Just a little help
i am almost finishing the salas sslv to power the BA-3. I think that i need to run the shunt in two wire mode. then, 1 wire comes out from the +F to the positive V+ of the Ba-3 in one channel, and 1 wire from the +S to the V+ to the other channel. Same thing with the negative reg.
don't know what to do with the 0s from the reg. there are 4. 2 from the positive reg and 2 from the negative. they are shunted because of the 2 wire mode. i think i will have to connect one 0 from the positive and another 0 from the negative to one channel ground and do same thing with the other channel. is it right?
i am almost finishing the salas sslv to power the BA-3. I think that i need to run the shunt in two wire mode. then, 1 wire comes out from the +F to the positive V+ of the Ba-3 in one channel, and 1 wire from the +S to the V+ to the other channel. Same thing with the negative reg.
don't know what to do with the 0s from the reg. there are 4. 2 from the positive reg and 2 from the negative. they are shunted because of the 2 wire mode. i think i will have to connect one 0 from the positive and another 0 from the negative to one channel ground and do same thing with the other channel. is it right?
An externally hosted image should be here but it was not working when we last tested it.
subir fotos
Hotrodding question
Joca,
Each Mosfet in the BA3FE is biased around 50ma and each Jfet around 8ma so it is around 116ma per channel. As per Salas' response when I asked a similar question about hotrodding, take the current/channel x 2 plus 200ma, so in this case 116 x2 +200=432 is considered hotrod.
I aimed for around 400ma and landed up with 360ma per regulator.
Why are you going two wire instead of 4?
nash
Ok. Thanks.
I asked some posts back if i have to hotrod for the ba-3 preamp. If i am not making wrong deductions, and if the ba-3 is 50ma hungry per channel, the total is 100ma and if this is correct don't i need to have a 200ma margin? So, 300ma is the total that i need to have. Just thinking. If so, the 300ma value exceeds the 200ma in the manual for a normal use. That's why i asked if i have to use the Mur820 . This thinking could be totaly wrong if the 200ma spare current is too much.
I am sorry, i am really in the beggining of my learning curve and my function is not exponencial!
Joca,
Each Mosfet in the BA3FE is biased around 50ma and each Jfet around 8ma so it is around 116ma per channel. As per Salas' response when I asked a similar question about hotrodding, take the current/channel x 2 plus 200ma, so in this case 116 x2 +200=432 is considered hotrod.
I aimed for around 400ma and landed up with 360ma per regulator.
Why are you going two wire instead of 4?
nash
No.
They are not parallel paths. What passes through one jFet has to pass through the other jFET. Similarly what passes through one mosFET has to pass through the other mosFET.
An output bias of 50mA and an input bias of 8mA adds up to a quiescent consumption of 58mA. There is no part that passes 116mA !
One channel of shunt regulator could be set to 68mA and the amplifier will operate, with no output to the load.
You do need to add on the effect of the load. Both capacitive load and resistive load. Allow at least as much for BOTH loads.
eg.
2Vac (~3Vpk) into 5kohms is equivalent to 0.6mApk
Allw 1mApk for capacitive cables and RF attenuation and you have a requirement for 1.6mApk to the load.
Now add 68mA & 1.6mA to arrive at 69.6mA for the two CCSs for one channel.
Last edited:
Thank you for the correction Andrew. i have erred. I confused positive, negative and channels.
Would it not be correct to say that each positive regulator and each negative regulator sees a quiescent consumption of 116ma(58 x2)?
No.
each supply rail becomes a series connected supply when the amplifier is quiescent.
The load gets none of the quiescent output.
The Power ground gets none of the quiescent output.
All the +ve supply rail current has to return through the other supply rail back to the Source.
Remember ALL current must return to Source, there are no exceptions.
each supply rail becomes a series connected supply when the amplifier is quiescent.
The load gets none of the quiescent output.
The Power ground gets none of the quiescent output.
All the +ve supply rail current has to return through the other supply rail back to the Source.
Remember ALL current must return to Source, there are no exceptions.
Salas,
Question on: Zobel Network 4.7uf-10uf + 1R.
I noticed the quality of cap makes difference here.
I need to consider size vs quality. Most of the times for the cap with same capacitance: the better the quality the bigger the size.
However I thought that capacitance of this cap may depend on load, does it? In my case is only 30mA at the most. If it was true then I could potentially use lower capacitance here (lower in size, but with better quality).
Can you explain how does capacitance of this cap impact performance and if it is load dependent? Maybe there is some equation for this ...
Thanks,
Question on: Zobel Network 4.7uf-10uf + 1R.
I noticed the quality of cap makes difference here.
I need to consider size vs quality. Most of the times for the cap with same capacitance: the better the quality the bigger the size.
However I thought that capacitance of this cap may depend on load, does it? In my case is only 30mA at the most. If it was true then I could potentially use lower capacitance here (lower in size, but with better quality).
Can you explain how does capacitance of this cap impact performance and if it is load dependent? Maybe there is some equation for this ...
Thanks,
I have been listening to my BA3 balanced pre with the Salas SLV1.1 regs for a few months now and have been very pleased with the sound.
So in true DIY spirit I decided to increase the CCS from around 350ma to around 450ma. I remember TeaBag/ Michael writing that he found that upping the CCS made the single biggest improvement in sound in his build. I have been listening with the upped CCS for about two weeks and he is right! Violins seem more refined and extended, piano definitely more solid and dynamics much better. The music seems to sound fuller with more energy for lack of a better description. Very worthwhile improvement doing this if you can afford the extra heat. This BA3 pre now sounds like it is on steroids.
Salas could you help explain technically why this happens?
Thanks. nash-(now onwards to building the DCG3)
So in true DIY spirit I decided to increase the CCS from around 350ma to around 450ma. I remember TeaBag/ Michael writing that he found that upping the CCS made the single biggest improvement in sound in his build. I have been listening with the upped CCS for about two weeks and he is right! Violins seem more refined and extended, piano definitely more solid and dynamics much better. The music seems to sound fuller with more energy for lack of a better description. Very worthwhile improvement doing this if you can afford the extra heat. This BA3 pre now sounds like it is on steroids.
Salas could you help explain technically why this happens?
Thanks. nash-(now onwards to building the DCG3)
Thanks Salas. I will try 4.7uf then.
BTW In addition to what is written in previous comment by nashbap, to squeeze out more hi-fi juice from SSLV I found out that Mundorf 4 pin cap works beeest for leds filtering (same applies for reflector) better than elna silmic, black gate and even small value films.
For Zobel network the best cap was Oil impregnate ones. It gave the smoooothest sound. However oil cap mod worked only if SSLV was used in gain stages - it did not work for buffers (DCB1) nor for powering of any dac chip sections like AVCCs, Vref - it those applications it just did not bring any improvement over regular non-oil film cap.
I put four SSLV BiB boards together, two for plus rails and two for minus rails.
Each rail is to supply 4 channels in an 8_channel preamp (Juma's bf862), but I have not gotten that far yet. Just finishing the PS section (in a separate box).
I loaded all SSLV boards with a 300 Ohm/5Wresistor (for 23Vdc reg supply). I trimmed the pots and three boards hold the Vout dead on.
> One board however is just killing me. First it appeared the voltage changed as I moved the output wires around and i though there was something wrong with one of the leads. so I jumped the gun and changed all four leads. no change.
So it dawned on me to check the voltage where the output wires solder on the board (duh) and, unlike the other three boards, the voltage from say -F to either "0" point is different!. the same from -S point to either "0' point, which brings me to believe that something on the board is not all right.
any ideas where to go from here?
Each rail is to supply 4 channels in an 8_channel preamp (Juma's bf862), but I have not gotten that far yet. Just finishing the PS section (in a separate box).
I loaded all SSLV boards with a 300 Ohm/5Wresistor (for 23Vdc reg supply). I trimmed the pots and three boards hold the Vout dead on.
> One board however is just killing me. First it appeared the voltage changed as I moved the output wires around and i though there was something wrong with one of the leads. so I jumped the gun and changed all four leads. no change.
So it dawned on me to check the voltage where the output wires solder on the board (duh) and, unlike the other three boards, the voltage from say -F to either "0" point is different!. the same from -S point to either "0' point, which brings me to believe that something on the board is not all right.
any ideas where to go from here?
It's good that you are testing the modules in stages.
Far easier to find the problem and sort it now than when all 8 channels are in a box.
Check for broken soldered joints, or a cracked trace.
Try touching each component and giving it a bit of a wiggle. An analogue meter connected during this fault finding is far better than a digital meter.
Maybe use a capacitor coupled speaker. You will hear it clicking if the DC is only momentarily broken.
Far easier to find the problem and sort it now than when all 8 channels are in a box.
Check for broken soldered joints, or a cracked trace.
Try touching each component and giving it a bit of a wiggle. An analogue meter connected during this fault finding is far better than a digital meter.
Maybe use a capacitor coupled speaker. You will hear it clicking if the DC is only momentarily broken.
it is a mystery! I resoldered all the points and it is still the same. the voltage changes on just pressing with two fingers on the output wires, or by moving them around (4 wires are twisted together; mind you they have already been changed so it is no fault of any wire). pressing on the board itself or tugging on any component has no effect.
I decided to move on and just solder them to the output connector on the chassis since when they are still i.e not moving and not being pressed the voltage is relatively stable; so I hope the preamp boards will be able to work.
the other three boards hold it spot on; no change with either pressing on the wires or moving them around. so the one odd board definitely has something hoaky going on.
I'll wait for Salas to pitch in; maybe he can shed some light?
I decided to move on and just solder them to the output connector on the chassis since when they are still i.e not moving and not being pressed the voltage is relatively stable; so I hope the preamp boards will be able to work.
the other three boards hold it spot on; no change with either pressing on the wires or moving them around. so the one odd board definitely has something hoaky going on.
I'll wait for Salas to pitch in; maybe he can shed some light?
Last edited: