RIAA Ideas From Thorsten

0 Members and 1 Guest are viewing this topic. Read 7826 times.

John Chapman

  • Facilitator
  • Posts: 825
    • http://www.bentaudio.com
RIAA Ideas From Thorsten
« on: 21 Jun 2004, 03:21 pm »
Hello!

The 10K RIAA has generated some interest but it is still for advanced diy guys as no 'canned' schematics or kits have cropped up yet. Thorsten (who has been the force behind many of the S&B products getting made) was kind enough to send along some ideas and allow me to post them here.

I will be at least a few months before I get to it but hope to do a stage with the RIAA in the fall. I expect once we get a bit further along a handfull of schemtics will appear and hopefully someone (maybe Brian at diyhifi?) will do a kit.

Here is the info from Thorsten:

---------------------------

I have a Phonostage on the Drawing Board that will use 7308(ECC88) (2pcs).

First stage self biased with a 150R Cathode resistor and operated with a CCS
load (will have to try at few) at around 5mA/50V.

This drives an UNLOADED 10K RIAA which is DIRECT COUPLED to the Anode of the
input Valve.

Coupling to the output stage via 10nF...100nF (Silver Mica) & 3M3 Grid Leak
Resistor with the output Valve grid leak biased at 15mA/65V with CCS Load to
give a nice low output impedance to drive a TX-102 via a 2u2 or larger
capacitor (Zout < 3k).

The Supply Voltage should be around 100...120V/40mA and 6V DC for the
heaters (Yes, 6V NOT 6.3V).

Anyway, I have that in mind for a simple to make Phono.

The +100..120V/40mA Supply and 6V/0.7A can be generated from an external 12V
DC supply (2 - 3A will do) with a simple 7806 regulator for the heaters and
using Steve Bench's LVPS shown here:

http://members.aol.com/sbench101/#Power

Another option would be to daisychain 5 pcs fairly cheap 1 Watt DC-DC
converters 5V or 12V to +12V/-12V (giving max 42mA load current). Suitable
6V or 12V switched mode or linear DC supplies (6V >2A if the DC/DC
converters are used) are generically available too.

====================================================================

Alternatively the 10K RIAA can be looped straight into the Euridice
Phonostage, available here:

http://img17.imageshack.us/img17/5087/euridicephonostage417A.jpg

Simply use a 2.2uF or larger value coupling cap from the first stage to the
10K RIAA and terminate the RIAA with 10K, the Network between the stages of
73k, 6k81,47n, 16n,221k is removed.

As 417A's are becoming rare and expensive, they can be substituted directly
(rewire sockets obviously) with the E180F/6688 (russian type 6Z9P/6Z9PE)
triode connected (G2 & G3 directly to anode) which is available from a range
of sources and even Siemens and TFK versions remain affordable (for how much
longer is another question).

I think you should also be able to "loop" the 10K RIAA into the "Groove
Thang" from Bottlehead, here with a 10K load and a 2.2uF coupling cap to the
input of the RIAA, equally with 2.2uF coupling and 10K load it can be used
in the Arthur Loesch Preamp either after the first but likely better after
the second (6GK5) stage.

Finally, you should be able to wire the 10K RIAA directly into Gary Pimms
Phonostage (with or without Output Transformer):

http://home.pacifier.com/~gpimm/phono.htm

Simply remove R4, R5, R6, C1 and C2, wire the input of the RIAA to the
(Mu-Follower) output of the CCS with a 10K Series Resistor, output goes
straight to C3, ground to ground.

Basically, as long as the first stage (or stage preceeding the EQ) of a
conventional passive equalised RIAA stage runs something a little meaty
(6GK5/6ER5/6ES5, 417A/5842 et al, 6922/7308/ECC88 et al) with at least 3 -
5mA current the 10K RIAA can be looped in, instead of the original RIAA on
the condition that it is coupled in with at least 2.2uF coupling capacitance
and loaded with 10K, as in this case the RIAA circuit is insensitive to
variations of the source impedance.

If the stage preceeding the EQ can be made to have pretty accuratly 10K
output impedance the RIAA can be connected directly (no coupling C required)
and left unloaded with the second stage in gridleak biasing using small
value Teflon or Silver/Mica coupling Cap's. This is preferred, but usually
hard to achieve without a complete re-design. By using a Valve in the first
stage that has a Anode impedance (at the chosen operating Point) of 5K or
less it is usually possible to add an unbypassed cathode resistor of
suitable value to raise the effective Anode impedance to 10K, but this is
really strictly for experienced people.

I hope that gives some ideas.

------------------------------------------------------

Thanks to Thorsten for sending this along!

andrew

  • Jr. Member
  • Posts: 44
RIAA Ideas From Thorsten
« Reply #1 on: 26 Aug 2004, 07:54 am »
Just a thought, how about an ECC99?

   cheers...

andrew

  • Jr. Member
  • Posts: 44
RIAA Ideas From Thorsten
« Reply #2 on: 22 Sep 2004, 05:10 pm »
Finally, found time to model Thorsten's suggested circuit in SPICE.

Quote
First stage self biased with a 150R Cathode resistor and operated with a CCS load (will have to try at few) at around 5mA/50V.

This drives an UNLOADED 10K RIAA which is DIRECT COUPLED to the Anode of the input Valve.

Coupling to the output stage via 10nF...100nF (Silver Mica) & 3M3 Grid Leak Resistor with the output Valve grid leak biased at 15mA/65V with CCS Load to give a nice low output impedance to drive a TX-102 via a 2u2 or larger capacitor (Zout < 3k).


So far results look good, further analysis is required. At this point, however, I'm a bit puzzled about the 10-100nF silver mica. Using a standard 10K RIAA passive circuit ion place of the RIAA-in-a-can, a 10nF appears to give a bass lift below 150Hz, 3db up and at around 50Hz. Does the RIAA-in-a-can compensate for this? What am I missing?

   many thanks,