JFET SRPP RIAA Preamp - Page 3 - diyAudio
Go Back   Home > Forums > Source & Line > Analogue Source

Analogue Source Turntables, Tonearms, Cartridges, Phono Stages, Tuners, Tape Recorders, etc.

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 21st October 2005, 08:36 AM   #21
AndrewT is offline AndrewT  Scotland
diyAudio Member
 
Join Date: Jul 2004
Location: Scottish Borders
Hi Wrenchclone,
Why do you take the output from inside the GS resistor loop i.e. from above R115? Rather than from the bottom of R115?
At the input end you have shown R106 inside the loop and R205 outside the loop. Which is correct?
__________________
regards Andrew T.
  Reply With Quote
Old 21st October 2005, 10:56 AM   #22
Previously: Kuei Yang Wang
 
Join Date: Nov 2002
Konnichiwa,

Quote:
Originally posted by wrenchone
The simulation results inducate that one should not be using an SRPP as the front end of a passively equalized RIAA amp.
Coulda toldcha....

Here is what you WANT to do.

1) Set up the input as cascode, J-Fet to BJT.
2) Degenerate the J-Fet's source sufficiently that with 50mV RMS input the current shifted is linear.
3) Add the RIAA Network in parallel to the CCS used as DC load for the Cascode, the entire RIAA network is of the shunt type.

If you have enough gain in the cascode and enough supply voltage all you need after this circuit is a buffer, so it could be made "single-stage".

If you go 2-Stage anyway you can probably omit the CCS load and just use the RIAA Networks Shunt resistor (the 3180uS) to supply the supply current.

One might even conceive a "split EQ" version of this, with the 3180/318uS EQ in the Drain connection of the first stage and the 75uS in the second stage.

Using 2SK170BL with cascode gives around 10mA/V transconductance once degenerated, at a current in the 2-5mA range. If we select for around 4mA and use 3K3 as Load we need a 330R in series with around 1uF. The resultant gain @ 1KHz would be around 10db. We need a suitable shunt capacitor for the 75uS EQ, I'd expect around 330nF should work, best shunt this through some simulator, calculating networks by hand according to lipshitz is too hard work for me on a friday.

Then add a second stage with a higher current (and Gm) selection (say 10mA @ 30mA/V) and a 1K load. Done.

We did a somewhat "inbetween" this shunt EQ with cascode and the traditional EQ version of the l'Pacific. PSU via PP3 NIMH Cells X 3, Cascode BJT's from the first battery at around 8V, main supply around 24V. Worked great... I suspect the Shunt EQ would have been even better.

Sayonara
  Reply With Quote
Old 22nd October 2005, 03:32 AM   #23
diyAudio Member
 
wrenchone's Avatar
 
Join Date: Feb 2004
Location: Silicon Valley
KYW - been there, done that (well, sorta). See my thread called "Open Loop Follies". The RIAA preamp I just replaced had a cascoded 2SK170GR front end with a current source loaded source follower (circuit (1) shown earlier in this thread). I used a JFET cascode, though. The circuits are in the cited thread. I don't really have any problem with a simple common source amp at the input. The distortion is very small (<0.01%) for low level operation. My big concern was with the second stage, which had about 0.4% distortion for about 1V output using the same type of gain cell. The distortion is overwhelmingly 2nd order, but it irks me none the less. This distortion level seems unavoidable wih a simple common source open loop amp, cascoded or not. Using the SRPP was an attempt to get around the high level distortion at the second stage. I also wanted to see what it would take to get one to work, as an earlier attempt at an SRPP preamp using JFETs was unsuccessful.

For a passive equalized RIAA amp, I like to split the gain in two pieces, usually with a gain of about 40 on the first stage and 30 for the second stage. This gives me aroound 40 dB gain at 1kHz, which is right about what my system wants to see. I wouldnt feel too comfortable putting all the gain in the first stage.

I'm also fairly adamant these days about using readily available devices (in the US, at least) in my recent circuits, especially after the fulminations from some pundits that insist that the only way to get decent sound is to use a 2SK170 or the like. I don't have anything against these devices (I have them, like them, and have used them), but I don't like absoute statements like that, especially without any real supporting data (and, may I add, contrary to my experience).

The next pass will most likely have a common source amp with cascoding (JFET) and degeneration driving a current souce loaded source follower. With even the fairly lowly PN4393, you can get a follower with ~50 ohms output impedance or less. This helps to get around any effects of output impedance on RIAA accuracy. I will also sum a current source into the drain of the gain FET (circuit (3) shown earlier in this thread). Adding the current source allows one to use high IDSS devices like the PN4393 or J110 without excessive degeneration, and still be able to adjust for optimum gain and voltage centering at the output. Since the gain FET runs with more bias current, the distortion is lower, and the cascoding keeps the FET from running too hot.

I'm thinking of still using an SRPP for the second stage with an optimized capacitively coupled load resistor, followed by another current source loaded source follower.

These topologies will allow me to build the preamp with JFETs I can order straight from Mouser and/or Electronic Goldmine.

For the correspondent that was asking about the rationale behind the output connection on the circuit, do a seach for SRPP on Google. You'll get a lot of information about the theory behind the SRPP (the Tube Cad article is especially good), which is a poor man's push-pull with gain. Most of the discussion will focus on tube applications. If you shift the output connection to the drain of the bottom FET, you get a common source amp with current source loading, an entirely different beast (I'm working on one of those with global feedback for a line amp application - yet another thread).
  Reply With Quote
Old 14th December 2005, 04:28 AM   #24
diyAudio Member
 
wrenchone's Avatar
 
Join Date: Feb 2004
Location: Silicon Valley
Well, I originally set out to do a preamp with a cascoded JFET common cathode for first stage and an SRPP for the second stage (the boards are (9/10 stuffed), but some recent simulation results made me lay these boards aside to try something different.

Attached is a schematic showing a MOSFET common source amplifier with cascode and a current source to bias up the MOSFET without all the bias current going through the load resistor. I simulated one of these using a Zetex ZVN3306 (Orcad had the device model) for active device, and high level distortion (~2V out into a 20k load) was around 0.05%, overwhelmingly 2nd harmonic. This was without the benefit of an output buffer. Needless to say, I'm trying this circuit out for the second stage of my RIAA preamp. Once I have the boards up and running, I'll start a new thread. I'm using the Supertex VN0106N3 instead of the Zetex devices, as I have loads of them in my parts bin.
Attached Images
File Type: gif comp1.gif (6.4 KB, 2860 views)
  Reply With Quote
Old 14th December 2005, 07:15 AM   #25
AndrewT is offline AndrewT  Scotland
diyAudio Member
 
Join Date: Jul 2004
Location: Scottish Borders
Hi,
what are Q2 & Q4 doing for you? They appear to be outside the signal route, just that 22k from the signal input.
Is it some kind of non adjustable bias for the amplifying stage?
__________________
regards Andrew T.
  Reply With Quote
Old 14th December 2005, 05:32 PM   #26
diyAudio Member
 
wrenchone's Avatar
 
Join Date: Feb 2004
Location: Silicon Valley
Funny you shoud ask - I screwed up, and there's an important piece missing out of the schematic I posted. There needs to be a second current source. The attached schemaitic is the one I used to do my simulation, and it has all the pieces...

M1 is the gain FET. M2 is used to set the overall bias current in M1. The ratios of R3 and R6 set the ratio of bias currents between M1 and M2. Ideally, one would use a current source in place of R10, and M1 and M2 would be matched and thermally coupled.

J1 is used to pin the drain voltage of M1. This provides the benefits of cascode operation, and also reduces the drain voltage of M1, reducing power dissipation and allowing use of a TO-92 device with relatively high bias current. The small device will need a lower bias current to get it into a "sweet" operating region. The small devices also have lower input capacitance.

Current source I1 is used to maintain the bias current in M1 without having it all flow through gain resistor R1. This allows one to have a greater range of choices for gain resistor R1 and still be able to center the output voltage near 1/2 the supply voltage for maximum symmetric output swing.

Since most of the bias current for M1 flows through current source I1, the power dissipation in J1 is relatively small. Also, since the voltage across I1 is pretty much constant, one can place a resistor in series with it to limit the power dissipation in whatever real-world device is actually used.

I suspect that most of the benefits of this topology are derived from operating M1 at a relatively high value of bias current, so that the signal current is a relatively small fraction of the bias current. The high bas current in M1 also places it in a more linear operating region.
Attached Images
File Type: gif mosamp.gif (7.4 KB, 2260 views)
  Reply With Quote
Old 14th December 2005, 08:30 PM   #27
diyAudio Member
 
wrenchone's Avatar
 
Join Date: Feb 2004
Location: Silicon Valley
Attached are the simulation results for the circuit in the previous post. Output level was 3V p-p into 33k at 10kHz excitation This output level was with 40mV excitation, so overall gain was 37.5 (31.5 dB).
Attached Images
File Type: gif results.gif (6.6 KB, 1937 views)
  Reply With Quote
Old 15th August 2008, 10:13 PM   #28
lineup is offline lineup  Sweden
diyAudio Member
 
lineup's Avatar
 
Join Date: Dec 2005
Location: the north
Nice work, wrenchone

Will be a great help for my own project with
JFET SRPP
(Shunt Regulated Push Pull)

Regards Lineup
__________________
lineup
  Reply With Quote
Old 30th September 2008, 03:07 AM   #29
diyAudio Member
 
wrenchone's Avatar
 
Join Date: Feb 2004
Location: Silicon Valley
I've done a little more looking at the JFET SRPP preamp circuit, which caused me to revive this thread. A possible solution to the variable gain caused by the variable load presented to the first stage SRPP by the passive equalization network is to use a capacitively coupled fixed load, followed by a buffer.

For some reason, I didn't try that approach when this thread was more alive. Some recent simulations showed that the approach had promise, so I hacked an older board into shape using the circuit shown below. Simulations seemed to indicate that one could get by with a DC coupled fixed load on the SRPP stages to define a fixed gain. In practice, a cacacitively coupled load is needed so as not to disturb the DC balance of the SRPP stages. The SRPP stages used PN4303 JFETs (National) matched for VGS at 2 mA drain current. The output buffers are PN4393 loosely selected for drain current of ~ 10 mA.

This circuit will go into my living room preamp test bed for listening tests. If I like it, it may go to Burning Amp with me this year.
Attached Images
File Type: gif jfet_srpp_riaa2.gif (58.9 KB, 1541 views)
  Reply With Quote
Old 30th September 2008, 03:12 AM   #30
diyAudio Member
 
wrenchone's Avatar
 
Join Date: Feb 2004
Location: Silicon Valley
The circuit I just showed was the old circuit - let me see if I can select the new one instead. All the comments in the last post apply - this one is the real deal. What is ahown is two channels of RIAA with a common regulated supply. I'd like to point out that the gain is optimized for MM or high output MC cartridges.
Attached Images
File Type: gif fet_srpp2.gif (63.5 KB, 1575 views)
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
jfet srpp preamp yves Solid State 12 23rd June 2012 05:50 PM
Riaa Pacific Jfet maxpou Analogue Source 23 15th April 2006 05:51 PM
All JFET open-loop RIAA pre-amp Mr Evil Analogue Source 29 12th April 2005 02:26 PM
Help wanted on SRPP preamp jtcc1015 Tubes / Valves 8 29th November 2004 04:44 PM
Preamp with 6sn7 & 6sl7 SRPP ??? Gregouille Tubes / Valves 4 18th June 2002 10:41 PM


New To Site? Need Help?

All times are GMT. The time now is 11:54 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2