The simplistic Salas low voltage shunt regulator

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
diyAudio Chief Moderator
Joined 2002
Paid Member
Why don't you listen to it first and then make the change so to compare subjectively too? Just use 1 LED instead of three and a sinked BD140, emitter towards R1. Only make R1 3R3 or lower to maintain over 150mA in the CCS. Measure the drop across it to determine its value for the current you like. Keep it at double your audio circuit steady state need.
 

Attachments

  • bd140.gif
    bd140.gif
    11.8 KB · Views: 900
Member
Joined 2005
Paid Member
Spartacus said:



5.6V is actually rather low for the Buffalo. The supply feeds an LM4562 opamp that acts as a regulator, and it'll perform better with more volts across it. You should be looking at around 7V or more.


I think the LM4562 opamps are for balance to single-end conversion in the I/V. The regulators are LT1763 which is said to config for 1.2V output. These aret low dropout type. TP recommends the digital supply not to exceed 7.5V otherwise these SO-8 reg will get too hot.

While I am planning to rework the shunt reg per received advices, I put the Jung/Didden back on for reality check. It confirms my initial finding.

Best
Prakit
 
prakit said:



I think the LM4562 opamps are for balance to single-end conversion in the I/V. The regulators are LT1763 which is said to config for 1.2V output. These aret low dropout type. TP recommends the digital supply not to exceed 7.5V otherwise these SO-8 reg will get too hot.

While I am planning to rework the shunt reg per received advices, I put the Jung/Didden back on for reality check. It confirms my initial finding.

Best
Prakit

Hi Prakit
sorry to slightly off topic question here-
What are your impression of Borbley Shunt Regulator and whare are you using ? I am thinking of trying out this shunt on a Nakamichi 1000ZXL playback amp as it handles low voltages like a Phono section and require a high qulaity power-supply.
thanks in advance for your resposnse
Salas - appreciate your help in designing a simple practical Hi quality design
SK
 
Salas shunt regulator vs. SuperTeddyReg
- rectifier 4 x MBR1045 Schottky + 4700uF Rubycon
- raw DC input 16V
- Salas shunt reg running at 4.97V, SuperTeddyReg set to 4.97V
- Salas shunt reg shunting 240 mA
- DAC consumption is 56 mA, constant load

My first impression is that Salas shunt regulator is playing with more energy, more emotions, but SuperTeddy reg have smooth silky highs & vocals are more forward.
Difference is very small, in blind A-B test I will fail to recognize which regulator is powering DAC.
 
prakit said:



I think the LM4562 opamps are for balance to single-end conversion in the I/V. The regulators are LT1763 which is said to config for 1.2V output. These aret low dropout type. TP recommends the digital supply not to exceed 7.5V otherwise these SO-8 reg will get too hot.

While I am planning to rework the shunt reg per received advices, I put the Jung/Didden back on for reality check. It confirms my initial finding.

Best
Prakit


The LM4562 is also used as a regulator for the analogue side, with the LT1763 prividing a reference as well as regulating the digital part of the DAC.
 
@salas
If I can trust my ears, with your shunt reg is deeper stage, more 3D, easier to imagine & place vocals.

Also very relaxed sound, much better than 3-pin regulators. Less tension, less nervous.

Or just my ears playing games with me and component between my ears need to break in :D
 
diyAudio Chief Moderator
Joined 2002
Paid Member
Tell us again in 48h. Congratulations for your build. Photos and more information about the whole system will be appreciated. We are trying to gauge the range of applications and performance here. There is a transistor CCS substitution you plan and maybe V2 ahead. Plenty of fun.;)
 

iko

Ex-Moderator
Joined 2008
If it's indeed constant load as he says, I doubt v2 would change anything and I'd say that the only difference with the superteddyreg is one of tone. If there was some variation in the load, the superteddyreg wouldn't fare so well, due to its higher output impedance.
 
Administrator
Joined 2004
Paid Member
Hi Andrew,
You can certainly create a current source or sink using either a current output DAC or voltage output DAC. You may run into trouble with a current output DAC simply due to the voltage compliance of the output(s). You can even use a digital "volume control" to control current.

What I'm thinking of is varying a reference voltage in a current generator, monitoring the emitter (source) resistor to read the current. An error amplifier applies the correction needed to keep the current constant.

A DAC may be used, but needs a micro controller at least in order to control it. If you use a PIC micro-controller with at least one DAC output channel (motor control series for example), you can eliminate the need for a DAC chip and possible "glue" logic.

If you use some of the audio volume control IC's, or a "Digi-pot" (industrial part meant for closed case calibration), they often use a simple up down pair of switch contacts. That would make the interface much easier. They may also retain memory if the power is cycled. :) That solves the stored current level issue. For a current indication, just measure across the current sense resistor and scale it properly. Just a simple voltmeter display kit will do that easily enough.

Hi stormsonic,
I can see how the "SuperTeddyReg" may reduce noise effectively, but it doesn't regulate that well. The author has also made some errors in logic based on what he believes is the speed with which a linear regulator can correct the output voltage. Where the regulation is far superior to his circuit, most three terminal regulators will also allow a fair amount of high frequency noise through. I think that is actually the issue he has a problem with. His idea regarding the possible time lag between event and correction is vastly exaggerated. If this were actually the case, these parts would not be useful, and yet they sell by the train car. They must work well for industrial applications, which includes sensitive temperature measurement applications.

The fact that he finds the LM329 series parts too noisy suggests that the circuit is not really designed for low noise. Can a lower noise reference be had? Yes. Better ones are available right now, however he only needs to design the current source differently to correct the noise issues. Personally, I use red, low voltage LEDs. The transistor is going to contribute more noise than the LED for sure.

Hey Salas,
Looks like an interesting thread. I'll have to find some time to read it completely.

-Chris
 
diyAudio Chief Moderator
Joined 2002
Paid Member
Hi Chris,

Thanks. Its just about a basic principle not easily oscillating DIY shunt trying to be cheap and easy for a large gamut of voltage and current applications. Recently members have been experimenting with it for DACs. Its interesting, because I had not those in mind when I made it. Ikoflexer's extended bandwidth and even lower Zout V2 circuit maybe is even more suitable for DACs by being faster?
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.