advertisement


howto: replacing the ALWSR pre-regulator by a VBE

Teddy
The confusion or lack of understanding is mine, it is not a criticism of your work (all of which is far beyond me), what I am trying to understand is does it matter where the resistor connections are taken from i.e between the capacitor terminals (mmk) or from the output side (red capacitor version).
Question relates to the positive version.
 
Teddy,

My first post since registering.

1. If I was to use a recently serviced (and modified) HiCap to feed the ALWSR board with prereg, will the final performance of the regulator still be enhanced by removing the pre-reg and fitting the VBE stage, given that the HiCap (and its 317s) already sits between the mains and the input of the ALWSR board?

2. Using the HiCap, if I was to build the VBE stage on a separate board, and continue to use the ALWSR with the pre-reg, would the final performance be even better? ie making use of the HiCap, VBE, pre-reg, then main regulator stage. (I realise the voltages and power losses are adding up here).

I would be grateful for your advice.
 
norwood said:
what I am trying to understand is does it matter where the resistor connections are taken from i.e between the capacitor terminals (mmk) or from the output side (red capacitor version).
Question relates to the positive version.
Click to expand...

Hi,

I'd really love to help, but still I don't understand the question. Can you be more specific? What do you mean by the output side?

The 62K is connected between the transistor's collector to the large capacitor, then another 100K is connected from the same point on the large capacitor to the smaller capacitor. At this point the PCB link between the small capacitor and the transistor base is cut, and the connection is made through a 118R, bypassing the link that was cut.

The other side of both capacitors is connected to the ground.

The 1.5M is connected in parallel to the large capacitor.

Does it help?
 
kubota1000 said:
Teddy,

My first post since registering.

1. If I was to use a recently serviced (and modified) HiCap to feed the ALWSR board with prereg, will the final performance of the regulator still be enhanced by removing the pre-reg and fitting the VBE stage, given that the HiCap (and its 317s) already sits between the mains and the input of the ALWSR board?

2. Using the HiCap, if I was to build the VBE stage on a separate board, and continue to use the ALWSR with the pre-reg, would the final performance be even better? ie making use of the HiCap, VBE, pre-reg, then main regulator stage. (I realise the voltages and power losses are adding up here).

I would be grateful for your advice.
Click to expand...

1. Yes, but a little less. The VBE (capacitor multiplier) filters the high frequencies that neither the hicap nor the pre-reg filter.

2. Possibly, and theoretically yes, but it assumes that the pre reg doesn't add noise, which I'm not certain about...


Teddy
 
Teddy, another possibility:

Could the VBE stage be connected between the ALWSR pre-reg and SR stages, ie, inside the constant voltage (say 5V now) loop provided by the pre-reg, but still directly feeding the SR stage, as is preferred?

If it's possible, any down side?

A thought. Is there a possibility that the VBE stage, when 'inserted' in this manner, will be detrimental to the SR stage performance, because there's now two sequential stages within the pre-reg loop (VBE and SR), which may prevent the SR stage from seeing the intended constant voltage, or does the VBE not interfere with this aspect because of its long time constant...?

(I am attempting to incorporate the very obvious benefits of the VBE stage ahead of the SR stage, while maintaining the constant-voltage benefits of the pre-reg for the SR stage).

Again, I would be most grateful for your advice.
 
A thousand thanks, again, to Teddy and Martin and all who helped to make this circuit better known.

Know I understand a little how it works :p (after building it)

I have already lost 2kg since my first VBE :)
Today I resucitated my M-Audio superDAC with the first of four gyrators :cool: Allways the same, additive (and may I say addictive) effect.

Teddy, I will try the home-made Darlington. I remember someone mentioned bc547c was not ideal (to be paired to D44H11), low current maybe...any advice about alternatives?

Regards,
M
 
kubota1000 said:
Teddy, another possibility:

Could the VBE stage be connected between the ALWSR pre-reg and SR stages, ie, inside the constant voltage (say 5V now) loop provided by the pre-reg, but still directly feeding the SR stage, as is preferred?
Click to expand...

It's an interesting option which I never tried. Obviously you'll have more dropout. I've tried however an LT1086 before the VBE but didn't notice an improvement.

Another option that I've tried is to use the VBE in a tracking-pre-reg configuration, that is, instead of connecting it to the ground I connected it to the output of the SR. IIRC I used a series of 3 LEDs in parallel to the VBE capacitor to ensure a fix dropout on the SR. It worked, but no real advantage...


Maxlorenz - The bc547c has the advantage of high hfe, around 500. I don't know of a better option, but you can try ZTX651 and alike.
 
The ztx651 works very well with a d44h11. I used this combo with a 1K3 and 47u MMK to decouple the front end of my NAP140 boards from the main psu. Dropout was small, < 2V if I recall correctly, but it was a while ago (NCC200s now :D)
 
If you really want low dropout, use a small-signal NPN (BC547, ztx 651 etc) with a PNP pass transistor configured as a Sziklai pair. You can get the dropout down to around 0.8v ;)
 
PS to make this work well, add a smallish (470 - 1Kohm) resistor between the pass transistors base and emitter to speed turn-off; and add 22-50R in series at the driver transistor's base connection to kill any tendency to oscillate at VHF.

You can use PNP with an NPN pass transistor for a negative rail version.
 
martin clark said:
PS to make this work well, add a smallish (470 - 1Kohm) resistor between the pass transistors base and emitter to speed turn-off; and add 22-50R in series at the driver transistor's base connection to kill any tendency to oscillate at VHF.

You can use PNP with an NPN pass transistor for a negative rail version.
Click to expand...

Take a look here.

BTW, do you recognize the amplifier circuit? ;)
 
I just fabricated an ALW pcb with the VBE to the side.I'm using an Infinicap 10uf(kind of large for the board),and a Wima .1mf.The ALW SR before was a little whitish sounding but was not noticed until I did the change.Night and day difference.The treble is clear,clean and natural soundstage-very explansive.Bass is better-more impact and tighter than before.I plan on using better resistors throughout the 2 circuits but for now-fantastic improvement.I'm using them in a Sony scd 777es to supply the +/-7 vdc to the current pulse chips.Soon to add to the opamps(opa627's off I/V chips).A total no brainer mod.Thanks.Anyone want photos?

Mike
 
Well guys after first trying the VBE mod a while ago using mosfets I was not impressed tbh, I didn't like this coloration that was in the bass.

Today I was sorting through my box of components and found a Tip132 Darlington, seeing as though I had a 5v ALW reg spare thought I'd give it another chance but using the Darlington.
I put in what parts I had knocking about and used it to supply the SB3:D well what a difference this one makes! I'm quite impressed and that colored bass isn't there either, the power in the midrange is very impressive!
Nice one Teddy:cool: Its strapped to a heatsink and after a few hours it only gets warm

236049804.jpg
 
The SB3 needs relatively high current (1A).

IMO the MOSFET-VBE is best for small loads (pre amp/CD).
The high current in MOSFET VBE cause too much dropout- and BJT have much lower dropout.

Avi
 
Last time I tried the Mosfet VBE's was for both I/V and analogue supplys in my modded A5
 
AR_sound said:
The SB3 needs relatively high current (1A).

IMO the MOSFET-VBE is best for small loads (pre amp/CD).
The high current in MOSFET VBE cause too much dropout- and BJT have much lower dropout.

Avi
Click to expand...

Given the Darlington has a lower output impedance and lower dropout than the mosfet, what are the real advantages in using a mosfet in the VBE role?
 
You must log in or register to reply here.


advertisement


Back
Top