Pearl Two

Hi all,

I just ordered the Pearl 2 boards from Passdiy, and selected components from Mouser/DK based on the spreadsheet posted earlier in this thread. I don't know much about electronics, but I have successfully built the Hagerman Piccolo, Bugle, and power supply, so I think I'll be able to build this as well wthout a problem.

In the notes for the Pearl 2, Wayne Colburn say that a wall wart can be used for a PS for the Pearl. What do you think of using a pair of these?

Power Supplies, Plug Supplies, 30V, 400mA HP DESKTOP SUPPLY, 15890 PD

Thanks,

Mark
 
Hi all,

I just ordered the Pearl 2 boards from Passdiy, and selected components from Mouser/DK based on the spreadsheet posted earlier in this thread. I don't know much about electronics, but I have successfully built the Hagerman Piccolo, Bugle, and power supply, so I think I'll be able to build this as well wthout a problem.

In the notes for the Pearl 2, Wayne Colburn say that a wall wart can be used for a PS for the Pearl. What do you think of using a pair of these?

Power Supplies, Plug Supplies, 30V, 400mA HP DESKTOP SUPPLY, 15890 PD

Thanks,

Mark

Pearl II needs symmetrical supply @ 24V. You'll have to buy two of them and hook them accordingly :)

Best,

nAr
 
Thanks - can you elaborate on the way they would need to be connected? I assume symmetric means +24v and -24v.

Hmmm ... I would say, output in series, minus of the 1st one connected to the plus of the other one, thus creating virtual ground between them two ... but it also depends a bit of what's in the supply, that the pic doesn't show ... some internal arrangments/components could be an issue. It could work fine too :)

nAr
 
Last edited:
Some questions for the experts...

I started stuffing the Pearl 2 boards today and I have a few questions -- please excuse my ignorance!

  • Should I use a jumper in C15?
  • The bill of materials posted earlier in this thread specifies a variable cap for C9. Since this doesn't fit the holes in the board, what's a good way to wire it in? Also, what's the purpose of the variable cap, as opposed to the 5pf cap specified?
  • R12 is marked 909. The BOM specifies 12a at 1k and 12b at 10k. How should these be wired to achieve 909 ohms?

Thanks! I still need to select a PS and an enclosure.

Mark
 
I started stuffing the Pearl 2 boards today and I have a few questions -- please excuse my ignorance!

Should I use a jumper in C15?

No, leave it open. And use C9 @ recommended value for a start.

The bill of materials posted earlier in this thread specifies a variable cap for C9. Since this doesn't fit the holes in the board, what's a good way to wire it in? Also, what's the purpose of the variable cap, as opposed to the 5pf cap specified?

You can add some legs to a variable murata cap, like this :

An externally hosted image should be here but it was not working when we last tested it.


Big thing is that it can allow to fine tune on square wave, with a plastic screw driver of course :)

R12 is marked 909. The BOM specifies 12a at 1k and 12b at 10k. How should these be wired to achieve 909 ohms?

Wire 1K and 10k in parallel, it does the 909R; or use better 0,1 % 909R resistor, preferred :cool:

nAr
 
Hi Andy or Nar ,

would this be a good starting place for a PSU ? /www.quasarelectronics.com/velleman/k8042-symmetric-1a-power-supply-kit.htm

Thank you , Rich

Yes, I think it's a nice one to start with :)

Seems to be LM317 / 337 based, and ajustable. You'll only need to add xformer.

For later upgrading, I suggest to use either a shunt reg as the one seen with the Salas Buffer, or a symmetric one based of the Pearl original PSU, like below.





I use this power supply with a new UGS preamp, it is very stable, dead quiet, excellent regulation, provides excellent sonics. Layout takes care of ground lines that have a "star" like pattern, this helps to lower the noises created in ground nodes.
I'm very happy with those supplies :) They can virtually be used on any voltage, with small resistance adjustments. I plan to use it again but @ 30V for upgrading of my Aleph Ono clone ;)

Best,

nAr
 
Pearl 1 channel imbalance

RE: 9 db difference left / right when using Dynavector 10x5, not seen with V15 -III

I think the issue has to do with differences in gain between the JFET input stages, which, for some reason is exaggerated when using the moving coil Dynavector (even though it is a moving coil, 10x5 is a 'high output' MC,it has a 2 mv output, not much lower than the Shure) Although the Dynavector has an output voltage similar to the Shure, I bet the CURRENT is much lower. (Although maybe I am NUTS, who knows)

Using a signal generator as input, I measure about 1 dB more gain in LEFT channel vs RIGHT

Anyway - the parts I have in this Pearl were a bag of parts assembled by a guy who never got around to building it; I bought the whole kit of parts, less transformers and chassis, DC umbilical, etc.

In this kit, the JFETS (2SK178BL) all had their tops painted white with whiteout, and marked with a 2 digit number an a number of dots. I have no idea what these are- gain readings? Transconductance? 'ON' resistance?

Should I have use these numbers in some way? Look at the schematic area, below, I have noted the numbers written on each FET in BLUE for Left, RED for right- should I combine these in some different order?
fets.png


PHOTO OF RIGHT CHANNEL FETS
RIGHT-FETS.JPG


PHOTO OF LEFT CHANNEL FETS
LEFT-FETS.JPG
 
You should measure the V drop on each source R, and calculate X 4 resulting idle current for the quad. Then, same measures on the other channel. if you have current imbalance it could explain the mismatch with a particular cartridge.

If the jfets are well matched, they should at +/- 0,02 mA between the quad. And stereo quads should behave the same ...

Hope it helps.

nAr
 
Yeah that might be it. I guess I could measure the current for each FET and then use matrix alegbra to work out which transistors could be used in each quad to give the closes match between the two channels, or use some brute force recursive thing in Excel to figure which two sets of four to use. Hate to unsolder a bunch of FETs, I hate unsoldering. Never looks as nice afterwards. But this could be it.
 
SSLV1.1 BiB

Hi Nar ,

got a PM from Salas regarding PSU . This is a design you mentioned above .
. Very versatile and a proven design . I have B1 boards and the Pearl kit ,so building something that can power both would be good practice . I also like the idea of being able to use it for a new Aleph amp later on .

Are the PCB available do you know ,or anyone know ?

thanks , Rich

BTW, I also forgot to say thank you for the help .... sorry !
 
Last edited:
Yeah that might be it. I guess I could measure the current for each FET and then use matrix alegbra to work out which transistors could be used in each quad to give the closes match between the two channels, or use some brute force recursive thing in Excel to figure which two sets of four to use. Hate to unsolder a bunch of FETs, I hate unsoldering. Never looks as nice afterwards. But this could be it.

You could also slightly vary R Source to get more precise currents between the quads. Even if it's practical, in theory it's not as good as one could think, plus it induces unsoldering too :(

I would suggest you use patience and unsolder carefully the 8 jfets, then testing them to know their real IDSS. If needed, you can mix them with some new ones that would have a good matching value. K170BLs are becoming rare and prices go up nowadays.

Hi Nar ,

got a PM from Salas regarding PSU . This is a design you mentioned above .
. Very versatile and a proven design . I have B1 boards and the Pearl kit ,so building something that can power both would be good practice . I also like the idea of being able to use it for a new Aleph amp later on .

Are the PCB available do you know ,or anyone know ?

thanks , Rich

I thinks Salas rather prefers using a shunty than the design I attached ;) My mistake, if I recall properly the B1 has more than a layout, original boards don't have Salas PSU, and there is a DCB1 version that has the shunt built in ... not sure what you want to use here :confused:

BTW, I also forgot to say thank you for the help .... sorry !

No problem :cool:

nAr
 
Last edited: