Thank you for the update LC. I am looking forward to using your VSSA boards.Hi Odysseas
Last 30 VSSA PCB sets are waiting to be send out as soon as I'll get assembled PCB's from manufacturer.
I aligned numbers of received ALF mosfets and PCB's since there were inquiries up to the last piece. Cannot help for the time schedule since it doesn't depends just from me.
Remaining 7 GB members to receive VSSA as soon as possible:
milandks 4 PCB set
audionootje 4 PCB set
ua3grn 6 PCB set
Odysseas 4 PCB set
edbk 4 PCB set
woofertester 4 PCB set
Metallicus69 4 PCB set
Regards L.C.
Assembling three First One-s for the exhibition, haven't sleep for a few days.
I'll be back
Hey then good fight with solderiron....Assembling three First One-s for the exhibition, haven't sleep for a few days.
I'll be back
Nice exhibition and audio talk/gossip........
.
..And good night when bedtimes come.....................
...
In sim only, in reality 15 MHz oscillator
How much practically maximum ULGF in VSSA if we try another compensation?
Last edited:
To bimo and analog_sa, the only way to get an answer to your question is to... try by yourself.
Maximum ULGF is only the one you can get stable with your real build, ...experimenting.
Fets input VS BJT ? The one you prefer (on your listening experiences). While you can prefer a single path feedback network with fets ? An other amp. Fets are better for EMI/RFI, while a little more 'grainy' ?
May-be they will be at their best with different gain ?
High input impedance ? What the need, as we prefer it as low as possible, preamp side ?
Gain variations ? Same response than ULGF. More open loop gain, more feedback, less distortion, less stability... Listen and chose your best compromise, the only way...
Or take for granted that L.C. had spend a lot of time to do all those experiments, and save time using his amp as-it
That i will do, on my side, with may-be a little try to increase PSRR, with a VAS/Input rails external regulation and tune the low pass filter FC in my real hifi system configuration.
Maximum ULGF is only the one you can get stable with your real build, ...experimenting.
Fets input VS BJT ? The one you prefer (on your listening experiences). While you can prefer a single path feedback network with fets ? An other amp. Fets are better for EMI/RFI, while a little more 'grainy' ?
May-be they will be at their best with different gain ?
High input impedance ? What the need, as we prefer it as low as possible, preamp side ?
Gain variations ? Same response than ULGF. More open loop gain, more feedback, less distortion, less stability... Listen and chose your best compromise, the only way...
Or take for granted that L.C. had spend a lot of time to do all those experiments, and save time using his amp as-it
That i will do, on my side, with may-be a little try to increase PSRR, with a VAS/Input rails external regulation and tune the low pass filter FC in my real hifi system configuration.
Last edited:
We do? That's news to me.
The preamps i use seem to always prefer lighter loads. Not to mention the cost of output coupling caps. Rather 0.47uF than 4.7uF
And in case one opts for bi/tri-amping the input impedance gets even lower.
Any thoughts about the reasons professionals prefer balanced 600 Ohms ?
Unbalanced prefer lot lower impedance.
Lowest the output impedance, better the EMI/RFI immunity and ground leakages influences.
A good preamplifier output is protected with some <= 150 Ohms serial resistance. It is to reject EMI from feedback in it, safe short circuit, keeping-it able to drive parasitic capacitances of any cable+load.
10K is high enough to do not provide a noticeable attenuation at the power amp input, and kind of normalized, nowadays. The old 47K was too high, and gave too much problems with hum etc...
You know, guitar amps ? ;-)
With fets and high preamp output impedance, you will risk a reduced response curve, due to addition of the cables capacitances and fet input parasitic ones.
Don't try to higher the amp impedance, just overkill your preamp current ability: the best sounding one will always be the one which can drive any headphone at killing levels
I don't understand you input about multi amping, don't you use an active filter ?
Last edited:
I use various preamps and some are tube with no cathode followers as i hate them for the sonics.
Not likely i would change a design which works really well in order to satisfy some or other engineering dogma. If fact - plain absurd. My main power amp has a fet differential and 100k impedance, as do my tube amps. Problems with hum? Apparently none.
Maybe digital crossovers - some day. Active analogue crossovers are of no interest to me. Passive crossover triamping works plenty well and most importantly - works with available commercial speakers. I have no interest in building speakers.
Not likely i would change a design which works really well in order to satisfy some or other engineering dogma. If fact - plain absurd. My main power amp has a fet differential and 100k impedance, as do my tube amps. Problems with hum? Apparently none.
Maybe digital crossovers - some day. Active analogue crossovers are of no interest to me. Passive crossover triamping works plenty well and most importantly - works with available commercial speakers. I have no interest in building speakers.
To add some thoughts about BJTs VS JFets for input power amplifiers stages.
As usual, there is some pros and cons in both devices.
JFets are more difficult to bias in a traditional way (consequence of the low gate current). And have lower transconductance (Gm).
Means you cannot degenerate them too much. So, they need to be paired. Quite an expensive process. They can have some advantages, like higher Ft (High Gm is not good for slew rate) and lower noise.
They usually have less Drain/source max voltage, so, they ask to be Cascoded to afford most amplifiers high voltage rails.
The main disadvantage of BJT is their 'S' curve (Compressive) witch can bring some FM distortion. But remember that Current feedback amplifiers are 'expansive' by nature, so it cancels this inconvenience ?
About overall distortions, JFets can be better. While, if you degenerate BJTs for the same Gm, you can expect similar results.
Looking at all this, one in the other, including price, BJT seems a clever choice witch seems shared by the majority of audio designers. I believe L.C had tried both and made his choices, in regard of his 'simple is beautiful' quest, according to the listening results ?
If your source is 10K, you can expect problems.
As usual, there is some pros and cons in both devices.
JFets are more difficult to bias in a traditional way (consequence of the low gate current). And have lower transconductance (Gm).
Means you cannot degenerate them too much. So, they need to be paired. Quite an expensive process. They can have some advantages, like higher Ft (High Gm is not good for slew rate) and lower noise.
They usually have less Drain/source max voltage, so, they ask to be Cascoded to afford most amplifiers high voltage rails.
The main disadvantage of BJT is their 'S' curve (Compressive) witch can bring some FM distortion. But remember that Current feedback amplifiers are 'expansive' by nature, so it cancels this inconvenience ?
About overall distortions, JFets can be better. While, if you degenerate BJTs for the same Gm, you can expect similar results.
Looking at all this, one in the other, including price, BJT seems a clever choice witch seems shared by the majority of audio designers. I believe L.C had tried both and made his choices, in regard of his 'simple is beautiful' quest, according to the listening results ?
The line impedance is fixed by the source (preamplifier).
If your source is 10K, you can expect problems.
Last edited:
So, after this long and pointless discussion we are back to where we started - even if a preamp has low output impedance but requires cap coupling, a power amp with high input impedance is just great
After all a pair of nice 4.7-6.8uF couplers can easily be more expensive than a pair of power amps.A builder here has already replaced the inputs with k170/j74 in a low voltage VSSA. In order to avoid cascoding and more drastic changes to the pcb i proposed k246/j103. Will probably just build it on a proto board.
Almost no new professional equipment has been designed for the 600 Ohm interconnect system since about 1970. This has become a legacy design.
Both modern unbalanced & balanced interconnect system prefer a lot lower output impedance.
I can agree, but not 'almost'. (And we still use a lot of old good equipment from 1970, like Neuman mikes ;-)
What i wanted to enlight is all professional (not home studios) measurement and specs of equipments are given for 600 Ohms LOADS
. We agree on the benefit of lower impedance.But this is far out of topic.
For you ? Really sorry.
I thought it could bring some useful datas for some readers. I'm so stupid.
You are free to use valves and vinyls, even in 2013. I don't, like most of us.
This said, i would be VERY interested to read comments about the results of a Jfet VSSA. And measurements.
Last edited:
More pointless generalisations.
Never seen a coupling cap in solid state equipment? And did i even mention vinyl??
Did-you have problems in your personal life, to be so disagreeable ?
You asked questions, i tried to answer them, just to help. I had spend time to it.
Strange way to thank for this effort.
Did i provided false inputs ? I don't think so.
Did -you yet know what i pointed ? How can i know ?
Are-you the only one to read this thread ? I don't think so.
This thread is about VSSA.
If you want to change the VSSA input stage in order to use JFETs, you will need more than one little change, probably including bias of the input stage and all its consequences, including VAS degeneration value. No more a VSSA, anyway, witch look simple, but is a very precise and optimized balance of each stages gains and devices.
If i was L.C. my answer to your initial question would have been:
VSSA is provided with a 10K input impedance (kinda standard). You don't like it ? Buy or build an other amp. Period.
From your side, i was very amused by: "After all a pair of nice 4.7-6.8uF couplers can easily be more expensive than a pair of power amps."
Condensateur polypropylène MKP4 4.7 µF 400 VDC
Not good enough ?
I love people using tubes and buying out-priced special caps handcrafted by virgins in pure liquid unubtainium.
Welcome in my ignore list, in company of other very nice and well educated gentlemen.
Last edited:
Hi L.C.,
Another VSSA is finished.
I am really impressed by the sound, and in the same time by the silence when it should be The first moment I connected the speakers I thought I have a broken cable, a destroyed connector, or something missing in the chain. I could hear nothing with my ear "glued" to the speaker. it's really the first time for me to experience this.
As posted in other sister threads, I use a toroidal transformer custom built for audio, 500VA with separate outputs for each rail, each channel. Each output goes to a dedicated discrete rectifier bridge (4x IXYZ soft recovery) that feeds a capacitor multiplier (Mr.Evil/PMI) via 2x15.000 uF on each rail. It might be an overkill, but I prefer to have more than too little.
With complete separate ground lines, I kept the GND - separating resistors on board, and i replaced by a wire each D-R group that feeds the VAS. The case is connected to each GND by a lifting DDRC group for each channel, coupled in the same point where the transformer's screen meets the chassis.
I would like to thank everyone that constructively contributes to DIYaudio discussions, and especially to Andrej for providing such a masterpiece to this nice community.
Cheers!
Another VSSA is finished.
I am really impressed by the sound, and in the same time by the silence when it should be
The first moment I connected the speakers I thought I have a broken cable, a destroyed connector, or something missing in the chain. I could hear nothing with my ear "glued" to the speaker. it's really the first time for me to experience this.As posted in other sister threads, I use a toroidal transformer custom built for audio, 500VA with separate outputs for each rail, each channel. Each output goes to a dedicated discrete rectifier bridge (4x IXYZ soft recovery) that feeds a capacitor multiplier (Mr.Evil/PMI) via 2x15.000 uF on each rail. It might be an overkill, but I prefer to have more than too little.
With complete separate ground lines, I kept the GND - separating resistors on board, and i replaced by a wire each D-R group that feeds the VAS. The case is connected to each GND by a lifting DDRC group for each channel, coupled in the same point where the transformer's screen meets the chassis.
I would like to thank everyone that constructively contributes to DIYaudio discussions, and especially to Andrej for providing such a masterpiece to this nice community.
Cheers!