Renovation of Celestion 66 Studio Monitors

[DarrenHW]

Darren - Yes just the coils mounted on their circumference so rectangular in plan view.

So it was a stupid question . Maybe this is too, but why 5 instead of 3?

A PP works in the series location only in the midrange filter, don't add any ESR compensating resistors though.
The 24/30uf Elcap used in this location was LL so less ESR anyway.

Interesting, I'll have to try it.

In this location I've tried both PP and several make of electrolytic and combinations of both and different makes of PP.
The Ansar is my favourite of what I tried, and is not too big either. A Solen PP for instance, sounds similar to an Alcap Lytic in this location, very little difference between the two.

Did you notice any difference in bass? I thought it noticeably tighter when using a PP instead of an Alcap but this was in a different pair of 44's so not sure if this was due to differences between the two (different cabinet construction, wiring and sound deadening).

If you use a PP you will need the L-Pad in the mid range or the mids are too forward.

I found the same myself, I thought the speakers had far better driver integration using the Alcaps, I wasn't sure if this was due to using a 27uF PP (as per the DIYAudio mod) as I used 3 x 10uF Alcaps (to keep with original values) in the second pair.

[Reffc]

So it was a stupid question . Maybe this is too, but why 5 instead of 3?



Interesting, I'll have to try it.



Did you notice any difference in bass? I thought it noticeably tighter when using a PP instead of an Alcap but this was in a different pair of 44's so not sure if this was due to differences between the two (different cabinet construction, wiring and sound deadening).



I found the same myself, I thought the speakers had far better driver integration using the Alcaps, I wasn't sure if this was due to using a 27uF PP (as per the DIYAudio mod) as I used 3 x 10uF Alcaps (to keep with original values) in the second pair.

A classic case of what happens when electrolytics are replaced with polys in certain positions , where adding L-pads to tame perceived brightness does not address the crossover frequency slope shift that the ESR and ESR slope with frequency gives with electrolytics compared with polys. In the lower mids, the differences may only be 0.5 Ohms or so but that's enough to throw phase integration out by shifting crossover point upwards. Also, it means adding more components into the signal path (l-pad) when sticking to the Alcaps avoids this. Sticking with close tolerances, matching manufacturers tried and tested exact values ("near enough" can often not be near enough), using like for like in the LF and lower MF positions for electrolytics good layout and quality of inductors. If these things are adhered to, the end result should be as good or better than when the speakers left the factory.

[Qwin]

Reffc is right Darren, any deviation from original values or cap type is going to effect x-over point/slope.

So my choices in the mid range are a compromise but one I feel I can live with.

"Maybe this is too, but why 5 instead of 3?" There has allways been 5 coils in the Celestion 66 circuit.
I am just re-using the originals.

I found the 30uf Ansar PP cap gave more body/depth/texture to the mids, the Solen I tried did not and I prefered the Alcap Lytic over this.

Strangely the bigest impact I found on percieved bass was in the HF caps. I am using the Seas tweeter now but the same thing applied when I had the originals in use. I know the various parts of the filter inter act but I could not hazard a guess as to what is happening. Also sound is a mixture of frequencies, a drum strike has bass content but it also has higher frequencies that give it shape and snap/slam difining its edges.

My experimenting went like this, two Sonicap Gen1's for the tweeter values was very clean and transparent, maybe too forward in the treble, but very focused, Piano sounded very realistic at the top, the bass was there but nothing special. Two ClarityCap ESA's and the treble softened, too much for my liking and piano notes lacked that percusive hammer sound you get higher up the scale, mids were very realistic male voices had great texture and realism and Bass had more presence and depth. I liked the best bits of both but not the bad so started experimenting with mixtures of caps. I put a ClarityCap ESA in the large value HF and Sonicap for the smaller value. This was much better and kind of got the best of both, but still felt a little too forward, Joni Mitchell's voice was shrill and piercing at times. I reversed the makes eg Sonicap large Clarity small which was worse. So I kept the Clarity cap large value and dropped an Ansar Supersound in for the smaller, now we are cooking, It had natural tone to piano, good percussive action on the keys, voices both male and female sounded right and bass texture and slam seemed to jump up a notch. This voicing did it for me, I never listen to how good something sounds but how realistic it sounds, you can kid yourself that something that sounds pleasent and easy on the ear is better, but when you put something on that you can relate to and have a referance point for like acoustic instruments and voices you spot that something is not quite right. This mix can be quite spookey at times, I have my Satalite TV's audio playing through my system via a DAC and the other day I went to answer my mobile which was on the coffee table in front of me, to discover the sound was actually on the TV program. When I was in the kitchen on another occassion I thought someone was knocking at my front door, to dicover it was on the radio 4 play I was listening to. I know we all strive to get this kind of realism, but there are times when it raises the hair on the back of my neck. Don't get me wrong, my speakers are sounding good, I can still hear the resonance on certain tracks though, which means I tend to avoid playing them, this is why I am still faffing about. I could hear the same thing on Marco's speakers and it has even been measured on the diyaudio thread, if you use PP's with ESR compensating resistors it seem to bring it out to the point were it was visible on plots. It was at this point I gave up on using them and looked for an alternative approach.

[DarrenHW]

"Maybe this is too, but why 5 instead of 3?" There has allways been 5 coils in the Celestion 66 circuit.
I am just re-using the originals.

Another stupid question then, sorry .

Glad you're making progress with your crossovers and that the 66's are winning you over.

[Qwin]

They won me over first time I heard them.

I just want them to be the best they can be, as I will probably have them for a long time.

The basic character, which is what I like, has not changed, I am just tweaking the fine detail.

[Qwin]

These are the circuit boards I made from "Paxolin" sheet.

Brass eyelets with 3mm through holes have been fitted for use as point to point solder contacts. The six large holes just off centre line up with the six binding posts I have already fitted to the cabinet. The boards will be secured using 16mm M4 stand off's, this will position them just above the block of wood on the rear face of the lower compartment, that Celestion added for mounting the binding posts. Plus give additional clearance to accommodate the plastic spacer bar that each pair of binding posts I have fitted is secured by.

[YNWaN]

As ever, very nicely made. I'm surprised that you bothered with the eyelets though - you could just have small holes, thread the lead and wires through, and point to point solder it all together on the other side.

[Qwin]

YNWaN - I have learned the hard way, that even at this late stage, expect to modify further.

This arangement allows easy part changing without disturbing the main circuit, speaker connections etc.

I've even changed some parts on this type of board whithout re-moving the board from the cabinet and just worked through the driver aperture.

No rear access is required so keeps it simple and tidy after multiple mods and takes some of the strain off the parts as they are soldered to a fixed point. If you do as you sugested you can end up with loose parts that wiggle about as they have no direct connection to the board.

[YNWaN]

If you do as you sugested you can end up with loose parts that wiggle about as they have no direct connection to the board.

Well I would fix the components separately to the board - but if this solution works for you, was just wondering. .

[Qwin]

I'm waiting till after the hollidays before building the new boards, I want to try some Visiton caps in the lash up first to see how they sound.

Meanwhile I have been trying the circuit Tri-Amped.




This is the Power Amp I purchased second hand.
It's a Rotel RMB-1066, I wouldn't call it HiFi more like MidFi, but it’s a cheap way into Tri-Amping or Active crossovers. This model was mainly sold for home cinema use, where folks wanted to get into separates and split the amplification from the receiver/control section. It has several features that make it attractive for multi channel music use though.

First it is set up as 3 pairs of channels, so it's like having 3 Stereo Power Amps in one box

Second, each pair has it's own gain control, the pots are adjusted using a screwdriver and are located between the front fins, as indicated on the picture. For three way speakers this allows adjustment of each driver pair from one pot, so left and right bass drivers can be adjusted together as an example. Once drivers are initially matched for sensitivity the volume is controlled from your pre-amp.

Third is the RCA inputs, one pair of inputs for each pair of channels, again more like a stereo amp x3 which is good for going Active.

Lastly the input pairs can be linked by sliding a couple of switches so that one pair of inputs serves all three amp sections, saving on cable runs etc and useful for Tri-Amping from a single pair of outputs on a Pre-Amp.

This is a very versatile piece of kit, you can also bridge any pair of channels to increase the power in that section but reducing it to single channel. So you can have 6,5,4,or 3 channel output depending on which pairs you bridge. Power output is 60W/8 Ohm x6 or 150w/8 Ohm x3 (Bridged).

It doesn't sound bad at all, quite nice really and it has cleaned up the Tweeter circuit, no distortion getting through at all, now each one has it's own amp channel.
The mids still have the same distortion, which pretty much confirms the source of the problem, what ever it is, lies there. Lets hope when I move the coils appart it resolves it. I have taken the L-Pads out from the circuit and have adjusted the driver levels using the Amps gain controls, sooo easy to get the voicing the way you prefer it.