The problem with Baxandall’s circuit is the overload margin at HF is poor. There are many Stereophile reviews of solid state RIAA EQ amps proudly flying their passive EQ flags with 9 dB overload margins. That is just not enough to cater for ’hot’ cartridges, hot recordings and the overall dynamic range of recorded material to say nothing of surface noise clicks and pops that may be overlaid on that. Use passive in tube circuits where you have high supply rails and plenty of swing, and use all active a la Lipshtiz for low voltage (ie =-15V rails) opamp or discrete designs.
Shure did a lot of work on this in the 1970’s and the recommended OL margin is 26 db. Consider the requirements made up as follows: 12-14 db for music dynamic range + 9 dB to cover hot carts and hot recordings. That’s 20 dB to which a further safety margin of 6 dB is added to cover clicks and pops etc.
Shure did a lot of work on this in the 1970’s and the recommended OL margin is 26 db. Consider the requirements made up as follows: 12-14 db for music dynamic range + 9 dB to cover hot carts and hot recordings. That’s 20 dB to which a further safety margin of 6 dB is added to cover clicks and pops etc.
You respond to a point I was not making!
I agree, BTW, that overload margin is important. the only point I made was that if you separate the time constants it is easier to obtain an accurate result.
It's not an approach I use.
I agree, BTW, that overload margin is important. the only point I made was that if you separate the time constants it is easier to obtain an accurate result.
It's not an approach I use.
Back in the 1980's when such things still mattered, I was asked to build a phono preamp. It was two inverting stages using a quad op-amp. I make no claims to it being exceptional, but it worked very well with no high frequency overload problem and imuned to ESD. I was also asked to repair an audiophile discrete phono preamp, that frankly was junk and way overpriced.
I am getting weird results when I put at the input of my other RIAA sims. No doubt it's because of my tremendous limitations in working with LTspice.
Until now I have been anti-RIAA passive inputs in this way, as suggested by Rod Elliott and others. Like this one.
What do you think?
Here you are.
I have used the universal amp 2 and entered all OPA627 parameters into this model.
BW=16Mhz, SR=55V/usec, Noise: 4.5nV/rtHz, Knee point: 400Hz, Offset: 130uV
Hans
I have used the universal amp 2 and entered all OPA627 parameters into this model.
BW=16Mhz, SR=55V/usec, Noise: 4.5nV/rtHz, Knee point: 400Hz, Offset: 130uV
Hans
26dB = times 20
"Shure did (...)"
What nominal output level does this information refer to? This can mean anything and nothing from +/-9V to +/-16V to +/-30V Rails and even more.
HBt.
I noticed you changed the input capacitor C2 and R7. I can't: 220n and 1M5 serve my needs perfectly.
I was the one that asked Marcel to redesign the RIAAs I was going to use to include a servo and work as HP filter for the very low (3Hz and up) frequencies the Thorens TT outputs. One design had been based on an Erno Borbely discrete two stage published on The Audio Amateur, and the other a Walt Jung two stage passive design published on his opamp book.
My demands had been not to put a cap before first stage, but now I wish to listen to this single stage Marcel did, which seems to sim so well. And when built compare the audio of the three of them.
What do you think of the passive anti-RIAA filter I am using, compared to the Laplace filter you use?
The results I get from your "Laplace" input are different from the ones I get from my RIAA filter, and I wonder which is closer to the real output.
I didn't know how to adjust the settings of "the Laplace" to sim noise and THD.
Do you think the settings I'm using on my sim are correct?
Laplace is the most accurate but can only be used in the .AC mode.
For the .Tran mode you will have to use the RC model.
Hans
For the .Tran mode you will have to use the RC model.
Hans
Good, then it means my RC sim settings are correct?
I would like to simulate other things for the RIAA preamp, but I don't know which and how. Which would you suggest?
I would like to simulate other things for the RIAA preamp, but I don't know which and how. Which would you suggest?
Something that puzzles me is the "hump" I get on several on my noise sims on several of these RIAA versions.
I thought it might be related to the opamp model I was using, so I took my single IC Marcel version, with 220n and 1M5 producing the right filter for my needs. Then I used the OPA627 model you made.
Then ran the noise sim... to get the same curve hump.
Is there something wrong on my sim, settings or whatever?
Please run it and check, and tell me what you get.
I thought it might be related to the opamp model I was using, so I took my single IC Marcel version, with 220n and 1M5 producing the right filter for my needs. Then I used the OPA627 model you made.
Then ran the noise sim... to get the same curve hump.
Is there something wrong on my sim, settings or whatever?
Please run it and check, and tell me what you get.
I see nothing special.
That hump has everything to do with the rumble filter where the gain left of the hump is dropping rapidly.
Hans
That hump has everything to do with the rumble filter where the gain left of the hump is dropping rapidly.
Hans
You should expect to see noise levels following the frequency response fairly closely unless some parts have frequency-dependent noise models (1/f noise etc).
In a good MM RIAA amp the 47k resistor and cartridge series resistance are often the dominant noise sources. Spikes in the noise graph at high frequencies often indicate marginal stability as they are due to insufficiently damped poles.
In a good MM RIAA amp the 47k resistor and cartridge series resistance are often the dominant noise sources. Spikes in the noise graph at high frequencies often indicate marginal stability as they are due to insufficiently damped poles.
Kind of resurrecting this, but that is similar to Bob Cordell's VinylTrak design (Linear Audio V4, September 2012, and supplementary information here https://www.cordellaudio.com/preamplifiers/vinyltrak.shtml ). He drives the 75uS time constant from precisely the above. The rest of the time constants follow as a conventional feedback arrangement.
This gives not only accurate RIAA, but also huge overload margins. Ref 5 mV MM or 500uV MC, 31dB at 1kHz and 41dB at 20kHz
You can get all 14 volumes of Linear Audio from the DIY Audio Store here https://diyaudiostore.com/collections/books/products/the-complete-linear-audio-library or individual printed volumes from Amazon. Or if you just want an individual article for $2.99 . See https://www.linearaudio.net/
I was lucky enough to subscribe when it was being published, so I have a complete set. Including the two additional volumes, Audio Power Amplifiers, and correspondence between Douglas Self and Peter Baxandall on audio power.
I was lucky enough to subscribe when it was being published, so I have a complete set. Including the two additional volumes, Audio Power Amplifiers, and correspondence between Douglas Self and Peter Baxandall on audio power.
Last edited:
