowitz (Stryke) somewhere here posted in some detail on this subject.

dave

 

[Dr. Ram]

Keep in mind that if you are worried about errors in measuring the T/S parameters, what really matters is not the error in the parameter per se but what is the resulting error in your enclosure design that uses these parameters. For example, you may encounter Vas errors of 10% or more but 10% is not as terrifying if the resulting SPL errors for the enclosure are 0.5 dB (as is often the case).

 

[forr]

For some drivers, I found about 3% differences (fs and Qts) between vertical and horizontal moutings.

 

[mashaffer]

Thank you all for your comments. Ramkumarr, that point is something about which I was wondering. In simulations that I have run it seems that one can often vary parameters a fair amount without much change in performance.

When you think about it even individual samples of the same driver vary somewhat yet most people don't realign each cabinet for the individual driver. In my case most of these drivers will be used either as Open Baffle, Sealed or Tapped Horn.

Suppose that what I should do is model which the measured parameters and then run a simulation with each parameter varied by 10% each way and see what the tolerance for variation amounts to.

I suspect that sealed and OB will both be very forgiving. The TH however may be more sensitive.

 

[planet10]

When you think about it even individual samples of the same driver vary somewhat

20% variation is considered normal.

dave

 

[LineArray]

Hi,
i experienced that the driver should be
broken in for a while, to stabilize the drivers
parameters.

Vas lowers dependent on the suspension materials
used in the driver.

I had a driver with a plastic surround, which contributed to
stiffness of suspension. The driver lowered its Fs
about 15% as far as i remembered in the first half hour ...

It is a difference with many drivers at which excursion
you measure.

I tend to measure somewhat below the limit of linear
excursion.

 

[mashaffer]

All of the drivers I have been working with are used (most of them very used).

One thing that I should mention is that I am testing at very low volume levels under the theory that these are small signal parameters. Maybe I am taking this too far though. Usually the peak voltage across the driver itself (at fs) is about 1V or slightly below. Should I be using more signal?

 

[claudio]

When using the added mass method to measure the VAS, I prefer to place the driver horizontal.

About the voltage to feed the DUT, assuming you are using a sinewave signal,
0.5 Vrms are considered small signal.

 

[LineArray]

Should I be using more signal?

Hi,

i would say it depends on how we want to use the
specific driver.

If it is used as a horn driver in a home installation driven
at moderate levels, the small signal parameters are
more important.

For low efficiency application in normal boxes or open
baffles the large signal parameters are important IMO.

Kind regards

 

[Dr. Ram]

If you want to measure small signal parameters (e.g. T/S parameters) then the smallest signal that gives you consistent and repeatable measurements is the signal to use. That usually means less than - typically much less than - 0.5W. That the parameters may have been measured at 50mW does not affect that fact that they will give you good modeling results even at 1W, but perhaps not at 10W. That specific upper limit depends on the driver.

-Ram.

 

[John_E_Janowitz]

Here's a post I put up on our forum some time ago on properly measuring T/S parameters and all the efforts we went through to do it.

https://www.aespeakers.com/phpbb2/viewtopic.php?f=4&t=927

John

 

[richie00boy]

Lots of good info in that post. I do have one question though. If the driver is facing upwards, yes on the updard stroke it will be working against the weight of the cone and gravity, but this will be cancelled out exactly by it being assisted by gravity and the weight of the cone on the downward stroke, will it not?