Salas hotrodded blue DCB1 build

[Salas]

Yes, and be careful not to touch adjacent ones with the probes.

 

[Erlend Sæterdal]

Erlend Sæterdal

What are you using around 6 capacitors ? sound better ?

This is Cotton but whool and Silk is better. Yes I think plastic degrade sound.

 

[zonker92]

Rare to hear such an error report, not inherent, not a known issue. I would suspect:
-Signal path connections/solder joints/wiring.
-Relay. Tap it with the back of a screwdriver in case it latched.
-Rare to expect a weak JFET in the audio section quad that went out, but touch each of them after system warm up to make sure you got no cold one. Or compare votlages across each.

Everything seems OK: solder joints, connections, wiring, voltage across the JFETs, tapping the relay, voltage at the power eyelets.

I hope it's OK to check the DC resistance across the input and output connections points on each channel, with the power off, because I did that, and the right channel shows about 1.2K and the left about 220K.

 

[BigE]

Looking at the data sheets, the variation current given Vds is low, but increasing Vgs will increase Ids. Maybe time for some yellow LEDS?

 

[zonker92]

Everything seems OK: solder joints, connections, wiring, voltage across the JFETs, tapping the relay, voltage at the power eyelets.

I hope it's OK to check the DC resistance across the input and output connections points on each channel, with the power off, because I did that, and the right channel shows about 1.2K and the left about 220K.

OK, the problem seems to be the little three-pin connectors at the inputs and outputs: I bypassed them and (fingers crossed) everything seems to work fine. I had a bear of a time crimping on the little springs clips on the ends (before soldering them), and every time I pushed them onto the pins some of the wires would bend and come loose and loose their connections.

At this point I may just switch those connectors to soldered-on wires.

 

[Salas]

OK, the problem seems to be the little three-pin connectors at the inputs and outputs: I bypassed them and (fingers crossed) everything seems to work fine. I had a bear of a time crimping on the little springs clips on the ends (before soldering them), and every time I pushed them onto the pins some of the wires would bend and come loose and loose their connections.

At this point I may just switch those connectors to soldered-on wires.

Congrats. It was suspect case 1, connections.

 

[Salas]

Looking at the data sheets, the variation current given Vds is low, but increasing Vgs will increase Ids. Maybe time for some yellow LEDS?

Green also, but don't use blue or superbright. Much noisier.

 

[zonker92]

Congrats. It was suspect case 1, connections.

Yup. Thanks for your help!

 

[BigE]

Thanks Salas! I tried orange, but not enough V drop. Green tomorrow. Why not just a resistor?

 

[Salas]

Soft for the purpose.

 

[Erlend Sæterdal]

Connections

I always threw thoose connectors away from the beginning.

Anyway is there difference in sound when changing from 5 ohm resistors to 2,2 ohm ?

 

[BigE]

Soft for the purpose.

Sorry, I don't understand... Soft?

 

[Erlend Sæterdal]

Sorry, I don't understand... Soft?

Maybee not " Superbright" Big E

 

[BigE]

I always threw thoose connectors away from the beginning.

Anyway is there difference in sound when changing from 5 ohm resistors to 2,2 ohm ?

Looking at the irf9240 and 240 datasheets, it looks like that would change Vds,. Creating a larger voltage drop across the group of 3 LEDs would change Vgs, having greater effect on Ids, which I *think* would have more of a sonic effect.

The 9240 and 240 have slightly different curves, so matching Vgs between the two is not going to give you matching Ids. ( Currently, the polarity with lower Vgs is showing a higher Ids.) I will report my results using one or more Green LEDs.

 

[Salas]

Larger VF allows more voltage across Rset which imposes more Id that ups Vgs as a consequence, is the right description of the mechanism. Vrset=VFsum-VGS.

 

[BigE]

Maybee not " Superbright" Big E

Soft using a resistor? That might mean a resistor could cause a variation in Vgs, while the Vf of the LED triple is a "hard" value?

 

[Salas]

The 9240 and 240 have slightly different curves, so matching Vgs between the two is not going to give you matching Ids. ( Currently, the polarity with lower Vgs is showing a higher Ids.) I will report my results using one or more Green LEDs.

And then you got each MOSFET having its own bit different Id/Vgs response as a sample. Perfection is a game of persistence. Especially here that it comes with no big prize, its more of a noble sport. Transconductance to really mismatch takes large chunks of Id differences between such MOSFETS in this bias region of interest.

 

[Salas]

Soft using a resistor? That might mean a resistor could cause a variation in Vgs, while the Vf of the LED triple is a "hard" value?

LEDS are active low impedance low noise voltage sources. A resistor is more susceptible to external variables.

 

[Erlend Sæterdal]

I use 0ne green led in each section. I changed to 3,3 ohm Kiwame and leds light is increased.

 

[Salas]

Optical blending of color and luminosity power with its neighbors. The current through them is the same like before the 2SK170 on their tale was actively sinking.