LCR RIAA


WORK STILL IN PROGRESS !!!
This project is in the making. I am placing info here as I am working, so it is by no means a ready or error-free design. This page is more of a means to provoke some comments from people who have traveled this path before me and can maybe share some knowledge.

 

Yep... A new preamp; again... Well, almost. This project is for a preamp, but not just for line-level use as I have always built. Recently I have joined the realm of analogue playback, yeah, I got a record player. I wouldn't be myself if I did not want to make my own RIAA correction for it. I'm new to RIAA, so I don't claim to know what I am talking about. Yet, even here the idea of using inductors is very appealing to me. I'm still figuring out the math and looking if my gain will be sufficient to still drive the rest of the system.

 

The filter

I don't want to re-invent the wheel, so I did what most people do... I ripped the schematic for the Tango EQ600 unit. It has gotten some rave reviews and the one-stage idea is very interesting. I put the circuit to the test (in theory) using MicroCap to simulate the response in combination to an inverse RIAA filter. The result is shown below.

 


Tango EQ600 on the simulator

 

The line may look wobbly, but this is over a 0.5dB deviation from 1Hz to 40kHz, after that it start to drop off slowly. Not really bad I think. The nice thing about a simulation is that you can change the parts within a few seconds and see the results, without ever heating up your soldering iron. Yes, this is purely software and there is no telling what reality may look like. Even so, I tinkered a bit with the values and got an even flatter response, as seen below...

 


Modified Tango circuit, better response (on paper)

 

This line is pretty much straight from 1Hz to 100kHz, about 0.1dB deviation here and there. But... how will it test in real life? I don't know... I hope to have some parts here soon to try it all out. The changes are minimal to the original values, as you can see below.


The 600 ohm filter

I noticed that the load is crucial to LCR corrections. Removing the 600 ohm load completely destroyed the correction in simulations. Even raising it to 620 ohms has enormous effects. The circuit will need to be tuned in real life because of this. The 600 ohm load also requires that the stage in front be able to handle this load, hence a stage is needed that itself has 600 or even lower output impedance. Without resorting to cathode followers the only solution left is using a step-down transformer. There are no tubes that have both a high gain and a plate impedance of 300 ohms or less.

This raises the question why the filter has to be 600 ohms. This low impedance means the values that are used in the filter are very low, but it also puts some strain on the choice of driver circuits (either step-down or cathode follower). I would like the load to be more in the area of 5K, this would mean I can drive with a lot less step-down or even directly with some tubes.

A 5K filter opens up the bag of tubes, but also poses another problem. Using a tube directly means DC into the filter. The LCR will not handle this very well as the chokes are made for AC only (saturation) and it's filtering will be destroyed (specification wise). Adding a capacitor in front of the LCR does the same as you are adding another part to the whole filter, so this will not work either. Adding an RC filter behind the LCR offers a different load, which is crucial to the LCR. In other words, the only way to do it right is to use a step-down inter-stage or a cathode follower.

Sticking with the inter-stage, I revised the circuit to be used on a 5K load. I will probably use a 6K:2K inter-stage to drive the LCR. This will only pose a 1:0.6 step-down in signal while still driving the circuit properly. The values of the chokes in the LCR filter have increased quite a bit, yet the capacitors have only decreased. I will place the circuit once it has been tested.

 


RIAA section of the preamp

 

The tubes

437A
The 437A is a well known tube I think. I chose it for the highish mu, low Rp and high gm. It is the first tube and will perform a good part of the gain and also decouple the signal from any AC by means of an inter-stage. I had these tubes left over from the spud attempts (see singularity), they are put to good use in the RIAA now.

EC8010
The EC8010 is not as well known as the other candidates I had like the 437A, 417A and 6S45. Yet it is in the same league as these high mu/gm tubes. The EC8010 has a mu of 60 and transconductance of around 28mA/V, giving a plate resistance in the neighborhood of 2K2. The benefit of this tube over the others is that it has the highest mu, and I will need about all the gain I can get for the RIAA. Biased at 140V/25mA it is very similar to the 417A/5842. I had originally intended to use the KC3, a directly heated tube. It has a mu of 30, but because of it's DHT nature and high mu, I am afraid that microfonics will be a killer at this position in the chain.

 

5842 EC8010 WE437A 6S45P KC3
Ua 150 140 140 150 135 V
Ia 26 25 29 40 3 mA
gm 24,000 28,000 43,000 45,000 2,500 uA/V
Rp 1800 2150 960 1160 12 K
mu 43 60 41 52 30 -
Pd 4.0 4.2 7.0 7.8 ? W

The candidates and their specs

 

The stages
There are really five stages that have effect on the overall gain of this amplifier. Two transformers, two tubes and a RIAA filter.

Input
At the input I am using a step-up transformer with a 1:10 ratio. The cartridge I am using (Benz Micro Glider) needs a load of around 1K, this requires a secondary of 100K. With a 1mV input signal I should get around 10mV output (without losses in the step-up transformer of course).

First stage
Here I use the 437A with inductive load (IT) which gives a gain of around the tube's mu (37-42). With 10mV input I should get around 400mV of output.

Step-down
This is the load of the 437A tube. To get a lower impedance I have used a step-down IT with an impedance ratio of 2:1. With a 5K load, the primary offers a nice load for the 437A. With 400mV input there will be a 200mV output.

LCR
The heart of the RIAA amp, a 5K LCR filter. I will loose around 20dB of gain here (around 10x). So with the 200mV input I will be left with 20mV output. This is quite a bit more than with a normal RC RIAA filter (-60dB normally).

Second stage
Here the EC8010 amplifies the signal from the LCR filter. The EC8010 will give a gain of around 50-60, which means the 20mV input will give an output of 1.2V.

 

23-03-2003
Right now I am in the process of ordering the chokes and other components for the RIAA/Preamp. Chassis design is also turning up some nice features. It will be made out of a solid piece of wood, machined and finished to give a nice look. The transformers will be potted in wood as well, using bee wax. The end result should be quite pleasing to the eye and ear.

 

01-01-2004
This project was removed during an update sometime, don't know why... It's back up again. Maybe I'll do some work on it this year. Haven't really been doing any DIY or phono listening lately so I never did any work on this project since then.


Jim de Kort