Hello All,
No worries, all this is for entertainment.
This Pearl 3 project is perfect for examining, dissenting and measuring.
Examining the assumptions, the riaa equalization curve should be included in the equations. PREAMPGAIN is not constant across the bandwidth. Reminds me of the old days with Pascal and ADA programs.
I like using FFT plots with a sound floor reality check; placing a real measurement at 1kHz. The FFT dB scale can be adjusted with a relative number so the FFT "sound floor" matches the real 1kHz measurement. (credit to jackinnj)
Many of the FFT plots I have seen are dominated by power supply noise. Real world you can hear power supply noise where there should be silence.
Thanks DT
Unfortunately, there are other paths for hum ingress beside power supply. All the eforts to cleanse the PS can be in vain due to minor mistakes in grounding scheme or wiring.
So MM cartridges with 5mv output needs better PSRR to get same relation between cartridge noise and PSU noise......
If there are three orders of magnitude margin designed into the power supply filtering built into the Pearl 3 why did you design a external power supply with multiple layers of filtration to use with it? Given a separation between the transformer(s) bridge rectifiers and the pre-amplifier is a good idea. Now you are using affective logic (foolish play) to argue against your own work.
I will give you the benefit. This is DIY after all. You are as entitled as the next person in poking, prodding and designing or measuring (in my case).
I believe you are correct that the signal/noise ratio is important. Just talking out loud, How low of a s/n dB before detail is lost in the noise?
Thanks DT
The Pearl 3 is a gift.
the correct way to measure PSRR is to remove the input capacitor. I am measuring the cap multiplier in isolation. before you run the test do a calibration run.
I wouldn't say it is eye-opening, it's just there ... because Pearl 2 uses 3.3mF and Pearl 3 uses 220uF plus a 100 ohm resistor (R21) on the base of Q8.
Jack, maybe consider simulating with a Jacked-around .MODEL of the ZTX851, setting "VAF" (-1 * Early_voltage) to 267 instead of 158. Since our actual measurement of Early voltage is 267 not 158. See post #291 of this thread. With higher ZTX851 Early voltage, the input stage's noise filter circuit (which includes Q8) will get better. And perhaps this may embolden daydreamer designers to wonder "could we drop in a different Q8 chosen for extremely high Early voltage?". Or perhaps not.
_
_
Last edited:
I have only now been tuning into this thread, after receiving one board at BAF.
It appears this build requires 2 boards, so if anyone else received just one board, you are welcome to this one for postage costs.
It appears this build requires 2 boards, so if anyone else received just one board, you are welcome to this one for postage costs.
Here's what an ideal RIAA-conforming playback curve looks like, when the gain is set to 76 dB. There's one order of magnitude higher gain at 10 Hertz, and one order of magnitude lower gain at 22 kilohertz. For simplicity I've omitted the lowest LF pole / warp filter / DC offset nuller.
_
_
Attachments
Here's a faceplate concept I've been playing with. Made in front panel designer, using a font that's free for non-commercial use. Except numerals, which are not free. I might see if Gianluca can make one for me.
Did I give you just one board? I hope not. If I did send your address and I will get you on if you want.
Chassis update. Jim and I are working with Gianluca one a 1 box prototype with a wall between PSU and RIAA - and a 2 box version. Drawings were tweaked and tuned over the past week, and the final revs were sent to Gianluca yesterday.
I expect the they’ll be produced in the coming days. With any luck and the help of a speedy FedEx system we will be able to show you some metalwork within about a week.
I expect the they’ll be produced in the coming days. With any luck and the help of a speedy FedEx system we will be able to show you some metalwork within about a week.
Indeed it is!
No teaser images yet until we have items in hand and can test configuration(s). I think it’s going to look amazing.
No teaser images yet until we have items in hand and can test configuration(s). I think it’s going to look amazing.
Hi Wayne,
No, the boards were making it down my row and someone handed me just one. I didn't know it was a 2 board build, so yes, please! I'd love to get the second board and build this. Putting the project on my white board for parts ordering and progress is what had me looking at the board again, then checking this thread.
Thanks so much, Wayne!
Greg
Hi Wayne your very gifted and the pearl 2 is still quite a gem as it still keeps true to the magic formula, can pearl 3 beat it? For this approach also consider adding a line driver thats not part of the equalization.
Last edited: