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

[rothwellaudio]

I'm sorry if my observation offended you. I'm not a member of the flat earth society and certainly don't believe that people shouldn't be asking questions.
Ask Wyn as many questions as you like about Laplace functions and modelling tonearms and turntables - it's up to him how long he continues to write answers.

 

[USMC Spike]

That was a nice try rothwell, your snide remark, then and your "oversevation" is bigoted.
If you take the time to go back and read the AD797 help desk thread, you'll can read
Wyn's comment for yourself. I was responding to that remark about modeling tonearms and turntables
in this thread, as follows:

"Why had mentioned other important aspects of the turntables and their cabling; (i.e., inductance, arm resonance,
wiring impedance/resistance etc. this is just recalling from memory)...

...Since I've not done this before and have a vague idea of the concepts etc. Help would be appreciated.
Turntable, cartridge, tone arm, cable and other measurements to define
what I actually have."

Wyn's comment seemed to notice that AK members were not focused on these other aspects of good tone.

He concluded mentioning something about taking a measurement of the whole turntable from
cable to cable to capture that measurement.

I am also sure that Wyn has many other "tricks-of-the-trade" that he's discovered over the years too.
Just so that you know rothwell, I don't have any notions of grandeur here, as I am just a tech,
and willing to learn what I can from someone who's proven himself to be a master in the art.
Your comments almost sound like petty jealously--it is what it is.

I still appreciate you submission of the SilmicII caps showing polarity.


That's for my kid who's lectured about electronics.
Here she is:



And, rubbing the belly for good luck:



Reminding me where the engine compartment is:



And, finally playing some relaxing music:

 

[USMC Spike]

As I find the references I mentioned I'll post them here. It's a nice break from doing Physics problems too.
Here are the links:
1. LINK

 

[USMC Spike]

Update, Uplate, going through everything a few more times.
Don't have much at the moment ans Physics is trying to eat my lunch.
Most of my free time during the day is spent at campus studying...
or maybe I should say looking at the book solving the problems
and me trying to follow along.

There is a time or two when I figure something out.
Otherwise it's just looking at it and going what did I
get my self into?

Mid terms have passed and so have aye,
at least for the time being...something like
an 88 percent. That includes the -16 points
on the midterm exam that I didn't have time
to stay in the class and finish because I have
another class that starts as soon as the first
class ends....No walking time between them!

Otherwise all the parts for this and a few other
projects should be arriving at the end of the
week. Knowing how the weather is not holding
up...I think it will be another week for things to arrive.

If I can pull my head out enough during spring break
I should be able to knock the MM board out and trouble
shoot it, take base level measurements etc.

Wyn, would you care to share your PS design
with the regulators, fed from the switcher?

That would probably get me going quicker then
trying to figure out one on my own.

I've got a smattering of some of the LT Regs.

In the worst case, I can always plug in a power supply
to the board and rails and dial in +15V and -15V.

Wyn, thank you for all your insight and assistance.

 

[USMC Spike]

Update

Got the board up and running on the bench.
Power is limited to +- 12.541 on the rails.

I haven't done the numbers but here is the raw data.

997Hz, 10mV
3.25mV in R
0.565.09 V out R

3.25mV in L
0.565.40 V out L

997Hz, 3mV
1.08mV in R
0.180.05 VAC out R
5.510 mVAC out R 1st stage

1.08mV in L
0.180.18 VAC out L
5.8513 mVAC out L 1st stage

The phono pre is just sitting on the bench
in the board holder used for soldering.
All is open with temp wiring, etc.
and I have 1-inch solid copper wire antenna
for the audio inputs and outputs, and power Inputs.

THD is better than -60 dB.
THD is better than -70 dB A-weight.

That's all for now.

 

[USMC Spike]

Some clarification:

I won't have any listening tests until I get the power supply and
voltage regulation squared away along with getting my TT up and
running along with a cartridge installed in it.

Once I dial in the +- supply voltages on the rails, they match
down to the mV level from + to - and left to right.

with 3.25mV input on the L & R channels the output
for each channel with within .31 uV. .56540VAC-.56509VAC.
for a 0.05% level difference between channels.

Less than 60dB non-weighted
Less then 70dB A weighted.

This with 18awg solid copper wire "posts" for hook up
on my panavise board holder on my bench.

I could fool around with changing devices but i'm not
sure what benefits would happen.

I just wonder how different these devices would be
with 17VDC rails? I suspect they would be slightly
less distortion.

 

[wyn palmer]

So the gains look about right. The noise is higher than I expected even accounting for the fact that the input level is c. 4dB less than the rated value.
I'm also confused by your terminology/organization so perhaps I'm not reading things correctly.
What's the offset at each stage?
The distortion should be significantly better than -60dBc at the levels you are operating at, especially as it is not at the rated output level.
How are you measuring the distortion and noise?
17vDC should not make a significant difference at the signal levels you are operating at, and as the LME has an absolute max supply voltage of 17v I wouldn't recommend that you change supplies.

 

[USMC Spike]

Hi Wyn,

I assume you are asking for the DC offset voltage on the opamp pins?
For the LT1115 both opamps measured with HP34401A:
Both +12.541VDC
Both -12.541VDC
NET = Zero Offset.

For the first stage
Both +12.494VDC
Both -12.494VDC
NET = Zero Offset.

From the Oscillator of my HP339A,
set to 10mV, 977Hz output, xVAC
IN R = 3.25mV
out R= .565,09V

IN L = 3.25mV
out L= .565,40V

From the Oscillator of my HP339A,
set to 3mV, 997Hz output, xVAC
IN R = 1.08mV
out/in = 0.180,09VAC (p6) 0.180,05VAC (p3)
out R = 5.510mVAC

IN L = 1.08mV
out/in = 0.180. (p6) 0.180,18VAC p3)
out L = 5.513mVAC

Distortion figures were taken from the Shibasoku 725D
distortion analyzer, indicated on the buttons, it was late
and Its been a while since I read the meter movement.

The distortion numbers I posted were from numbers above
the button LEDs, Red in dB. Therefore, my statement,
less than, 60dB and 70dB. Those numbers are at least 25dB
conservative. To find out how much, we have to look on the movement.

Then moving to the meter movement,
for the 60dB indication on the button and the needles was
moving about .1 to .2 % indicator (the top most scale) and
on the .1 or .2 marks right of zero.

From this I can say with confidence, -60dB LED, plus viewing
on the red scale that ends with -20 dB corresponding to about
1.1% on the tops scale. Then EYEBALLING the top scale
at .5% should easily add another -5 dB to the total. So that
puts us at, at least -85dB.

Now, I've forgotten how to calculate the % from the LED and Scale.
as -80dB would correspond to 0.01%, then, multipling that by
.01% (0.01% * .01% = 0.0001% or .02%(0.01% * 0.02% = 0.0002%).
If so that would be better than -100dB.
Probably (0.01% x .1% = .001% = -100dB) unweighted.
THD A Weighted that is -110dB! which isn't bad just plugging
in four devices randomly.

What say you?

 

[USMC Spike]

Pretty much the same results this evening as well.
There is no offset from the board, when I measured
an offset, I went back to the variable power supply
I use for the Positive side and it drifts about 5 - 8 mV.

Numbers slightly better than yesterdays.

Same 10mV, 997Hz Input.

with A weighting, THD is -100dB.

See attached.

 

[wyn palmer]

Pretty much the same results this evening as well.
There is no offset from the board, when I measured
an offset, I went back to the variable power supply
I use for the Positive side and it drifts about 5 - 8 mV.

Numbers slightly better than yesterdays.

Same 10mV, 997Hz Input.

with A weighting, THD is -100dB.

See attached.
View attachment 1464160
View attachment 1464161

View attachment 1464162

The offset I'm referring to is the dc offset at the output of each of the opamps. Remove the ac signal and measure the voltage at the outputs. The opamps have very high psrr. In essence having the supply voltages matched is irrelevant.

 

[wyn palmer]

Pretty much the same results this evening as well.
There is no offset from the board, when I measured
an offset, I went back to the variable power supply
I use for the Positive side and it drifts about 5 - 8 mV.

Numbers slightly better than yesterdays.

Same 10mV, 997Hz Input.

with A weighting, THD is -100dB.

See attached.
View attachment 1464160
View attachment 1464161

View attachment 1464162

The distortion seems ok, if convoluted, however the noise remains a mystery.
How are you measuring that? I need to know equipment, measurement conditions and any assumptions you are making. A xxdBc S/N ratio needs to have the signal levels specified and also the source impedance. A raw integrated noise voltage measurement should be made with a shorted input and either an unweighted 20-20kHz band or A weighting applied.

 

[USMC Spike]

Hi Wyn,

No assumptions on my part, just plug and play.
Signal source is HP339A, 600ohm output impedence.
connected to either left or right channel + and - posts
through banana plug thru 50ohm RG/58 a/u cable with clip on
connections.

The output is clip on connections of +, - on the output posts
through RG/58 a/u 50 ohm cable with BNC input to the
shibasoku 725D. Input and results are shown in post
289 above. I went back and looked at the pic,
AK doesn't like large images, even as thumbnails
so it made it fuzzier than it should be.

My DC offsets are as follows:

as labeled on the board:
Rt Channel
OA1 -0.518mV
OA3 -0.163 V

Lt Channel
OA2 +0.456mV
OA4 +0.133V

Thoughts?

 

[wyn palmer]

Hi Wyn,

No assumptions on my part, just plug and play.
Signal source is HP339A, 600ohm output impedence.
connected to either left or right channel + and - posts
through banana plug thru 50ohm RG/58 a/u cable with clip on
connections.

The output is clip on connections of +, - on the output posts
through RG/58 a/u 50 ohm cable with BNC input to the
shibasoku 725D. Input and results are shown in post
289 above. I went back and looked at the pic,
AK doesn't like large images, even as thumbnails
so it made it fuzzier than it should be.

My DC offsets are as follows:

as labeled on the board:
Rt Channel
OA1 -0.518mV
OA3 -0.163 V

Lt Channel
OA2 +0.456mV
OA4 +0.133V

Thoughts?

DC offsets are fine. I still really don't know what your noise measurements are.
I expected them to be several dBs better than you stated even allowing for the fact that you are providing a S/N at a 4dB lower output level than was used as a basis. The 600ohms source resistance also degrades the output S/N if the measurement gear doesn't de-embed it.
In any case, it seems to be acceptable, but I'm big into results versus expectations...
How is the RIAA compliance?

 

[USMC Spike]

Hi Wyn,

What should I change the source impedance too? I've got an HP Meter that can change
the impedance etc. Same for the output voltage...where do we want to go for it?

I don't know how the measurement gear would emphasize the 600ohm source
impedance unless I turn off the 600ohm resistance button, but I'm thinking that is
what it does when illuminated.

Also looks like I have to resolder one of the 3300pf caps...appears to have a
sub optimal solder joint.

I haven't checked for RIAA compliance yet.
just burning the phono pre amp in at the moment.
and will do a sweep for compliance when I get
that set up.

Now that I don't have any signal, with only a shorted input
I look again at the noise figures on the distortion analyzer.

 

[wyn palmer]

Hi Wyn,

What should I change the source impedance too? I've got an HP Meter that can change
the impedance etc. Same for the output voltage...where do we want to go for it?

I don't know how the measurement gear would emphasize the 600ohm source
impedance unless I turn off the 600ohm resistance button, but I'm thinking that is
what it does when illuminated.

Also looks like I have to resolder one of the 3300pf caps...appears to have a
sub optimal solder joint.

I haven't checked for RIAA compliance yet.
just burning the phono pre amp in at the moment.
and will do a sweep for compliance when I get
that set up.

Now that I don't have any signal, with only a shorted input
I look again at the noise figures on the distortion analyzer.

I expect typical harmonic distortion levels of about 0.0011% at 1kHz into a 600 ohm load, based on the TI specs. for the LME alone, so the distortion is good and there seems to be negligible contribution from the caps unless the TI spec is wrong, and semiconductor manufacturers rarely pad anything as important and as predictable as distortion.
In any case, it's fine.
Simulations indicate an unweighted S/N ratio with your 600ohm source and load and a 565mv rms output, of 73dB, and about 80dB A weighted.
The S/N ratio is about 2dB better with a 50 ohm source, and about 4dB better again with the output signal at the rated level of 1v RMS.
SINAD is dominated by the noise.
The DC offsets are well within the expected range.

 

[USMC Spike]

So I'm finding specs for some reason I couldn't find them.
The imput impedance I'll check again and use the 100Kohm default range.
I will also try again and get the signal level to be 1V RMS.

I'll perform that with the HP339A with 600ohm impedance.
I'll use my nosier source that does have 50ohm source resistance
from my Leader LFG1300s.
The wave options I have are sine, triangle, and square.

Grounds are connected to the power ground lug.
Signal connections to the TT ground lug.
I'm thinking I have not connected the signal and measuring
devices to ground, they are floating....perhaps proper ground for
these instruments on the TT ground lug should be in order.

My modded HP339A will perform 1khz sine wave to THD > -120 dB,
that should be somewhere between -120dB and -130dB. We couldn't
measure below that reliably at the time.

I've got homework that I've postponed, etc and physics has
started to eat my lunch as they say, so my postings will be
intermittent.

 

[wyn palmer]

So I'm finding specs for some reason I couldn't find them.
The imput impedance I'll check again and use the 100Kohm default range.
I will also try again and get the signal level to be 1V RMS.

I'll perform that with the HP339A with 600ohm impedance.
I'll use my nosier source that does have 50ohm source resistance
from my Leader LFG1300s.
The wave options I have are sine, triangle, and square.

Grounds are connected to the power ground lug.
Signal connections to the TT ground lug.
I'm thinking I have not connected the signal and measuring
devices to ground, they are floating....perhaps proper ground for
these instruments on the TT ground lug should be in order.

My modded HP339A will perform 1khz sine wave to THD > -120 dB,
that should be somewhere between -120dB and -130dB. We couldn't
measure below that reliably at the time.

I've got homework that I've postponed, etc and physics has
started to eat my lunch as they say, so my postings will be
intermittent.

You don't need to change the load R from 600 ohms. The amp is fine and although the load acts like an attenuator of c. 1dB it doesn't affect the output S/N ratio.
Similarly, the output level doesn't have to be 1V, just as long as you know what it is and what either the integrated noise or the signal to noise is.
Spike, it's not clear to me that the gear you have actually measures the noise performance, just the harmonic distortion.
I perform my measurements using my RME ADI-2 pro FS ADC/DAC and I can measure noise. Do you have a comparable ability?
If you don't post for a while that's fine, after all this is for your benefit.

 

[USMC Spike]

I do have a QA400, that has FFT and will show noise floor.

Hmmm, I can take the calibrated output from my 725D and using the analysis mode
port it to my iphone, then use the peterson's IStrobosoft tools to view the Spectrum.
That or as I said the QA400 FFT to look at the noise floor. My R&S UPD took a dump
and won't start any longer. It need new caps and resoldering of the switch. It's a common
problem with these units.

go to about the 3:15 mark for their scope and spectrum features.
Link to peterson istrobosoft; LINK

For the QA400 the here is a link: LINK
The new version is the QA401.
I have the older QA400. The price QA401
more then doubled, so perhaps in the future.

 

[wyn palmer]

I do have a QA400, that has FFT and will show noise floor.

Hmmm, I can take the calibrated output from my 725D and using the analysis mode
port it to my iphone, then use the peterson's IStrobosoft tools to view the Spectrum.
That or as I said the QA400 FFT to look at the noise floor. My R&S UPD took a dump
and won't start any longer. It need new caps and resoldering of the switch. It's a common
problem with these units.

go to about the 3:15 mark for their scope and spectrum features.
Link to peterson istrobosoft; LINK

For the QA400 the here is a link: LINK
The new version is the QA401.
I have the older QA400. The price QA401
more then doubled, so perhaps in the future.

Any updates?

 

[USMC Spike]

None, other than to mention both my FFTs don't work.

I do have a couple of other distortion analyzers though,
not sure how to config for THD + Noise.

I can set up for 1KHz, ? mV RMS out into a Bruel & Kjaer
Frequency and Measurement Bridge to bring down the
1KHz signal to say -40db into the MM board and measure
the output.

I haven't looked up there yet and have to run have a school
lunch date with my six year old kid.

Spike