Simplistic NJFET RIAA

[Salas]

You will need two 34-36VAC feeds for two bridges and main capacitors. 1.41 is not reached dead on bcs of losses. Maybe 1.33. You can test with 470R 10W dummies across the raw DC outputs for final voltage level at a consumption reminding the FSP channels main CCS. Up to 50V is safe.

 

[LucasAdamson]

Oh bum. I need another transformer then.

 

[Salas]

Manufactor speces are usually all over the place.

If Marra will measure AC across his new Tx secondaries will much likely be >than their 30VAC nominal for instance.

 

[Salas]

Oh bum. I need another transformer then.

Ask Marra how to get one like his. We know it makes 46.5VDC when loaded by FSP on UK mains with no hum issues. Beaten path tactic.

 

[stajo]

If Marra will measure AC across his new Tx secondaries will much likely be >than their 30VAC nominal for instance.

Likely it is speced for 10 times its present load or something.

 

[Marra]

Am I right in thinking that with a 230v - 2x18vAC transformer (from an old, disbanded project), on a 240v supply (UK) I would be looking at 37.6vAC x 1.42 = 53vDC less diode losses etc., at the Folded shunt boards? Is this too high?

That's what I had when I tried with 18V secondaries.I think the recommended is 46V but you could always use dropper resistors if using the GB boards but why waste voltage as heat.
I bought a toroidal from Air Link with 30V secs. which actually gave close to 34V as where I live the mains never drops below 240.
Air Link have a quick turn around;ordered saturday arrived this morning;they must keep a good stock.

 

[Salas]

Did it give bit more than 34VAC+34VAC or bit less?

 

[Salas]

P.S. Was that secondary VAC off load or when loaded with the FSP?

 

[Marra]

That was loaded with the FSP.
I couldn't wait until tomorrow and spent a couple of hours listening to lp's.I think the Decca likes it

 

[Salas]

That's nice bcs the Decca is peculiar. So, given the trafo was loaded, it should have about 1.35 full wave rectification ratio DC/AC in this app. Is it 80-100VA rated model?

 

[LucasAdamson]

My calculations for 30v are 30 /230 *240 = 31.3VAC * 1.41 = 44.1VDC less diodes etc.

A 33 VAC Tx comes out at 48.5 VDC. Hmmm

 

[Salas]

But the Airlink 230/30+30 gave 46.5+46.5VDC because we barely load it to its true current ability. I would order same secondaries and VA model from same manufacturer as Marra if in your shoes. Because all unknown factors are already tested. I would only ask them if they offer magnetic shielding peripheral band as an extra option.

 

[LucasAdamson]

Thank you. Indeed it is to be a DCB1 buffer preamp with the FSP as integrated RIAA. Akouo select & volume with a Muses 72320 chip.

It doesnt show at the pic but the boards are actually piled up to the final height so the legs are about right. Ill solder them when its fixated.

I am to fodder the alubox with thin ply and fodder the heatsinks even more so they only radiate to the outside.

The DCB1 will have its own foddered compartment with heatsinks with the same principle.

Plus a separate PSU box with a lot of different voltages in the umbilical cord connecting the two.

Stajo,

It looks like a very practical solution. My only question is, if you are to have a separate power supply unit feeding the unit by an umbilical, isn't it better to bring DC into the pre-amp with the umbilical, rather than AC? The DCB1 has onboard rectification. Are you planning to move the rectification off the DCB1 board and into the PSU box?

 

[LucasAdamson]

My matched pairs of 2SK170BL are as follows:

6.60 mA pair
6.89 mA pair
7.07 mA pair
7.39 mA pair
7.47 mA pair
7.52 mA pair
7.62 mA pair
8.12 mA pair

They were, however, all tested with 9v, back when building DCB1, F5, etc.

I also have a near match at 9.80 / 9.96, which when I retested at 22v, came out at 10.31 / 10.50 which shows that voltage is critical to evaluating IDSS.

Should they all be retested please, and if so at what voltage?

 

[Salas]

Use 10V for your re-testing confirmations if you got a variable PSU. Leave each one sample enough minutes on power until it stops climbing on the reading.

 

[LucasAdamson]

OK. Yes they can sometimes climb for about 5-10 minutes as I remember, which complicates testing. Doesn't the actual voltage they will encounter in the circuit matter though?

 

[stajo]

Stajo,

It looks like a very practical solution. My only question is, if you are to have a separate power supply unit feeding the unit by an umbilical, isn't it better to bring DC into the pre-amp with the umbilical, rather than AC? The DCB1 has onboard rectification. Are you planning to move the rectification off the DCB1 board and into the PSU box?

Yes. I make a pre reg filter C-R or C-L in the PSU-Box and plug directly on to the onboard C.

 

[Salas]

OK. Yes they can sometimes climb for about 5-10 minutes as I remember, which complicates testing. Doesn't the actual voltage they will encounter in the circuit matter though?

Practical test:

10V K170BL 7.48mA (3 minutes are enough)

9V same sample 7.44mA

They will not encounter more than 9-10V VDS in circuit.

 

[Marra]

That's nice bcs the Decca is peculiar. So, given the trafo was loaded, it should have about 1.35 full wave rectification ratio DC/AC in this app. Is it 80-100VA rated model?

I went for the 120VA as it was only a couple of pounds difference and added an interwinding screen.Didn't bother with the Goss band as the power supply is in a separate case so I didn't think it was necessary.

 

[AndrewT]

Will do once I sort out photobucket.

Attach your pics.