mik

Recent house move caused two problems ~

1. when I disconnected my speakers from the amp I forget whether they were bi-wired into the same output, or separate outputs.

2. My impedence switch on the amp got knocked during the move - not sure where it should sit.

The manual for the amp is long gone (NAD 3240PE amp with Tannoy 609's if it matters....circa 12yrs old....blimey! )

I remember there was a distinct warning regarding bi-wiring but can't remember the details (oops).

Should I have high/low lines connected into separate outputs (and run amp switched to "A+B" speaker outputs?).

Should impedence be set to 4ohms as opposed to 8?

Much obliged,
mik

dr_ming

1 - Same o/p. Otherwise the A/B speaker switch will give odd results! A purist would probably say use different outputs, but in reality, the extra 3 inches of wire back to the A/B switch makes little difference.

2 - If the speakers are 8 ohms set it to 8 ohms (similarly 4 ohms > 4 ohms). Bi-wiring makes no difference to the speaker impedance.

Hope this helps.

ozzy

... and if you've no idea what the speakers are rated at, then you can contact Tannoy in Scotland.

They should be able to find details of that model of speaker if you can't find anything on the net.

Tannoy Ltd
Rosehall Industrial Estate
Coatbridge
North Lanarkshire
ML5 4TF
Tel: 01236 420199
Email: elaine.gardner@tannoy.co.uk

Stefan

mik

Thanks folks.

Ozzy - I'm in Glasgow, so have already been to the Tannoy plant (10 yrs ago?) ~ to get voice coils on woofers replaced after approx 6months 'cos I overdrove them

609's are 8ohms in linked mode, and since they are bi-wire-able I'm certain they're matched 4 & 4 for woofer & tweeter.

I've currently got outputs on different amp channels, so to listen you need to select "A+B". I've therefore set the amp to 4ohms, but not sure if this is right

Not driving them with any real volume at the moment for amp safety.

Maybe I should just contact NAD?

ozzy

Yeah, that's worth a try.

Maybe one of the AV shops in Glasgow would help you out. Sound & Vision are pretty good.

I did some work in Tannoy myself about 4 years ago (I'm from Lanarkshire). Small world eh

Stefan

hades

It's not as simple as a 4 Ohms plus 4 Ohms = 8 Ohms in this case. Simplistically, the woofers (and their share of the crossover) will provide 8 Ohms below the crossover frequency (probably a couple of Khz or so), and much higher impedance above that frequency. The tweaters (and their share of the corssover) will provide 8 Ohms above the crossover frequency, and much higher impedance below that frequency. Therefore, collecting them in parallel gives a flat 8 Ohms across the board, whilst connecting just one or the other gives 8 Ohms over a lesser frequency range. The crossover will be designed to make sure that the impedance stays roughly even across the range 20Hz to 20Khz (that's very roughly in practice).

IIRC, from owning a NAD 3020i many years ago, the 8 Ohm setting is fine for the range 4-8 Ohms, the 4 Ohm setting is for 4 Ohms and under. Therefore, I would recommend setting the switch at 8 Ohms if your speakers are indeed 8 Ohm rated.

Dr_ming: I don't know of purists who say they should be bi-wired to different outputs. A purist would probably complain that you're introducing an unnecessary switch into the circuit

Mike P

You sometimes get a warning for the speakers in a bi-wire/bi-amp set up.

My kefs had a bridging plate between the lo and high frequency input terminals which has to be removed if bi-wiring/amping. I guess if you don't you can damage the crossover in the speaker.

hades

Shouldn't damage the crossover in any way, but if you're bi-amping and leave the plate in, the amps may tend to fight each other, in an expensive manner. Also, if you don't take the plate out when biwiring, you've just lost the benefit and so wasted half the money you spent on speaker cable

mik

Don't worry - I was smart enough to remove the joining plate

Contacted NAD ~ although they have several "classic" manuals on-line, the 3240PE is too old.

they are however sending me a manual for it by post

Top service NAD!

dr_ming

Hades,

True. But then 'purists' will always find:

a) something to moan about.

and

b)something to spend incomprehensible amounts of cash on

Turbo_Steve

And the benefits of bi-wiring with the same type of conductor are almost microscopic anyway.

Bi_amping has real benefits. Bi-Wiring is "just running thicker cable".

I've yet to find a single "audiophile" who can hear the difference.

NOTE: Biwiring using DIFFERENT types of cable is a known and proven technique, and I am not knocking it. But buy "4 conductors the same" is just a good way to sell more cable.

hades

Steve. I disagree, and I could even give a scientific reason why biwiring with equivalent conductors can have benefits. I have found audiophiles who can tell the difference.

I do agree that the effects are much smaller than bi-amping, but the principle is pretty much the seem.

chiark

I don't class myself as a serious audiophile, and biwiring make a huuuuuge difference. The scientist in me can't understand it either, but it ain't just placebo...

Turbo_Steve

Big 'ole world.

If YOU can hear the difference then that's what counts.
But I can't see a single principal advantage in bi-wiring: You still have to move the same number of electrons: you're just lowering the resistance.

Apparently the concept originally came about because higher frequencies experience less loss in solid copper, and lower frequencies in multi-core, so a carefully matched set of differing conductors would provide a flatter and faster response over longer cable runs.

chiark

that's what pee's me off - in my head, I think that there shouldn't be a difference... But my ears do tell me otherwise fairly reliably.

To be honest, I thought the whole cable makes a big difference was a huge con, until I had a good test in a very helpful hi-fi place. There can be a massive difference.

I'd say biwiring isn't as much difference as cable can make, but it's noticable.

Turbo_Steve

Chiark....as an experiment, could you try keeping the bars in place on the back of the speaker, and just running it as a "thicker wire" with less resistance instead of actually "bi-wiring" (if you see what I mean) and see if there is still a difference?

I HAVE noticed that the lower resistance of more conductors (even if they are commoned) results in a better sound.

It's all subjective, I know.

hades

IMHO, it all boils down to fourier (sp?) tranforms - imagine splitting the signal into a big selection of sine waves at different amplitudes and frequencies. Scientific concept, but it actually models how things work in the real world.

You flow big currents at low frequencies, resulting in fairly large magnitude magnetic fields around the centre of the current flow (i.e. in the wire). Try and pass another electric signal (sine wave at different frequency) through this field, it too will have a magnetic field, and the two will interact - result is distortion of the purity of the signal. Incidentally, at high frequencies, high rate of change of smaller current also results in significant magnetic flux. The magnetic field density will be much higher around the centre of the current flow than even a few mm away.

By biwiring, you are flowing much reduced current at high frequencies down the wooder wire, and much reduced current at low frequencies down the tweeter wire. Therefore you are reducing the interaction between the different frequency components of the signals and hence the distortion. That's my theory, anyway, and I'm sticking to it

I have applied similar principle to the construction of a few bits of hi-fi, I believe with success - the highlight being demonstrating to a well respected hi-fi dealer that my home made pre-amp sounded nicer than his £2.5k favourite!

hades

Btw steve, it's the other way around - higher frequencies experience lower losses in multi core cable. All down to electro-magnetic fields again, and is call "skin effect". At higher frequncies, only a thin layer at the surface of a conductor actually passes current. Lots of thin wires can therefore each contribute to the conductance at high frequencies, whereas most of a thick (single core) wire can't.

Sorry to get scientific, and not meaning to have a pop. I must have been sober the week they were teaching us that bit of electro-magnetism at uni!

Turbo_Steve

Sorry...concur 100% with regards to the conductor types. Muppet hat now removed!

However, as one end of the bi-wired conductors is commoned, surely any distortion introduced into the waveform would be present across the entire run of that conductor?
Speaker cables are chosen for their 'losslessness' (he said inventing a word ) so surely and distortion would not be lost either? Am getting to the absoloute limits of my understanding here, though.

hades

Not quite. If it's commoned at the amp end only, you're introducing the distortion over the very short run from the output transistors in the amplifier to the speaker terminals - say 100mm. Typical speaker cable length is say 5m - i.e. 50 times that, so the magnitude of the distortion is much bigger.

Of course, if you common both ends (ie leave the plates in the conductor), then you're not splitting the frequency components, so the benefits should be much less.

You're right in that with lossless cables, you won't lose distortions any more than you lose the sounds you want. What we really need is "intelligent" cable that just takes out the bits we don't want

Turbo_Steve

LOL...Lossless cable:

Put amps right next to speakers (active speakers?) with D/A converters in with error checking. Send data to amps. No loss!

actually, I think Meridian do this?




I don't understand still: cables are commoned at the amp end, but distortion is only introduced into the one feed? Why?


[Edited by Turbo_Steve - 10/30/2003 9:51:23 PM]

hades

Sorry, that's not what I meant. You get distortion introduced over the 100mm of run from the output transistor to the speaker binding posts. That is fed down both "halfs" of the biwire away from the amp.

If you're biwired, you don't get the current at bass frequencies interfering with treble signal in the cable, and you don't get current at treble frequencies interfering with bass signal (remember magnetic field strength is proportional to current, not voltage)

However, if you don't bi-wire, you introduce more distortion along the lengths of the cable run. If the cable is 5m long, there is more distortion than you'd got from the 100mm in the amp.

If that doesn't make sense, it just proves I made a good career choice not to teach!

I think you may be right that meridian do have a system where you transmit digital signal, and only convert it to analogue inside the speakers, hence much shorter speaker cable runs within the speakers. They probably still have some separation, to take the power amp away from the speaker magnet.

They also probably suffer from the DAC, the preamp and the power amp all running off the same power supply - this will have finite (non zero) output impedance and so introduces distortions due to minor voltage supply fluctuations. Don't get me started on power supplies in hi-fi!!!

Turbo_Steve

Ah. That all makes more sense definitely clearer the second time.

And as for power-supplies. Yes. VERY complicated. Head F_ck.

dr_ming

For copper conductors, skin depth (in metres)is given by:

0.066 / square root(frequency Hz).

At 22.05kHz (the max frequency that can be recorded onto a standard CD), skin depth is 0.466mm. i.e. for any conductor smaller than 0.93mm in diameter, skin depth has no effect.

Note: the formula above is simplified to remove various constants that pertain to the conductor material etc. Trying to type all the greek symbols is too much effort. Do a search on the internet if you want the full detail.

hades

Saved me looking it up!

Working from memory, skin depth isn't a "square edge" type fall-off. More like a crossover - as you move from the surface, the conductance falls off exponentially? (long time since I did this stuff, correct me if I'm wrong). I know you said you simplified - and quite right too - Assume the above figures are for something like a nominal 3dB roll off?