A few posts back, you said you were. Perhaps you could clarify your earlier misstatement?
Indeed, a 1nV/rtHz FET opamp would be a nice thing there!
there is also the issue of DC bias current flowing in lots of turns wrapped around some iron
even fractional uA seems to excite our gurus even when flowing in MC
even fractional uA seems to excite our gurus even when flowing in MC
Perhaps it was designed to be used with an amplifier which also has a matching one turn transformer in series with ground in phase to out the voltage??
jn
SY, JCX,
I mentioned the highish input bias as an issue, but it can be dealt with. John will likely use a servo and I might do so in a final version, but a good coupling cap does the work for me at the moment.
Since the MM's I have had are usually are around 1KOhm or even a bit lower, that is pretty ball park optimal for the 797.
On JCX's point about the "heavy bias bjt input current noise flowing in the MM Z creates more Vnoise than the 797 vn" let me first straighten out a semantic point. The noise produced by current flowing through the MM Z is called excess noise, not current noise. From analogy with what is measured in real life resistors, I would expect a wire wound resistor made with defect free copper wire to have exceedingly small excess noise in the uA region we are talking about.
Now, have a look at the goodies. The 797 not only has very low voltage noise, but also a low 1/f corner for both voltage and current noise. With RIAA empasizing lower frequencies, that is what you want. Stellar distortion, load driving capability (allowing small value R's in the feedback network), CMR and PSR are other great qualities of the 797.
In short, I don't dig the deregatory remark about "but when the preamp is selling for big bucks, one might expect attention to detail". You imagine problems that aren't really there, and neglect all of the good reasons one might have to use the 797.
It's not a perfect world, and I share John's wish for an OPA134 with <1nV voltage noise, but it isn't there.
I mentioned the highish input bias as an issue, but it can be dealt with. John will likely use a servo and I might do so in a final version, but a good coupling cap does the work for me at the moment.
Since the MM's I have had are usually are around 1KOhm or even a bit lower, that is pretty ball park optimal for the 797.
On JCX's point about the "heavy bias bjt input current noise flowing in the MM Z creates more Vnoise than the 797 vn" let me first straighten out a semantic point. The noise produced by current flowing through the MM Z is called excess noise, not current noise. From analogy with what is measured in real life resistors, I would expect a wire wound resistor made with defect free copper wire to have exceedingly small excess noise in the uA region we are talking about.
Now, have a look at the goodies. The 797 not only has very low voltage noise, but also a low 1/f corner for both voltage and current noise. With RIAA empasizing lower frequencies, that is what you want. Stellar distortion, load driving capability (allowing small value R's in the feedback network), CMR and PSR are other great qualities of the 797.
In short, I don't dig the deregatory remark about "but when the preamp is selling for big bucks, one might expect attention to detail". You imagine problems that aren't really there, and neglect all of the good reasons one might have to use the 797.
It's not a perfect world, and I share John's wish for an OPA134 with <1nV voltage noise, but it isn't there.
Last edited:
I stand corrected. It would be interesting to know how much the noise figure is changed with the "equivalent." Again, a few dB of noise floor is not earth-shattering, but why not make the very best if you can?
I admit I don't pay much attention to data sheets and there is much more money in medical/telecoms. When I did pay attention most digital audio discussion was engineer to engineer. It would be unconscionable to have a product with such an obvious flaw.
It's either
5534 for lowest cost and some good all round tradeoffs
797 for superior noise performance
Discrete parallel single ended JFET input for lowest noise
LM49xx for lowest distortion and noise performance somewhat less that 5534
Take your pick. All require tradeoffs and none deliver #1 in everything.
1,2& 4 are sensible engineering solutions. 3 and others like it interesting design challenges and wha audiophiles apparently want and are willing to pay for.
5534 for lowest cost and some good all round tradeoffs
797 for superior noise performance
Discrete parallel single ended JFET input for lowest noise
LM49xx for lowest distortion and noise performance somewhat less that 5534
Take your pick. All require tradeoffs and none deliver #1 in everything.
1,2& 4 are sensible engineering solutions. 3 and others like it interesting design challenges and wha audiophiles apparently want and are willing to pay for.
Last edited:
I'm afraid this is incorrect, 2pA/rt-Hz of current noise will make 2nV/rt-Hz of voltage noise in a 1K resistance no matter what the resitance is made of. Excess noise is the excess current noise resulting from a DC voltage applied to the resistor. Measurement of excess noise in a cartridge coil would probably destroy it.
https://dcc.ligo.org/public/0002/T0900200/001/current_noise.pdf
Scott, for this specific case, is current noise correlated or uncorrelated with voltage noise?
I prefer: 2pA/rt-Hz of current noise will make 2nV/rt-Hz of voltage noise in a 1K Z no matter what the impedance is made of
Because 'the best' is defined by more than one dimension.
Just focussing on noise is either an oversight or an attempt to tilt the discussion.
So what defines the very best? OPA211 or, with even better specs on paper, the LME49990? You will only be able to find out by using these opamps in actual circuits and measuring them and, if so inclined, listening to the results.
Scott, you misunderstood what I posted.
I had already dealt with the current noise issue and a 1K Rs being slightly beyond optimum for the AD 797. A 1KHz resistor has about 4nV/rt-Hz thermal noise, dominating the 2nV opamp current noise, which adds just a bit over 10 percent to total noise here. I call that ball park right, good enough to let other design considerations prevail.
JCX was pointing to the excess noise created by the bias current in the cardridge, which you agree with me, is nonsense.
Last edited:
I understand that it is not the ultimate solution to cartridge loading but why not a variable capacitance switchable loading as I have on an older HK integrated amp? It has a normal position whatever that is and plus 100, 200, or 300 pf of switchable capacitance. Couldn't this option be used to add an adjust-ability to many phono preamps, perhaps with a low initial value and smaller steps for fine tuning?
your misinterpertation of my comment is where the nonsense came in - while I didn't spell out the bias current issue - I thought most would recall distortion is an issue with coil current biasing ferromagnetic cores - not that I really think it is a biggy when the MM transducer magnetics must have relatively large air gap compared to transformers and "large" input bias may be 1 uA
I completely distrust comments that uA bias currents in MC can have noticible magnetic effects - there is lots of air, usually fully saturated pole pieces, hi coercivity magnets giving near free space incremental permeability for any material within many times the diameter of the tiny moving air coils
Only in the 0 to 50 Kelvin temperature range. Down there, defect free will lower the collision rate significantly.
Room temp, no.
jn
- Status
- Not open for further replies.