Pacific RIAA phono pre: failure!

[Malefoda]

I'll do my homework and I'll be back

 

[Malefoda]

On each Test Point (TP) voltage on left/right channel. Houston, we've got a problem.
Death has taken a Fet? And worst, it's sending DC to my cartridge :'(

 

[merlin el mago]

Bad news.

 

[Malefoda]

You mean I have to order matched pair on fleabay again? It's fried?

 

[merlin el mago]

Take off & measure Idss to be sure.

 

[Malefoda]

No idea, close to supernatural here: I must work, it does not...

Hi Kevin & fellows, I've been told "don't buy a phono amp, for far less money and few time you can build a great amp!" Lies! DIY stands for "Dig In Yard", to bury yourself of course!
Well you seem more optimistic than I am, hope we will find...

Check the tone arm wiring, and does the problem swap to the other channel when you swap the left and right inputs?

Nop, it's with the phono amp, changed TT and channel, still following the amp channel. Wich channel is ok I don't know, they're not the same...
TT is innocent.

Left and right inputs both return to ground somewhere on your nicely built boards? (No opens)

I'll do pictures but : it's a source selector with amp included for the phono, both channel having a dedicated single coper wire linking the whole left and the whole right RCA's ground to the main grounding point (the outside point for the TT ground wire), the DC socket with it's gnd only connected to that gnd point, the gnd track of each phono channel going only to that "star" point. The pics will say it all.

Measure dc voltages on fet drains and make sure channel to channel that they match to say within 5 - 10%, if they do not, try to match some fets as variations in transconductance, pinch off voltage, etc. may effect the channel to channel gain match. I would not expect the differences to be that large unless the fets were grossly different.

In my above message. I've found a lonely 2SK170 to replace the one under 1V. (in fact maybe it's the 11V one that is dead?! I have no idea how it should measure!) So with Fet out, over 20V there, with the spare Fet... 0.6V, the same :'(

Even better would be to measure the signal levels directly in each stage, verifying that levels match after the first stage, after the passive eq, and finally the second stage.

I may do it a bit later, I'm not that familiar with the spectrum/network analyzer we have at work. And that non voltage matching should be fixed first I guess.

Post a reasonably sized picture of each side of the PCB and let us have a look, perhaps something different between the two channels will stand out.
Finally have someone else take a look at your handiwork.

Second picture is the ps wires bent down for clear view. Last one is of usless stuff :'(''
I've shown it at work, no idea, they're not in such things anyway.

Merlin, I may build a little fet tester and check it out. If I can manage to make a working tester, 3 parts so complicated for me under curse!


Ok time to listen to CDs then...

 

[osscar]

I am soldering a number of pacific RIAA (p2p and pcb) - never had any problems (using both N and P channel JFET without gate stoppers) source resistors 47R and i use drain resistors - 4k7 - it will reduce gain slightly. maybe your Fets are fakes ?

 

[kevinkr]

Yeah, based on the voltage measurements I'm thinking there is something fishy about those fets. I'd expect something like half supply on the drains ideally speaking, what Idss family do these parts come from?

 

[Malefoda]

They're all fets bought on fleaBay, the seller states about these 2SK170BL:
"This is Original Toshiba FETs made in Japan.
The price is for 4(four) items - 2 very well matched pairs: I declare +/-2% to be on a safe side but really I select no more than +/- 1% (Vgs = 10V)."
I only needed 4 matched, so bought that
As for the spare one its is unknow...
Want me to measure something? Got these three-legged bugs out of pcb is quite tricky...
If any of you know where to find a matched pair for my first stage that'll be nice (or offer me a deal in MP )... I can exchange first stage fet for the second stage's one on the harsh left channel to check voltage again an see if it's Fet related.

Thanks!

 

[osscar]

there should be about about 7-10 V on drain (depending from PSU voltage 24-27V ) and about 170mV on source with 50r resistors for both jfets.

good luck!

 

[Malefoda]

The picture on the Spencer's offer thread shows the real Toshibas, I can tell that the one I got had straight legs, not pre-bended... smells more and more fakes.
I'll try to get some from him. And when get them will check capacitance, resistors and volatges without the fets in order to check if they'll settle in the very same environment.

 

[merlin el mago]

You can trust in Spencer, I'm a happy customer.

 

[Malefoda]

He asked me "You can select the idss you want" so what I do need?

 

[osscar]

i use about 8mA idss, but IMHO you can use any from BL range 6 - 10mA (if i remember correctly). In original Pacific circuit they use 5mA idss with 2k4 at drain and without source resistors.

 

[Malefoda]

Thanks

 

[kevinkr]

You can use a higher Idss range, but that then requires the use of source resistors which may significantly increase noise. I was going to recommend 5mA Idss based on the value of the drain resistors, a lot of other good members beat me to the punch..

 

[Malefoda]

Hi Kevin, Spencer will send me matched 6.5mA ones, do I need to add these source resistors? In fact there are resistors everywhere, where can I add some more
And if of any use I've put the amp on online circuit simulator.
https://www.circuitlab.com/circuit/gmk34z/pacific-riaa-prephono/

 

[Malefoda]

Ready...

Here I'm FET-read... voltages all OK, I can only pray now... I've removed the 100R on each source, hope the Toshiba's won't oscillate!

 

[Malefoda]

A final thanks for support fellows, it's now working like a charm =)

 

[merlin el mago]

Congratulations.