DM 70s

[TopMarc]

I just posted in the general loudspeaker section and it was suggested by a couple of members that I also post in the ESL section - I have DM70s and being hybrid I am always a little schizophrenic about what they really are and where to post )

I just rebuilt two units (both refoaming the woofers and rebuilding the ESL panels) but I'm considering rebuilding them again because both hybrid system have sonic "holes" in the mid-base frequency range - its either caused by soft woofers that has lost their color or by ESL panels that dont reach down to the crossover of 400 Hz as they should. The fact that both speakers are behaving the same trends me to believe that its the ESL units. I cant be too sure of this because in rebuilding them I changed/updated a number of things.

Does anyone have any ideas about what tension I should be using to preload the Mylar prior to bonding the stator/airgap to them? I stretched the Mylar on a flat table using adhesive tape.

Should this tension be applied uniformly (ie in both directions) or only in the vertical plane? THe reason I ask is that I have noted that when the assembled stator/airspacer/diaphragm is bent over the curved wooden chassis considerable effort is required and I am starting to believe that this increases the tension in the horizontal plane. (I did place adhesive on the airgap spacers in between each of the 11 panels so it is anchored at regular intervals)

What do you guys think? I understand the tension will effect the fundamental resonance frequency of the diaphragm, do you think it willl also have any meaningful effect on the lowest frequency being produced by the panels? Could this be causing the hole?

I want to state that the speakers, when working properly (some years ago), were a perfect fit to my hearing and I do not worry about what some call a "design flaw" with the hybrid nature of these units. If I can get them working as I remember them some years ago then I would be happy.

Your job, should you wish to accept it, is to help me decide if its the ESL panel. I'll keep the moving coils stuff on the general loudspeaker section so as not to muddy the water here.

Thanks for thinking about this.

Marc

 

[TopMarc]

Replying to my own post

Sorry for replying to my own post but I just got this reply to my questions in the general loudspeaker section and wanted to add it, and my response, to this thread in the hope that the collective knowledge in here can offer some guidance.

THanks

Marc

Now that you've explained it more fully, the problem could be the ESL re-build. With the ER Audio kit, did you replace the membrane with the same thickness of mylar as originally provided by B&W. Rob prefers to use 3.8 micron mylar which is very thin by industry standards, it provides excellent speed and transient response but may not be giving the lower mid/upper bass that a thicker membrane provides.

It may simply be that the thinner mylar gives a frequency response that you are not used to, perhaps lacking the body in vocals that you are accustomed to with normal DM70?

Ah ha.

You might have hit the nail on the head. The ESL panels sound perfect on their own and so do the base drivers. When paired they leave a gap. I was searching for a reason and I think you may have it, the diaphagm is thinner and has moved the response curve up in frequency. I feel an experiment coming on!

So what thickness should I use? 6 micron seems to be the choice of the quad guys for treble panels and 12 for the base. Which would be better given the 400 Hz crossover? Could I get 12 micron to handle up to 18/20 KHz? What tension should I try? Where would I go to get it in the USA? Area 51?

I love you guys....

 

[maudio]

Thinner mylar itself should not make much difference to lower frequency behaviour. It's just lighter improving hf response.

But thinner mylar will likely result in lower membrame tension = lower free air resonance. So it could be that the resonance frequency of the rebuilt panels is different from the originals. And that the original design used the peak to lift output near the crossover point. Not uncommon practice in esl's.
Depends also on how much you stretched it. In general thicker film allows for more tension, moving the resonance peak up but also making it flatter and wider.

It's a good idea to measure resonance frequency before taking an original panel apart so that you know what to aim for. Oh well, too late I fear

It's difficult to say how much tension is ok. I have 4 um mylar that I stretch 0.4..1 % (in length, both axis). I have some old 6 um that needs at least 1.5 % and will withstand almost 4. Bottom line: there are huge differences between various brands and batches. Trial and error

In general: <=6 um is ok for fullrange. 12 um should only be used for bass panels, won't do very well at 20 khz.

You could try playing with the EHT on the membrame.

 

[TopMarc]

Originally posted by maudio Thinner mylar itself should not make much difference to lower frequency behaviour. It's just lighter improving hf response.

I have conflicting opinions here - one member suggests membrane thickness effects frequency response and another suggests it makes little difference. Not trying to be antagonistic but where does truth lie in this issue. If I used approximately the same tension in 6 micron Mylar as I have in 3.8 micron Mylar (I assume that what eraudio sent me is 3.8 based on another member's input) will the thicker membrane result in better low frequency response for the panel or a higher output at lower frequencies? I would not mind sacrificing some of the high end for more low end since I now find myself using the HF cut on the amp (Audiolab 8000A) on some material -- something I never used to do with these speakers. This suggests to me that I have room to play with the response curve.

But thinner mylar will likely result in lower membrame tension = lower free air resonance. So it could be that the resonance frequency of the rebuilt panels is different from the originals. And that the original design used the peak to lift output near the crossover point. Not uncommon practice in esl's.
Depends also on how much you stretched it. In general thicker film allows for more tension, moving the resonance peak up but also making it flatter and wider.

Its a long time since I did any maths at this level but could someone point me in the right direction to try to figure out how i would get the DM70 panel to resonate somewhere closer to 400 Hz or am i better off with several trial and error experiments?

In general: <=6 um is ok for fullrange. 12 um should only be used for bass panels, won't do very well at 20 khz.

Would 9 micron be an alternative or is that too heavy?

You could try playing with the EHT on the membrame. [/B]

I dont have the tools to measure the HT output -- I simply rebuilt the Cockroft Walton ladder out of new components. I assume they work ok by the fact that the HF output on the panels is amazing. Were you suggesting that I might increase the HT voltage to increase output?

Thanks
marc

 

[moray james]

thinner is better

thinner film will result in better bass extension as it will stretch more easily (farther) than a thicker film and require less applied force to do so. Thinner film will also result in better high frquency extension as it has reduced mass and so mass related roll off will happen at a higher frequency. Overall a lighter diaphragm will result in greater efficiency.
The confussion about thicker films I think arrises from the fact that Quat used a thicker Saran film for the bass panel of the 57 bass panel. This was chosen soley on the basis that that was the thickness that was necessary to achieve the required stretch characteristics with that film. In other words The desired quality was stretch. So if you reduce your mylar thickness down to 3-4 microns it will be able to stretch enough to be used for bass. Such thin mylar was not commecrially availabel at the time that Peter Walker designed the Quad 57.
That's how I had it explained to me and I think it makes sense.

 

[maudio]

thinner film will result in better bass extension as it will stretch more easily (farther) than a thicker film and require less applied force to do so.

I agree with Moray. I't my experience that long stroke panels are only possible using thin film with moderate tension.

I wouldn't use 9 um for mid/high range panels. 6 um is about the limit if you don't want to lose too much detail in the highs. Look at Quad's 63: it uses 3.8 um film. It's the mass of this thin film that is responsible for the drop in output near 20 khz (according to Peter Walker). Imagine what happens with a 3 times heavier film

One disadvantage of thin film is that due to the reduced mechanical tension the membrame will be less mechanical stable, so a thin film panel will withstand less EHT bias beofre the membrame collapses in to one of the stators. One solution is to keep panel width small. Or support the membrame with silicone dots (as Audiostatic does).

 

[TopMarc]

I agree with Moray. I't my experience that long stroke panels are only possible using thin film with moderate tension.

This seems most logical to me and the bit that has me most interested is the "moderate tension". I think the tension on my panels was OK before I bent them around the curved chasis at which time it went up causing the drop off in mids and lows. I will now rebuild the panels with the material bonded to the other stator which will not put additional tension on them during assembly.

One disadvantage of thin film is that due to the reduced mechanical tension the membrame will be less mechanical stable, so a thin film panel will withstand less EHT bias before the membrame collapses in to one of the stators. One solution is to keep panel width small.

Fortunately the B&W panels are relatively small being some 2 inches by 3 inches in area with eleven panels making up the total array so I dont think their is much of a chance of the lower tension causing the diaphragm to collapse into the stator,

Thanks all for your assistance.

Marc