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
*With that low voltage squeeze, semis Cxxx go up. From 10V Vds to 3.3V Vds the MOSFET triples its Crss for instance, and its a high capacitance one to start with. Try 2.2 Ohm + 3.3uF Zobel. Use a lower capacitance low noise JFET of enough mS like 2SK117 (what is 177?). Priority is about the Zobel.
 
*With that low voltage squeeze, semis Cxxx go up. From 10V Vds to 3.3V Vds the MOSFET triples its Crss for instance, and its a high capacitance one to start with. Try 2.2 Ohm + 3.3uF Zobel. Use a lower capacitance low noise JFET of enough mS like 2SK117 (what is 177?). Priority is about the Zobel.

Thanks Salas.
I will try and report later.
Oh, mistyped again, 2SK117 is correct on the schematic.
 
#5322
I made prototype Reflektor suggested by Salas.
Zobel value is changed from 0.1ohm+20uF to 2.2ohm+(2.2uF+0.47uF+0.47uF).

I use four Reflektors, so I replaced one of them that is connecting SI5317(Jitter Cleaner).
I am not good to explain impression of sound in English, but I say that it becomes more musical than before:D
Thanks again Salas. Your contribution is great.

Next step I will change three other Reflektors.
 

Attachments

  • P1050715.jpg
    P1050715.jpg
    131.2 KB · Views: 616
diyAudio Chief Moderator
Joined 2002
Paid Member
Kazuo-san, did you you replace all others by now? A photo of the total project would be nice. Also do you know a store in Japan that we can look for NOS Toshiba, Hitachi, Sanyo semis, Tantalum and YAM resistors, sending few and accepting Paypal?
 
Kazuo-san, did you you replace all others by now? A photo of the total project would be nice. Also do you know a store in Japan that we can look for NOS Toshiba, Hitachi, Sanyo semis, Tantalum and YAM resistors, sending few and accepting Paypal?

I replaced all Reflektors and I got the used oscilloscope Tektronix TDS350 (200MHz, 2ch). I have been evaluating on these few weeks.

Sound becomes changed drastically. Bass and mid are clear, smooth, deep.
But I find one serius problrem. Treble is harsh. Although I changed Zobel C (2.3uF, 3.8uF), no effect. I guess that the cause may be big Cxxx of 2SK2586 in low Vds area as Salas mentioned.
Of course, I have observed that it does not oscillate.

From these result I am convinced that we are going in the right direction, but I must explore other low Cxxx MOS FET. Now I am purchasing some MOS FET from DigiKey.


Salas, I didn't noticed in your PM, sorry to be late.
I am glad you are interested in the items. But I use these in my junk box.

YAM was discontinued long years ago. You can get these in stock from Michael Percy Audio.
2SK2586 can be bought at Chip One Stop www.chp1stop.com
As for the other item, I have no information, unfortunately.
I believe you can find better items than that I use.

Best regard.
Ozawa
 
diyAudio Chief Moderator
Joined 2002
Paid Member
Low Vgs Mosfets like the 2SK2586 Renesas you used to go under 5V reg normally have very high transconductance but high capacitance always comes with it. Congrats on getting a good scope, it will prove very valuable. The fact that you saw its stable plus it sounds good on all regions except HF points to suspecting the Mosfet's capacitance, yes.
See another low Vgs Mosfet example (they call them TTL level some times). This is STP55NF06L from ST Micro. Yours is 2V Vgs(th), this one is 1.7V. Looks better for capacitance. Pay special attention to Crss for this application when choosing candidates to test. Maybe you will find more ideal. Thanks for replying about the NOS Japan parts.
 

Attachments

  • 1.png
    1.png
    32.2 KB · Views: 462
  • 2.png
    2.png
    30.6 KB · Views: 452
Low Vgs Mosfets like the 2SK2586 Renesas you used to go under 5V reg normally have very high transconductance but high capacitance always comes with it. Congrats on getting a good scope, it will prove very valuable. The fact that you saw its stable plus it sounds good on all regions except HF points to suspecting the Mosfet's capacitance, yes.
See another low Vgs Mosfet example (they call them TTL level some times). This is STP55NF06L from ST Micro. Yours is 2V Vgs(th), this one is 1.7V. Looks better for capacitance. Pay special attention to Crss for this application when choosing candidates to test. Maybe you will find more ideal. Thanks for replying about the NOS Japan parts.
Thanks Salas for good advices.
Your voice makes the Gate open in front of me:)
 
diyAudio Chief Moderator
Joined 2002
Paid Member
If you got wire leads between rectification/main filtering to reg's input decouple the reg's input with 0.1u to 1u to ground. Tie the top shelf mirror BJTs face to face. Make it tight especially around the input CCS. Scope it for stability. Good luck.
 
Of course, I have observed that it does not oscillate.

I have to correct the previous report.
The attached photo shows that my regulator is oscillating.(Vertical X10) :ashamed: First I thought that it was normal fluctuations.

Salas. I was searching low Cxxx MOS FET. The device you propose is the best as far as I found out. Now I am purchasing STP55NF06L. I will report later.
 

Attachments

  • P1050730.jpg
    P1050730.jpg
    120.8 KB · Views: 410
diyAudio Chief Moderator
Joined 2002
Paid Member
~50ns period, that is very high. 20MHZ. Also low in amplitude. Can it relate to any clock fundamental or harmonic since you use them for digital? Anyway put 0.1R at the +side on the DAC's particular power feed point and see if that waveform damps away.
 
diyAudio Chief Moderator
Joined 2002
Paid Member
This is very high a frequency, if it is just a matter of moving it maybe you need shielded coax and/or to shorten/redress the wires loop. Or to dampen it as I mentioned. Maybe your new Mosfet with less capacitance will make it less sensitive, but it may very well be it just picks something over the wires so try to address that as it is made now. I see much loose wiring in #5326 as it is in test configuration now. In any account make sure that when the whole configuration is assembled as a final system the layout is economical for wire loops and there are no waveforms.
 
Hi Salas, built Kazuo Ozawa version (5V) of your reg. A few observations:
1. I am using the reg to power a shunt LDR on my balanced pre-amp. I was not expecting much, given that it's just a shunt. The shunt reg is being fed by a "ALW super reg" which takes raw dc (18V) and down regulates it to 10V. Your shunt reg is then fed the 10V to produce 5V output.
2. I was surprised by the mid-range, which seemed more present/live/real using this reg in this position. I listened till 130am this morning and only went to sleep because I had to be up at 530....
3. Very keen to try the reg with the transport, DAC but....
4. I noticed at start-up that the shunts voltage output would initially spike to around 5.8-5.9V before quickly settling at 5.00V. Is this normal? I do not believe my DAC's can live above 5.5V....
5. Second q: What is the reg doing that makes it sound so good? Is it filtering noise in a way that other regs cannot?
6. Thanks for the circuit, great performer!
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.