Converting the AD797 MC Phono Pre to a MM Phono Pre - A Mixed Design.

[wyn palmer]

What role does the U5, U6 (AD795) play?

Is that some type of servo, feedback leveling?

Anyway, I guess I got lucky the first time removing the NE5532.
Because after there, no matter what I put in U1, it didn't work.
It couldn't open any of the library files for anything. Go figure.

I try it again some other time.

I've attached the NE5532.asy and .sub.
The extra opamp is there to provide an alternative to having a DC blocking coupling cap at the output while not having a feedback integrator to set the LF cutoff. Basically, it reduces the offset by a fixed ratio so that the max is, if I remember correctly, 4mv.
I was asked to produce a design that did not have the series output cap, nor a cap in series with the resistor to ground at the opamp inverting inputs nor a design that uses a lossy integrator in the overall feedback (the requestor really disliked large, and potentially non-ideal caps in the signal path, apparently) !!!! This approach has the benefit of only introducing a large cap in a location which only has subsonic signal present on it, and which can have a "better" cap put in parallel to extend the rejection to higher frequencies than the cap SRF.

 

[EngineerNate]

@USMC Spike - Instead of screenshots, there's a dropdown menu in LTSpice where you can copy the graph to your clipboard and drop it directly into paint and save as a PNG or JPG. Should give you better quality.

 

[USMC Spike]

Nate, let us see if this is any different. I'd be surprised if it is, but who knows.
Me thinks it issue is actually from they way AK processes the images.
It must be on a low resolution mode, maybe 72 dpi. What ever it is,
it makes the image less bright also.


Here is the Dropdown capture via:Tools:copy bitmap to clipboardasteToMyEasyToUsePaintShopProProgram:SaveAs a jpg.

Enquiring minds want to know:


Just as I suspected. AK kind of fuzzies the image when uploading a file.

There is another option....I wonder if it will add a pic using the Sun&Mountain icon, next
to the my little girls favorite emoticon icon. Lets try it and see the result below:

Image URL. Nuts, this might require a two step process.
Get the image up here first (into AK) then, try the Sun&Mountain Icon and
paste the image number into it where it asks for the URL:

Survey Says:

It didn't like it. And since the "" is an actual command for HTML, I can't get around it
to edit it...go figure.
[IMG]http://[ATTACH]1380017[/ATTACH]

 

[USMC Spike]

This is the result of a direct upload using AKs Camera Icon, "Media Gallery Embed".

 

[USMC Spike]

Don't even ask, I have no idea why I got every emoticon. However, my little girl will scream with joy when she see's
those cute little emoticons in my post. Being a dad is fun, I love that kid.

AND

I'm enjoying it knowing that in eight more years, she'll hate me for a few of them.
She is five now and a kindergarden (can't think of the right word). But I've told here
and my better half, that when she is bridgeable [(SP) elidgible], she should run for student body president,
and continue running for stuff through out here life. Everyone loves here and she is fair.

How many 5 years old do you know that when you give them that little 6 pak of
peanut butter crackers, she tells her classmates, to come on over and
I'll give each of you a cracker and we can all eat one together.

The book, everything I need to know I learned in Kindergarden, comes to mind.
(cit. please) Okay, time to play buy a car fun and games.

 

[EngineerNate]

Easy there.

I host all my images on imgur and just embed them into my posts.

For example, here's a distortion FFT I made:

Just copy the imgur URL for the image and paste it into the image embed tool when you make a post.

 

[USMC Spike]

Your greens and blues are definitely an improvement.
Thanks Nate!

 

[USMC Spike]

@ Nate,

Are we talking the same thing here? The embed icon is the two pieces of film looking thing.
AK says we can only embed media from the following places.

edited

 

[EngineerNate]

One over.

 

[nj pheonix]

At your leisure if you could tweak design to a 2.5mv input it'd be appreciated.

 

[wyn palmer]

At your leisure if you could tweak design to a 2.5mv input it'd be appreciated.

I already have a 2.5mv design- that was what I tweaked for the 5.5mv design. I designed it a while ago "on commission" so to speak.
I've attached an essentially unmodified .asc file. It includes a cartridge model (I don't remember what type it was for), an A weighting network for noise measurements, a cutter model, a cantilever resonance model, a pick up arm model, and a bunch of ancillary stuff. There were other, more exotic, models as well, intended to do things like modelling the cartridge back emf and other even more incomprehensible things, but I removed them.

 

[nj pheonix]

Thank you

 

[nj pheonix]

I guess I have to plug that into LT spice

 

[wyn palmer]

I guess I have to plug that into LT spice

Do you want it in some other way?

 

[USMC Spike]

Wyn, another LT Spike question. In the MixedRIAA asc post #191,
How do I link one of the outputs on the left
with one of the phono amps?

Maybe I have the answer. For the Mixed Active LPOLE,
do I just type a new name in the input box?
That is, for the A-weighting network, I'd type "AWeightout"
where the input currently says "IDEALRIAAPREMP".

That worked for a little bit, but then after clicking on outmixed1 a few times
it reverted to just a horizontal dotted line at 0dB.

 

[USMC Spike]

For some of the variables on the left side of the mixed model schematic,
are those for any specific device? Are they just generic?
How would I, or anyone for that matter, find the specifics of them?

Since the boards haven't arrived yet, job 1 here for me should be to
1. put together a BOM for this.
2. Measure my devices and see how how this compares to specs from data sheets.
3. I assume there are models in the data sheet for to build something to measure it?
4. Any suggestions will be helpful as I've not done this before.
That is measuring OpAmps for noise harmonics, S/N and THD, or THD +N.

First I'll stare with 1V @ 1kHz.
compare with 1V @ 997 Hz a prime number.

 

[nj pheonix]

I guess I have to plug that into LT spice

My thought was a schematic.
If that's a problem or bother , I'll have to take a crash course on LTS.

 

[wyn palmer]

My thought was a schematic.
If that's a problem or bother , I'll have to take a crash course on LTS.

I can, of course, supply a single channel schematic that you can visually convert to whatever format you wish.
However, just think of how much fun you'll have getting to know LTspice.

 

[wyn palmer]

Wyn, another LT Spike question. In the MixedRIAA asc post #191,
How do I link one of the outputs on the left
with one of the phono amps?

Maybe I have the answer. For the Mixed Active LPOLE,
do I just type a new name in the input box?
That is, for the A-weighting network, I'd type "AWeightout"
where the input currently says "IDEALRIAAPREMP".

That worked for a little bit, but then after clicking on outmixed1 a few times
it reverted to just a horizontal dotted line at 0dB.

I think you need a crash course on noise sims.
To perform a noise sim you need to specify the signal source (v3 in this case) and the node at which the noise is to be measured (v(outmixed1)) in the simulation command that you access using the Simulate pull down menu, Edit simulation command, Noise tab.
Output would be v(outmixed1), input v3, type of sweep- Decade, points per decade- I use 50, start frequency 20, Stop frequency, 20k.
Click Ok then Simulate, Run.
A black window appears. Left click on the output node that you defined and you get the output noise spectral density (V(onoise)).
Press Ctrl and left click on V(onoise) on this plot and you get the integrated RMS noise over the specified frequency interval (say 200uv) The ratio of the output signal (1vrms for the specified input, either with an input of 5.5mv at 1kHz etc. or 2.5mv for the "new" design" ) to this is the signal to noise ratio, usually expressed in dB- i.e. 20 log (1/(2*10^-4))= 74dB.
However, noise is usually not specified over an unweighted bandwidth as it's not a realistic reflection of the sensitivity of the ear.
The unweighted one is often called Z weighting (Zero weighting) and the most common one is the A weighting.

Putting the weighting network between the source and the amp is not what is supposed to happen.
The way that the simulator figures out what the noise characteristic is, explains why you need to specify the input and the output.
Basically, the simulator calculates, at each evaluation frequency, the gain from the source to the output and to each of the noise generating elements in the circuit. It then determines the gain from each element to the output and adds the right noise source for each element at the circuit location of the element, then works out the contribution of every source at the output and adds all the contributions up.
So, to use the weighting network you wire the preamp output to it (through a noiseless R- the value is unimportant- just copy the 6.04k noiseless R) and specify the measurement node to be the output of the weighting network (V(AWEIGHTOUT)) which was configured to be unity gain at 1kHz.
In this case, the Aweighted output is 5-6dB lower (about 110uvrms) than the Z weighted one and the amp has about an 80dB S/N ratio.

 

[nj pheonix]

I can, of course, supply a single channel schematic that you can visually convert to whatever format you wish.
However, just think of how much fun you'll have getting to know LTspice.