advertisement


Snell K restoration

L

limono

pfm Member
I brought home a nice pair of early Snell K I with vifa drivers; M21WG paper woofers and D19TD. Suspension on woofer is crackling but I was able to play the pair on my 300b and sort of "asses" the sound. I like the woofers part. It has a nice tone and I really like the paper's cone feel and already ordered proper surrounds. The cheapest vifa poly dome plays clean although I'm pretty sure manufactured in 83' drivers lost all of it's ferrofluid in the gap (maybe it's good since Audio Note version is made without it for better transients). Sure,I couldn't crank up the speakers but even without it, the highs are evidently "generic" sounding , flat and "splashy". I would really like to change the tweeters but it's a sticky bit if it comes to Snell . I could buy new more or less exact vifa D19TD-05-08 which might have been improved over the years and are better sounding than 30 years ago for lunch money (~$30 a pair) I could try to improve using newer DX19TD ($60 matched pair ) which seem really more suitable to cross to 8" with it's lower 800hz fs (compared to 1700 hz d19td) I have a pair of Hiquphon OWI tweeters which I promptly popped in. They work OK (much , much more refined driver ) but are little shy at 87dB compared to 89dB vifa . I could maybe ask PQ for AN K tweeters but price would be high (south of $200 a pair probably).
Everybody knows, that the problem is matching and individual crossovers in Snell speakers. It was more relevant 30 years ago when drivers were not very consistent (my OWIs are 0.5dB apart ) . Consequently the crossover differs one from another slightly. HF filter caps measure 7.6uF and 8.9 uF and curiously one is made with parallel electrolytic caps and other with one mylar and one el cap. (same with cap across the woofer ) I don't think values drifted and ESR is rather low at 0.05 Ohm. I do not have much against el caps in crossover (Audio note use all el caps in their speakers up to $10k after all ) but would probably want consistent caps in both speakers .
My question would be. Does anybody have basic schematic of Snell (AN) K crossover with "theoretical" values around which they "tweaked " to get "master curve" right . Anybody replaced tweeter with something better and got around tuning? I don't have ARTa just yet but maybe it's a good excuse to buy calibrated mic and software and start messing around :) The surrounds arrive probably in 2 weeks so I have time to "ramble around:)
Tia, L
 
Thank You
I will have to think long and hard if I want to spend $200 for basic Vifa D19TD tweeters. Maybe I will once I refoam the woofers and like the speakers enough..
Regards, L
 
Contact Guy Sergeant, who posts here as Pure Sound. Quite an expert on these, and has some spare parts I think.
 
You'd have to be crazy to spend that on D19's.

There are other 94mm faceplate tweeters that would be far superior.

Sounds like you're well tooled up; why not just measure the new tweeters in situ and adjust the XO components to suit?
 
Rusty
Reasonable point > once I change surrounds (Audiofriends from Netherlands) and break-in the drivers and maybe match suitable capacitors and change zip-cord wire inside I'll think of rolling some new tweeters .
I exchanged emails with Martin at Audio Note service and he said that it's not given that the tweeters lost ferrofluid and they measured some equally old ones and they were fine , and holding up the original specs. Will report back in a couple of weeks :)
Thanks, and Regards, L
 
I've got the surrounds yesterday and proceeded to change them immediately (kind of strange since all my projects takes years of deliberating and decades to half finished state in which they remain gathering dust :) The surrounds appear to be slightly wider than original ones but it's not a deal breaker.I learned that AN have the surrounds in stock too. They subjectively work properly (I still didn't buy this woofer tester gizmo) and speakers sound actually very pleasant in my small bedroom. I think I will keep the tweeters for a time being enjoying them for what they are and only try to source capacitors for the crossovers. I don't think the one I have there are shot just they are mismatched with one speaker having mix of mylar and electrolytic caps and the other only electrolytic ones. Maybe try some silver ribbon wire later on. All and all for the speaker which according to modern speaker design fashion has all the bad choices regarding cabinet and drivers used it sounds splendid.
Rgrds, L
 
The unmatched capacitors are not a mistake. Each crossover was adjusted in production such that the pair of drivers exactly matched the performance of a reference pair of drivers & crossover on the day of production. It was not uncommon for the crossover in one speaker to be slightly different to the other. However the pair matching was second to none. Now that you've had to change the surrounds all bets are off though. The cabinet shape (wide & shallow) was also deliberate.

There are few if any modern speakers of a similar size that are better than fully functioning Type K's imho.
 
Yes , I know the parameters changed but it's unavoidable. Even if I'd buy those expensive replacements that are listed on Snell site they are likely different type of drivers really.
I know about the matching process , and that components value in crossover can differ slightly (Martin from AN said they were matched incredibly tight) .What I meant was that one tweeter filter uses parallel electrolytic caps and other mix of electrolytic and mylar. If I read correctly Peter Snell was always on a verge of insolvency and was using the parts he had on hands. The drivers itself were probably the cheapest decent quality ones he could use. Parts Express still sells vifa d19 for $16 . One has to wonder what really happened with "progress of speaker design" Anyway , PQ was remarking few times that he will set up somebody with a jig to tune those old Snells and breath a fresh life into them alas it never happened. When I will have opportunity I will measure and see how do they match. Both sound very close to each other on few mono tracks but I wouldn't certainly hear 2-3 db difference if it was there. Then , they still need 100-200 hrs of break in of new surrounds. So far I like them. I'd risk saying that in that room I like them as much as LV Avatar and Proac Response 2.
 
limono said:
If I read correctly Peter Snell was always on a verge of insolvency and was using the parts he had on hands. The drivers itself were probably the cheapest decent quality ones he could use.
Click to expand...

I think you do him a disservice there, because...

limono said:
So far I like them. I'd risk saying that in that room I like them as much as LV Avatar and Proac Response 2.
Click to expand...

I think he knew what he was doing.
 
Surely he knew what he was doing but the fact that he was struggling financially and was using whatever parts he could get remains , no disservice meant .
Rgrds, L
 
Even for the K's they'd buy large quantities of the bass mid driver and the tweeter and pre-select them before using the 'good' ones in production. The others would be sent back to Vifa and sold elsewhere.
 
So Guy , how one gets "fully functioning type K " speaker nowadays? (Except for buying $$$ AN K version ;)
 
. One has to wonder what really happened with "progress of speaker design" Anyway , 2.[/QUOTE]

Replace those nasty polyamide domes with something better and you'll soon see!
 
RustyB said:
. One has to wonder what really happened with "progress of speaker design" Anyway , 2.
Click to expand...


Replace those nasty polyamide domes with something better and you'll soon see![/QUOTE]

Rusty
I'd really like to keep this speaker Snell . I'd really would like to keep that vigorous midrange and lively character of a speaker pretty much intact without annoying "splash" on top end. I'm not sure it's doable with OWI tweeter. I sorta started that thread with hope that somebody before me already did that hard work and have ready to go solutions. There must be plenty of model K around . I do not believe that there is one speaker which survived with original suspension intact , so it's not an unlikely scenario. Having said that Seas Prestige 19ttf looks like reasonable candidate. I think it's and updated version of old d19dt and would guess that's what AN is using in their version (custom without ferrofluid ).
 
The OW1's 87 dB sensitivity is usually an advantage, typically corresponding with system sensitivity and removing the need for a resistor or L-pad.

However, seems like the K's are too sensitive and would need the OW2. Ka-ching!

The Seas would be a sensible option.
 
Since I have more time on my hand these days I went ahead and put those OWI's instead of vifa tweeters. I shorted 1.6 ohm resistors in series and put 12 Ohm across the tweeters since it's OWI is not damped by ferrofluid.
I took measurements of both crossovers elements and they are as follow.
Speaker A .
LF filter 2.3mH and 12uF
HF filter 8.8uF 1.6 Ohm variable resistor and 0.31mH inductor
Speaker B
LF filter 2.42mH and 11.4uF
HF filter 7.5 uF , 1.7 Ohm 0.58mH
Yes , it's apparent that OWI is much better tweeter and although my implementation is crude and is not result of any measurements the results are liveable, more so than original tweeters although with additional 1 Ohm at the tweeter terminal it was much more listenable . I don't mind if the frequency response will be tilted down towards HF about 2-3 dB.
 
You don't have to worry about the damping.

What you do have to worry about is the difference in the lower end response. Because the OW1 has a chamber behind the dome, it has a different roll-off, earlier and more gradual, than an equivalent tweeter without a chamber.

Compensating for that can be done by ear, increasing the value of the series cap until it sounds right.
 
limono said:
So Guy , how one gets "fully functioning type K " speaker nowadays? (Except for buying $$$ AN K version ;)
Click to expand...

That's tricky and I don't think that AN's type K is even trying to be like the Snell one any more.

A year or so back I had 3 or 4 pairs here so had a look at and measured the values of the crossover components in all of them. One of the speakers was actually functioning and the bass driver, although the surround was fragile, did still function as it should so I got my Neutrik chart recorder & calibrated Mic out & did a slow sweep several times & printed off a few reference charts.

At around the same time I acquired a load of Heybrook Trio cabinets, a box of similar volume although slightly deeper & slightly narrower. I began some experiments with new production (by Scanspeak) Vifa M21 WG-09-08 drivers. I have alot of these as I supply them to HB1 owners who either don't want to go down the re-foaming route or who have tried it & not liked the results). I already had a quantity of unused D19's. To begin with I had some air cored inductors made basically to the average value of those in the various samples I had together, ditto the HF inductor. I did the same with the caps, measuring what the actual values were & deriving an average for each position. I decided to go straight to film caps for these values. I also looked at the effect of the variable resistor in the HF circuit and what value it typically had to be set for. Putting the new parts together it was possible to get an FR close to but not quite as even through the midrange as Snell appeared to have managed. That may have been because their pre-selection of drivers rejected the worst case bass units & tweeters and I wasn't doing that, but it may just have been that I did need to start adjusting values slightly to see if their exact transition could be replicated. The other interesting thing about the K appears to be how well it measures off axis particularly at HF where this tweeter is usually much maligned. So, I did get a working set of speakers going that were fairly close, I could certainly live with them. I do have another set of Snell K's here with new M21's fitted, not ideal, but a sort of reference and I recently bought some samples of the Scanspeak Classic D2008/8512 to try instead of the D19. Its a project I must return to. The crossover appears to occur at about 2100 Hz which is somewhat lower than might be expected and quite some way from where JPW & Heybrook used to cross the M21 in their designs AP2 & HB1. The thing with all such projects is finding the time to do it properly. I suspect the modern M21 bass drivers are probably made more consistently than perhaps they were being made 25 years ago. Much more to do.
 
Sounds like fun!

Interested in your comments re the HB1's. Were they not happy with the bass, the mids, or both?

Sealed box drivers should be re-foamed with thin, high-compliance surrounds so as to maintain the original Fs as closely as possible.

The HB1's are a bit of a knife-edge design, any reduction in bass could I imagine easily render them unlistenable.
 
You must log in or register to reply here.


advertisement


Back
Top