newbie06 said:
Perhaps I was a bit too direct or even heavy-handed. What I wanted to impress upon you is that at your level of understanding, messing with things like switch-mode power supplies can be downright dangerous.
One thing I will repeat though is that these forums really aren't for absolute beginners such as yourself. There are a lot of people here with agendas whose incomes and/or egos depend on finding beginners like you and filling them with all sorts of ideas (such as modding switchmode power supplies for better audio/video performance) that a more experienced person would laugh off.
Here are some resources to help you to get started:
http://www.science-ebooks.com/electronics/table_of_contents.htm
http://www.epanorama.net/links/basics.html
There are a lot of web sites, but I suggest you avoid the audio specific ones until you have some knowledge and experience.
You can learn a lot from books at your local public library. Try The ARRL Radio Amateur's Handbook and Horowitz and Hill's "The Art of Electronics". If they look too advanced, there are plenty of other, more basic books. Don't be discouraged that engineering books don't have much to say about the "sound" of one diode vs. another in a switch-mode power supply. There is a reason they don't say much about it...
As you get deeper into electronics, if you really want to understand what is going on, you will need to learn some math (maybe you already have that) and physics. Almost nothing that is worth-while is easy.
After you've read a little and understand a little, then come back and post some questions and you'll find people willing to help. Just like when you're in school, always question what you are being taught. My 9th grade science teacher, Ralph Vena, taught me one thing that has proven time and again more valuable than anything else I have ever learned: "a good student is eternally skeptical".
In the meantime, try to resist the temptation to poke around in switch-mode power supplies. Those things are difficult even for experienced engineers.
I_F
singa,
I'm just trying to say that there's no capacitor in the C2 slot even though the schematic says
there should be one. As soon as I take my camera back from my friend, I'll take a shot for you to see.
I_Forgot,
No hard feelings so no worries.
Thanks for the valuable resources. Those links seem easier to digest than the darn book I have with me now.
Maybe it isn't obvious, but I hold no one responsible for suggesting me do this or do that. I just thank
them for their suggestion and work it out in my own time/space.
Anyhow, thanks for the good links/resources.
I'm just trying to say that there's no capacitor in the C2 slot even though the schematic says
there should be one. As soon as I take my camera back from my friend, I'll take a shot for you to see.
I_Forgot,
No hard feelings so no worries.
Thanks for the valuable resources. Those links seem easier to digest than the darn book I have with me now.
Maybe it isn't obvious, but I hold no one responsible for suggesting me do this or do that. I just thank
them for their suggestion and work it out in my own time/space.
Anyhow, thanks for the good links/resources.
Hi,
I was a switchmode power supply designer, though not an offline guy. So I do know what I'm talking about here. First off, I Forgot is dead on in his somewhat abrasive reply that an electrical engineer like myself had to forgo several years of girls and fun to learn how to design one of those things, and we really do know what we're doing, and we really did pick those parts for a good reason. In fact we picked all the parts on the board for a good reason so you shouldn't change them willy-nilly, even just to learn.
Mind you, there are a lot of reasons why we pick each part. One is cost. When building a Ferrari we select better components than when building a Yugo. So you can improve your Yugo, or your $20 DVD player, by replacing some of the parts, but it will never be a Ferrari. But it will be better and it will be a learning experience.
The components C1 L1 C2 form a filter on the input. It is much easier to not place a part on a finished board, especially a production PCB, than it is to need one and not have a place to put it. Therefore when laying out boards we frequently place component footprints that we don't really expect to need. In the case of C2, a measurement of input line ripple was probably performed and the designer determined that it was acceptable (read: able to pass FCC tests) without it.
What would happen if you put something in footprint C2? Depends. If you pick the right part, you might notice very slightly improved performance of the system. If you pick the wrong part, you might burn your house down or hurt someone. Do you have a 'scope to see the difference? If not, I submit that it is not a good way to spend your time.
In addition, you don't have nearly enough information with this schematic to make significant changes. What is the value of zener diode D111 (yes, it's more than 12V but what is it...)? You would benefit from a parts list. What is the turns ratio of transformer T1? You would benefit from a specification for the transformer. What is IC102, perhaps an LM340? These are all things which the original designer had available to him.
Power supply ripple is attenuated by selecting good components in the signal chain. A good amplifier will have a high PSRR, or power supply rejection ratio. The higher the PSRR, the less ripple will impact the output. Assuming the power supply works properly - and I submit to you that it does, and works well for its intended application - your best time-to-performance-improvement ratio will be found by looking at the signal chain.
If you want to learn how to play with circuits, I suggest you get an oscilloscope and learn to use it properly, and take care with measurements both on the line side as well as on the isolated side of line-powered circuits. If you aren't careful about grounding you can hurt yourself even on the LV side of the circuit. Once you can use the scope (and these things don't have to be expensive, look on ebay) probe around with the schematic in hand. You will see ripple on pins 8 and 10 of CN101. Is it too much? Who knows? (The original designer sure does, he made a tradeoff there). If it's less than 10% it's probably fine.
I think it's interesting that the transformer is coupled to the FLAC(A) and (B) inputs. I don't know what those pins do but perhaps they are used to cancel out (or reduce the effects of) power supply ripple.
I really hope that you see my point here. I am an electrical engineer with experience in similar designs to this, and I'm telling you that you need to have more information to play with it. I'm telling you first that an offline SMPS is not a good place to get started or even gain experience. Further, I'm telling you that you really need more tools if you want to go this route, because you won't even know if you're getting an improvement or why. Finally, I'm admitting that I myself don't know enough based on what I've seen to start making changes.
Don't hurt yourself. Don't change the board. Understand it first. Once you feel qualified to make changes, consider the following notes:
NOTE: Don't touch anything below C5 on the left side of the circuit. This is a huge minefield. If you must play, D102 and D107 could have faster trr times if you can figure out their ratings and find better parts. Look up LC filters, L101 and C109 might be tweaked for lower 5V/6V ripple. The ESR of C105 and C112 could maybe be reduced for lower ripple if you can find a rated vented cap to replace them with. I don't even know what goes between A-B because that circuit's cut off. Maybe an LED to show power on. Probably nothing worth messing with.
I tell you, that "bad capacitor" fiasco with computer power supplies has made more people want to mess with capacitors for no good reason than anything else. Which reminds me - a lot of times parts are selected for reasons that don't show up in specifications. Power supply capacitors are notorious for this - their ESL is never, ever specified and is very important at high frequency.
One last time - BE CAREFUL. Your personal safety is your own responsibility (damn sure not mine).
I was a switchmode power supply designer, though not an offline guy. So I do know what I'm talking about here. First off, I Forgot is dead on in his somewhat abrasive reply that an electrical engineer like myself had to forgo several years of girls and fun to learn how to design one of those things, and we really do know what we're doing, and we really did pick those parts for a good reason. In fact we picked all the parts on the board for a good reason so you shouldn't change them willy-nilly, even just to learn.
Mind you, there are a lot of reasons why we pick each part. One is cost. When building a Ferrari we select better components than when building a Yugo. So you can improve your Yugo, or your $20 DVD player, by replacing some of the parts, but it will never be a Ferrari. But it will be better and it will be a learning experience.
The components C1 L1 C2 form a filter on the input. It is much easier to not place a part on a finished board, especially a production PCB, than it is to need one and not have a place to put it. Therefore when laying out boards we frequently place component footprints that we don't really expect to need. In the case of C2, a measurement of input line ripple was probably performed and the designer determined that it was acceptable (read: able to pass FCC tests) without it.
What would happen if you put something in footprint C2? Depends. If you pick the right part, you might notice very slightly improved performance of the system. If you pick the wrong part, you might burn your house down or hurt someone. Do you have a 'scope to see the difference? If not, I submit that it is not a good way to spend your time.
In addition, you don't have nearly enough information with this schematic to make significant changes. What is the value of zener diode D111 (yes, it's more than 12V but what is it...)? You would benefit from a parts list. What is the turns ratio of transformer T1? You would benefit from a specification for the transformer. What is IC102, perhaps an LM340? These are all things which the original designer had available to him.
Power supply ripple is attenuated by selecting good components in the signal chain. A good amplifier will have a high PSRR, or power supply rejection ratio. The higher the PSRR, the less ripple will impact the output. Assuming the power supply works properly - and I submit to you that it does, and works well for its intended application - your best time-to-performance-improvement ratio will be found by looking at the signal chain.
If you want to learn how to play with circuits, I suggest you get an oscilloscope and learn to use it properly, and take care with measurements both on the line side as well as on the isolated side of line-powered circuits. If you aren't careful about grounding you can hurt yourself even on the LV side of the circuit. Once you can use the scope (and these things don't have to be expensive, look on ebay) probe around with the schematic in hand. You will see ripple on pins 8 and 10 of CN101. Is it too much? Who knows? (The original designer sure does, he made a tradeoff there). If it's less than 10% it's probably fine.
I think it's interesting that the transformer is coupled to the FLAC(A) and (B) inputs. I don't know what those pins do but perhaps they are used to cancel out (or reduce the effects of) power supply ripple.
I really hope that you see my point here. I am an electrical engineer with experience in similar designs to this, and I'm telling you that you need to have more information to play with it. I'm telling you first that an offline SMPS is not a good place to get started or even gain experience. Further, I'm telling you that you really need more tools if you want to go this route, because you won't even know if you're getting an improvement or why. Finally, I'm admitting that I myself don't know enough based on what I've seen to start making changes.
Don't hurt yourself. Don't change the board. Understand it first. Once you feel qualified to make changes, consider the following notes:
NOTE: Don't touch anything below C5 on the left side of the circuit. This is a huge minefield. If you must play, D102 and D107 could have faster trr times if you can figure out their ratings and find better parts. Look up LC filters, L101 and C109 might be tweaked for lower 5V/6V ripple. The ESR of C105 and C112 could maybe be reduced for lower ripple if you can find a rated vented cap to replace them with. I don't even know what goes between A-B because that circuit's cut off. Maybe an LED to show power on. Probably nothing worth messing with.
I tell you, that "bad capacitor" fiasco with computer power supplies has made more people want to mess with capacitors for no good reason than anything else. Which reminds me - a lot of times parts are selected for reasons that don't show up in specifications. Power supply capacitors are notorious for this - their ESL is never, ever specified and is very important at high frequency.
One last time - BE CAREFUL. Your personal safety is your own responsibility (damn sure not mine).
You've all convinced me. I will leave my PS alone until I learn a lot more.
I'd like to thank you all for taking your time to reply to this thread.
jimmyswimmy, I find it interesting that you recommended using a lower value for C105 and C112 because I was about to swap them with higher cap values based on some info from an article posted by singa earlier on this thread.
Anyhow, I'll look elsewhere for fun and will leave the PS alone for now.
I'd like to thank you all for taking your time to reply to this thread.
jimmyswimmy, I find it interesting that you recommended using a lower value for C105 and C112 because I was about to swap them with higher cap values based on some info from an article posted by singa earlier on this thread.
Anyhow, I'll look elsewhere for fun and will leave the PS alone for now.
Hi newbie,
Glad you decided to skip the PSU for awhile. It's pretty interesting stuff but can be very subtle. Just to clarify, I was referring to using a capacitor with lower ESR - not lower capacitance. There's a big difference!
Anyway, as I suggested before, see about getting your hands on a scope and maybe even a copy of Horowitz and Hill. It's not my favorite book but it's pretty accessible and avoids heavy mathematical treatment of things. There's a decent section in there on power supplies.
Good luck!
Glad you decided to skip the PSU for awhile. It's pretty interesting stuff but can be very subtle. Just to clarify, I was referring to using a capacitor with lower ESR - not lower capacitance. There's a big difference!
Anyway, as I suggested before, see about getting your hands on a scope and maybe even a copy of Horowitz and Hill. It's not my favorite book but it's pretty accessible and avoids heavy mathematical treatment of things. There's a decent section in there on power supplies.
Good luck!
Glad to see someone interested with cheap DVDP SMPS.
Regarding the omitted C2,it's no surprise they just leave it out as the manufacturer doesnt want spend for it as long as they feel that there aint no big difference whether it's there or not.
Newbie06,dont mind post the picture of the internal,to see the transport and decoder board???I am interested.
.Regarding the omitted C2,it's no surprise they just leave it out as the manufacturer doesnt want spend for it as long as they feel that there aint no big difference whether it's there or not.
Newbie06,dont mind post the picture of the internal,to see the transport and decoder board???I am interested.
thanks for the info, jimmyswimmy
Leolabs, see links for the images. It's such a pain to post images in this forum.
[img=http://img191.imageshack.us/img191/5628/1gx6.th.jpg]
[img=http://img191.imageshack.us/img191/8397/2wr1.th.jpg]
[img=http://img143.imageshack.us/img143/917/3cb5.th.jpg]
Leolabs, see links for the images. It's such a pain to post images in this forum.
[img=http://img191.imageshack.us/img191/5628/1gx6.th.jpg]
[img=http://img191.imageshack.us/img191/8397/2wr1.th.jpg]
[img=http://img143.imageshack.us/img143/917/3cb5.th.jpg]
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.