B&W DM4 With Poor High Frequencies

[fionasdad]

I bought some second hand B&W DM4s (early 1970s) which have very good mid/low frequency reproduction but have lost their high frequency performance (muffled/reduced cymbals, guitar string scratch, speech sibilance, amplifier hiss etc) compaired to the fantastic top end I remember hearing on others (the reason for buying them!). The high frequency drive units (Coles 4001 & Celestion HF1300) are producing sound in their appropriate frequency ranges, ie they are functioning & not blown. Have applied frequency sweep to speaker inputs & viewed signals with scope through xovers to drive unit contacts and appropriate signals appear at the drive units. Can the drivers themselves lose their upper frequency performance over time (diaphram gets brittle etc)? I replaced the HF1300s with a second hand set bought off ebay with no marked difference so either they have the same problem or its the 4001s or xovers. Could the capacitors in the xovers have aged? they appear to be good quality polypropylene film not electolytic. The diaphrams on the 4001s look OK, no cracks etc. I see you can get replacement diaphrams but the coil wires are very fine & I'm not sure I'd like to take a soldering iron to them. I see you can pick up DM4s on ebay for ~£100 so not sure if buying new 4001s at £45 each is worth the punt!

I have tried connecting both types of driver in place of the Trio tweeters in some old Trio (Kenwood) LS-202s and they are not nearly as good at the top end as the orginal Trios (this may not be particularly valid as the Trios are 4ohm on a much more basic x-over).

I might try experimenting with shorting out coils & capacitors in the xover but dont want to damage the drivers by putting too low frequencies up them, so advice on this appreciated.

Any thoughts?

 

[Mooly]

Memory can be a funny thing. A very real (and perfectly normal) reason could be your hearing that has lost some of its HF response. Sadly that's a fact of life as we age.

Do tweeters age as such... I don't know but its possible I suppose. Those with more speaker experience could chime in on that.

Don't go shorting out components in the crossover as that is a surefire way to kill the HF drivers. If you need more treble then a simple tone control could be the answer.

 

[fionasdad]

Thanks Mooly for quick reply.

I hear what you say about the aging ears (ha ha) but the old Trio LS-202s have a much better top end (my 14 year old daughter is my bench mark & she said the DM4s were dreadfully muffled compaired to the Trios as have other people who've got her own hifi setup).

I got the shorting out caps & coils idea from an earlier post in this thread I think, but will take your advice.

 

[Mooly]

From what you say it does sound as if they are below par.

It might be worth you starting a new thread on this, as this one is quite old. I could copy your first post over to a new thread if you wanted... its a pretty much perfect candidate for that.

 

[fionasdad]

Yes sounds like a good idea, might get more attention. Thanks.

 

[system7]

I think this thread is pretty good on DM4's as it goes. Do we really want to repost everything?

I'd be looking for corroded connectors here, and maybe reflowing some solder. Corrosion will kill the sound.

But nothing stopping you replacing the Coles 4001 with something like a H0414-08 19TAF/G and a bit of crossover work. Be worth measuring the DC resistance of the HF1300 and 4001 units. 8 or 16ohms? Maybe they have been clumsily replaced. The DM4 crossover looks like a 16 ohm one to me.

 

[Mr. dB]

Be worth measuring the DC resistance of the HF1300 and 4001 units. 8 or 16ohms? Maybe they have been clumsily replaced. The DM4 crossover looks like a 16 ohm one to me.

The HF1300 used by B&W was a 15-16 ohm driver, and the Coles 4001 was "21 ohms".

 

[Mooly]

New thread split from,
http://www.diyaudio.com/forums/multi-way/32034-b-w-dm4.html

 

[system7]

And here's some relevant images from Malefoda, who had this speaker:

And two of my own archive pica below. 16 ohm mid and tweeter it is then.

 

[fionasdad]

Thanks to all for responses.

Will check the connections & reflow the solder joints to see if that helps.

I have checked the DC resistance of the 2 drive units, 4001 = 23 ohms & HF1300 = 12 ohms so looks like they are the correct units. I was wondering if the unusually high impedances might be something to do with the problem. My amplifier is not a high power & intended to drive 4-8 ohms, I'll see if I can get some other amplifiers to test with.

Interesting idea about replacing the 4001s with the cheaper modern 8 ohm replacement, I had wondered about doing this, as a substitution test for the 4001s. Presumably I need to modify the cap values (C6 & C7) in the xover to lower the xover point, suggestions? I looked at the SEAS website, lots of tweeters, why the H0414-08 specifically?

Any mileage in replacing the xover caps for the 2 high drive units? If so recommendations?

 

[fionasdad]

Attached is B&W original circuit diagram

 

[system7]

Thanks to all for responses.

Will check the connections & reflow the solder joints to see if that helps.

I have checked the DC resistance of the 2 drive units, 4001 = 23 ohms & HF1300 = 12 ohms so looks like they are the correct units. I was wondering if the unusually high impedances might be something to do with the problem. My amplifier is not a high power & intended to drive 4-8 ohms, I'll see if I can get some other amplifiers to test with.

Interesting idea about replacing the 4001s with the cheaper modern 8 ohm replacement, I had wondered about doing this, as a substitution test for the 4001s. Presumably I need to modify the cap values (C6 & C7) in the xover to lower the xover point, suggestions? I looked at the SEAS website, lots of tweeters, why the H0414-08 specifically?

Any mileage in replacing the xover caps for the 2 high drive units? If so recommendations?

That schematic is useful. B&W and myself must be the only people in the World who aren't secretive about schematics.

B&W Group North America Service & Support - Home

I picked the SEAS 19TAF-G because it is a 19mm dome and a known quantity. Easy to fit if we go that route. Maybe just put an 8.2R resistor in front of it and use the existing 16 ohm crossover. Would have to model that, but it's more efficient so some leeway. I can't see much mileage in changing capacitors since they are stable polyesters, but nothing stopping you putting in 250V polypropylene which are smallish.

I'd be looking for corroded connectors here, and maybe reflowing some solder. Corrosion will kill the sound.

But nothing stopping you replacing the Coles 4001 with something like a H0414-08 19TAF/G and a bit of crossover work...

Not really much possibility of upping tweeter level on that filter. Maybe it's designed for the lowered top end rolloff known as BBC dip.

 

[KatieandDad]

Are you 100% the fault is with the tweeter ?

The midrange unit produces most of the audible high frequencies and were known to break up internally.

You can still buy the HF1300 on E-Bay http://www.ebay.co.uk/sch/i.html?_f...TR0.TRC0.H0.XHF1300.TRS0&_nkw=HF1300&_sacat=0

 

[system7]

Hey, KatieandDad, nice to hear from you.

It crossed my mind that the HF1300 might be dulled too. It actually goes up to about 14kHz with the L3 rolloff coil removed. Which was good enough for people with the Celestion Ditton 15. The B&W/Spendor style Coles 4001 supertweeter just adds a bit of air.

TBH, it's a pig to replace the Celestion HF1300 though Coles still make it. It's rather unique. Maybe this one on a 2.5kHz crossover?

TW 70 - 8 Ohm

16 ohm can be dealt with.

 

[xrk971]

A frequency response measurement would be worth 1000 words and help plan course of action.

 

[system7]

A frequency response measurement would be worth 1000 words and help plan course of action.

Me, I'd just listen to it! Put my hand over the two tweeters in mono to see what is being contributed by each driver.

fionasdad has already tried replacing the HF1300 tweeters.

I suspect he is running into the polite, and frankly, dull sound of seventies BBC Bextrene speakers.

This is a third-order BW3 design which I like. All the flat power dispersion. Not much bafflestep so best close to a wall for bass. Which might make the top end sound restrained.

To me an old speaker is just a box awaiting new drivers. But it's worth reading Troels on this top-end combo. The HF 1300 was near enough a cone tweeter. Spendor BC1

I usually look at what will fit in modern replacements. A question of measuring the holes and cabinet size.

 

[KatieandDad]

There is no point in replacing any of the crossover components, polycarb cap's don't age like electrolytics.

Put your ear next to the HF1300, if the cone has started to break up you might hear it crackle.

 

[system7]

I wouldn't bet my life on polycarbs or polyesters being immune to age.

It's cheap enough to replace parallel C4 and C5 with a new polypropylene 3.9uF for instance. After all these are just plastic films with a few atoms thick of corrosible aluminium. Aluminium turns to oxide faster than almost anything else, being a lively alkaline earth.

Nothing lasts forever. Maybe it'll fix it. Maybe it won't. Time will tell. And TBH, I certainly prefer the idea that tweeters and glue degrade faster than anything else.

 

[fionasdad]

Interesting experiments tonight;

Firstly used more powerful amp, Cyrus One, no real difference, still poor high frequencies. Amplifier miss matching theory out the window.

Second, swapped white & grey driver leads round on the 2 tweeters to change the xover/driver relationships (see attached photo), big difference, now the 4001 output is louder than the HF1300 (whereas normally the HF1300 is much much louder than the 4001, which is barely audible) So looks like the 4001s are working OK, when driven from the mid filter output. The HF1300 is quieter but definitely louder than the 4001s were when connected to the hi filter. Looks like the HF1300 is more efficient at utilising the signal from the hi filter

The HF1300 is now reproducing the higher frequencies due to being connected to the hi filter, but I guess not as well as the 4001 would if working properly as the HF1300 rolls off at 13-15kHz I think.

Now I think the overall hi frequency sound is better too, comparable to the Trio (will need to get the 14 year old ears to check tomorrow!)

Not quite sure what to make of this, it suggests to me that the hi filter is not matched correctly with either of the tweeters somehow. Cant be the coils, may be the capacitors attenuating somehow?

Investigate further tomorrow....

 

[irext]

Beg, buy, borrow a capacitance meter (they aren't expensive these days) and measure the caps. You will have to remove them from the pcb to do this properly. I wouldn't recommend leaving the HF1300 and the 4001 swapped for long. You may do the 4001 in. I wouldn't expect much audible o/p from the 4001 in normal operation. Most of us can't hear much above 14khz anyhow. If the caps are out of tolerance buy some inexpensive polyprops to start with. If this cures the problem you can go with more upmarket ones if you wish. Were the HF1300s you bought new? They may be down on o/p also if they are the same age as yours.
Cheers.