Using the HYPNOTIZE as a general shunt reg PCB

Status
Not open for further replies.
Jump to Latest
#21
Nice. I used 2A for the 13V model. Don't know about the cap on the amp, I guess its a worthwhile experiment to get it out and use 1uF film there for decoupling only. I would try up to 15.000uF smoothing cap for the rectification stage for 2A constant current. I guess you will find your best sonics by changing some values and types for that. Depends a lot on the impedance of your speakers and how loud, music types you listen to, etc.
 
#22
From what Borbley' suggestion - very large stiff powersupply caps will benfit this application ( Shunt regulators). You should try as much as 30000 MFD. Also good quality Cap will show its merit
kannan
PS I am planning to make one set of +/- 24 V very soon using this board
 
#23
Thanks for your help, I put 13000uF on it, couldn't tell a lot of difference, but I guess with the bench supply output caps and the 4700uF on the amp PCB, who would know what happens. It sounds fantastic as is, so I will get a power tranny, build the rectifier and lay it out as it should be, go for approx 13V @ 2A, let it break in fully and tweak from there. If it sounds as good as it does now when its finished, I will be very happy 🙂

Kannan, I will be attempting a +12/+20V supply for a Monica/Mojo DAC next, proly in the next month or two. It looks quite feasible to make the negative reg a positive one, to accommodate dual positive supplies, using Salas' method of tying gate to source on the 2SK170 before inserting them backwards (for the neg rail positions that is). Very versatile...
 
#24
Salas

would you please recommend good brand smoothing cap and value also rectifier to be used to power dac chips (with Hypnotize or with LM317 / 337)
I am planning to get Nichicon but they are so many types (KG, FG, KW, PW)
with rectifiers people say hexfred but what brand and model
 
#25
KG

Rectifiers that fit directly on the Hypno PCB? Then the MUR are OK. Or some TO-220 quad (or dual if 2 in package) with some makeshift add on perfboard adapter? Then see about Stealth series. Whatever you gonna get, see it is soft recovery always, that parameter is important.
 
#26
I tried a SMPS 24V 3A to power the Hypnotize for the T-amp. Set it to 21V, adjusted the CCS to 2A and output voltage to 13.8V. Removed the 4700uF from the TA2020, put in 10uF but should have been 1uF. The SMPS would not start with 13000uF at the front of the Hypno so I reduced it to 6600uF then no problems. Sound is the best so far. Very nice to listen to, more detail, very good indeed.

The reason I tried a SMPS to power it, is that I can't find a 150 to 200VA EI tranny with the correct output voltage, and the SMPS was $22 delivered. So decided to try something different and see what happens. Nothing getting too hot, not heatsink, nor SMPS.

Ian.
 
#28
I retraced my steps this afternoon, back-tracked a bit, did some comparisons. Removing the local big cap is what makes it sound much better. I have 2 identical T-amp PCB's, one with the big cap and one without big cap, makes it quick to compare. I ask my wife to help when doing A/B comparisons, 4 ears better than 2. Then we found the bench supply sounds better than the SMPS, a more realistic sound, so I guess I should look for a suitable tranny. The heatsink will take a bit more power yet, so there is still a bit of room to play with input voltages.

So the SMPS seems not the best way to go, nor the big cap on the T-amp, as you suspected I believe...thanks for asking those questions, I learned something 🙂
 
#33
Theoretically yes, but it will be very difficult to match the sides with the LED references and 0.3x resistors. I would use a Vbe controlled source for that. How much more is the 6A than the bias needed per side? You better be economic, so the dissipation is mainly kept at the load.
 
#35
jleaman said:
I think i might try this project for my aleph mini's @ 20V Pos/Neg.

Wonder if it will improve the sound at all.


J'
Click to expand...

Based on Salas first post, in may be do-able, but one would have to be creative and get the diodes powerful enough, and off board and PCB, since the holes dont except MUR860's, etc. By default the limit is 1A, and this is based on the MUR120 diodes which are a perfect fitr. I've seen at least one proto builder find a way to put MUR860s in anyways.

If you do it, I'll do it...At some point.
 
#36
Tea-Bag said:
Based on Salas first post, in may be do-able, but one would have to be creative and get the diodes powerful enough, and off board and PCB, since the holes dont except MUR860's, etc. By default the limit is 1A, and this is based on the MUR120 diodes which are a perfect fitr. I've seen at least one proto builder find a way to put MUR860s in anyways.

If you do it, I'll do it...At some point.
Click to expand...

If i do it, ill wire diodes off board and beside the fet's.

J'
 
#38
It comes from 3LedsVf minus Vgs divided by R1. That 2.2V is a measured ballpark from various builds, so to project some R value, since there is tolerance both in Leds, their local NJFET CCS, IRFP Vgs. You always should measure the drop on the set resistor(s) R1 in the end and divide by Ohmic value so to really know the running current in each build, and adjust if necessary.
 
#40
Salas said:
It comes from 3LedsVf minus Vgs divided by R1. That 2.2V is a measured ballpark from various builds, so to project some R value, since there is tolerance both in Leds, their local NJFET CCS, IRFP Vgs. You always should measure the drop on the set resistor(s) R1 in the end and divide by Ohmic value so to really know the running current in each build, and adjust if necessary.
Click to expand...

Everyday I learn something you tech:worship:
 
Status
Not open for further replies.
Top Bottom