Bob Cordell's VinylTrak

Finally....

DSCN7538.jpg


DSCN7536.jpg


DSCN7540.jpg


DSCN7541.jpg


DSCN7526.jpg


DSCN7530.jpg


Very nice JOB!!!!

Leo
 
Matching not needed :)

Hey Dontsov,

Since none of the experienced builder here has answered you yet, I'll give it a try. It is better than nothing, right? :)

In general, matching in audio will *always* give you a better performance. However, there is a point of diminishing returns. If I would build the VinylTrak, I would not match Q10/11 resp. Q29/Q30. My reasoning is as follows:
I assume you would match beta of the BJTs. I am very sure that the performance of the VinylTrak is mainly limited by the matching of (the Idss of) the input JFETs. That is why Bob uses dual JFETS (sry, can't find the thread righ now). Compared to the variation between the individual FETs, the input/output current variation in the current mirror due to unmatched beta of the BJTs will typically be negligable, especially since beta is already very high for the 2N5089. Therefore, I would rather spend my time listening to the preamp than matching transistors. :)
BUT, if you already have matched the transistors, of course use them. There is nothing to loose here. ;)

By the way, Bob found later that a 3-transistor current mirror gave better noise performance, see http://www.cordellaudio.com/preamplifiers/vinyltrak.shtml. I added the picture below.

I hope it helps :)
JB

fig1.jpg
 
Member
Joined 2006
Paid Member
@ Bob Cordell

Hi Bob

I am studying the schematic of the VinylTrak published in linear Audio Vol4 page 139...
Everything is quite clear to me but one transistor:

Q9 is a common base inside the loop of the wilson mirror..... would you care to explain it's function ?

Best regards

Ricardo
 
AX tech editor
Joined 2002
Paid Member
With all respect, you should read the article. Or do a 'find' on D1 for instance.

The absence of NFB allows the 75us RIAA time constant to be implemented with a hand-selected combination of 2 polystyrene shunt capacitors totaling 0.02uF at the high-Z VAS node. This means that the amplifier need not ever swing the amount of voltage that would otherwise occur at high frequencies if the amplifier had a flat gain of 10. Nodes D1 and D2 are connected to ground through relay contacts for conventional equalization. The contacts are opened for damped EQ, inserting a zero at 8kHz via R19.

Jan
 
With all respect, you should read the article. Or do a 'find' on D1 for instance.

The absence of NFB allows the 75us RIAA time constant to be implemented with a hand-selected combination of 2 polystyrene shunt capacitors totaling 0.02uF at the high-Z VAS node. This means that the amplifier need not ever swing the amount of voltage that would otherwise occur at high frequencies if the amplifier had a flat gain of 10. Nodes D1 and D2 are connected to ground through relay contacts for conventional equalization. The contacts are opened for damped EQ, inserting a zero at 8kHz via R19.

Jan
Yes,yes,find it.thanks Jan again
 
Can I use the 2sk489 to replace for BOTH the 2sk389 and LS844?
But check the datasheet the 2sk489 noice is 1.8nv and 2sk389 only 1.3nv,still better than 2SK489.

Which grade 2SK389 should be used.A,B or C?

The LSK389 has significantly lower voltage noise than the LSK489. They are both very good parts, but optimized for different applications. The LSK389 is a large die with high transconductance that leads to very low noise. However, as a result, its capacitances are larger. The LSK489 is a smaller die with quite low capacitances and smaller transconductance, and yet still achieves low noise performance, just not as low as that of the LSK389. Use of the LSK389 is particularly important in the MC part of the preamp. The LSK489 can and should generally be used instead of the older LS844.

These parts are readily available from Linear Systems. Look for the LSK parts rather than the 2SK parts. The old Toshiba 2SK389 is no longer available. I usually recommend using the B version of the LSK389.

Cheers,
Bob
 
The LSK389 has significantly lower voltage noise than the LSK489. They are both very good parts, but optimized for different applications. The LSK389 is a large die with high transconductance that leads to very low noise. However, as a result, its capacitances are larger. The LSK489 is a smaller die with quite low capacitances and smaller transconductance, and yet still achieves low noise performance, just not as low as that of the LSK389. Use of the LSK389 is particularly important in the MC part of the preamp. The LSK489 can and should generally be used instead of the older LS844.

These parts are readily available from Linear Systems. Look for the LSK parts rather than the 2SK parts. The old Toshiba 2SK389 is no longer available. I usually recommend using the B version of the LSK389.

Cheers,
Bob

Bob,thanks for you detail input.that is clear