j-FET / Triode Phono front-end

What people are working on at the moment
Post Reply
User avatar
Joe Roberts
User
Posts: 16
Joined: Fri Jan 20, 2012 10:02 pm

#61

Post by Joe Roberts »

1.5k is a step in the right direction in terms of driveability.

7k looks better yet. 0.1 caps anyone?

Silk Audio in Thailand makes 10k canned networks. We had a couple floating around...jc has em now, I think... don't know if he ever listened to them.
Laurence
I am not a user, I am a free man
Posts: 73
Joined: Wed Oct 05, 2011 6:21 pm

#62

Post by Laurence »

Are Silk the only ones making 10k these days? On a search only talk of S and B 10k is back in 2003. Was it a failure? Is the Silk 10k a failure?

Why haven't people like Dave and Tribute succeeded to make a good 10k EQ? You can't tell me they haven't attempted it. But nothing come to market?

Should they maybe try a compromise better than 600 ohm less than 10k?
User avatar
Joe Roberts
User
Posts: 16
Joined: Fri Jan 20, 2012 10:02 pm

#63

Post by Joe Roberts »

We never tried the Silk 10k. Might be very good. I know their 600 ohm unit measures well. They seem quite careful and quality oriented.

Doesn't Allnic make a 10k?

Dave Slagle will build you anything you want (and pay for). As mentioned, the 7k uses standard caps which make it easier to get the values needed. Slagle has the designs worked out for a number of impedances.

Silbatone uses Intact Audio inductors and we build our own modules with our choice of parts. In jc's new FET/tube design we are using pure silver/nickel core Slagle inductors we supplied 5N wire for. That might be interesting. Certainly esoteric.

I think the parts options are there...but so is the 600 ohm fixation, for some strange reason.

At least. I don't get it. I blame Tango for re-popularizing the old Pultec 600 ohm filter.
Andrew
Eternally single
Posts: 4206
Joined: Thu May 24, 2007 2:18 pm

#64

Post by Andrew »

the 7k uses standard caps which make it easier to get the values needed. Slagle has the designs worked out for a number of impedances.
Yes, that's how I understand it too, 7k is standard caps, so is 1k5 for that matter. I think Dave will wind you anything you want.
Last edited by Andrew on Sat Jan 21, 2012 10:38 am, edited 1 time in total.
Analogue, the lost world that lies between 0 and 1.
jc morrison
User
Posts: 56
Joined: Thu Jan 19, 2012 11:06 pm

#65

Post by jc morrison »

hey joe, we did try the silk 10k... it was ok. a little bass shy but ok. it was the first LCR i've used other than the old pultec module. i prototyped the C-100 with it. slagle's was just much much better.

it is harder to make the high impedance filters... more turns and more to go wrong. the higher the impedance, the more you invite unwanted crap slipping into the signal path. things get bigger and harder to shield, etc. dave makes a bunch of taps so one can make final adjustments. it may be controversial to say it here, but i never heard the difference of +/- .5dB. sue me. i tried to convince him to forget the taps... less of a chance for crap to go wrong, in my book. i guess there is a point of diminishing returns somewhere not far above 10K. it just becomes more antenna and parasitics than is fun to mess with.

i am a big fan of standard parts... there are several convenient points of that between 600 and 10K. no particular advantage of 250 ohm or 600 ohm and mainly a pain to drive for small signal tubes. no problem for solid state. for tubes, EL84's and 6V6's are ok, and that is what was used in the past... both in cathode followers. i think a cathode follower is better than a transformer if you already have a tranny in the circuit doing something else. too many trannies and the drama level gets out of hand (:

about buffers, i love them. my work in the effects business got me to relax and go with the flow. i have made gadgets with a dozen buffers in one function. thank goodness for cathode followers. they also make pretty wonderful output stages... i understand this isn't a particularly popular stand to take in the hifi world. i don't have a problem. they can be optimized like any other circuit.
jc morrison
User
Posts: 56
Joined: Thu Jan 19, 2012 11:06 pm

#66

Post by jc morrison »

i think it is a perfectly rational strategy to follow a high impedance stage with a buffer. if you have to drive a network afterwards, especially so. i know the diy world has it as a value to "simplify" above all in a quest for more "purity". of course i think this is BS, even though it comes from a good intention.

there is nothing simple about driving a network from a varying and indeterminate impedance (depending upon the cartridge Z and the stage output Z). of course it can be done! but is it necessary or simple. no. i say it is better to drive the network from a shorted or an open source impedance than something vaguely in between or worse: varying with signal or frequency.

of course a 600 ohm network could be driven from a current source too! not any harder than driving it from a 60 ohm impedance... 600K should be pretty good. a better ratio than 60:600 anyway. and any of the VHF pentodes you are considering can be optimized for that. but it can make for additional layout problems... it's a different kind of work, but absolutely works as well. the output off the network ABSOLUTELY should be buffered though, to keep the network load constant.

when filters are concerned, there is an increased need for constancy... this is easily provided by a buffer. one of the things they do best.
jc morrison
User
Posts: 56
Joined: Thu Jan 19, 2012 11:06 pm

#67

Post by jc morrison »

yeah, the 600 ohm thing is not significant in itself. and not trivial for most commercial receiving tubes. i couldn't afford the tango and really wanted it. i have had a few of the pultec networks which were used with older transcription systems (such as gates). gotham audio in nyc used to sell pultec to radio and TV stations, so the surplus showed up on canal street. i still have an EMT 930 that came from gotham. that preamp (which sucks) has a 600 ohm LCR network in it identical to the pultec in every way.

but, to make it work with tubes (it was germaniums in the pultec i yanked the thing from... i don't remember what they were but it might have been 2N404s?) took a while. i also tried to forgo the cathode follower. i went interstage (UTC, "you take a chance") but it sucked. by that time we had all built arthur loesch's preamp and it killed everything around. i wanted to beat it. very badly. i was sure i could by going lower impedance... and LCR. it took a long time to outdo that pre, and it was RC RIAA.

i don't think i will be able to agree that the problem is the resistors, although they definitely affect the sound more in simple circuits. but the caps are another thing. the biggest improvement came from isolating the poles of the RIAA. splitting the job into two stages, and isolating the one from the other. hence my point about constancy and filters. that made things much more solid and yet relaxed at the same time.

a really good way is a hybrid split: passive on the 75uS and active on the bottom (feedback). that can be made to sound almost indistinguishable from an LCR. always roll the top first, at a lower signal level, then the bottom. you have to keep an eye on the clicks and pops.

1.25K is much easier than 600 ohms for a 6SN7 or 12AU7 type tube... although not perfect. parallel both halves in a CF with a current source or choke as a load, of course much better. as i suggested earlier, a 6V6 run hot with a current source load will do a decent job with 600 ohms, and that is the old school way. it's just sensible.

if you are making or ordering your own, i'd strongly suggest something above that for tubes.
User avatar
Joe Roberts
User
Posts: 16
Joined: Fri Jan 20, 2012 10:02 pm

#68

Post by Joe Roberts »

The Pultec PQ-10 used a parallel 12AU7 CF to drive an RIAA network that was a small octal plug-in.

I could never find out the design Z or the schematic...anybody have this info?
Andrew
Eternally single
Posts: 4206
Joined: Thu May 24, 2007 2:18 pm

#69

Post by Andrew »

This one :?:
Attachments
Pultec PL10 or PQ10?
Pultec PL10 or PQ10?
Analogue, the lost world that lies between 0 and 1.
jc morrison
User
Posts: 56
Joined: Thu Jan 19, 2012 11:06 pm

#70

Post by jc morrison »

ahh, i wasn't allowed to post a link yet. but they have a tubed P-10 at vintageaudio_net. this is not what i remember. i plundered solid state units. the LCR is in the grey hammertone box with the cable. it IS the same. it's potted, with a switch on top. it plugs into an octal socket.

i have no idea what the pre was. wasn't interested.
User avatar
Joe Roberts
User
Posts: 16
Joined: Fri Jan 20, 2012 10:02 pm

#71

Post by Joe Roberts »

OK doing some research I find an argument that the plug in in the PC-10 is a damn RC network...not LCR.

Google "PULTEC PC-10 : La synthèse" on melaudia.

Really, driving 600 ohms with a parallel 12AU7 CF is lame.

A 2uF cap is not good for 600 ohms.

The can is too small for an LCR filter.

I think the Tango is from a Pultec design that was a filter intended to go between 600 ohm I/O amps, as I argued earlier.

Maybe jc's solid state modules were 600ohm LCR but really can all this filter junk fit into a 1x1x3" can?
jc morrison
User
Posts: 56
Joined: Thu Jan 19, 2012 11:06 pm

#72

Post by jc morrison »

the can i am talking about is 2-1/2 by 2 by 4 inches, give or take... has a cable leading out and a jones octal plug. they're heavy, and meant to be mounted through the turntable platter itself. there's a dial switch with the different slopes etched into it.... including bypass (flat).
User avatar
Joe Roberts
User
Posts: 16
Joined: Fri Jan 20, 2012 10:02 pm

#73

Post by Joe Roberts »

Hey on the topic of FET/tube stages, lest we forget the Berning TF-10:

davidberning com /support/tf10

This is the first one I know of.

(had to break that link up to get it posted, since I'm a newb here)
Laurence
I am not a user, I am a free man
Posts: 73
Joined: Wed Oct 05, 2011 6:21 pm

#74

Post by Laurence »

jc morrison wrote: a really good way is a hybrid split: passive on the 75uS and active on the bottom (feedback). that can be made to sound almost indistinguishable from an LCR. always roll the top first, at a lower signal level, then the bottom. you have to keep an eye on the clicks and pops.
Now then this sounds like what my budget can afford. I have been watching this thread thinking I just can't justify the cost of the Silk's and somewhere above someone said the Slagles are better so I wouldn't be happy buying the silk's.

I can manage the first section as an RC as Morgan Jones goes into plenty of design detail. But I don't have the first idea how to create an active second section. Can anyone point me in the right direction? I would prefer to use tubes than opamps.

My thoughts on the first stage would be more old fashioned, I would be using SUT then 12AX7 or 6SL7. I know you will all say use an ECC88 or so but no thank you I have my reasons. Noise is not the only issue in life. I happen to think that these chosen tubes work well after an SUT in a two stage RIAA EQ. Sukuma San also seems to think so.

The high transconductance favourites of recent times all have stability and consistency issues. The 12ax7 and 6sl7 are well known for their tonal qualities in many arenas. For me better tone is more significantly weighted in the decision process than noise avoidance. Make me a silent amp which is void of emotion and I will never listen to it.
Laurence
I am not a user, I am a free man
Posts: 73
Joined: Wed Oct 05, 2011 6:21 pm

#75

Post by Laurence »

I found this:

Image

R1 R2 and C1 are what I am looking for. Why this writer has put his HF EQ second is an error right?

So it seems with an OPAMP I could just copy that portion I want put it after my 12ax7 RC network and I have a hybrid right?

This guy seems to have the right idea but he is back to front and hasn't yet learned what the tone of tubes brings out in music. however before I critique I should build it as is. Might just be in the right ball park.
Post Reply