Replacing an 8 ohm driver with two 4ohm drivers

Jump to Latest
R

Richard_James

Member
Joined 2017
#1
What would happen if you replace an 8ohm midrange driver in a 2 way speaker system with two 4ohm in series. The thiel and small parameters are very similar. Except efficiency of the original 8ohm driver is 89db and 4ohm drivers 88 db or so. Driver size is 6.5".

Drivers are used in closed box .
 
#2
Combined nominal impedance of two 4-ohm drivers in series is the same as the nominal impedance of a single 8-ohm driver. But here similarities stop. It is the same as replacing one 8-ohm driver with other, totally different 8-ohm driver - it doesn't matter whether their TS parameters are similar or not. Frequency response certainly will be different, together with the impedance (although both are nominal 8-ohm), so the crossover will output very different frequency responses. You have to modify the crossover.
Edit: Now I realize that it is 2-way loudspeaker, so it is not a midrange driver as you wrote, but midbass driver! Because TS parameters are similar, you need double box volume! So, it is much better to use one 4-ohm midbass driver and modify the crossover.
 
Last edited:
#4
I see. I suspect that there will be less power used by each 4ohm driver but together they will output the same SPL as one 8ohm driver. The two 4ohm drivers will move less than one 8ohm driver. Less movement means less distortion..

Is this the right way to think about it?
 
#5
Richard_James said:
What would happen if you replace an 8ohm midrange driver in a 2 way speaker system with two 4ohm in series. The thiel and small parameters are very similar. Except efficiency of the original 8ohm driver is 89db and 4ohm drivers 88 db or so. Driver size is 6.5".

Drivers are used in closed box .
Click to expand...
I think the main problem will be the doubled required box volume that Sonce wrote about.

Assuming that 88 dB is 88 dB at 1 m at 1 W into the nominal impedance, so at 2 V RMS rather than 2.828 V RMS, the sensitivity will be 2 dB too high rather than 1 dB too low. At 2.828 V, each loudspeaker gets 1.414 V, each produces 85 dB at 1 m, and on axis the signals of the two loudspeakers add in phase, so together they produce 91 dB at 1 m.
 
#6
norman bates said:
I've wondered this 2, like replacing 1 of the 8 ohm rs180p 8 with 2 of the rs180p 4 (wired in series) in the cat raven design.
https://ampslab.com/blog/2018/02/28/raven-cat-1/Yea the qts fs is different, yea the dispersion would be different, but i think it could work in this case.
Opinions ?
Thoughts ?
Click to expand...

Richard_James said:
I see. I suspect that there will be less power used by each 4ohm driver but together they will output the same SPL as one 8ohm driver. The two 4ohm drivers will move less than one 8ohm driver. Less movement means less distortion..
Is this the right way to think about it?
Click to expand...
You need to double the box volume, and possibly to modify the crossover - I am surprised frequency responses are not identical, although both versions use the same cone. Also, 4-ohm version is not exactly half of the impedance of the 8-ohm version. Nominal impedance is just that - nominal, not exact.
Yes, two drivers means less distortion when pushed hard, but do you really need that? After all, RS180 has the same active cone area as conventional 6.5" driver - much better than usual 5.25" or 4" drivers.
 
#7
Richard_James said:
What would happen if you replace an 8ohm midrange driver in a 2 way speaker system with two 4ohm in series.
Click to expand...
In addition to what Sonce already said, there is another problem. Moving from a single driver midrange to a dual driver midrange means you need to design an MTM, and an MTM with a 6.5" driver is is not a good solution as the distance between the dual mids (with the tweeter in between) is generally too large for a normal crossover point.

Ralf
 
#8
It is not necessary to design MTM. MMT is another design. Different polar response. One can play tricks with the new small tweeters slightly offset.
Just looking at efficiency does not tell you much as there is additional gain when in close proximity.
Anyway, as any idea. you need to mount them in a prototype and measure the response and impedance in-box. Then design the crossover to match. If you are not into designing a crossover than leave it be.

The RS drivers can be quite nice. Used a pile of them, but very hard to deal with their breakup. Think low steep crossovers which then get very difficult to get the phase right over wide enough a range. I have gone back to paper. Seas and SB. CSS and Purify have my eye. Paper really is a marvelous engineering material. If it does not sound exotic enough, call it "critically doped cellulous fiber composite".

Less movement means less distortion, but only as you approach Fs. If you have measured many drivers, you will see they stay pretty well behaved up to 1/2 X-max at about twice Fs. Below there, it skyrockets. A second driver only gives you a little over 3 dB, so don't think it will magically play louder or cleaner. Look where your sub is coming in and see that is a problem.
 
#14
Here's a link to the page that originally explained the method, it's a bit hard to find (& remember the title!):
http://community.fortunecity.ws/rivendell/xentar/1179/theory/dddllqd/dddllqd.html
The guts of it is here: (& yes, it does work , I've used it several times, and they don't need to be 'Low Q' drivers)
the "Compound Driver" approach
... modelling the two drivers as a new single driver with combined parameters would give at the VERY LEAST some reasonable indication of the possible behaviour of the combo...
Adding the Equivalence Volume (Vas) of both Driver and adding the Cone Surfaces (Sd) is part of mathematically assembling the new "compound" Driver. The new Resonance Frequency, Qes/Qms/Qts becomes the geometric mean of the two drivers values, so basically square the actual two values, add them up, divide by two and than take the square root. The Voice Coil DCR (Re) simply becomes the result of parallel resistors. (or series if that’s what you choose to do)

The original author is Thorsten Loesch
 
Last edited:
#15
tried winisd.
how does it know if the speakers used are connected in parallel or in series
here are the parameters of the woofers. i had difficult time finding parameters of the old polk mw6503 8ohm driver. looking to build new boxes and put two dynaudio 17w75xl 4ohm in series instead of one polk mw6503 8ohm. dc resistance of one 6503 and two `17w75xl 4 ohms turnes out to be identical. around 6.4ohm
17w75xl mw6503
qts .4 .31
vas 20L 85L
fs 40hz 29hz
re 3.0 ohm 6.54ohm
le .17 na
xmax .006m 3.180m
z 4 ohm 8 ohm
qms 1.9 1.77
qes .5 .38
spl 0 0
pe 130w 35w
bl 4.8 5.24
dia .177m .175(this is my aproximation of diameter)
sd .012 m2 .013m2


i attached a graph of 2 17w75xl drivers in 13L box. but not sure how winisd modeled it.. in series or parallel.
the other graph is of one polk mw 6503 driver in 13L box

is there any difference in these parameters to warrant a major dramatic change required in cross overs?

does anyone see major issues. i can build new boxes to accomodate different drivers
Screenshot (3).png
 
#16
re:'any difference in these parameters to warrant a major dramatic change required in cross overs?" - the crossover is designed around the impedance plot of the combined drivers, not the parameters

Re:'how does it know if the speakers used are connected in parallel or in series" you input the COMBINED parameters, as described, and model them as a single 'compound' driver
 
#17
I see best plot for dynaudio when using in combined mode is 3L in closed box. Strangely small volume

Is impedance plot something easily obtainable? I rarely see impedance plots for drivers. 17w75xl seems like a popular driver. Should have the plot.
Screenshot (5).png
 
Last edited:
#18
Impedance curve is right on the datasheet for that driver already, but they've measured it in a ported box rather than free air or infinite baffle, so you might not be able to use it as-is for certain crossover design applications. For example, if it were going to be the LF in a 3-way, you might be crossing over near to where the box-induced changes in impedance are relevant. For most other applications, it should be fine as-is.
WinISD models the combined acoustic response of the two drivers as if they were identical, just with the higher SD & VAS as already discussed by others in this thread, but it always displays the impedance response of a single driver.
The best way to work with it in terms of series/parallel wiring IMO is to always use the "driver input voltage" field rather than "system input power", then you always know what's happening to each driver, and you just need to remember to double it if you're planning on wiring in series.
 
You must log in or register to reply here.
Top Bottom