advertisement


pfm Ogre I

orangeart

KJF Audio Ltd.
Ok here goes,my first speaker build!

Firstly I've no real idea what I'm doing so please don't blindly follow my recipe and then slag me off for designing a crap speaker! You were warned :)
I had intended on building James's PFM SII but alas James hasn't had the time to get it finished. This build is based loosely on his design ethos, sealed and dynamic floorstander.

I have borrowed elements of his box design but have ommitted a vertical brace as my box isn't as wide so no real room. It is also slightly smaller at 33l where James's is 42l. The box is closer to the box on this page, that whole site is a great resource in fact. My dimensions are slightly different because the 25mm mdf that the front panel requires in both the aforementioned boxes comes in 8x4 sheets which is more than large enough for all the panels to be cut from, so that is what I've done I've kept the intenal dimensions the same and adjusted the panel sizes acordingly. I hope the extra thikness will make a stronger and more acousticaly inert structure.

I also spent a lot of time reading over this design which is the basis for the popular madisound kit of the same name. I am using a modified version of this X-over to get started. The drivers I am using are the scan-speak 18W/8531G00 and the D3004/662001

The frequency response of the 66200 is flatter toward the top end and more extended than the 6600 used in the ZRT, the lower end is almost the same and most importantly, the impedance curve is almost exactly the same. All of this means that the cross over doesn't need re-designing, there is a impedance flattening notch in Zaph's design that'll do exactly the same job here and as the lower FR is almost identical the crossover point can remain the same. The only change I have made is to lower the value of the padding resistor from 6 ohm to 5 ohm because of the sensitivity difference.

Because I didn't have room for the vertical brace as per James' original plan i decided to add rigidity by rebating the rear and side panels as per the photo.

IMAG0088.jpg


The rebates should be 12.5 mm as the board is 25 mm. However as you look at the image the board on the left is the side panel and on the right the back panel. I actualy made the rebate of the side panel 13mm wide x 12.5mm deep, and the back panel 12.5mm wide and 13mm deep. This gives a 1mm overhang at the rear, allowing me to run a router with a fluted bearing bit along the back ensuring an absolutely flush fit to save sanding and filling later. This took some doing as I don't have accurate plunge control on my router so had to use a lot of scrap wood and a set of callipers and swearing!! Here are my boys helping with the clamping to line all the panels up for routing (note the power tools are not plugged in :)

IMAG0051.jpg


I have also rebated the side panels to allow the internal bracing and top and bottom panel to sit in their own rebate, adding further to rigidity.

IMAG0068.jpg


The panels are all then glued and screwed together. I don't think screwing is really needed but as I don't own enough sash clamps to hold a drying box together I had no choice. I'll have to fill and smooth the counter sinks later.

IMAG0069.jpg


The drivers, cable, stuffing, foam, crossover components, spikes etc are all coming from Wilmslow Audio. They have their own very similar design called the vogue and were happy for me to use this kit as a base for my component changes and adjusted the prices accordingly. This allowed me to get a very keen price for a kit of parts rather than buying a la carte, a significant saving. I went there in person to pay for the bits and Terry was a very helpful fellow, well worth speaking with if anybody does venture to recreate the PFM Ogre I. They also make cabinets as well but while they were well finished there wasn't anywhere near enough bracing IMHO for a project with this level of driver in. They sounded a bit like a drum with the old knock test. James had set the bar with his PFM SII box for stiffness and I wasn't going to settle for anything less. That said, if your woodworking skills aren't up to the job you could get one and just add some extra bracing, like this.

IMAG0066.jpg


The cabinets have been lined with bitumen/acoustic foam laminates and now feel very solid when knuckles are rapped on the side. They sound denesly hollow but with none of the drum sound of the aforementioned Wilmslow boxes.

IMAG0065.jpg


That is as far as I've got for the moment, the drivers will arrive late this week and then i can move on, I'll add more as I get more done. If there is interest I can post panel dimensions and provide a diagram, may even be able to make a few boxes!!

Finally, to prove that helping daddy in the shed isn't just the preserve of the lads here is my other little helper with a cabinet, this was after spending two hours helping me make a circle cutting jig for the driver holes - not just a photo op:D

IMAG0070.jpg
 
I've edited the OP and added some photos as promised, i'll add to it as things happen. If you have any questions please ask - i'm no speaker designer so i'll probably mot be able to answer!! he he.

Stefan
 
Stefan,

Firstly, my apologies for not finishing the PFM-SII. Loads have been going on in my life to distract me from the project, and I fully understand your impatience. I still hope to finish them off before the drivers around which I've designed them becomes obsolete.

Secondly, I applaud your initiative. The first DIY loudspeaker is always a bit special. I note that you're using the 18W/8531G00 in a sealed box whereas John Krutke is using it in a ported enclosure. In truth, the 8531G00 is probably not ideal for sealed box operation, and I think you'll have a relative lean sounding bottom end. If you haven't already committed to purchasing the drivers (and routing the baffles), you might like to consider the Scan-speak 18W/8535-00 instead. It's an older driver, but much better suited to sealed box operation.

I will watch your project with interest and help where I can. The best way to learn about loudspeakers is to build your own and get stuck right in. Good luck.

James
 
Lord of the Ergo's on board, nice! Hope you don't mind my play on Ergo - Ogre. seemed apt as you are very knowledgeable on the subject and your speakers are all well received whereas I'm a blundering fool with a router and some rather too expensive drivers!

Firstly, my apologies for not finishing the PFM-SII. Loads have been going on in my life to distract me from the project, and I fully understand your impatience. I still hope to finish them off before the drivers around which I've designed them becomes obsolete.

No worries James, just haven't got the room to do them indoors so it had to be a summer project, drew enough inspiration from your thread to get me going anyway.

Secondly, I applaud your initiative. The first DIY loudspeaker is always a bit special. I note that you're using the 18W/8531G00 in a sealed box whereas John Krutke is using it in a ported enclosure. In truth, the 8531G00 is probably not ideal for sealed box operation, and I think you'll have a relative lean sounding bottom end. If you haven't already committed to purchasing the drivers (and routing the baffles), you might like to consider the Scan-speak 18W/8535-00 instead. It's an older driver, but much better suited to sealed box operation.

Interesting, John does suggest that using it in a sealed box would be OK so long as there is sufficient volume, he suggests the Madisound 20l enclosure as a minimum. I am locked into this drive purchase now but would appreciate if you could give an explanation of what drive specs I should take particular note of when considering ported or sealed designs in the future. Are we talking about the bass rolling off early when you say thin? Also if i used a different driver here i would have the problem of having to model (or pay for modelling) a different crossover which i wanted to avoid on my first outing. I am considering making the baffle removable so if the bass really is to thin I can always revert to a ported design, and although calculating port tuning is way beyond my current knowledge i am sure I can get if figured out. TBH my preference would be to build a sub at some point in the future for the lower couple of octaves, I've never really got on with the sound of ported designs.

I will watch your project with interest and help where I can. The best way to learn about loudspeakers is to build your own and get stuck right in. Good luck.

Thanks James, I'll look forward to your help, can't wait to hear them really!! The next step will be an active set-up!

Thanks

Stefan
 
... would appreciate if you could give an explanation of what drive specs I should take particular note of when considering ported or sealed designs in the future.
The general rule of thumb is the fs/Qes ratio. Anything under 50 would be ideal for sealed, whereas anything over 100 should be ported. The 8531G00 is somewhere in between.

Are we talking about the bass rolling off early when you say thin?
Yes.

I am considering making the baffle removable so if the bass really is to thin I can always revert to a ported design, and although calculating port tuning is way beyond my current knowledge i am sure I can get if figured out.
That's not a silly idea. For a sealed design, you don't really need to line the inner walls. Just stuff the cabinet with fibreglass batts or dacron. Of course, if you go ported, then a lot of the stuffing will need to be removed to allow the port to breathe properly, in which case the wall linings are probably useful.

James
 
The general rule of thumb is the fs/Qes ratio. Anything under 50 would be ideal for sealed, whereas anything over 100 should be ported. The 8531G00 is somewhere in between.

Interesting, I did read something about this somewhere and did have a look at the figures at the time. As you mentioned it I thought I'd go back and look again. I noticed a bit of a strange anomaly in the specs for this driver. In the PDF for this driver the fs is 28 and the Qes is .39 which gives a ratio of 71 - indeed in the middle of the range you suggest as no-mans land between the two ideals. However I had a ratio of 58 in my head from somewhere which while still in that no-mans land is closer to the sealed ideal we are looking for. When I first started looking at scanspeak drivers I downloaded their speaker toolbox spreadsheet for advanced parameters to play with. Putting the parameters into that gives a very different story.

advanced%252520parameters.JPG


As you can see the calculated fs and Qes from the advanced parameters are very different to those published in the PDF and do indeed give the ratio of 58 I seemed to remember. What gives? which should be believed? I have also put pictures below of the response plots this tool box gives using these figures

ported (30hz tune with 40% stuffing)

port%252520tuned%252520to%25252030Hz.JPG


and sealed (100% lambswool stuffing)

sealed%252520%252528.JPG


They track almost identically to 50 Hz +/- 1dB By 40Hz the sealed design is 3dB down on the ported design and by 30 Hz (the port frequency) there is a 5dB difference and finally by the magic 20 Hz the difference is back to around 3dB

Assuming that the Advanced parameters model is correct those figures don't seem to terrible to me unless I'm missing something? James? If in the future I feel I'm missing that 5dB I think I'd be inclined to add a sub and keep the tighter sound of the sealed design.

Really enjoying this!!

Stefan
 
Hi Stefan, welcome to the wonderful world of speaker building.

I often find that measured parameters bear little resemblance to the published specs.

Sometimes that is due to newness, so if your drivers are new, try running them at high excursion for 12 hours or so.

You may find that will dramatically change the T/S values, for the better.

Having measured hundreds of drivers, my position now is that I would never embark on a box design until I'd measured the actual drivers fresh from an overnight burn-in.
 
Never seemed to get round to taking any piccies along the way. Spare time always seems precious and the actual building took precedence over the imagery!

Anyway they are not quite finished but they are 90% there. Enough for me to be able to get a listen. The rest is finishing touches, tweaking and costmetic.

Hooked them up last night and despite the fact that I guess they have a lot of running in to do they sounded great. The treble is maybe a little high but after it settles down i'll try swapping the padding resistors. BTW what make of resistor do folk find works well for series applications?

Although the treble is a little high at the moment i can hear that it is of a quality i have never heard before, very transparent and revealing. The mid woofer is the first paper cone i've ever had. I've previously had speakers with either metal or doped paper drivers so it'll take me a little getting used to before I can really evaluate the quality of the sound.

I'll need to take some measurements now i am up and running and would appreciate some advice on the easiest and cheapest solution to measurement. Does anyone have a calibrated mic and an RTA I could borrow?

Piccies and stuff to follow.

Stefan
 
If the low end is a bit thin with the sealed boxes, you can always move them closer to the rear wall...
 
Very keen to hear more on this, sound like a great build. It will be good to know if you hear that bass roll off and how you intend to finish the cabinets.

Also what tools did you need to make the cabinets?
 
seconded, a nice family friendly size too.
keenly awaiting more pics, inspiration to possibly start something similar..

Very keen to hear more on this, sound like a great build. It will be good to know if you hear that bass roll off and how you intend to finish the cabinets.

Also what tools did you need to make the cabinets?
 
Been a long while since I posted sorry.

Been using these speakers every day for a while now. For the most part they are great but there are a few rough edges. The quality, clarity and timing are generally great especially at normally listening levels. At first I thought the treble was set a little high, but that isn't the case.

The upper midrange, female voices, brass, piano etc seems to become very hard and louder than the rest of the piece when played at louder levels, and in more congested passages. For example adele sounds great at the begining of most of her recordings but when the band come in and she starts to give it some welly it is unlistenable, the instuments which previously had space and air around them become congested and difficult to follow and the voice gets very hard and looses all subtlty.

At first I suspected that I just had the crossover point to low, I have blindly followed the zaph audio crossover design which puts the point at about 2.4 Khz, looking at the numbers myself it seems that around 3.1 would be the sweet spot for this combo. As I intend to go active with these in the near future that will allow me to test various crossover points easily, instead of designing various iterations of the passive crossover, especially given the price of components.

Despite thinking that the crossover point is wrong I suspect other things at play. Stuffing and internal reflections/standing waves. The cabinets are stuffed 100% with wool insulation at the moment. I've got my head around scanspeaks advanced driver Excel app and I've realised that the right amount of stuffing for the ideal system q of .707 is about 50% stuffing and slightly less dense than I have in currently. That's not the end of the story though. I have noticed that the hardness and congestion is far worse on axis, when I am sitting else where in the room it's markedly better. I think that the cone maybe acting as a window for what ever modes of reflection and standing waves are inside the cabinet. Unfortunately I have no access to measurement equipment so it's a bit trail and error.

I intend to get active first which will help me sort the crossover point, then adjust the stuffing and add some diffraction to the bottom, top and behind the driver.

I'd appreciate some input and would really appreciate the loan of a Mic and RTA when the time comes if anyone has any kit? Pretty please :)

We're definately not far away with these, they mostly sound absolutely stunning, just guilding I think rather than any massive design floor.
 
Can you post your XO details so I can see or model what the transfer characteristics might be?

James
 
Looking at JK's LP filter for his ZRT design, it does not look like he has incorporated a shelving circuit for the rising response of the 18W8531 above 1kHz. Given that this is far less pronounced off-axis, this is what I'd suspect first. You could try small incremental increases in the inductor value. It's less likely to be internal reflections coming back out the cone. You could address this remote possibility by putting some loose stuffing behind the woofer.

Without measurement data, it's going to be a bit tricky to solve this problem. But I would make one change at a time, and not try to change both HP and LP filters simultaneously.

James
 


advertisement


Back
Top