Why does the SI need a high-pass filter? If I'm sure no DC is coming from my source, then I don't need a cap to block the DC, right? So the best input cap is no input cap, right? Is the cap necessary to adjust gain?
Thanks for any explanations and sorry if this has already been discussed...I couldn't find any info on it.
Thanks for any explanations and sorry if this has already been discussed...I couldn't find any info on it.
Gmorris said:
The input cap is needed to block dc coming from the amp. There is a 2 - 3 volt bias on the input. This is needed to allow the push pull output and a single rail supply.
The standard cap is 0.33 ufd. This value needs to be raised to at least 1 ufd.
To answer yor question, it is needed.
George
Hmm, with such a loss, you should hear that...
I'm curious to know if that measurement is done using batteries powering the amp ? Insufficient power seems to be more of a logical reason for such a drop, which also has been reported by different sources to be the case with this amp.
I'm using a regulated PSU and I'm really not noticing that kind of a drop
I'm curious to know if that measurement is done using batteries powering the amp ? Insufficient power seems to be more of a logical reason for such a drop, which also has been reported by different sources to be the case with this amp.
I'm using a regulated PSU and I'm really not noticing that kind of a drop
So the dc is actually coming from the amp and not the source signal?
Replacing the input cap would seem to make the most noticable difference in sound but reports of people frying their amps makes me a bit nervous. I'm assuming the micro-caps and resistors make over-soldering more likely if one is not very careful. Does replacing the input cap have more or less potential of frying the amp or do all modifications have equal potential? I was thinking about using 24-ga wire as leads to fit in the holes for the original input cap....will 24ga fit?
Thanks again for any information.
Replacing the input cap would seem to make the most noticable difference in sound but reports of people frying their amps makes me a bit nervous. I'm assuming the micro-caps and resistors make over-soldering more likely if one is not very careful. Does replacing the input cap have more or less potential of frying the amp or do all modifications have equal potential? I was thinking about using 24-ga wire as leads to fit in the holes for the original input cap....will 24ga fit?
Thanks again for any information.
Gmorris said:
I didn't mod the SI board, but I did a tripath eval board and the input caps were surface mount, not through-hole design. So getting the Auricap leads soldered to the surface mount pads took magnification and patience. I'm thinking that the SI input caps are surface mount as well, but I could be wrong.
Capacitor replacement
Do not even think of soldering in a normal cap to the SM cap attachment points. What others are doing is jumpering these micro pads, remeoving the input rfi filter, and connecting an input cap external to the board.
I have done a couple boards using larger SM ceramic caps. The sonics are very nice. Tried 4.7 ufd and 2.2 ufd. Liked the 2.2 better, but more for less turn of thump. The 4.7 ufd caused more thump on power up.
The sound of the SI stock is nice, but it will always roll off the bass no matter how powered or what the impedance of the speaker. My speakers only run down to 40 hz flat, but it was very noticable. Larger value caps and better power supply made the bass very good. But the magic is the high frequency reproduction.
The mids might be better with large film caps on the input. Have not tried, but my next thing is using some off board flat foil air cored inductors on the output. This may get me where I need to be.
My speakers are 15 ohm, so I shucked the 0.15 ufd caps from + to -. Currently using 10 uH chokes and 0.15 ufd ceramic caps from each otput to ground, like the evaluation board.
Next boards with be 15 uH chokes and 0.1 ufd caps from outputs to ground. The higher impedance speakers really need a little large value choke. A 0.01 ufd ceramic caps is strapped across the binding posts.
The output filter is an area where I think there is still improvements to be made in the sonic qualities of the 2024. We have ALL focused on the input stage, and neglected the output filtering.
George
Do not even think of soldering in a normal cap to the SM cap attachment points. What others are doing is jumpering these micro pads, remeoving the input rfi filter, and connecting an input cap external to the board.
I have done a couple boards using larger SM ceramic caps. The sonics are very nice. Tried 4.7 ufd and 2.2 ufd. Liked the 2.2 better, but more for less turn of thump. The 4.7 ufd caused more thump on power up.
The sound of the SI stock is nice, but it will always roll off the bass no matter how powered or what the impedance of the speaker. My speakers only run down to 40 hz flat, but it was very noticable. Larger value caps and better power supply made the bass very good. But the magic is the high frequency reproduction.
The mids might be better with large film caps on the input. Have not tried, but my next thing is using some off board flat foil air cored inductors on the output. This may get me where I need to be.
My speakers are 15 ohm, so I shucked the 0.15 ufd caps from + to -. Currently using 10 uH chokes and 0.15 ufd ceramic caps from each otput to ground, like the evaluation board.
Next boards with be 15 uH chokes and 0.1 ufd caps from outputs to ground. The higher impedance speakers really need a little large value choke. A 0.01 ufd ceramic caps is strapped across the binding posts.
The output filter is an area where I think there is still improvements to be made in the sonic qualities of the 2024. We have ALL focused on the input stage, and neglected the output filtering.
George
Re: Capacitor replacement
George,
I'm with you on the outputs. I just don't know enough about choke selection, so I haven't replaced mine yet. I have a huge chunk o' sprague orangedrops for the output caps, though.
It also seems like the diodes are pretty far away from the pins. The datasheet says that these should be as close as possible. I dont know how easy it would be to move them closer, but it's at least a thought to potentially improve things.
Panelhead said:
George,
I'm with you on the outputs. I just don't know enough about choke selection, so I haven't replaced mine yet. I have a huge chunk o' sprague orangedrops for the output caps, though.
It also seems like the diodes are pretty far away from the pins. The datasheet says that these should be as close as possible. I dont know how easy it would be to move them closer, but it's at least a thought to potentially improve things.
A few things:
I can understand why some might be skeptical of the improvements made by replacing the input caps, as this little amp has received such high praise in its stock form. The praise surprised me, as I found the stock S.I. to be thin and shrill.
As for the question of power, yes, better power does help the amp. But I made measurements and listening tests using stock batteries, a regulated supply, an SLA and a marine deep cycle battery. The AAs are notably weak, everything else worked well. But bass and dynamics did not improve much until the amp was modified.
Improving the input circuit and all he connections made world of difference. Bass become solid and full, detail improved noticeably, as did dynamics.
Most of my modifications were made at the same time, so it is impossible to say what changes counted most. Maybe all of them together. after the 1st round of mods, the amp was everything it was reported to be, and more.
The second round of mods concerned the output filters. I found that better inductors helped the feeling of power and dynamics a bit. The improvements were not as great as those after the inputs mods. Replacing my handmade aircore inductors with larger toroids seemed to free up the sound a bit, as well a educe a bit of harshness in the sound. I think that switching noise was getting back into the input.
My 2 cents worth would be: Change the input cap, replace the onboard power cap, replace or eliminate the volume pot, use better connectors.
This amp can sound very, very good - bu not in stock form. At least not to me.
My modifications and impressions can be found elsewhere on this forum and on my website.
I can understand why some might be skeptical of the improvements made by replacing the input caps, as this little amp has received such high praise in its stock form. The praise surprised me, as I found the stock S.I. to be thin and shrill.
As for the question of power, yes, better power does help the amp. But I made measurements and listening tests using stock batteries, a regulated supply, an SLA and a marine deep cycle battery. The AAs are notably weak, everything else worked well. But bass and dynamics did not improve much until the amp was modified.
Improving the input circuit and all he connections made world of difference. Bass become solid and full, detail improved noticeably, as did dynamics.
Most of my modifications were made at the same time, so it is impossible to say what changes counted most. Maybe all of them together. after the 1st round of mods, the amp was everything it was reported to be, and more.
The second round of mods concerned the output filters. I found that better inductors helped the feeling of power and dynamics a bit. The improvements were not as great as those after the inputs mods. Replacing my handmade aircore inductors with larger toroids seemed to free up the sound a bit, as well a educe a bit of harshness in the sound. I think that switching noise was getting back into the input.
My 2 cents worth would be: Change the input cap, replace the onboard power cap, replace or eliminate the volume pot, use better connectors.
This amp can sound very, very good - bu not in stock form. At least not to me.
My modifications and impressions can be found elsewhere on this forum and on my website.
panomaniac said:
After reading gazillions of reviews, you are probably the first person in the entire audio world to find the stock SI 'thin and shrill'
I mean, I buy there can be made improvements doing mods on the basic low cost design. Sure. But, especially with a beey PSU, (which doesnt improve much ?) calling this amp 'thin and shrill' smeems to me either the case (a problem ?) with this inividual amp or some sort of bias from the listener.
edit: typos
You may believe what you like, I believe my ears.
It may be that I my 1st Sonic Impact had problems not shared by others. But my second one sounds just like it. Also, it may be that my system does not work well with the S.I., I have mentioned that elsewhere. My speakers are not very efficient, so I work the S.I. pretty hard, right to its limits.
After 2 months and over 30,000 hits on my little website about the S.I., not one person has disagreed with my findings, measured or subjective. You'd think someone before you would have. In fact, I get a lot of emails agreeing with my findings, some of them a bit sheepish, others a bit unhappy with all the hype. But all interested in the little "wonder amp."
That does not make me an expert, it just means that so far no one has told me that there is no need to modify, stock is just fine.
Please remember that the "gazillions" of reviews are headed up by the widly quoted and linked to TNT and Six Moons reviews. Both found the S.I. lacking in the bass department. But they enjoyed the amp more than I did. Good thing, too. It led me to the S.I. which I like a lot, in a modified form.
As for bias, Don't have one. It's Class-T, you know!
But seriously - If the amp's sound had pleased me right off, in stock form, I couldn't have been happier. Alas, it did not. That's why I set about to modify it. The surprise was that the input modification had a much larger effect than the output modification. The opposite had been my guess.
There you have it. Believe what you will. If your stock S.I. rocks your world, good for you!
It may be that I my 1st Sonic Impact had problems not shared by others. But my second one sounds just like it. Also, it may be that my system does not work well with the S.I., I have mentioned that elsewhere. My speakers are not very efficient, so I work the S.I. pretty hard, right to its limits.
After 2 months and over 30,000 hits on my little website about the S.I., not one person has disagreed with my findings, measured or subjective. You'd think someone before you would have. In fact, I get a lot of emails agreeing with my findings, some of them a bit sheepish, others a bit unhappy with all the hype. But all interested in the little "wonder amp."
That does not make me an expert, it just means that so far no one has told me that there is no need to modify, stock is just fine.
Please remember that the "gazillions" of reviews are headed up by the widly quoted and linked to TNT and Six Moons reviews. Both found the S.I. lacking in the bass department. But they enjoyed the amp more than I did. Good thing, too. It led me to the S.I. which I like a lot, in a modified form.
As for bias, Don't have one. It's Class-T, you know!
But seriously - If the amp's sound had pleased me right off, in stock form, I couldn't have been happier. Alas, it did not. That's why I set about to modify it. The surprise was that the input modification had a much larger effect than the output modification. The opposite had been my guess.There you have it. Believe what you will. If your stock S.I. rocks your world, good for you!
Again, I do believe the SI can be improved my modifying things on both the input and output side.
In fact, it's obvious on some points and I've ordered Art Taylor's SI+ chassis
as a seemingly excellent starting point.
And of course it could very well be the combination with your setup
causes the SI to sound totally different than somewhere else. I didn't
consider that, my bad.
But based on what I've heard (and the reactions of the people I've
listened with), just cannot call the standard SI 'thin and shrill', in any imaginable way.. so hence my post out of sheer amazement.
btw a quote from your site where you state something different than 'thin and shrill' ?
Summery:
Out of the box the Sonic is a good sounding amp, plenty of mid range detail, not a lot of power or bass. It seems to do really well on percussion, e.g. bells, vibraphone, wood block, bongos.
But hey, with no one of your respondents disagreeing with you finding the standard SI 'thin and shrill'. I rest my case
sorry for the off topic
In fact, it's obvious on some points and I've ordered Art Taylor's SI+ chassis
as a seemingly excellent starting point.
And of course it could very well be the combination with your setup
causes the SI to sound totally different than somewhere else. I didn't
consider that, my bad.
But based on what I've heard (and the reactions of the people I've
listened with), just cannot call the standard SI 'thin and shrill', in any imaginable way.. so hence my post out of sheer amazement.
btw a quote from your site where you state something different than 'thin and shrill' ?
Summery:
Out of the box the Sonic is a good sounding amp, plenty of mid range detail, not a lot of power or bass. It seems to do really well on percussion, e.g. bells, vibraphone, wood block, bongos.
But hey, with no one of your respondents disagreeing with you finding the standard SI 'thin and shrill'. I rest my case
sorry for the off topic
I agree in hindsight. I followed the lead (photos) of a pro modder but failed to understand why he did it that way.
I now think he used the SM pads for the input cap mod because he had added bybees to the input leads and that real estate had been taken. It would have been simpler for me to have done it Panelhead's way, but it worked nonetheless (on the evaluation board).
Sometimes good things happen despite my best efforts to fail.
Actually, I hate to rock the boat, but I've been giving the input capacitor some more thought.
First of all, I've recently completed my rather arduous mods:
1) Removal of the input caps (replaced offboard with 2uF polypropylene)
2) Remove of L1/L2
3) Removal of 100pF caps between input and ground
4) Feedback resistor changed to 20k
5) Reservoir cap swapped to 1000uF Panasonic FM
6) WBT-like RCA's, binding posts installed
7) Put the whole shabang in a context-engineering brushed aluminum case.
After doing this (and ruining another SI board by lifting some of the pads.. UGH), I've decided that this might be the way to go:
1) Replace the SMD input caps with *small* Black Gate NX Hi-Q caps. The 47uF/6.3v caps that I used recently for upgrading my DAC seem like they would fit. Smaller values should fit even better.
2) Replace the feedback resistors with 20k SMD devices
3) Replace the resevoir cap with 1000uF Panasonic FM's
4) Swap the chokes for something better
I believe that using the BG-N's would allow for a higher-quality cap to be installed than just another ceramic. I'm not sold on the whole BG cap thing, but it did improve my DAC over the cerafine that was in the same position. So, I don't think it can hurt versus a ceramic. Plus, I really do think they're small enough to go on there without destroying the pad, and that it'd probably be easier to solder. A few drops of hotglue after the installation should hold them from causing any problems.
I have on SI amp left (I've killed 1, and dissected another) to modify, so I may research which BG would fit best to replace it.
Mike
First of all, I've recently completed my rather arduous mods:
1) Removal of the input caps (replaced offboard with 2uF polypropylene)
2) Remove of L1/L2
3) Removal of 100pF caps between input and ground
4) Feedback resistor changed to 20k
5) Reservoir cap swapped to 1000uF Panasonic FM
6) WBT-like RCA's, binding posts installed
7) Put the whole shabang in a context-engineering brushed aluminum case.
After doing this (and ruining another SI board by lifting some of the pads.. UGH), I've decided that this might be the way to go:
1) Replace the SMD input caps with *small* Black Gate NX Hi-Q caps. The 47uF/6.3v caps that I used recently for upgrading my DAC seem like they would fit. Smaller values should fit even better.
2) Replace the feedback resistors with 20k SMD devices
3) Replace the resevoir cap with 1000uF Panasonic FM's
4) Swap the chokes for something better
I believe that using the BG-N's would allow for a higher-quality cap to be installed than just another ceramic. I'm not sold on the whole BG cap thing, but it did improve my DAC over the cerafine that was in the same position. So, I don't think it can hurt versus a ceramic. Plus, I really do think they're small enough to go on there without destroying the pad, and that it'd probably be easier to solder. A few drops of hotglue after the installation should hold them from causing any problems.
I have on SI amp left (I've killed 1, and dissected another) to modify, so I may research which BG would fit best to replace it.
Mike
1) Replace the SMD input caps with *small* Black Gate NX Hi-Q caps. The 47uF/6.3v caps that I used recently for upgrading my DAC seem like they would fit. Smaller values should fit even better.
2) Replace the feedback resistors with 20k SMD devices
3) Replace the resevoir cap with 1000uF Panasonic FM's
4) Swap the chokes for something better
I believe that using the BG-N's would allow for a higher-quality cap to be installed than just another ceramic. I'm not sold on the whole BG cap thing, but it did improve my DAC over the cerafine that was in the same position. So, I don't think it can hurt versus a ceramic. Plus, I really do think they're small enough to go on there without destroying the pad, and that it'd probably be easier to solder. A few drops of hotglue after the installation should hold them from causing any problems.
I have on SI amp left (I've killed 1, and dissected another) to modify, so I may research which BG would fit best to replace it.
Mike [/B][/QUOTE]
Mike,
I like it! Let us know how it works.
You might consider using two smaller value FM caps soldered directly to the power pins of the 2024 also. This is how the EB-2024 is done. I have done a couple this way.
The smallest BG's are the NHX-Q 22 ufd 6.3 volt. They should work great and are less than 2.00 each.
A problem I ran into is the charge pump thing that allows bias on the input. As the coupling cap get larger in vlaue, there is a power up thump. 4.7 ufd gave a fairly sizeable transient, the 22 ufd might be a real pop. My crossoverless 100 dB efficient speakers do not like power up nasties.
If I had a hot glue gun I would try though, I agree completely that they got to be better for coupling than the XL7 multi layer ceramics. A cap I want to try somehow is the stacked film and foil Seimens cap. They have a nice 2.2 ufd with blue conformal coating on the end caps. But they are bigger than the 2024 chip. Afraid of lifting the mounting pads.
George
2) Replace the feedback resistors with 20k SMD devices
3) Replace the resevoir cap with 1000uF Panasonic FM's
4) Swap the chokes for something better
I believe that using the BG-N's would allow for a higher-quality cap to be installed than just another ceramic. I'm not sold on the whole BG cap thing, but it did improve my DAC over the cerafine that was in the same position. So, I don't think it can hurt versus a ceramic. Plus, I really do think they're small enough to go on there without destroying the pad, and that it'd probably be easier to solder. A few drops of hotglue after the installation should hold them from causing any problems.
I have on SI amp left (I've killed 1, and dissected another) to modify, so I may research which BG would fit best to replace it.
Mike [/B][/QUOTE]
Mike,
I like it! Let us know how it works.
You might consider using two smaller value FM caps soldered directly to the power pins of the 2024 also. This is how the EB-2024 is done. I have done a couple this way.
The smallest BG's are the NHX-Q 22 ufd 6.3 volt. They should work great and are less than 2.00 each.
A problem I ran into is the charge pump thing that allows bias on the input. As the coupling cap get larger in vlaue, there is a power up thump. 4.7 ufd gave a fairly sizeable transient, the 22 ufd might be a real pop. My crossoverless 100 dB efficient speakers do not like power up nasties.
If I had a hot glue gun I would try though, I agree completely that they got to be better for coupling than the XL7 multi layer ceramics. A cap I want to try somehow is the stacked film and foil Seimens cap. They have a nice 2.2 ufd with blue conformal coating on the end caps. But they are bigger than the 2024 chip. Afraid of lifting the mounting pads.
George
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.