SG-9500 muting JFET

One of the channels in my SG-9500 takes significantly longer to un-mute on power-up. Each channel has a JFET that blocks signal until its gate sees a certain voltage. Since both gates are fed from the same source it seems clear one of of my JFETs is going bad.

The part, 2SK30A, is long obsolete and I haven't found any NOS source. UTSource has a clone/knockoff which would probably be fine, but shipping from them is either very slow or expensive. Anybody know of a sub that would work for this application?

BD is still there?? Every link I've ever followed to them has been dead!

Thanks!

grandpaslab said:

One of the channels in my SG-9500 takes significantly longer to un-mute on power-up. Each channel has a JFET that blocks signal until its gate sees a certain voltage. Since both gates are fed from the same source it seems clear one of of my JFETs is going bad.

The part, 2SK30A, is long obsolete and I haven't found any NOS source. UTSource has a clone/knockoff which would probably be fine, but shipping from them is either very slow or expensive. Anybody know of a sub that would work for this application?

Click to expand...

NOPE, it's NOT your JFETS:

First, swap the JFETS between the channels to be ABSOLUTELY sure. I think you will be surprised.

You see, there are CSSA Sanyo sky blue capsistors (capacitors that RELIABLY fail to become resistors) isolating them from Q7's DC.

c33 (x2), C46(x2) electrolytic 4.7 10v cssa

In my judgement, those are far more likely to fail than the JFETS.

A recap is in order, AT LEAST all the cssa caps!! There are four in the unit.

mouser.com 647-UKL1E4R7KDDANA 4.7uf 25v low leakage capacitor, electrolytic
awr-095 power supply

awr-095 c8 1000 35v cea 102p 35 647-UPW1V102MHD 1000 35
awr-095 c9 1000 35v cea 102p 35 647-UPW1V102MHD 1000 35
awr-095 c10 100 25v cea 101p 25 647-UPW1V101MPD 100 35
awr-095 c11 100 25v cea 101p 25 647-UPW1V101MPD 100 35
awr-095 c12 100 10v cea 101p 10 647-UPW1V101MPD 100 35 **npairv**
awr-095 c13 100 10v cea 101p 10 647-UPW1V101MPD 100 35 **npairv**
awr-095 c16 1 50v cea 102p 50 647-UPW1H010MDD 1 50
awr-095 c17 10 35v cea 102p 35 647-UPW1H100MDD 10 50 **npairv**
awr-095 c18 22 16v cea 220p 16 647-UPW1H220MDD 22 50 **npairv**

awg-036 1 of 2 equalizer boards

awg-036 c31 220 16v cea 221p 16 647-UPW1V221MPD 220 35 *for margin*
awg-036 c32 220 16v cea 221p 16 647-UPW1V221MPD 220 35 *for margin*
awg-036 c34 47 10c cea 470p 10 647-UPW1E470MDD 47 25 **npairv**
awg-036 c35 22 10v cea 220p 10 647-UPW1H220MDD 22 50 **npairv**
awg-036 c36 100 10v cea 101p 10 647-UPW1V101MPD 100 35 **npairv**
awg-036 c40 22 10v cea 220p 10 647-UPW1H220MDD 22 50 **npairv**
awg-036 c41 22 10v cea 220p 10 647-UPW1H220MDD 22 50 **npairv**
awg-036 c42 22 10v cea 220p 10 647-UPW1H220MDD 22 50 **npairv**

awg-036 c33 4.7 10v cssa 4r7m 10 647-UKL1E4R7KDDANA
awg-036 c47 4.7 10v cssa 4r7m 10 647-UKL1E4R7KDDANA

awg-036 q8 2sc1312 512-KSC1845FTA
awg-036 q9 2sa725 512-KSA992FBU
awg-036 q10 2sc1312

**no price advantage in reducing voltage**

unusual small signal jfets

markthefixer said:

NOPE:

First, swap the JFETS between the channels to be ABSOLUTELY sure. I think you will be surprised.

You see, there are CSSA Sanyo sky blue capsistors (capacitors that RELIABLY fail to become resistors) isolating them from Q7's DC.

C46 electrolytic 4.7 10v cssa

In my judgement, those are far more likely to fail than the JFETS.

mouser.com 647-UKL1E4R7KDDANA 4.7uf 25v low leakage capacitor, electrolytic

unusual small signal jfets

Click to expand...

Mark, thanks for the intel. Ain't no capsistors on my board, tho! I've already done a full re-cap--those 4.7's are now WIMA stacked film. My money's on the JFET.

DC leakage BACKFEED from the output jack on that channel could also be the culprit. There is no DC blocking caps from an external influence.

Swap your output rca plugs and see if the "same" channel does it. That means an external influence

If swapping the RCA's moves the problem, than it is internal to the unit.

THEN swap the jfets and see if the problem moves.

IF NOT, then something else in the circuit is causing the problem.

A bad jfet will not do a "time delay" like that. Especially when both gates are fed by the same signal - which is a heck of a lot faster than your (and my) ears...
Check the diodes as well. 1s2473. THAT could in some way change the shared gating signal.

MY money's on something other than the jfet.

Did you follow my link? the only detraction is that their minimum quantity is 20.

That circuit isn't subjected to high DC voltages, so a normal (30v) JFET would do in this muting application.

markthefixer said:

DC leakage BACKFEED from the output jack on that channel could also be the culprit. There is no DC blocking caps from an external influence.

Swap your output rca plugs and see if the "same" channel does it. That means an external influence

If swapping the RCA's moves the problem, than it is internal to the unit.

THEN swap the jfets and see if the problem moves.

IF NOT, then something else in the circuit is causing the problem.

A bad jfet will not do a "time delay" like that. Especially when both gates are fed by the same signal - which is a heck of a lot faster than your (and my) ears...
Check the diodes as well. 1s2473. THAT could in some way change the shared gating signal.

MY money's on something other than the jfet.

Did you follow my link? the only detraction is that their minimum quantity is 20.

That circuit isn't subjected to high DC voltages, so a normal (30v) JFET would do in this muting application.

Click to expand...

Hey, thanks for the link. That looks like a handy site. I sense a stocking-up coming on...

DC backfeed seems unlikely, since the EQ has exhibited the same problem when attached to different receivers.

I did test the diode, it tests OK.

I do grok your logic--if one JFET is leaky, it should drag on the voltage to both, since they're connected to the same source.

/headscratch

I'll do some more testing tomorrow.

Don't forget it is a voltage DIFFERENCE in the jfet from the source-drain to the gate. If the dc bias on them source - drain area, is different, they will behave differently.\

Could be something on the board that has gotten leaky between traces.

Hey, this is interesting. When I said I tested the diode, I had tested the only diode on the board. This morning, with my brain somewhat more engaged, I realized that the diode I tested has nothing to do with the mute circuit. I was looking around the JFET for anything that could maybe cause the DC bias issue, so I pulled the only 2 parts that seemed like they could be involved--the 470pF drain-to-gate cap, and a carbon comp resistor that was in series with the gate. Which I then tried to locate on the schematic only to realize that the component that's supposed to be where this resistor is is D3, the 1S2473.

So somebody has modded this? I guess the CC resistor should have clued me in. This resistor is labeled as 700 ohms but has too many ohms for my cheapo meter to measure. Other channel has the same mod, though I haven't yet pulled that resistor.

Why would someone remove the diodes and put in 700 ohm resistors?

MTF has it right..

Nothing is ever completely fool proof, because fools are so ingenious.

Click to expand...