SSLV1.1 builds & fairy tales

[evengravy]

Hi,

I wonder, is there any xls spreadsheet to help with values for this version, I made a board but not exactly sure if the v1.1 xls is suitable for calculating the values.

I am looking to use it on a JFet preamp, current draw is 60mA at 24v. Filtered DC is sitting at 38.5V. Any tips for values greatly appreciated. Thanks gents.

http://www.diyaudio.com/forums/powe...w-voltage-shunt-regulator-46.html#post1852710

 

[Salas]

That one was a a V1 flexible routing to can do several things. Post the exact circuit you are going to do on it to give you some guidelines.

 

[evengravy]

Thanks Salas,

Essentially a haptone preamp x2, one per channel. Current consumption per channel is around 40mA but can be lower depending on how it is set up (30mA lowest per chan):

https://www.google.co.uk/url?sa=i&r...HdjLeFJEJ0NoSbtJJRVED3bw&ust=1472294065290999

 

[Salas]

Use about 150mA CCS. There is a setting resistor. If you will use the three LEDS way it could be R1=10R 2W. With 24V Zener is going to give 24.6V DC output.

 

[RCruz]

Here we have my go at the V12R with tracking CCS built for 10VDC out.

I reduced R4 R5 to 800 ohm so to have 5mA there (on the ccs)

R6 goes down from 10k to 1k8 so I have 3.2mA under the mosfet driving buffer.

Are these good choices ?

 

[RCruz]

D3 is a 10v zener... not sure if it will work... maybe I should use a 6v zener ?

 

[Salas]

R6 could be higher value no matter less mA, say 4k7, so to keep the open loop gain higher.

Do try a lower voltage Zener.

 

[RCruz]

R6 at 5k6 passes 1mA... good

zener at 4.7volts seems to work.... now breadboard

 

[RCruz]

Too much current in the zener (2.8mA) after startup

Can it work with a 12v zener for 10 vout ?

 

[Salas]

Too much current in the zener (2.8mA) after startup

Can it work with a 12v zener for 10 vout ?

Depends on Vin-Vout available and on R15 value. Its a trial thing for some certain set-up. I would use the JFET CCS tail at such low rail voltage though. Has no start up condition glitch to solve. If not having the high Vgs (OFF) Q3 cascoding part handy, OK use BJT tail and tune the Zener kickstarter.

Admittedly it takes an effort.

 

[RCruz]

IMO the zener voltage should be slightly higher than Vin-Vout but lower than Vin.... after that R15 should be adjusted......

 

[selfy]

Folks, I usually use two resistors for start-up in similar implementation. My goal, however, was to use one SK117 less

 

[Salas]

But surely Ricardo is perfectionist and wants to keep very high current source impedance on the non voltage regulated side

In what equipment application you use 1.2R ?

 

[selfy]

LTspice simulations showed almost no significant difference in the css impedance with or without the startup resistors. I am not quite sure about the real world measurements since I usually use a DMM and my ears only

I haven't tried V1.2 so far. I am using V1.1 (BiB) but with modified current source in my XMOS/AD1865 DAC (codenamed Moby Dick).

 

[Salas]

Ring of two BJTs is not a bad choice (I see your goal about not sparing Toshibas) but prefer IRF9610 for Q1 at this current level. For Q8 its the IRF9530 that remains best. You don't need the IRF9530 power for Q1 though and its slower. Lower parasitic capacitance Q1 part gives sense of silkier extended highs. I had used many combinations so I know but it just maybe proves not so noticeable on every different nature of client circuits it powers.

*You simulated with and without R4 just the modified BIB or the actual 1.2R with and without 2X330K? They both have so much PSRR in simulation that will make no interesting difference anyhow.

 

[selfy]

Salas, thank you for your advise! I know IRF9610's are better for the CSS and I do use them in most of my implementations. It is just that IRF9530 is four times cheaper and readily available...

 

[AndrewT]

Here we have my go at the V12R with tracking CCS built for 10VDC out.

I reduced R4 R5 to 800 ohm so to have 5mA there (on the ccs)

R6 goes down from 10k to 1k8 so I have 3.2mA under the mosfet driving buffer.

Are these good choices ?

D3 is a 10v zener... not sure if it will work... maybe I should use a 6v zener ?

I have lost count of the number of times I have recommended that the CCS part alone be assembled and tested before a Builder adds on the shunt part of the regulator.
Did you do that?
If not then disable everything to the right of R6 and disable M2 (pull out the drain lead).
Test the CCS without a load.
Then add on a 100r load and test again.
Then add on a second 100r and test again.
Post your results.

 

[Salas]

Salas, thank you for your advise! I know IRF9610's are better for the CSS and I do use them in most of my implementations. It is just that IRF9530 is four times cheaper and readily available...

Its probably because they use it in car audio amplifiers so more produced and spread

 

[selfy]

*You simulated with and without R4 just the modified BIB or the actual 1.2R with and without 2X330K? They both have so much PSRR in simulation that will make no interesting difference anyhow.

I simulated many different combinations. My first implementation of the BiB was with the CSS Ricardo is trying to improve. Here is a picture of an old implementation of mine and the two 220k SMD resistors I added to fix the startup issues:

PS: Why it takes like 10-15 minutes to actually see my post here?

 

[Salas]

I simulated many different combinations. My first implementation of the BiB was with the CSS Ricardo is trying to improve. Here is a picture of an old implementation of mine and the two 220k SMD resistors I added to fix the startup issues:

PS: Why it takes like 10-15 minutes to actually see my post here?

Because you have very few posts the system still puts you in the pipeline of approval. Its a software thing to avoid spam posters. Robots or human.