i have not absolute problem to read paper or philosophy of others.
But I not to accept "as xx-person says that .." because I want to evolve.
well, the answer is contained in your post. we know the benefits of the current feedback but is not able to control an offset well (as self-offset) just because it has no gain in DC. a remarkable coupled with current feedback can be a dc servo-control. this does not alter the parameters in AC (or right in the segment of work, eg. below 5Hz). if it is well developed, the amplifier is a rock. Obvious that the problem is not 1-3 mV, as you have said. but the stability in function of the variants, the first is the thermal, suppose that the amp has + /-70V .. becomes ungovernable.
I have recommended to put the trimmer but do not know the last diagram of this amp, if already have a trimmer haha!
Is better that i not touch this key..in future
Last edited:
Hi
I'll be placing an order with Mouser in the near future for this project and a couple of others I have in progress.
Is there anybody else in Australia who is building VSSA and wishes to jump on my order to save shipping?
I'll be placing an order with Mouser in the near future for this project and a couple of others I have in progress.
Is there anybody else in Australia who is building VSSA and wishes to jump on my order to save shipping?
Capacitors for VSSA PCB
10 uF/63 V, MKT, EPCOS, 2 pcs
- 2200 uF/6,3 V, FG, Nichicon, 2 pcs
- 1000 uF/50 V, FG, Nichicon, 4 pcs
Hi Ranchu, i jump on that. I would like 4 of the nichicon 2200uf 6.3V if that is ok with you.
Thanks.
Hi Roberto
I am hobbyist so leave deep tech talk to you specialist, especially when it concern DC servo's.
But i think i can read funktion of your trimmer that is simple and match simpliness of VSSA design, seems to do Wonder for symetri and Groupdelay at lows.
If i do it right when simulated makes Wonder for electrical symetri and it makes the groupdelay in Lows caused by the 2200uF cap way better.
Am i right it sits there between rails and makes floating input ground ? if it Work reallife it must be better because the input ground and the return ground for speaker (heavy currents) are totally seperated.
But it seems to to make signal to noise 14dB worse, but still nice.
Ricky
I am hobbyist so leave deep tech talk to you specialist, especially when it concern DC servo's.
But i think i can read funktion of your trimmer that is simple and match simpliness of VSSA design, seems to do Wonder for symetri and Groupdelay at lows.
If i do it right when simulated makes Wonder for electrical symetri and it makes the groupdelay in Lows caused by the 2200uF cap way better.
Am i right it sits there between rails and makes floating input ground ? if it Work reallife it must be better because the input ground and the return ground for speaker (heavy currents) are totally seperated.
But it seems to to make signal to noise 14dB worse, but still nice.
Ricky
Last edited:
No way. If the group delay is better, it means the impedance of the current source had lowered on one side. Not necessary good news.
In VSSA, input ground is referenced to...ground. Adding this trimer should reference-it to some mv in the opposite direction of any offset.
With the risk to introduce ripple.
If i understand well, the offset trimmers yet exists (in a different place, but with the same effect).
May-i insist on one point ? It is possible that in few cases, with a lot of bad luck, one has some active devices (input or VAS) far from average, in the opposite direction, making a problem with DC offset. That happens sometimes on few units in any industrial fabrications. Those amps go to SAV to get those device changed.
Don't try to modify this nice and compact circuit, carefully designed, you have all the chances to make them not so good. Any DC <+-5mv has to be ignored. Better try to tune the pots, according to L.C. instructions, in order to minimize it after half an hour of power (with closed enclosure).
Last edited:
Hi Ranchu, i jump on that. I would like 4 of the nichicon 2200uf 6.3V if that is ok with you.
Thanks.
No worries - happy to. A prepaid satchel is $8.25 which is an expensive way to post a few small caps. I'll ask the post office if it is cheaper to send it in a padded envolope. In any case I will definitely advise you of your postage costs before I plan an order to confirm whether you want to go ahead or not.
I see you are in PMI cap multiplier group buy - as am I. I'm putting together a Mouser BOM to stuff these boards, so if you want to order parts for your boards I'd be happy to at the same time.
I recently bought some boards from Junie who I'm sure will happily vouch for me.
Thanks Christophe
I was not to mod the modules and layout from Andrej when i get them and the small mV offset i lived with for many years. I Intend to make one build as recommended all way, just fideling with ordering from Hypex NL, and Panasonic caps for tuneup, even it is Little expensive amongst the other typo. But will also try standard crc/Mr. Evil-PMI CM PSU/and maybe SMPS500R because i got to hear difference for myself to judge and get experience on same amp modules. It will take long time and maybe Roberto comes up with PSU that rocks VSSA SQ higher.
But Christophe if later i build my own VSSA i think of MARC's BG1-BJT with SANKEN MT-200 (4 poles), then use this trimmer. But i don't understand why the floating has to be referenced when Circuit is so symetrical.
Ricky
I was not to mod the modules and layout from Andrej when i get them and the small mV offset i lived with for many years. I Intend to make one build as recommended all way, just fideling with ordering from Hypex NL, and Panasonic caps for tuneup, even it is Little expensive amongst the other typo. But will also try standard crc/Mr. Evil-PMI CM PSU/and maybe SMPS500R because i got to hear difference for myself to judge and get experience on same amp modules. It will take long time and maybe Roberto comes up with PSU that rocks VSSA SQ higher.
But Christophe if later i build my own VSSA i think of MARC's BG1-BJT with SANKEN MT-200 (4 poles), then use this trimmer. But i don't understand why the floating has to be referenced when Circuit is so symetrical.
Ricky
Hi Ricky, this simple pic can help. independant of trimmer, show the ref of Ac input signal.
For noise you have ref with trimmer, it's not a trimmer but the long path up to big capacitors that is "hot" pole of input.
we right to discuss in general, the amplifier is very good.
I not have time for load it in my simulator, if we add the trimmer, it should probably be revisited by 2x15K also.
Attachments
Last edited:
Oh, please, if you'll do this contest, can-you publish your results and feelings ? I'm specially interested in regulated vs non regulated SMPS.
Too, if you can do-it, to know if the addition of a cap multiplier with SMPS bring some sonic improvement, as it do with linear ?
I will try to make the same kind of comparisons on my side, but it will be Linear vs Connex, with and without cap multiplier (+zener).
My mistake. Please, read: "As the input stage of the VSSA has a gain of ~1 at DC (while x10 for AC). I forgot to take care the VAS in the total gain of the amp. The VAS is full DC coupled.
Post 2228: Thanks Roberto for explanation i will try simulation and understanding later this evening. By the way is it your DPS-500/S with "CustomizationForU" that will suit a setup of VSSA, and if it is what is cons of using 1 verse 2 DPS-500/S for a stereo amp.
Post 2229: Hi Christophe yes i will report for others to hear my opinion/experience. It gets a Little expensive doing trails with the different PSU's, but then i will know which to use in final build which are 4 x mono for my two-way speakers. Under test i will use only stereo setup, patch in headphones at preamp then compare with output VSSA and also listen fullrangers in Sealed box and EQed.
Also interesting your idea adding cap multiplier after smps will test too, and like to hear your reports too.
But all this will take months, if i am Lucky i will have first setup of HYPEX+TuneupCaps running in 3 weeks time.
Ricky
Post 2229: Hi Christophe yes i will report for others to hear my opinion/experience. It gets a Little expensive doing trails with the different PSU's, but then i will know which to use in final build which are 4 x mono for my two-way speakers. Under test i will use only stereo setup, patch in headphones at preamp then compare with output VSSA and also listen fullrangers in Sealed box and EQed.
Also interesting your idea adding cap multiplier after smps will test too, and like to hear your reports too.
But all this will take months, if i am Lucky i will have first setup of HYPEX+TuneupCaps running in 3 weeks time.
Ricky
Last edited:
I tried to convince L.C. of the benefit we can have, to use a single PSU instead of two monoblocks. It is all about the behavior during high transients.
The cons is the reduction of channel speraration. Not an issue, on my point of view for several reasons. The first one is most of the records have the most power demanding instruments in the center (Bass and Kick drums), both when the mix was done with level stereo (pan pots) or phase stereo (pair of mics, because,, in real life, low frequencies have no directivity).
We have to remember vinyls only have something like 30db of separation, and everybody found it as satisfactory enough.
The practical result is an indiscernible reduction in the width of the stereo image. Notice that too much separation is not natural.
The good is, as any reduction of gain in one channel, due to high current demand on one side, will be more reflected on the other channel, you will have a better stability of the virtual position of the instruments, specially those at low levels. Better and easier localization is the goal.
For the simulation, let me know the result, also with the trimmer we add a total "R" in parallel to the 2.2 MF, this should pave the answer at the bottom, in theory.
Yes, it is customizzable the "S" version for this amp but you can save some money if you use the DPS-500/SD, as shown recently in the Wire-Amp thread. if you stay in the range of + / - 45V, you have left a good margin of power even connecting two amps with 1 dps-500/SD.
But everyone knows that I think otherwise, the problem is not the separation between the channels, but the retention of the current flow is asymmetric, in the envelope audio. so there is no doubt that the best condition is that each amplifier is free to modulate the current, without changing (such as soft), if a channel has the voice, and the other on bass or drums. this is just one of the facts.
A real surprise is the new release of the compact DPS-400 (in fact, I would send one at LC to hear the difference, if he agrees), right is unregulated but absolutely clean, manages to give a warm sound also with attack musical instruments very dry. sincerely would be perfect for this amp. voltage drop under burst is surprisingly small, especially at /-40V output. and ... have nice aspect for audio use.
I do not want to convince anybody, I think you should choose according to the measures, how a product is made, and efforts to keep the price affordable.
Other than that, I have no doubt that 2xdps-500/SD may be the top
Regards
Post 2232: Hi Christophe good pros and cons i agree, same is example in my present setup i have 2 stereo amp, one located at each 2-way speaker. Amps born with 2 separate secondary with own bridgerectifier + capbank for each mono block, but i have paralelled the secondaryes to bridgerectifier and joined the two cap banks and then split to amps, then the rail voltage regulation useage is equal for the amp/drivers in 2-way speaker, but not for right contra left speaker.
But i think the electrical data when using regulated PSU would benefit, i mean regulation and impedanz data is better with short leads between 1xAmp/1xPSU than long leads with split in 1xPSU/2xAmp setup.
Ricky
But i think the electrical data when using regulated PSU would benefit, i mean regulation and impedanz data is better with short leads between 1xAmp/1xPSU than long leads with split in 1xPSU/2xAmp setup.
Ricky
Post 2233: Hi Roberto thanks for explanation about SMPS, i see the top-way to go 1xAmp for1xPSU. It gets expensive, but for sure not if it is better in VSSA setup than other PSUs. For example i end at 2x130€ for the trails with recommended Hypex plus tuneup-caps.
Your datasheet for DPS-500/SD says it can emulate 2x100.000uF seems like an very solid powerline for an amp and thereby high SQ.
I will get back later with simulation results.
Ricky
Your datasheet for DPS-500/SD says it can emulate 2x100.000uF seems like an very solid powerline for an amp and thereby high SQ.
I will get back later with simulation results.
Ricky
I think two are spared from one GB member's canceled order, so yes one pair of VSSA PCB set still available.
Tomorrow I'll call Semelab again and try to find out concrete delivery time of my back order. It is really a shame that this is happening, but it is out of my influence.
Regards to all L.C.
BYRTT 8 PCB set
CaféNoir 6 PCB set
ColinAlex 8 PCB set
DavorXXL 6 PCB set
DuffyDawg 4 PCB set + 4 SMPS400
electromf 4 PCB set
badrisuper 2 PCB set
volavola 2 PCB set
JethroTull 2 PCB set
kp93300 2 PCB set
momomo67890 4 pcb set
ua3grn 6pcb set
mj777 2pcb set
mkusan 2pcb set
Odysseas 4pcb (2 pairs)
hcbonfim 2pcb set
Albrerta 2pcb set
dw1narso 2 pcb set (1 stereo) + extra output device pairs & isolators (for stereo if possible)
supernet 2 pcb set
milandks 4pcb set
tjencks 2pcb set
quanghao: 2pcb set
audionootje: 4pcb set (2x stereo)
tritosine 4pcb (2 pairs)
Bobo le Chat 2 pcb set
AP2 2pcb set+2pcb empity (i need this also)
Meganinja 2 set ( stereo )
mikvous 2 pcb set
Rick G 2 PCB set
kindhornman 2 boards
ravid 2 pcb set
siberia 2 pcb set
gnat 2 pcb set
potepuh 2 pcb set
metallicus69 4 pcb set
Ranchu32 2 pcb set
woofertester 4 pcb set
edbk 4 pcb set
Raj1 2 pcb set (or 6 if available)
CaféNoir 6 PCB set
ColinAlex 8 PCB set
DavorXXL 6 PCB set
DuffyDawg 4 PCB set + 4 SMPS400
electromf 4 PCB set
badrisuper 2 PCB set
volavola 2 PCB set
JethroTull 2 PCB set
kp93300 2 PCB set
momomo67890 4 pcb set
ua3grn 6pcb set
mj777 2pcb set
mkusan 2pcb set
Odysseas 4pcb (2 pairs)
hcbonfim 2pcb set
Albrerta 2pcb set
dw1narso 2 pcb set (1 stereo) + extra output device pairs & isolators (for stereo if possible)
supernet 2 pcb set
milandks 4pcb set
tjencks 2pcb set
quanghao: 2pcb set
audionootje: 4pcb set (2x stereo)
tritosine 4pcb (2 pairs)
Bobo le Chat 2 pcb set
AP2 2pcb set+2pcb empity (i need this also)
Meganinja 2 set ( stereo )
mikvous 2 pcb set
Rick G 2 PCB set
kindhornman 2 boards
ravid 2 pcb set
siberia 2 pcb set
gnat 2 pcb set
potepuh 2 pcb set
metallicus69 4 pcb set
Ranchu32 2 pcb set
woofertester 4 pcb set
edbk 4 pcb set
Raj1 2 pcb set (or 6 if available)
This is my first sim of VSSA, thanks to Dadod who provided the file and models to me.
Green trace is bandwidth
Dotted green: phase (-250° at 0db, 110° margin ! Rock stable)
Golden trace is open loop bandwidth (inverted due to my personal method): Cutoff frequency: 14Khz !!
Feedback ratio at 14Kz: 23dB (-4db comparing to 100Hz).
I don't see any noticeable phase error at 20Hz, means no group delay error.
wow !
Green trace is bandwidth
Dotted green: phase (-250° at 0db, 110° margin ! Rock stable)
Golden trace is open loop bandwidth (inverted due to my personal method): Cutoff frequency: 14Khz !!
Feedback ratio at 14Kz: 23dB (-4db comparing to 100Hz).
I don't see any noticeable phase error at 20Hz, means no group delay error.
wow !
Attachments
Last edited: