8 Ohm & 4 Ohm treble units

[Fred Shaw]

One of my Bandor 50mm treble units has developed an intermittent problem - it cuts out from time to time! The fault lies somewhere in the flexible wire connecting the voice coil to the input wire connector. Bandor units are hard to obtain but, fortunately, I have managed to source a pair. Unfortunately, the replacement units are 4 ohm and my units are 8 ohm. I am using these bi-wired speakers with a solid state amp so the amp should be able to cope with 8 ohm base units and 4 ohm treble. Is this discrepancy in resistance likely to make any audible difference? If so, could I put a 4 ohm resistor in series with the 4 ohm speaker to bring the total resistance up to 8 ohms? It could be that I am way off here as the DC resistance of the 4 ohm speakers is 3.2 ohms and the 8 ohm units 6.7 ohms, and I don't know how to measure/calculate impedence. As you can see I am slightly out of my depth here so any help will be appreciated.

Fred Shaw.

[Little Mick]

Hello

Excepting a better explanation the impedance you refer to is not based at DC rather AC i.e. as compared to an actual signal applied to it and is a nominal figure as obviously your tweeter operates across a bandwidth, as far as matching is concerned I would just try it and see what imbalance you get, as they are a pair it would only effect how the speakers are voiced, to note the voice coil on the higher impedance one will be lighter it could effect its power handling.

Stuart

[Richard]

Hi Fred, yes I'd been wondering about this and agree with Stuart that there should be little risk in trying the new units.

If they are louder I'd add resistance before the crossover elements in the treble section, a simple inline resistor, try circa 1R or 2R wire-wounds from Maplin, it shouldn't need a lot to make an audible change and if the results are promising you can then adjust to taste.

Here's a drawing of a manufacturer's bi-wired crossover where the treble level is adjusted with the 3R3; you'd fit a resistor at that point. You could even fit them temporarily outside the cabinet, simply inline between treble speaker wire plug and the speaker terminal.

[Fred Shaw]

Many thanks Stuart and Richard. I have installed the speakers in the cabinet and they seem to work fine. The treble does sounds a little forward though, so I will get some low value resistors and see what difference they make. I'm afraid any modifications will have to wait for a couple of weeks due to domestic matters, but I will report back with my findings in due course.

Fred.

[Richard]

Hi Fred,

I've been thinking about this a bit more. You won't do any harm running the 4 ohm units in place of the 8 ohm, and adjusting level before the xover if you can get a satisfactory result, but the filter point for the treble will be affected by the lower impedance, moving it up an octave all other things being equal. Putting a 4 ohm resistor in series after the crossover would restore the filter point but lower the Bandor output by 6dB.

Looking at the specs of your units here,
http://www.bandor.co.uk/bandor50r.htm

we see the 4 ohm and 8 ohm units are almost identical sensitivity at 85dB. Their impedance is similarly flat for a good length of the midrange and is double for the 8 ohm compared with the 4 ohm (compare graphs). Given that the speakers were good before, this should make for a simple change of the xover without needing to know much more about it.

The Bandors are very wide band with a low resonance around 80Hz. That means they might be crossed over a couple of octaves above that around 400Hz with a simple xover. If you have the xover or can tell us the bass units it would be interesting but if they are just 2 way speakers with the Bandors handling mid and treble it may not be needed.

Trace the treble section of the xover. It may be first order with just a series capacitor feeding it, or second order with a series cap followed by an inductor shunted to ground.

If so all you need do is double the cap value and (if second order) halve the inductor value. The bass section need not be touched. This will restore the xover point for the 4 ohm unit and the level will not be attenuated.

The cap will hopefully have the value on it and if there's an inductor measure it or just send them both to Falcon and ask for twice the cap and half the inductance.

[Fred Shaw]

Richard, many thanks for your suggestions for the Bandor speakers. The bass speakers are Bandor 150mm, 8 ohm units and the crossover is the one supplied by Doreen Bance with a crossover point set at 150Hz. I have a copy of the schematic for the crossover (4th order treble & 3rd order bass) but I can't seem to post it to the board. Despite the difference in treble speaker and bass speaker impedance, the speakers (to my restricted hearing) do sound pretty good, and, as speakers in my second system, are more than acceptable. However, it would be interesting to hear them with a crossover using the correct specification components. I will email you the crossover design Richard so you can see the layout.

Fred.

[Richard]

Hi Fred,

Yes got it fine thanks, it's a bit too large file size for the bb. It looks like a 3rd order xover for both drivers. The treble has 4 elements but the small cap makes an AB (Acoustic Butterworth) network which levels the response a little just after crossover.

I ran it through XSim with flat 8 ohm drivers and it shows a 6dB hump 200 - 300Hz which may not be surprising if it's been optimised for a box system with real drivers' increasing impedance near resonance.

I changed the tweeter setting to 4 ohms and sure enough there is now a 5dB dip in the treble after xover. I doubled both treble series caps and halved the inductor and the original response returned.

So without going further into it (which we cannot without lots more info) to get back to the old response double your numbers C2 and C3 to 200uF, and halve L3 to 2.5mH. The 2.2uF AB cap makes very little difference so leave it as it is.

If you are into these programmes or for anyone else following,
XSim is wonderful, http://libinst.com/Xsim/XSimSetup.exe

This calculator will give you values for flat drivers which agree nicely in XSim, http://www.diyaudioandvideo.com/Calc...akerCrossover/

FP Graph Tracer http://labs.fprawn.com/index.php/fpgraphtracer/ will convert the graph response images into .zma and .frd files if you want to get a bit closer using your actual drivers in XSim. (Remember to set the graph boundary lines, enter the xy limits, and tick log scale or not.) I found it necessary to draw over the response line heavier in photoshop first and then ran XSim with the driver files. The response flattened quite a bit and looked good but we don't know how the original graphs were made (baffle or box) or the boxes your drivers are in etc.

The bottom line again seems to be that if they were good before then doubling the series caps and halving the inductor will restore the response for the 4 ohm tweeters

Bandors with 4 ohm treble unit (C and L numbers are assigned by the programme and are different to Fred's drawing mentioned above),