It is to take it to chassis. Is also the reason why all the chassis panels must be in continuity with each other. The cable shield is effectively chassis. Making a dedicated line to the ground post is a band-aid if it’t not constructed properly, but if it’s quieter with it, use it.
If the XLR housing is plastic, then yes, experimentation is in order.
If the XLR housing is plastic, then yes, experimentation is in order.
if the XLR housing doesn't make good electrical connection to the chassis then a wire is not a bandaid... it's mandated.
and better to a chassis bolt than the tonearm ground post, no? one less (or fewer) connection(s).
of course the enclosure needs to be constructed properly to work as intended.
and better to a chassis bolt than the tonearm ground post, no? one less (or fewer) connection(s).
of course the enclosure needs to be constructed properly to work as intended.
Hello All,
The DUT (Pearl 3) is on the bench powered up in the rubber jaws of my PanaVise. There is a 220R resistor across the input because it was handy. The resistors are all Vishay RN types mostly RN60’s. The Op-Amp is a 1656. C6 and C7 are both nichicon 100uf. The poly caps are mostly Vishay. The LM 7815 and LM7915 are both installed on the PCB.
The low gain jumper is installed at R22.
The Keysight bench supply is in series mode with 34VDC output, or 19VDC and -19VDC to the on PCB regulators. The power is a regulated 15 + / - VDC. The LEDs are both on.
The jfets are closely matched 2SK170’s with an Idss of 7.4ma. The current measured across R10 is a total current of 12ma. The jfets are plugged into sockets on the PCB, later I will roll in some 209’s and others.
The bias current through Q9 is 4ma.
Coming up will be noise, gain and 1KHz FFT distortion measurements.
Thanks DT
The DUT (Pearl 3) is on the bench powered up in the rubber jaws of my PanaVise. There is a 220R resistor across the input because it was handy. The resistors are all Vishay RN types mostly RN60’s. The Op-Amp is a 1656. C6 and C7 are both nichicon 100uf. The poly caps are mostly Vishay. The LM 7815 and LM7915 are both installed on the PCB.
The low gain jumper is installed at R22.
The Keysight bench supply is in series mode with 34VDC output, or 19VDC and -19VDC to the on PCB regulators. The power is a regulated 15 + / - VDC. The LEDs are both on.
The jfets are closely matched 2SK170’s with an Idss of 7.4ma. The current measured across R10 is a total current of 12ma. The jfets are plugged into sockets on the PCB, later I will roll in some 209’s and others.
The bias current through Q9 is 4ma.
Coming up will be noise, gain and 1KHz FFT distortion measurements.
Thanks DT
You are correct that is a typo. I tried to go back to edit but it was too late. Thanks for pointing that out.
Yes they are 1000uf capacitors.
Thanks DT
bummer
now you don't have space for proper M26 GND bolts, due to increased caps size
now you don't have space for proper M26 GND bolts, due to increased caps size
Heat conducting epoxy and 16d duplex nails will hold it down.
I think.
See the attached FFT plot of P3 in shielded16ga. steel cookie tin. Power wires are not shielded. FFT 48K length and 5 averages.
Not a hint of power line harmonics.
Removing the 220R resistor at the input and connecting the cartridge, plus grove noise will change things.
Thanks DT
220R is ~ 1.9nV/RtHz at room temperature...you have to adjust for the gain of the P3 (66 to 70dB) -- you should get a flat line past the 1/f noise corner.
One notable DIYAUDIO'r uses a 60R (~1nV/RtHz) resistor for his "sanity test" in the low noise amplifiers he has published over the years.
The 50/60Hz is extremely difficult to rid even if you are using a battery supply and muShield! I leave it to the men and women at Keithley!
One notable DIYAUDIO'r uses a 60R (~1nV/RtHz) resistor for his "sanity test" in the low noise amplifiers he has published over the years.
The 50/60Hz is extremely difficult to rid even if you are using a battery supply and muShield! I leave it to the men and women at Keithley!
Did anyone make any calculation so as to find minimum voltage- and current noise requirement for input stage op-amp U1a in order that its noise contribution be insignificant?
Parts arrived yesterday, and I am considering placing an order to various op-amp to try out.
I think I have some "brown-dog" type boards that will allow me to compose a dual op-amp from two single SOIC types thus allowing different devices to be used for U1a and U1b. Any suggestions to what parameter to look for position?
Parts arrived yesterday, and I am considering placing an order to various op-amp to try out.
I think I have some "brown-dog" type boards that will allow me to compose a dual op-amp from two single SOIC types thus allowing different devices to be used for U1a and U1b. Any suggestions to what parameter to look for position?
In LTSpice, substitute opamps -- for U1a, run a noise simulation, 20Hz to 20kHz and plot ONOISE. Ctrl-Click on the cursor to give integrated noise over the bandwidth 20Hz to 20kHz.
I have had nice results with LT1113 and OPA2604, two JFET opamps.
Hello All,
Still sorting out some observations.
I did input a test signal into the P3 PCB on my bench with the 4 2SK170's (Idss := 7.4ma each) the gain measured 50.4dB.
I am not thrilled with the AC power line harmonics that apparently leaked into the input and or BNC test lead from the analyzer. I will sort things out a bit more prior to posting that FFT plot.
jackinnj mentioned that there is a sort of standard using a 60R resistor at the input of low noise amplifiers for a "sanity Test".
I repeated the FFT posted the other day, now this time with a 60R resistor across the P3 input. Again the there is no apparent AC powerline harmonic distortion leading to to the thought that the power supply including the on PCB LM7815 and LM7915 regulators are innocent.
Also the 6 foot long conductors from the bench power supply are not shielded yet there are no apparent power line harmonics in the FFT plot.
So it looks to that a little bit of AC distortion at the input with 50.4dB's of gain can impact the noise at the output.
Thanks DT
Still sorting out some observations.
I did input a test signal into the P3 PCB on my bench with the 4 2SK170's (Idss := 7.4ma each) the gain measured 50.4dB.
I am not thrilled with the AC power line harmonics that apparently leaked into the input and or BNC test lead from the analyzer. I will sort things out a bit more prior to posting that FFT plot.
jackinnj mentioned that there is a sort of standard using a 60R resistor at the input of low noise amplifiers for a "sanity Test".
I repeated the FFT posted the other day, now this time with a 60R resistor across the P3 input. Again the there is no apparent AC powerline harmonic distortion leading to to the thought that the power supply including the on PCB LM7815 and LM7915 regulators are innocent.
Also the 6 foot long conductors from the bench power supply are not shielded yet there are no apparent power line harmonics in the FFT plot.
So it looks to that a little bit of AC distortion at the input with 50.4dB's of gain can impact the noise at the output.
Thanks DT
Last edited:
Is there a version of the OPA1612 that is a DIP package, not SMD?
I've noticed the Burson op Amps are heat sensitive... they need at least half an hour of warm up to settle down their treble.
This I found with my DAC once I put the higher gain 252 SIT into the system. My set up now has a lot more gain, so I can hear noise more clearly. The opamps have a more forward treble until they warm up... I have not detected this in the P3 since I have not yet figured out its sound ( too many changes lately in the system ).
The default op amps in my preamp were LME49720NA. Which sound quite good.
I've noticed the Burson op Amps are heat sensitive... they need at least half an hour of warm up to settle down their treble.
This I found with my DAC once I put the higher gain 252 SIT into the system. My set up now has a lot more gain, so I can hear noise more clearly. The opamps have a more forward treble until they warm up... I have not detected this in the P3 since I have not yet figured out its sound ( too many changes lately in the system ).
The default op amps in my preamp were LME49720NA. Which sound quite good.
Last edited:
OPA1612 are SMD only.
If you were going to choose one opamp for Pearl 3 and never think about it again, OPA1612 would be high on the list.
LME49720 would also be on that list.
If you were going to choose one opamp for Pearl 3 and never think about it again, OPA1612 would be high on the list.
LME49720 would also be on that list.
Yeah, I am familiar with the choice of OPA1612. I was hoping to hear them before making my final choice and for that I can't do SMD.
Have you heard the LME49720 and OPA1612? How do they compare? The former sounds pretty good actually.
Have you heard the LME49720 and OPA1612? How do they compare? The former sounds pretty good actually.
Yes, you can, with an adaptor.
Yes. They both work beautifully. One is DIP, the other is SOIC, one is from the height of the Bob Pease era at National Semiconductor, the other a fairly modern Texas Instruments creation, and for other comparisons, please consult the datasheet for the information you are after. I agree with your opinion of the 49720.
Yes. They both work beautifully. One is DIP, the other is SOIC, one is from the height of the Bob Pease era at National Semiconductor, the other a fairly modern Texas Instruments creation, and for other comparisons, please consult the datasheet for the information you are after. I agree with your opinion of the 49720.
As promised, an SMD soldering tutorial is here.
Any feedback is welcome.
Any feedback is welcome.
Nice method and thank you for the video. Nevertheless I frefer Jim's (6L6), beautifully explained on online guide. 🍻