The problem appears to be the cartridge-input impedance noise mismatch that is caused by a relatively high In in the OP AMP selected, the LME49990. An AD797 or LT1028 would have the same problem, as they are also optimized for lowest En rather than lowest In. IF you put a low noise Jfet input OP AMP in place of the input IC then you would have lower overall noise. The selection of the input op amp is relatively optimum for most phono cartridges, BUT the Ortofon 2M black has one of the highest inductances in the industry of the modern cartridges today, 630mH, and 1.2K series resistance as well. I recommend returning the JC-3, or changing the phono cartridge.
Last edited:
I was afraid you were going to say something like that. Thank you for taking the time and writing an explanation, it makes perfect sense.
If I could bother your for one additional piece of information. Is there an elegant way to lower MM gain to around 40dB? If so, can you please point me to which resistor and the value it needs to be altered to? Thank you again.
First: if there is ANY chance the phono inputs have been "insulted" with large audio or static discharge, replace one chip and compare. High current in precision input devices may do no gross harm but raise the offset and hiss voltages.
______________________
{While I was typing this, John Curl posted similar thoughts more succinctly. But I did the math so I'm gonna post....}
With all respect to The Designer.....
I believe, on paper, it "does hiss" when the input is terminated in high MM phono impedances.
The LME49990 hiss specs are 0.88nV/rt-Hz and 2.8pA/rt-Hz. This gives an OSI of 314 Ohms. Looking into MC's <100 Ohms, you can't do much better than this (without a transformer). Looking into the 2M Black's MM impedance (1.2K <303Hz, 47K ~16KHz) the 303r OSI is far off the optimum.
Phono hiss is a pain to calculate. The source impedance is very variable (1.2K-47K) with frequency. The amplifier gain has a major slope. The ear says "hiss" mostly 1KHz-10KHz, though with RIAA bass-boost the LF random rumble has significance.
My cave-man approach is to say the source is 10K-20K (*) where it counts for "hiss". Sorta the split-difference between the 1.2K self-resistance and the 47K top-termination.
(*)Some MM carts run lower. In my day Grado boasted of this. Randomly picking their current $200 job: 45 mH and 475 ohms. This is under 2K impedance over most of the band. But your 2M black is similar to the early GE at 630mH, and much higher impedance over more of the band.
Taking 2.8pA/rt-Hz times 10K gives 28nV/rt-Hz. (We see the 0.88nV/rtHz is moot.) _IF_ the 28nV applied over 20KHz this would be 3.9 micro-Volts of hiss.
Taking 2.8pA/rt-Hz times the Grado's lower ~~2K impedance gives nominal 1 micro-Volts of hiss, significantly less.
Th LME49990's 10K source 3.9uV hiss is very significantly greater than its 0.124uV on a shorted input or ~~0.2uV hiss on a 100r input.
The 19 cent TL071 has much higher hiss voltage but "zero" hiss current. Under the same over-simplifications we would calculate 1.2uV of hiss. A third, or 10dB less hiss. Yes, a '071 can't approach 4-oh THD or drive a heavy NFB network; also it is un-sexy.
The NE5534 has OSI nearer 6K, which makes it borderline for MC (we always end up with a head-amp/iron) but not far out for MM work. Great but not "vanishingly low" THD, good drive, etc.
And for background: the whole 47K interface evolved around vacuum-tubes such as 12SC7 and 12AX7, where the self-hiss was OTOO 2uV maybe 1uV (quite similar numbers to TL07x, though different spectrum).
BUT!! What is the sound of a phono preamp with no phono-disk in sight? Yes, I have spent restful nights with a neglected recordplayer hissing softly, but that is not the intended purpose of a $1K-$10K+ phono system. We usually judge with the needle ON the vinyl.
I'm going to wild-guesstimate and say a 5mV/cm/sec needle on very good vinyl outputs 5uV of surface noise (60dB below nominal recording level). Much higher on many commercial stampings.
So as a ballpark this design hisses less than the records which it is used with. Not a lot less, but the surface noise dominates (on paper).
OTOH I know a guy who was "amazed" at a phono preamp which sounded "silent" yet made a big sound when the needle dropped in a groove. The electronic hiss _can_ be much less than the record surface noise. I done good with three $0.10 NPNs. TL071 and '5532 do not stink in this test. A well-proportioned JFET input could have in-air hiss 20dB down from in-groove hiss.
So... is the hiss really intrusive In Use?
(Some old recordplayers shorted the wiring when the arm was at rest; this would mask the sitting-idle hiss.)
______________________
{While I was typing this, John Curl posted similar thoughts more succinctly. But I did the math so I'm gonna post....}
With all respect to The Designer.....
I believe, on paper, it "does hiss" when the input is terminated in high MM phono impedances.
The LME49990 hiss specs are 0.88nV/rt-Hz and 2.8pA/rt-Hz. This gives an OSI of 314 Ohms. Looking into MC's <100 Ohms, you can't do much better than this (without a transformer). Looking into the 2M Black's MM impedance (1.2K <303Hz, 47K ~16KHz) the 303r OSI is far off the optimum.
Phono hiss is a pain to calculate. The source impedance is very variable (1.2K-47K) with frequency. The amplifier gain has a major slope. The ear says "hiss" mostly 1KHz-10KHz, though with RIAA bass-boost the LF random rumble has significance.
My cave-man approach is to say the source is 10K-20K (*) where it counts for "hiss". Sorta the split-difference between the 1.2K self-resistance and the 47K top-termination.
(*)Some MM carts run lower. In my day Grado boasted of this. Randomly picking their current $200 job: 45 mH and 475 ohms. This is under 2K impedance over most of the band. But your 2M black is similar to the early GE at 630mH, and much higher impedance over more of the band.
Taking 2.8pA/rt-Hz times 10K gives 28nV/rt-Hz. (We see the 0.88nV/rtHz is moot.) _IF_ the 28nV applied over 20KHz this would be 3.9 micro-Volts of hiss.
Taking 2.8pA/rt-Hz times the Grado's lower ~~2K impedance gives nominal 1 micro-Volts of hiss, significantly less.
Th LME49990's 10K source 3.9uV hiss is very significantly greater than its 0.124uV on a shorted input or ~~0.2uV hiss on a 100r input.
The 19 cent TL071 has much higher hiss voltage but "zero" hiss current. Under the same over-simplifications we would calculate 1.2uV of hiss. A third, or 10dB less hiss. Yes, a '071 can't approach 4-oh THD or drive a heavy NFB network; also it is un-sexy.
The NE5534 has OSI nearer 6K, which makes it borderline for MC (we always end up with a head-amp/iron) but not far out for MM work. Great but not "vanishingly low" THD, good drive, etc.
And for background: the whole 47K interface evolved around vacuum-tubes such as 12SC7 and 12AX7, where the self-hiss was OTOO 2uV maybe 1uV (quite similar numbers to TL07x, though different spectrum).
BUT!! What is the sound of a phono preamp with no phono-disk in sight? Yes, I have spent restful nights with a neglected recordplayer hissing softly, but that is not the intended purpose of a $1K-$10K+ phono system. We usually judge with the needle ON the vinyl.
I'm going to wild-guesstimate and say a 5mV/cm/sec needle on very good vinyl outputs 5uV of surface noise (60dB below nominal recording level). Much higher on many commercial stampings.
So as a ballpark this design hisses less than the records which it is used with. Not a lot less, but the surface noise dominates (on paper).
OTOH I know a guy who was "amazed" at a phono preamp which sounded "silent" yet made a big sound when the needle dropped in a groove. The electronic hiss _can_ be much less than the record surface noise. I done good with three $0.10 NPNs. TL071 and '5532 do not stink in this test. A well-proportioned JFET input could have in-air hiss 20dB down from in-groove hiss.
So... is the hiss really intrusive In Use?
(Some old recordplayers shorted the wiring when the arm was at rest; this would mask the sitting-idle hiss.)
Last edited:
Yes, I can definitely hear the hiss with the needle on the record playing. Obviously it's a lot more difficult to pick it out, but it's there. Especially audible in between tracks.
Last edited:
Quick update on gain, in case anyone else is experiencing JC3 overloading preamplifier's inputs. I had a chance to take a look at the circuit today and alter the gain.
Simplest way from what I can see is to alter R7, which is originally a Dale CMF55 220 Ohm resistor. Changing it to 665 Ohms gets you more than halfway from original MM gain of 47dB to new target of 40dB. This was enough to stop input overload on all of my preamps.
To get closer to 40dB one can use an even higher value resistor, but I'll leave that up to you. Resistor should be 1% high quality low noise metal film.
Unfortunately, changing gain did not lower the hiss. It seems like this phono preamplifier was primarily designed for MC cartridges. I'll either have to live with the noise, or as its designer John Curl suggested, change to an MC cartridge.
Thank you to everyone who took the time to respond.
BTW, this lowers the gain on the MC settings as well.
Simplest way from what I can see is to alter R7, which is originally a Dale CMF55 220 Ohm resistor. Changing it to 665 Ohms gets you more than halfway from original MM gain of 47dB to new target of 40dB. This was enough to stop input overload on all of my preamps.
To get closer to 40dB one can use an even higher value resistor, but I'll leave that up to you. Resistor should be 1% high quality low noise metal film.
Unfortunately, changing gain did not lower the hiss. It seems like this phono preamplifier was primarily designed for MC cartridges. I'll either have to live with the noise, or as its designer John Curl suggested, change to an MC cartridge.
Thank you to everyone who took the time to respond.
BTW, this lowers the gain on the MC settings as well.
Last edited:
- Status
- Not open for further replies.