The "Elsinore Project" Thread

Well after much listening my preferences were as follows;

33R resisters in parallel with drivers did sound better. Bass/Mids appear more defined.

As for the Bolserts's mod(with Joes amendments) to be honest I couldn't really tell the difference. That said this was with listening tests via just one speaker. Now both mk5 crossovers are connected I will attempt this again. My primary reason to remove the mod may end up just being in regard to improving sensitivity, unless there is any noticeable affect on sound quality.

Lastly with the series cap for the tweeter I currently have this at 2uf as per default. But based on the previous results I will be increasing this to 3uf in the coming weeks. Based on my setup I think this change will go a long way in increasing the dynamics.

As I concluded earlier the Mk5 is most definitely less sensitive than the Mk3. The payoff is the sound quality is most definitely improved. So all in all I am quite happy with this upgrade.

Cheers, Andrew
 
Hi Joe, i have checked every step of the way (multiply times). From components to wiring. I am sure that the mk5 crossover is correct.

I was comparing Mk5 vs Mk3 in room. I used pink noise with an spl meter both in listening positions and also 1mtr off axis. Each speaker was tested independently. I also reversed the crossover in case the room was having a negative effect. For both channels I had to adjust the Mk5 +2dB in order to match the levels of the Mk3. Would this be related to sensitivity or something else?

Could it be Bolsert's mod that may be effecting it? I am planning on converting back to stock for testing.

Should the Mk5 be as efficient or more efficient than the Mk3?

That said... maybe the mk3 had an error! - its been a long time since I looked at it. And I do recall I was in a bit of rush to get them together.

But seriously the Mk5 does sound far better, its just im now having to drive the amp harder to get the same SPL.
 
I was wondering if the value I changed to (3uf) was inline with what you might have considered?

I would not have envisioned anything as high as 3uF. Have you checked the correct phase of the Tweeter, as that would cause a noticeable suckout and make it noticeable dull, then under those circumstance a much bigger cap might actually improve it, but clearly not in any way as originally intended.

The other thing I might add, but I know I have been guilty of this myself, making a mistake and going over things many times and not see the fault; then somebody looks over my shoulder and spots it instantly, very embarrassing. But it is sometimes the way the brain works. Have you got another set of eyes that can go over it? Even if they don't find the fault, then at least you know.

I am only saying this, as it doesn't add up: If you can hear a drop in sensitivity, when in fact it should have actually improved a bit (I am recalling the modelling in my mind that both series chokes are now smaller values than before and this actually increased sensitivity as there was more upwards summing), then something is amiss?

I assume the drivers are the very same as before, right?

Cheers, Joe
 
Originally I thought it might have been the tweeter phase, however this only prompted me go back and double check wiring and also battery test the tweeter phase. Everything checked out so I left it as is. Ive just tried reversing the phase and as you stated it does sound more dull. It also sounds more spacial - not as focused.

I have had someone look over both my initial component layout and also the finally wiring. I might take photos of both and post so we can isolate any errors here.

Drivers are unchanged.

You have me worried now... could it just be I am now so used to the mk3's overly bright imaging that now it sounds dull/flat in comparison - but maybe more accurate?

I did a calibration with my avr yesturday and have just been going through the results. Interestingly the eq curve thats been applied to the mains is very flat - far better than the results I got with the mk3.

Also the mains speaker levels have been reduced by 1dB (mk3: -2.5dB mk5: -3.5dB). Since i believe calibration is based on a preset reference point then does this not indicate that they are in fact more efficient?

Im now at a loss as to why it measured less when I manually tested. I was only using a budget SPL meter however (RS 33-4050).

Would 3uF negatively impact any of the other dynamics?

Thanks Joe, i really appreciate your assistance.
 
While the mk3 was excessively energetic and bright the mk5 sounded muffled and less dynamic.

Hello BL4S7ER,

One possibility for the Mk5 sounding less bright is the tweeter to waveguide mounting method.
Depending on the size of the gap between the tweeter face plate and waveguide, you can get a -3dB dip in the 2kHz - 3Khz frequency range.

The best solution to tracking down the source of the differences is to find somebody who can help you take some impulse type frequency response measurements that exclude the influence of the room.
 
Last edited:
Hi Bolserst,

What size gap is suggested to remove the dip? Is it possible to take measurements in room that will exclude the influence of the room? Is this done via timing (ei measurement is taken between say 20-50ms of output tone)? Can this be done with my meter and pc software? or would the results not be reliable?

I should mention that when I described the Mk5 as sounding muffled - this was in direct comparison to the Mk3. Its hard to describe the level of the effect in words. There just seemed to be less presence. Altering the 2uF to 3uF increased the presence and to my ears sounded much better.

Cheers, Andrew
 
What size gap is suggested to remove the dip?
Sorry for my less than clear response concerning the gap. I will try to remedy.
With no/minimal gap there is no noticeable dip in the response.
As the gap between the faceplate and waveguide increases, a dip develops in the response.
The frequency and magnitude of the dip changes with gap distance.

Is it possible to take measurements in room that will exclude the influence of the room? Is this done via timing (ei measurement is taken between say 20-50ms of output tone)? Can this be done with my meter and pc software?
To exclude the room, you generally need to look at a 3ms – 5ms time window after the first arrival of sound at the microphone.
I know this can be down with software like REW, ARTA, Holm Impulse, etc.
What software do you use?

I should mention that when I described the Mk5 as sounding muffled - this was in direct comparison to the Mk3. Its hard to describe the level of the effect in words. There just seemed to be less presence. Altering the 2uF to 3uF increased the presence and to my ears sounded much better.
Understand. It could be that if you had lived with the Mk5 for a year and then built the Mk3 you would think it sounded overly bright or harsh.
In either case, without measurements it is difficult to pinpoint what the cause of the difference is, or if either Mk3 or Mk5 is not operating as Joe designed it.
 
Yeah thinking about it as well. Have heard mixed reviews about the hds weeter thats why im asking around about it being recessed if it affecs transient response or frequency response. Thinking of swapping it for a sb acoutsics ring radiator tweeter or modding the crossover to flatten the hf with a wide notch filter
 
What happened with the Hamlet design?

My apologies, got overtaken by events, including my ClioFW developing a Firewire connection problem and had to be sent to Italy for repair and now $450 later got it back.

When it gets a little warmer, will get the Hamlet work done - so I am next month. As there are many similarities with the Elsinores, I don't think there are going to be many problems and reinventing the wheels. So once started, it should come together quickly.

But it comes down to it most likely being a 4 Ohm nominal design (if it does not drop below 2.8 Ohm it qualifies as nominal), but the actual impedance 3 Ohm flat from 40 Hertz up. Also, it can be current driven by those who want to build the 40 Watt Transconductance Amp.

Cheers, Joe

.