DIY Audio

Testing different RIAA Amplifiers

A simple but comparable test of DIY and factory made RIAA amplifiers

Because I have done three years of electronic's in high school and also worked with home electronics in the -80's I want to be able to not only listen but measuring my different projects. These days I work as a metrology engineer and are everyday measuring and calibrating different instruments. So for me DIY audio is an engagement between building, listening and measuring. Lately I have also have good help from simulating circuits in the free simulating software LT Spice.

Take these tests as it is. A homemade trial to compare and evaluate different RIAA amplifiers. Happy reading.

FFT measurment with the PC and an external soundcard

I do all my live "real world" tests with an external M-Audio Audiophile soundcard. The tests on this page is done with 16 bit 96kHz sample rate and FFT sample size of 8192. That gives a bandwith to 48kHz.

As software I often use both the Visual Analyzer and RMAA (Right Mark Audio). There are pitfalls and also other measuring problems such as groundloops when you are trying to measure. What about wireless telephones and other such equipment in the house and unscreened naked projects on the testbench? And are the figures absolutely correct?

Anyway, this is a comparable test between four different RIAA amplifiers. Ftm. I only present measurements from FFT with the Visual Analyzer software to get THD figures. I have for every device also one channel, the red one, showing the reference. So the signal is fed from soundcard trough a Hagtech Inversed RIAA unit (DIY) and then the red channel looped directly back to the soundcard without going trough the measured amplifier and the green one going trough the measured amp, or the DUT (Device Under Test). Just to see that I'm doing OK. There is still some hum that's not allways coming from the PSU. Anyway the THD figures shown are comparable because the four devices was tested under the same condition.

Input is 3.5mV 1kHz on all tests. The output differs because of different amplification factors but also from inexact figures from the graphs, even if I did calibrate the software. The software comes for free so top notch isn't expected in this regard. Therefore I have double checked and done notations on the amplification factor under each testgraph measured with a brand new Fluke 289 true rms DMM.

The four candidates tested

The four candidates which are tested in this page are three DIY amplifiers and one Manufactured by a well known company.

1.) My latest DIY RIAA amp called The Pacific. It uses two amplifying FET (2sk170gr) stages with passive RIAA correction in between Read about my work and findings here. Tested naked without screened enclosure. This is work under progress. Should I cascode first stage or should I not? Ftm I'm simulating different approaches in LT Spice.
2.) My first DIY Solid State RIAA amplifier. The VSPS from RJM Audio's design. Amplifying and RIAA in one stage. Read about my own work on that one here.
3.) My first DIY tubed RIAA amplifier. The Tetra One from Tubecad. It uses two amplifying stages with passive RIAA correction in between, using the CCD (Constant Current Draw) technique I like and also use on my own modified TDA1541A DAC. First stage using NOS Telefunken ECC803S with no bypass cap. Second stage with NOS Telefunken (ribbed plate) ECC81, using small positive feedback instead of bypass cap. Described on John Broskies pages about this design. Tested naked without screened enclosure.
4.) The inbuilt phonoamplifier of a Luxman LV105U. It uses a FET differential amplifying stage and after that an IC with the RIAA network in the feedback net. I have modified it with a better IC compared to the original one. It now have a National LM4562 (same as LME49720) instead of the M5220p. I have lowered the +-22.5V to 18V to not get the LM4562 to be to hot. (It worked well with a heatzink on before that mod anyway.)

Result for the Pacific FET SS amplifier

Output at -10dB gives 41dB amplification on the Pacific.

Result for the VSPS SS amplifier

Output was at -10dB gives 41dB amplification on the VSPS.

Result for the Tetra One Tubed amplifier

Output at -8dB gives 43dB amplification on the Tubed Tetra One.

Result for the Luxman LV105U phono amplifier

Output at -13dB gives 38dB amplification on the Luxman LV105U phonostage. Same THD as reference!

And what is the subjective sound from the four candidates?

This is hard to answer and will allways be subjective. There are small differences between the VSPS and Luxman, They sound clean and dry for sure. Do I miss some openess and dynamic punch? Maybe, I don't know but the two other candidates, The Pacific and the Tetra One sounds more like in the other league. The more pleasant and musical one I think. I have finally learned to recognice second order harmonics and it could be pleasing in small amounts, adding sweetness and bite. Because it's in harmony with the original music It's not that obvius and easy to distinguish as odd order distortions. I was quite amased how much one can tolerate! Maybe because this time I listened a lot before I measured ;-). I think that this is the tubesound all are talking about. But then again, the Tetra One amplifier is all tube and quite clean with It's THD ten times lower then the Pacific, still sounding open, musical and, hands down, the best in this test ,even though it needs some screening with a nice inbuild.
So then we have the question about global feedback vs none or little. The VSPS and Luxman uses a lot of it. The Pacific and Tetra One none, just some local fb. Maybe it has a large impact on the openess and liveliness?. I guess I have to fiddle with my Pacific some more to find out....

Thank you for reading and please share your DIY findings as well. //Ove Tegnér