I had no resources for those kind of measurements but the noise was very low and the sound was good, even with directly connected headphones.
I did already...some pages back I provided a link... I'm at the phone, can't show you with asc files.
Last edited:
Will need more help to understand and implement that. Thanks!
I think I have been using too high input voltages to simulate the RIAA projects. It seems certain I need some kind of standard, either for input voltage (after anti-RIAA filter) at specific frequency, like 1KHz. Or an output voltage.
One intriguing passive RIAA preamp suggested on the LM4562 datasheet.
http://www.ti.com/lit/ds/symlink/lm4562.pdf
There was a 5.6dB fall-off from 3KHz to 20KHz. But I think there's a problem with the LM4562 LTSpice model. Replacing the chips with LT1115s resulted in a +/- .5dB FR.
Can anyone suggest how and where to setup values to compare the simulations correctly?
I think I have been using too high input voltages to simulate the RIAA projects. It seems certain I need some kind of standard, either for input voltage (after anti-RIAA filter) at specific frequency, like 1KHz. Or an output voltage.
One intriguing passive RIAA preamp suggested on the LM4562 datasheet.
http://www.ti.com/lit/ds/symlink/lm4562.pdf
There was a 5.6dB fall-off from 3KHz to 20KHz. But I think there's a problem with the LM4562 LTSpice model. Replacing the chips with LT1115s resulted in a +/- .5dB FR.
Can anyone suggest how and where to setup values to compare the simulations correctly?
Here we go, as a direction. But the link I provided has a more elegant solution.
Btw, I wonder why the resident gurus, who happily posted in this thread, are of no help with this.
I read about this on Elliott and Hagerman sites too
Reverse RIAA
iRIAA2 - Inverse RIAA Filter – Hagerman Audio Labs
Reverse RIAA
iRIAA2 - Inverse RIAA Filter – Hagerman Audio Labs
Thanks, Sottomano.
How do I implement that on one of my simulations? Before the Anti-RIAA?
You didn't Play around with the file and looked what it does, did you?
Otherwise you would have find out, that this IS the anti RIAA...…..what else if I may ask.
So yes, I would start the signal chain with it.
Using this, you don't have to mess with anti RIAA networks, and these cannot interfere with the cartridge impedance model. This of course goes only for simulation, and not for measuring.
(You could generate an Anti - RIAA Test Signal file with LT Spice and RMAA, but sorry I do not have the time to go through that. )
Edit: If it does confuse you, +IN_LAPLACE goes to the Phono pre or, if present, to the cartrige model network. And of course you have to add an AC Analysis comand for V1.
Otherwise you would have find out, that this IS the anti RIAA...…..what else if I may ask.
So yes, I would start the signal chain with it.
Using this, you don't have to mess with anti RIAA networks, and these cannot interfere with the cartridge impedance model. This of course goes only for simulation, and not for measuring.
(You could generate an Anti - RIAA Test Signal file with LT Spice and RMAA, but sorry I do not have the time to go through that. )
Edit: If it does confuse you, +IN_LAPLACE goes to the Phono pre or, if present, to the cartrige model network. And of course you have to add an AC Analysis comand for V1.
Last edited:
May I suggest taking a step back and brush a little on your understanding of basic electronics (including noise) and simulation tools?
To start with, the schematic you posted in #268 https://www.diyaudio.com/forums/analogue-source/91497-bc550-bc560-low-noise-riaa-26.html#post5685615 doesn't make much sense, since the frequency dependent RIAA correction is in the global feedback loop, the input stage bias is ridiculously low (a few tenths of uA/transistor), etc... Then the noise simulations you presented also don't make any sense. You don't seem to have a good understanding of AC and tran simulations and what to expect from each, etc...
If you need help, start a new thread and ask for. A little bit of study before engaging in electronic design never hurt anybody.
To start with, the schematic you posted in #268 https://www.diyaudio.com/forums/analogue-source/91497-bc550-bc560-low-noise-riaa-26.html#post5685615 doesn't make much sense, since the frequency dependent RIAA correction is in the global feedback loop, the input stage bias is ridiculously low (a few tenths of uA/transistor), etc... Then the noise simulations you presented also don't make any sense. You don't seem to have a good understanding of AC and tran simulations and what to expect from each, etc...
If you need help, start a new thread and ask for. A little bit of study before engaging in electronic design never hurt anybody.
No problem about that. Noise in simulations seems not to be as easy as other sims.
If you mean the constant current resistors R2 to R9, they keep the same current as the original Elektor preamp: around 200uA. They were increased in value from the original ones (68K) when the power supply voltage was doubled.
Perhaps it would be better if you said how to do things right and correct the simulations variables as they should be. The asc files are all there.
Unfortunately I'm not a LTSpice user, so I can't help you with the asc files.
May I though suggest start from the bottom up, with like simulating the RIAA correction network, check against the theoretical frequency response, understand why you can't put it in a feedback loop, calculate the attenuation, check the time domain response, etc... and leave the amplifier and its noise for later, after you get a good understanding of the basics?
So what are you suggesting?
Start at the beginning I won't. So the way I see it I will keep trying and asking for help from people that are willing to, which doesn't seem something you are.
Maybe for you it doesn't make any sense, but I try to understand and use the tools at hand, and counting on the generosity of a few people to put things in the right path.
For LTSpice I followed the advice of a few people, like Bob Cordell's pages about LTS basic operating procedures, and use the settings other people used.
Until now I have been following that way of doing things with power amplifiers, and others came along that helped me (and probably others) take things further.
Then I got interested in RIAA preamps because I wanted to build one for me. And by chance I found this topic, working on the first RIAA preamp I had built.
It has been some time that I have been looking for ways to simulate RIAA preamps. You say that I "do not seem to have a basic understanding of what I'm trying to do". That might be so, but I think I have gone past the "basic understanding" as I see it.
As usual, we don't agree. You insist on wanting me to do things your way, but this will not be the case.
If you are really willing to help, if not me the others reading this thread, and interested in this specific project, share your ideas on how to make things better. If my simulations are wrong, correct them, do new ones, teach other people how to improve this project.
If you are not, as you say, an LTSPicer, then how can you say mine are wrong?
Anything else is a waste of time.
Start at the beginning I won't. So the way I see it I will keep trying and asking for help from people that are willing to, which doesn't seem something you are.
Maybe for you it doesn't make any sense, but I try to understand and use the tools at hand, and counting on the generosity of a few people to put things in the right path.
For LTSpice I followed the advice of a few people, like Bob Cordell's pages about LTS basic operating procedures, and use the settings other people used.
Until now I have been following that way of doing things with power amplifiers, and others came along that helped me (and probably others) take things further.
Then I got interested in RIAA preamps because I wanted to build one for me. And by chance I found this topic, working on the first RIAA preamp I had built.
It has been some time that I have been looking for ways to simulate RIAA preamps. You say that I "do not seem to have a basic understanding of what I'm trying to do". That might be so, but I think I have gone past the "basic understanding" as I see it.
As usual, we don't agree. You insist on wanting me to do things your way, but this will not be the case.
If you are really willing to help, if not me the others reading this thread, and interested in this specific project, share your ideas on how to make things better. If my simulations are wrong, correct them, do new ones, teach other people how to improve this project.
If you are not, as you say, an LTSPicer, then how can you say mine are wrong?
Anything else is a waste of time.
Actually the 200 uA per pair tail current is already well above the noise optimum, because a real moving-magnet cartridge has much higher impedance than the reverse RIAA network due to its 500 mH or so inductance. Increase the tail current further and you will spoil the noise performance.
According to the schematics it's about 25uA per transistor (30V/160kohm/4/2) or 50uA/pair tail current. The BC550 data sheet https://www.mouser.com/ds/2/149/BC550-888526.pdf goes as down as 1mA Ic for the DC current gain chart, but assuming the average Beta is still 200 at 25 uA (which I doubt) that would be a total current noise of about 0.5pA/rtHz. For a 0.5H MM cartridge inductance at 10KHz (~32Kohm) that's already 16nV/rtHz flat noise, or about 1.2uV averaged in 20-20k (napkin calculation). Almost too high even for a MM with 5mV @ 5cm/s since the S/N is 70dB, too close to the vinyl surface noise for my taste.
An I haven't considered yet the flicker noise, the 1/f noise frequency corner (usually in the KHz range), the (rb+re)/4 noise and the 3dB noise loss because of the differential uncorrelated sources.
Perhaps with C class transistors with Beta=800 this schematic would make some sense (strictly from a noise perspective), although it is still lacking current balance resistors in the emitters (otherwise some of the transistors will certainly be blocked due to the Vbe dispersion), which in turn are worsening the input referred voltage noise. But IMO paralleling so many bipolar transistors here doesn't make any sense and IMO, in general, using bipolars for a MM is a bad idea. That's why we still have low noise jfets that would do a much better job.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.