Building a symmetrical PSU B1 buffer

[dvb projekt]

Did you try the F5 with another preamp in the same system configuration?

Let me switch it to my Audionet, just a moment...

 

[dvb projekt]

Same result with the audionet and with both cables... a bit more silent

 

[Salas]

Problem with F5 and your cables maybe. Up its gate stopper in its circuit. Better ask Nelson first.

 

[buzzforb]

ONly time i had hum, was exposed or broken wire, or bad grounding. Guess what you call DC interaction. Once, was partially broken wire that was making contact with its solder point, but in such a way to create small consistent hum.

 

[wintermute]

Hi dvb-projekt, just noticed in the F5 thread that you found someone else with the same issue.

I'd try disconnecting the preamp altogether and running a jumper wire from the earth on one input socket to the earth on the other input socket (assuming they aren't connected already) and see if that causes your hum.

You may be getting an earth loop between channels out one interconnect lead and back in the other.

Tony.

 

[dvb projekt]

Hi dvb-projekt, just noticed in the F5 thread that you found someone else with the same issue.

I'd try disconnecting the preamp altogether and running a jumper wire from the earth on one input socket to the earth on the other input socket (assuming they aren't connected already) and see if that causes your hum.

You may be getting an earth loop between channels out one interconnect lead and back in the other.

Tony.

Hi Tony,

i have tried it. No hum when the input cables are disconnected, also with the wire between both cinch GND's...

 

[wintermute]

ok well at least it rules that out as a possibility It looks like it is an interaction with the preamp(s) then.

Tony.

 

[triode_al]

Hi all,

i have a funny problem with my DCB1.

Beacuse the old amp was mono you must have separate wires. Now all at once you connect them to a same ground again: you have now a big loop (Spule) that catches the hum from the mains.
So: wrap the two interconnects around each other (like twisted pairs), tape the two wires together, and presto . . .
hope this helps

 

[dvb projekt]

Beacuse the old amp was mono you must have separate wires. Now all at once you connect them to a same ground again: you have now a big loop (Spule) that catches the hum from the mains.
So: wrap the two interconnects around each other (like twisted pairs), tape the two wires together, and presto . . .
hope this helps

Thanks for your suggestion, but the problem is also with other interconnects, see Post #2458 . A bit better but not silent...

 

[KatieandDad]

Wierd.

I meticulously matched the LEDs that make up the 5.4V drop at the input of the DCB1 using the actual 2SK170BL that would be used in that part of the circuit.

Everything tested perfectly 5.38V at about 9mA (I was using the higher current unmatched 2SK170's.

Now all is soldered together on the PCB I have nearly 100mV difference between +Vcc and -Vcc. The series IRFPs aren't fitted yet so I am measuring at the gate stoppers which are (obviously) open circuit.

With a 15-0-15V transformer I am reading +17.4V and -17.3V.

Is this going to be an issue as the IRFPs aren't matched and the shunt reg should take care of any differences ????

 

[Salas]

CCS current=((3xVf)-Vgs)/Rset. The strings of 5 are going to set the +/- Vo. CCS currents are to be in the same ballpark and Vo polarities near enough. If the DC offset is going to be low in the audio output that is the only important criterion.

 

[KatieandDad]

So the 5V (3 x LED) are only setting the current through the regs.

 

[KatieandDad]

I've got approx 5.38V and the series resistors are currently 22R

I'm aiming for about 200mA - I did ask earlier about the voltage across the series resistors but no-one replied.

Using my BJT knowledge. With Vb at 5.38V, Vbe=0.6V so the voltage across the Emitter resistors needs to be about 4.78V.

I=E/R so I would be 217mA with a BJT. If that is still true with a MOSFET, I just don't know.

 

[Salas]

So the 5V (3 x LED) are only setting the current through the regs.

Those, the particular Mosfet Vgs, and the particular setting resistor. PMOS and NMOS are not having the same Id-Vgs curve in most cases anyway, no same Mosfet is guaranteed to have the same curve, and so it goes. But not crucial at all.

 

[Salas]

IRFP9240 shows about 3.5V Vgs @ 0.2A and Tj 25C. The junction will go hotter, so something like 1.6-2V across Rset is probably expected given the variations. Most use 8R2-10R 5W for circa 200mA CCS.

 

[KatieandDad]

So my 22R 3Watters may burn up then ???

I'm still not sure about the Nichicom caps. They do test OK.

 

[Salas]

They will just give less CCS current than you expected. Around 100mA maybe.

 

[KatieandDad]

I can play around from that value.

Some guys have tried as much as 600mA.

 

[KatieandDad]

Well I've finished the PSU part of the DCB1.

I've got +/- 9.58V give or take 10mV or one digit on the DMMs.

Running current with the 22Rs is 63mA - a bit lower than intended but everything is running cool.

The voltage seems to be dropping slowly as warm them up, not much but it started at 9.6V and after a 10 min test they were 9.5V.

The IRFPs have no heatsinking, they are cool to the touch.

Next step is to fit the elusive matched 2SK170BLs and then it's finished.

 

[udailey]

Remember, they dont have to be matched. AndrewT has covered that multiple times.