SSLV1.1 builds & fairy tales

[TheShaman]

I agree.
It'd be a good idea to have provisions on the PCB for installing resistors in parallel.

 

[Siberia]

I want to use SSLV 1.1 for feeding 28VDC to my future Pearl II RIAA. Getting a 2x24 R-Core, I would get around 26VAC on my Tx secondaries because of local mains, and 26 X 1.33 = 34,5VDC after rectification. Is it enough for an extra voltage? Or are two 2x15 R-Cores recommended for bigger DCin?

 

[AndrewT]

No.

How many times does it need to be repeated that the Salas Style Regulator needs >=7V across the CCS?

 

[Siberia]

I'm aware what's written in the guide. It says, at least extra 5V, but 7-10V recommended. The reason I'm asking this because i get <7V across, but with two 2x15 I would get into 45VDCin, which may be too much for cooling it down. All I wat to know which way is prefered.

 

[Salas]

Burn as much as you want on a CRC pre filter so you have control and extra filtering? Longer pants have always the better chance to get cut to measure.

 

[AndrewT]

7-10V recommended. The reason I'm asking this because i get <7V across,.........

Have you checked what DC voltage with ripple voltage superimposed that will appear at the regulator power input when the mains voltage swings over the range from maximum supply to minimum supply?

It is the voltage amplitude at the minima due to the ripple that determines the worst case Vin at the regulator.

 

[Siberia]

Have you checked what DC voltage with ripple voltage .

No.

I admit that I don't know the maths behind it, but would love to know how you do it.

Explaining that it's

voltage amplitude at the minima due to the ripple that determines the worst case Vin at the regulator

is not enough for my knowledge level. Sorry.

Then I just realised that I would better use Pear II, omitting onboard regulator and feeding it with 24VDC of SSLV 1.1 Tx 2x24VAC is about right then for the purpose.

 

[Tea-Bag]

Siberia, I am using 2 x 24v transformer with Pearl2 onboard reg removed.

 

[Siberia]

OT. Quick question. I'm planning to remove U1, U2, and to replace them with 10R resistances making an additional CRC filter. I would connect shunt. reg. after this filter. Is that the way yours is made?

 

[Siberia]

And I'm I correct to suppose Tx 30VA is enough for that matter?

 

[Salas]

Its plenty enough. Sink the input Mosfet well for it.

 

[Siberia]

Thanks. Tx has been ordered.

 

[Siberia]

Projecting my maths to confirm if everything is correct. Shunts for Pearl II, Vout 24V, CCS 200mA.

DCin 32.4V
DCout 24V
Load consumption 0.04A
CSS 0.2A

Choosing plan C. 10V-25V Mosfet reg. 2 1.9V LED, 1.8 kOhm R303. 5K trimmer. Other parts jumper.

Q101 dissipation = (DCin-DCout) * CCS current setting
(32.4V - 24V)*0.2 = 8.4V x 0.2A = 1.7 W
25º : 49.2W = 14.7 ºC/W

Q106 dissipation = DCout * (CCS-LoadCurrent)
24V x (0.16A) = 2.4W
25º : 10.32W = 10.4ºC/W

R101=(LEDS Vdrop-Mosfet VGS)/ 0.2A=(2.1V+2.1V+2.1V+1.9V+1.9V-6.9V)/0.2A=16R, 5W

Rdummy=Vout/I=24V/0.04A=600R, 5W

 

[Siberia]

I suppose then, it's a good calculation rather than a fairytale?

 

[Salas]

The calc in first page is reasonably dependable to verify the plan.

 

[AndrewT]

Choosing plan C. 10V-25V Mosfet reg. 2 1.9V LED, 1.8 kOhm R303. 5K trimmer. Other parts jumper.


25º : 49.2W = 14.7 ºC/W


25º : 10.32W = 10.4ºC/W

R101=(LEDS Vdrop-Mosfet VGS)/ 0.2A=(2.1V+2.1V+2.1V+1.9V+1.9V-6.9V)/0.2A=16R, 5W

 

[Siberia]

Yes, Andrew, I see, I've screwed up the values. It should have been:

Q101 dissipation = (DCin-DCout) * CCS current setting
(32.4V - 24V)*0.2 = 8.4V x 0.2A = 1.7 W
25º : 1.7=14.7º C/W

Q106 dissipation = DCout * (CCS-LoadCurrent)
24V x (0.16A) = 2.4W
25º : 2.4W = 10.4ºC/W

As for R101, I follow the guide, which states that VGS "value has to be subtracted from what forward drop is available by the LEDS array. What remains is the voltage across R101(201,301). By dividing that drop with R101 value in Ohm, the CCS current setting is found.

Following that I've found it should be 16R. Is there something wrong with it?

 

[Salas]

Use the Excel calc in post one. Is there open office for OSX? If yes, it will run it.

 

[AndrewT]

Look at the Vgs vs Id graphs in the datasheet. Read our typical experiences of Vgs. I don't think 6.9Vgs is a good model for a shunt FET.

 

[Siberia]

sure thing, Salas, just wanted to know I've understood the logic behind the calcs