advertisement


PFM Special - a simpler and more affordable DIY loudspeaker design

Nice to get there, really interesting thread. They are never going to look as "neat" as a single box speaker.

I think there is probably some development to come in the design but it allows plenty of scope for the DIYer to make the visuals to their own taste.

This is a design I might be tempted into if I had some time:( as I like the idea of the open baffle.

Welll done James!

Cheers,
 
"... pretty costs a lot of time and money."

Indeed it does. In hindsight, I should have finished the bass baffle in Sapele so that at least it looked uniform from the front. Oh well, it's the sonic results that really matter in this case. The whole raison d'etre is simple construction, cheap money and mind-blowing performance. For those wanting more of a challenge and who have deeper pockets, the Ergo series beckon.

I've now spent half the morning listening to the PFM-Specials and can add the following comments.

(1) There is clarity in spades. I'm hearing some things in more detail than the current reference E-IIIs, notably in the mid-band. I think it's a combination of a faster driver and one that is not boxed up. Rim shots and the snap of percussion are certainly more vivid. But this adds a bit of a hard edge to bright recordings, but not offensively so in my opinion. The lower treble quality reminds me of a time I demo'd a pair of Intros, but perhaps not quite as ragged. After all, I have been accustomed to the fluidity and organicity of the best 19mm dome tweeter in the world*.

(2) How they sound depends a lot on what's behind them. There is a fair bit of reflection that happens on the backing wall (not surprisingly given a dipole design), and pulling the blinds closed the sound in a little, making it sound more intimate and more direct. As it is, I have a bookcase between them, but they are around 1m pround of the backing wall - so I'm not sure how consequential that is. I can't be faffed moving the bookcase to find out. What is certain is that these are free-standing designs. The will sound crap in more ways than one when placed hard against the backing wall.

(3) The Vifa 10-inch woofer is amazingly good for the money. It does not go as deep as my current Scan-speak 25W-equipped E-IIIs, but then again it's barely run-in. It does not lack impact to any extent, and is more than capable playing bass tunes that are so easy to follow. More importantly, I can't seem to sense any crossover artefacts. Mind you, I haven't 'measured' them yet - preferring to let my ears judge. The XO as built from careful measurement and modelling seems to be delivering the musical goods.

(4) The rhythmic and timing quality is about as spot on as I can imagine. Certainly on par with the E-IVs and I dare say a little bit more tightly defined than the E-IIIs. The way notes start and decay sounds entirely natural to my ears - very reminiscent of the first time I discovered Naim and realised music can be rather startling. My only reservation is the slight edge that I can hear on, say, John Fogerty's Blue Moon Swamp but not on Jack Johnson's Bushfire Fairytales. I'll give it a week or so of running in before I tweak the XO. But I can honestly say I'm absolutely chuffed with them, and if not for the Ergo series whose aesthetics are much more in keeping with my partner's preferences, I'd be happy to live with them.

(5) The dipole mid does paint a bigger sonic picture, more spacious if you like, but the absolute size of the musical subject is still preserved. I'm not getting mouths that are ten feet wide or a guitar that fills the room. Surprisingly, the localisation of the various instruments is still apparent, but they do play at a bigger venue. I think it's a fair amount of natural reverb going on with the design.

Those that are contemplating the PFM-Special need to know that the most difficult part is routing the driver holes and rebates. If you can do that (or get someone to do that for you), then the rest is a piece of cake. The whole can be built from MDF (except for the upper baffle - where I recommend something stiffer) and very cheaply too. The drivers cost USD318 in for all six (plus shipping), and the XO bits come in at less than USD100. I don't think you could buy anything new or second hand for the money that will touch them. But then, I am biased aren't I?

James

* Hiquphon OW-1
 
James,

As always, and more so.

The E2..E5 are natural progressions of each other. It's honestly exciting to see you now dare try different designs like this one and do not only go for proven designs so you should be doubly congratulated.

This and of course being easier to build, cheaper, sealed base boxes which sound at least decent :)

Omer.
 
"...honestly exciting to see you now dare try different designs like this one..."

Thanks, Omer. I was getting a little jaded with the standard 3-way approach but saw no need for a 4-way system. You could say that the dipole was the natural progression. What next?

Today, I couldn't resist but take a test sweep of the frequency response of the system. As I suspected, the region above 5kHz was a little hot. About 2-3dB too hot, so I've taken the simple step of changing the tweeter series resistor. It's certainly more mellow and easier on the ear, but I wonder if I have gone a touch too mellow. Just as well it's really easy to swap the resistor out for a different value in the drawer full I have. Here is the XO chamber complete with four inductors (two air core, two iron core), five capacitors, and seven resistors. Not a bad count for a second order three way system. Note I have deliberately left the excess component legs just in case they need to be recycled. Once I'm firm on the design, they will be snipped for a tidier look.

3040049-md.jpg


And here is how the right one looks in daylight from my listening position.

3040048-lg.jpg


James
 
I'm watching this evolve with even more interest than your prior creations, considering that I'll eventually build them myselves. I'm just wondering how they'll measure up to the Royd RR3.

-=> Mike Hanson <=-
 
Mike,

"I'm just wondering how they'll measure up to the Royd RR3."

No doubt that will be the $64K question for you. I can't comment as I have never heard the Royds but I think you'll be in for a nice surprise given the very small investment. Plus, there is something extra special about an item that you build with your very own hands.

I'll be publishing the XO schematic as well as the parts list and assembly plans within the next week or two, or as soon as I can persuade Steve to fire up his CAD system.

Then I'll have to seriously rethink the aesthetic and possibly consider a new cabinet design. But not before the E-Vs are completed, tuned, run-in and declared my finest work. So much to do ...

James

P/S: The Sapele upper baffle will darken over time and take on a colour and shade not dissimilar to the marine ply. The CD shelf in the background gives some clues to the final colour.
 
Low Pass filter for Woofer

3043174-lg.jpg


Band Pass Filter for Midrange

3043175-lg.jpg


High Pass Filter for Tweeter

3043176-lg.jpg


These are connected in parallel and where possible, the inductor gauge should be chosen to provide the stated DCR values. If this is not possible, a lower DCR inductor should be specified to which resistance may be added to bring the total DCR up to spec.

I'll post a parts list soon.

James
 
These bits are required to make up one channel, so for a stereo pair, you need to double the quantities.

Inductors

Madisound Air Core 19AWG 0.8mH (unwound to 0.75mH) $3.25
Madisound Air Core 19AWG 1.1mH $3.65
Madisound Iron Core 18AWG 7.0mH (unwound to 6.5mH) $7.80
madisound Iron Core 18AWG 9.0mH (unwound to 8.5mH) $9.45

Capacitors (Solen Chateauroux Polypropylene Fast Caps)

4.0uF $2.20
12uF $4.40
15uF $4.95
30uF $8.65
35uF $9.85

Resistors (Eagle Metal Oxide Non-Inductive 10W 5% - unless otherwise stated)

1.0 ohm $1.00
4.0 ohm $1.00
2.7 ohm $1.00
7.5 ohm x 2 $2.00
10 ohm $1.00
Wirewound Sandcast 25W 5% 8.2 ohm $0.60

James
 
The design started off with a 1-ohm series resistor on the tweeter. This yielded an exciting but somewhat fatiguing sound. Not so much exaggerated sibilance, but a more metallic edge than what I considered natural. The change I alluded to earlier was to increase this (based on remodelling with LspCAD) to 4.7 ohms, which took the level above 5kHz down by around 3dB. This made it sound much more laid back and easy to listen to. But somehow, it seemed to "lack air" - to quote Stevea, who came around last night for a quick listen and to take final dimensions for the CAD plans.

So, I changed the resistors to 3.3 ohms, which made it sound quite alive again, but I found it just a touch too edgey for my liking after Stevea left. With just the standard 3.7 ohm vs 4.0 ohm resistors left to try, I put in the latter and this is where I think the right balance is struck in my room. The tonal balance has both detail and warmth, vitality and grace, smoothness and articulation. I think I could be happy with this configuration. I will know for sure in a couple of weeks once it's all run in.

James
 
"So, I changed the resistors to 3.3 ohms, which made it sound quite alive again, but I found it just a touch too edgey for my liking after Stevea left. With just the standard 3.7 ohm vs 4.0 ohm resistors left to try, I put in the latter and this is where I think the right balance is struck in my room. The tonal balance has both detail and warmth, vitality and grace, smoothness and articulation. I think I could be happy with this configuration. I will know for sure in a couple of weeks once it's all run in."

From my own experience I can't stress enough that anyone considering building these (or any other speaker) should be well aware that there is plenty of scope for fine tuning the sound to suit the room and the ear. Perhaps the most rewarding part of building the E-III's was the time spent trying different levels for the mid and tweeter, dialing in the sound until it just hit the spot.

Mr Tibbs
 
" ... there is plenty of scope for fine tuning the sound to suit the room and the ear ... "

Absolutely. I'll even tell you which resistors to change and what values to experiment with. One of the privileges of complete control over the design to suit your particular taste and environment.

James
 
As far as speakers that "do air' are concerned I think I would always opt for speakers that didn't rather than those that did it badly as I think the latter might just add harshness instead of subtlety. The version I first heard certainly lacked air, while the version in operation when I left seemed to be doing its best but I was not totally convinced that it reminded my of the same CD I had been listening at home that afternoon – something I confirmed later. I think it was probably correct to drop the treble back a bit again – it was trying a bit too hard.

The PFM specials are now built as a solid model in my CAD system but before dimensioning them (the time consuming part), I need to check with James to make sure I have captured them as built as opposed to as designed (from which I built the model).

They were certainly very good value for money. They may not be at the level of the E series for critical and/or analytical listening but are an easy and enjoyable listen without any obvious flaws.

Steve
 
Steve,

"I think it was probably correct to drop the treble back a bit again – it was trying a bit too hard."

In my experience thus far, half an ohm on the tweeter circuit can make a significant subjective difference. What surprised me was that it sounded acceptable on 1-ohm initially, even if it was obviously bright. In fact, I found the change I made (from 4.7-ohms to 3.3-ohms) sounded by far the worst and the last change to 4-ohms perfectly spot on.

I played three CDs back-to-back last night with the volume dialled up pretty much and felt completely satisfied with the result. No earache, plenty of chilling, and lots of groove. I now only need to spike the cabinets to the floor to see if that adds significantly to the clarity of loudspeaker. As it stands, I'm happy with the voicing and I think it is in keeping with the rest of the range musically, if not aesthetically. I will think of a different construction method for those who want a more elegant product but with exactly the same musical and electrical character.

"They were certainly very good value for money. They may not be at the level of the E series for critical and/or analytical listening but are an easy and enjoyable listen without any obvious flaws."

Cheers for that, Steve. I think if I had used a Scan-speak or Hiquphon tweeter, the result would be very different. The Vifa tweeter is definitely the weak link in the design, but one of the three design criteria is affordability. Premium tweeters don't come cheap.

Then again, I would also be rather disappointed if the PFM-Special bettered the E-IIIs in every respect. Hierarchy must be maintained. As it is, the E-IVs have already upset the apple cart in terms of the fun factor. I'm hoping that the E-Vs will restore some balance.

As for the PFM-Special, whose other two design criteria are ease of build and superlative-for-the-money musical performance, I will leave it to the others to judge how closely I've met the objective.

James
 
Originally posted by James
.......I now only need to spike the cabinets to the floor to see if that adds significantly to the clarity of loudspeaker.....
James
I think I should bring my paving slabs up over the weekend......

Steve
 
Only if you promise to take them away after the trial. Which day of the three do you have in mind?

James

P/S: We should also for interest reinsert the E-IIIs back into the system and note the difference in presentation and groove.
 
I will most assuredly take them home afterwards.

I am free all day Monday, the other two days have various other scheduled activities to get in the way.

Steve
 


advertisement


Back
Top