Pearl 3 Burning Amp 2023

Question.... let's say I like the P3 better then the P2.

I only got one turntable and no plans on doubling up ( I'd rather upgrade to an Ortofon Black, Ekos SE and Lingo 4 ).

What do I do with the P2 then?

You guys are terrible, really terrible.

I think the day comes when I will have to ignore the hardware fetish and just spin records...
 
First of all, i like the Pearls 🙂.

Without the outputstage Q7 & Q6, in my opinion you dont need it, i prefer a sligth modifikation for the secondstage. OP-Amp. is the NE5534. The accuracy of the RIAA-Fit is a little bit (with selected and matched component) better then Mr. Colburns choice - original design of the EQ.

Greetings from Germany,
HBt.
 

Attachments

  • iPearl_3_RIAA.png
    iPearl_3_RIAA.png
    6.4 KB · Views: 180
Last edited:
  • Like
Reactions: Gricko
A question of taste? Some quick modification of the firststage, now with more openloopgain for better distortion and control at LF. I like a -3dB (approx 5Hz) cornerfrequency at the beginning of the chain, so i add firstorder highpass ...

OP-Amp. maybe TL071 or even better.

Again,
a big thank you to Wahne C. and Nelson P. for longlife inspiration, passion and very fine amplifier.

HBt.
(greetings from germany)
 

Attachments

  • iPearl_3_1_Stufe_Kaskode.png
    iPearl_3_1_Stufe_Kaskode.png
    9.5 KB · Views: 172
Last edited:
If you plan to use single opamps like NE5534A, you'll need to create a new PCB layout. Wayne's PCB layout assumes the builder will plug in dual opamps like OPA2134.

You might want to take a look at post #146 of this thread. Wayne posted Pearl 3's measured frequency response RIAA inaccuracy. Measured on an actual, physical Pearl 3 in the real world {not a circuit simulation}. Here's Wayne's plot from that Forum post:

_
measuredfreq-jpg.1223563
 
Last edited:
Hello Mark Johnson,
your are right (i know), but the idea is not only based on simulation. In the past i build many Phono-EQ's, measure and compare it. Also i hear Wahne's explanation and see the impressive frequency-response ...

What do you thing about my quick posted schematic, sligth variation of the original ciruit?


My latest build of the original Pearl works perfectly, no failure yet. For MC- stepup i choose the LT 1028 operationalamplifer ... in the decade 1990 i also often use SSM-2015 from PMI, but the 2SK170 jFet's are amazing and the simplicity of Wahne's (and Nelson's Aleph 5) ciruit become an all time favorite (in germany). I love the Aleph 5 and the Pearl.


HBt.
(i know very well about the real world; measurement and build, component ... and the difference between simulations, and so on)
 
It's an illusion to get a better overall-fit of +/- 0.25dB. The deviation in our realworld buildings of RIAA-EQ's with pluged pickup (and cord, or wire) is never better that figur. Sometimes better, sometimes worse - it depands on clever choose the circuit, the OP-Amp, the value of components ... material, device form, the routing, ... in my example of the inputstage with TL071 you get +0,3dB to -0,1dB, with NE5534 +0,1dB to -0,1dB -> never better! It is impossible to get better flat response as +0,3dB to -0,3dB with 1% tolerance resistors and 2% tolerance capacitor - "belive me" [quote Douglas Self].

Not only the best approximation (with no derivation) counts. In Wahne's prototype he used the best approximation to 100nF. In my deepest respect to Wahne's (knowledge and overall-measurement, and the AudioPrecisionSystem, ...), but how did /get Wahne the measurement(?).


It's a nice EQ, i like the approach and implement that topologie many times - nothing really new, but quite interesting every time.
 
The core of the Cambride Audio 651P, the split-RIAA-EQ, is theoretically (with nominally value of R's and C's) much better, namely +0,03dB to -0,01dB. The problem is the inputstage, in particular the MC-Stage.

The overallresponse is often not ever pleasing, despite it - a good "Entzerrer" but the old NAD PP2 is really better. Why?


Greeting,
HBt.
(i only use Denon DL-103 and AT-OC9, sometimes Ortofon MC)
 
In post #146 Wayne claims that the curve shown is measured on a completed unit (Pearl3).

Here is what Wayne writes about the choice of components and also a comment on XS Phono (I think in post #149):

"The spec is + - .1 max on the XS but typically better. This Pearl 3 I used 1% Dale RN55 and 2% caps measured to .1% because I could. I guess you could call it + - .05 dB. If you use the parts straight out of the bag with no matching even on the fets I doubt you would see much difference. But being diy you might as well chrome it or gold plate it like the boards that are coming will be."
 
^^ This is correct.

Instead of buying fancy expensive boutique components, it would be much more efficient and beneficial to buy a handful of the capacitors and resistors you need, and a Mega328 tester, and match the components channel to channel. They don't even need to be centered on value, just get the closest ones in the same positions L and R.

In Pearl 3, there’s an additional benefit that the EQ capacitors are all 0.1uF, (and a 100pF trim) so you don't need to buy a bunch of different values.

If you don’t have one of these, you genuinely need one. Put any non-IC thing into it and Mega328 will tell you what it is. Resistor, Capacitor, Inductor, Diode, Transistor, Mosfet, etc… it’s a fantastic tool. I have 2.

https://a.co/d/dxEs8uN
 
Vishay Roederstein MKP1837 are -+1% tolerance.
100 nF is 7.5mm long and 9mm wide (11 mm tall). Can those fit the P3 board (5mm spacing)?
Then I could prepare for P3 and get some of these at the local dealer so I have some to select from?
 
Those will be a little tight, but should work.

Alternatively, you could buy more of the Kemet PHE426HJ6100JR05 , as the kit is going to have (8) of them in it already… along with everything else you need to stuff the RIAA boards.
 
Yes, I can see that best they are longest "between the legs".
They are also cheaper so I will get some of those.
Funny that Vishay writes that the MKP1837 is not for new designs and propose MKP385 instead. But those are really expensive!
I go with KEMET!
 
  • Like
Reactions: 6L6
I could not resist and purchased a bag with 1000 pcs. 1000 were just so much cheaper than getting bags of 10 pcs.
Then I have for rest of my life. They probably have a longer shelf-life than I have.
I may be able to help other if they are living close. Then it should be no problem to measure some close to 100 nF.
In Vishay data sheet they show at which freq. they have min. impedance before it raises caused by inductance (about 5-7 MHz). Could not find that for KEMET but they go straight to 1 MHz. Should be OK (also for my OCD).
Same place I could get also 10 pcs. ZTX851 to be shipped end of this month. So I grabbed those also. Next batch of those is far into 2024!
 
At least it was possible to order 100 pcs (from seller movtec_21 from Italy). He has a lot more and also have 250 pcs packages (which is of course cheaper).
I also got confirmation via email now. Paid with PayPal which is good. If any troubles I can get money back 🙂
Now I have used enough money for today I think 🙂 ......time to go go bed......

1700346582963.png
 
  • Like
Reactions: Lvandoorn and 6L6