I don't know. That's why I asked. It's helpful to know what people did... when it works and when it doesn't.
All this grounding stuff seems fiddly (though it probably isn't).
well, for instance, the guide indicates that the boards are connected to the chassis... and you didn't do that exactly.
there seems to be a lot of ways to skin the cat and still get good results.
there seems to be a lot of ways to skin the cat and still get good results.
I tried your configuration: slight hum on the right, strangely no hum on the left, and very slight noise in both channels... with other configs the same.
Next step: chassis for the PSU...
Strange with the "No noise/hum" on one channel only.
The slight "hiss" cannot be avoided. At 70dB amplification small hiss is unavoidable.
First you should look at any..... as in ANY difference in connection between left and right P-3 board.
If you´re absolutely sure, there is none, then yes....... PSU in separate housing 👍
The slight "hiss" cannot be avoided. At 70dB amplification small hiss is unavoidable.
First you should look at any..... as in ANY difference in connection between left and right P-3 board.
If you´re absolutely sure, there is none, then yes....... PSU in separate housing 👍
No, there is absolutely no difference. The only one if I will is that the signal cables are crossing each other (without contact) in the right channel, but being shielded cables it can't be an issue.
Now I even removed the psu board grounding but no change.
Parts for the psu chassis ordered, and I will use a shielded, 1.5 m long cable.
However, I noticed a strange thing... I use a grounding block where all my components are grounded. Connencting the Pearl's signal ground to the grounding block causes a loud hum. Funny...
Now I even removed the psu board grounding but no change.
Parts for the psu chassis ordered, and I will use a shielded, 1.5 m long cable.
However, I noticed a strange thing... I use a grounding block where all my components are grounded. Connencting the Pearl's signal ground to the grounding block causes a loud hum. Funny...
OK.
Then there´s no hun on input either,
Try with some other RCA-cables from riaa to pre/amp then.
Or try to switch the existing ones between channels??
@Plott Optimal grounding arrangements with your setup will depend how the rest of your kit is grounded.
The alternatives are variations on the connection between mains earth and signal ground. From a low resistance to no connection.
Try them out and see what works best.
Btw when you say you use a grounding block for signal grounds, what does this do? If it is an additional connection between all the signal grounds, isn't it creating a signal-ground loop?
The alternatives are variations on the connection between mains earth and signal ground. From a low resistance to no connection.
Try them out and see what works best.
Btw when you say you use a grounding block for signal grounds, what does this do? If it is an additional connection between all the signal grounds, isn't it creating a signal-ground loop?
Last edited:
Thats a serious rcrcrcrcrcrcrcrrcrcrcrcrcrcrc you have there boydk, .. works well, and then feeding the pearl regulator, not much left, im unsure of pearl pssr, but with 50-60db gain its needed..
Hehe...... It´s works beyond expectations.
One of my previous builds had UDP-like psu´s adjusted to 19 volts before "hitting" the Pearl. No lack of PSSR.
Despite that, this build sounds better, and is my most well sounding and most noies/hiss-frre build to this date.
I´m satisfied beyond expectations.
All this fine tuning, component selection and matching, and settling for your preferred opamp really brings out
all the best in Wayne´s Pearl-3.
I really meant it, when I said, that this has to be the DIY-gadget og the decade
I tried several opamps but there are not much difference.
This opamp is completely different level than Muses.
https://shop.cqpub.co.jp/hanbai/books/I/I100545.html
https://www.marutsu.co.jp/pc/i/3186638/
This opamp is completely different level than Muses.
https://shop.cqpub.co.jp/hanbai/books/I/I100545.html
https://www.marutsu.co.jp/pc/i/3186638/
My other phono preamps are working without any issue with the cables.
Thats the spirit plott..👍 lovely chassis, i like the 3 separate rooms..
I wish mine had better space.. i went through same issues also.. i fixed it by putting a gnd wire on all parts of chassis and return all to star ground, everything now has same potential .
I wish mine had better space.. i went through same issues also.. i fixed it by putting a gnd wire on all parts of chassis and return all to star ground, everything now has same potential .
It was a joke, i. Pictured 64 capacitors all with r in between..
Sorry thought it was my post you referred to.
No point in double r .. i would say crc on half board, the rest directly linked, .. your low esr buffer
Sorry thought it was my post you referred to.
No point in double r .. i would say crc on half board, the rest directly linked, .. your low esr buffer
I had an ai calculate some scenarios..
First scenario.. 8 by 8 capacitors 35v 1000uf.. all interlinked.
Second scenario.. 8 by 8 capacitors all with 1r in between.
+-/15v 200ma current draw. Dc is infused with 200mv ripple at 45khz.
All-parallel board
64 × 1000 µF in one big lump gives you 64 mF and almost zero ESR. At 45 kHz the rail impedance is so low that the 200 mV ripple is beaten down to the low-µV/mV region, limited only by the regulator and the wiring. There’s essentially no DC drop, but the in-rush current is savage (> 100 A), so you need a soft-start resistor, NTC or relay.
CRC board
Split the bank into eight blocks of 8 × 1000 µF and put a 1 Ω resistor between each block (7 Ω total). Each RC section has a 1 ms time constant, so 45 kHz is far above the corner and each step adds about −67 dB. In practice you can hit −120 dB or better, which is overkill for audio.
Price you pay: the 7 Ω chain drops V = I R = 0.2 A × 7 Ω ≈ 1.4 V on each rail. That’s 0.28 W of heat per rail and your ±15 V shrinks to roughly ±13.6 V.
How much headroom?
Add at least the lost 1.4 V plus a little safety margin. Feed the CRC with about ±17 V (34 V total) so that, after the resistive drop and normal regulator overhead, you still have a clean ±15 V at the load.
In short
• Parallel bank: dead-quiet, no voltage loss, just tame the turn-on current.
• CRC chain: even quieter, but wastes about 1½ V and some heat; raise the raw DC to ≈ ±17 V if you insist on keeping ±15 V out.
First scenario.. 8 by 8 capacitors 35v 1000uf.. all interlinked.
Second scenario.. 8 by 8 capacitors all with 1r in between.
+-/15v 200ma current draw. Dc is infused with 200mv ripple at 45khz.
All-parallel board
64 × 1000 µF in one big lump gives you 64 mF and almost zero ESR. At 45 kHz the rail impedance is so low that the 200 mV ripple is beaten down to the low-µV/mV region, limited only by the regulator and the wiring. There’s essentially no DC drop, but the in-rush current is savage (> 100 A), so you need a soft-start resistor, NTC or relay.
CRC board
Split the bank into eight blocks of 8 × 1000 µF and put a 1 Ω resistor between each block (7 Ω total). Each RC section has a 1 ms time constant, so 45 kHz is far above the corner and each step adds about −67 dB. In practice you can hit −120 dB or better, which is overkill for audio.
Price you pay: the 7 Ω chain drops V = I R = 0.2 A × 7 Ω ≈ 1.4 V on each rail. That’s 0.28 W of heat per rail and your ±15 V shrinks to roughly ±13.6 V.
How much headroom?
Add at least the lost 1.4 V plus a little safety margin. Feed the CRC with about ±17 V (34 V total) so that, after the resistive drop and normal regulator overhead, you still have a clean ±15 V at the load.
In short
• Parallel bank: dead-quiet, no voltage loss, just tame the turn-on current.
• CRC chain: even quieter, but wastes about 1½ V and some heat; raise the raw DC to ≈ ±17 V if you insist on keeping ±15 V out.