bequerel said:
Bequerel,
May I suggest that you keep it simple for your first try, less to go wrong, regulating a supply will increase it's complexity a lot and that is supposed to be the easiest part to build. I'm not saying supply regulation doesn't have merit, its just like trying to build a class D amp as your first project. Its too big of a leap, unless your into that 😉
Glenn
Nuuk said:
Nuuk,
I don't really have an opinion. I have not looked at the implementation close enough to have an opinion on whether it is to big or to small. I was just struck by the difference in size between what looked like small copper sheet heatsinks about the height of the regulator on your implementation and the large black anodized heatsink that carlosfm showed in the photo on post 2 of this current tread.
http://www.diyaudio.com/forums/attachment.php?s=&postid=410290
Just an observation, not a judgement.
Moving_electron,
I’ve talked with Nick about this some time ago and he clarified that behind the copper strip is 6mm aluminum. These heatsinks will certainly get warm, but considering the fact each channel has its own regulators, they are enough for the start, and voltage drop and usual level of listening will determine if bigger heatsinks are needed or not.
Pedja
I’ve talked with Nick about this some time ago and he clarified that behind the copper strip is 6mm aluminum. These heatsinks will certainly get warm, but considering the fact each channel has its own regulators, they are enough for the start, and voltage drop and usual level of listening will determine if bigger heatsinks are needed or not.
Pedja
WOW, what happened in this threat. I`ve been away for two days and the threat grew 12 pages! 😱
Very nice to see that the threat is back on subject on the last few pages!🙂
I hope you`ll still post your secrets, carlos, or at least on private mail, because I`m about to convert to the carlosfm way.
I have a pair of Dali Grand Coupe, wich are 86dB 6 ohm, so they might need even more current than yours.
I`ve been running them with a vbigc for a year now and the sound is very good, espessially on slow jazz and female voices, but I miss some dynamics and "party factor".
My orginal plan was to make a pair of easy to drive fullrange speakers to go with the amp. But as I`m going to be a father again within a week and have a whole house to paint this summer, I don`t think I will find the time to build speakers.
But I can find time to rebuild my amp with the lm3886 and regulated supply. I will also use the opa627 buffers to try the difference between a good chip buffer and the tube.😉
I might also build a circuit like your preamp to be used, as it originally was ment for, as an output buffer after my passive tda1543 IV. Maybe this will add som speed and dynamic to the system aswell😕
The dali speaker is not a very dynamic speaker, but it can sound really good. Maybe I`ll try an active filter for it later.
Keep on the good work guys!!🙂 🙂
Tor Martin
Very nice to see that the threat is back on subject on the last few pages!🙂
I hope you`ll still post your secrets, carlos, or at least on private mail, because I`m about to convert to the carlosfm way.
I have a pair of Dali Grand Coupe, wich are 86dB 6 ohm, so they might need even more current than yours.
I`ve been running them with a vbigc for a year now and the sound is very good, espessially on slow jazz and female voices, but I miss some dynamics and "party factor".
My orginal plan was to make a pair of easy to drive fullrange speakers to go with the amp. But as I`m going to be a father again within a week and have a whole house to paint this summer, I don`t think I will find the time to build speakers.
But I can find time to rebuild my amp with the lm3886 and regulated supply. I will also use the opa627 buffers to try the difference between a good chip buffer and the tube.😉
I might also build a circuit like your preamp to be used, as it originally was ment for, as an output buffer after my passive tda1543 IV. Maybe this will add som speed and dynamic to the system aswell😕
The dali speaker is not a very dynamic speaker, but it can sound really good. Maybe I`ll try an active filter for it later.
Keep on the good work guys!!🙂 🙂
Tor Martin
One option I have is to increase my output voltage (thus decreasing the amount of voltage dropped across the 338's).
I went for 27 volts as I had 35 volt rated 47 uF caps in my box. If I got some 50 volt rated caps, I could go with 30+ volts on the output and reduce the amount of cooling needed.
As Pedja says, this will get me going for now.
Milan, unfortunately this is a scope-free zone! 🙁
Football over for another night - back to thinking about hi-fi. Bad luck Italy!
(Stop laughing Carlos).
I went for 27 volts as I had 35 volt rated 47 uF caps in my box. If I got some 50 volt rated caps, I could go with 30+ volts on the output and reduce the amount of cooling needed.
As Pedja says, this will get me going for now.
Milan, unfortunately this is a scope-free zone! 🙁
Football over for another night - back to thinking about hi-fi. Bad luck Italy!
(Stop laughing Carlos).Nuuk said:
I would do that.
Make it 30~31 volts on the output, no more.
Me?Nuuk said:Bad luck Italy!
Why?
Well... it would be difficult to win them in the final.😀
Pedja said:
Oh.
Yes, you're right, Pedja.
It's not very clear that 6mm aluminium.
And independent regulators per channel, it should be fine.
😎
Nuuk said:
Nuuk,
I think it's safe to say that you can never have too large a heatsink, unless it's bulk or weight (or cost 😀 ) impedes the practicality of the design. Heat and ESD are the two biggest killers of semiconductors, period. And in terms of lethality, heat beats out ESD by a few orders of magnitude.
By large I am also implying low thermal resistance 🙂
Glenn
I agree that it is far better to keep semi-conductors cooler rather than warmer.
However, we have to look at the practicalities here. The regulators ideally need to be very close to the chip pins. The same goes for a buffer/LPF if you are using one.
This means that even in a large case, we need to keep the regulated PSU and buffer circuits as small as we can in order that they can both come close to the pins of the LM3875 or whatever.
I have been thinking about this (since the football 😉 ) and I will almost certainly change my regultor circuits to drop less voltage. If I change the 2K4 to 3K I will get about 32 volts on the output, dropping only 5 volts instead of the 10 volts that I am with the 2K4.
Could somebody tell me how much voltage is going through the 47 uF cap. Is it just the 'dropped' voltage? If so then the 35 volt rated caps I am using will be OK.
However, we have to look at the practicalities here. The regulators ideally need to be very close to the chip pins. The same goes for a buffer/LPF if you are using one.
This means that even in a large case, we need to keep the regulated PSU and buffer circuits as small as we can in order that they can both come close to the pins of the LM3875 or whatever.
I have been thinking about this (since the football 😉 ) and I will almost certainly change my regultor circuits to drop less voltage. If I change the 2K4 to 3K I will get about 32 volts on the output, dropping only 5 volts instead of the 10 volts that I am with the 2K4.
Could somebody tell me how much voltage is going through the 47 uF cap. Is it just the 'dropped' voltage? If so then the 35 volt rated caps I am using will be OK.
carlosfm said:
I can tell you on a PM, send me one and I'll explain you.
Here I'm just gonna say: that PCB is too bad, miserable layout.
Commercial spys are here, so...
I don't wanna give them clues on how to make a better layout, but I also don't wanna ruin their business.
Seems like if one wants to put the caps closer to the chip's supply pins on BrianGT's boards then the caps could just be soldered onto the bottom of the PC board. Looks like only one would have to be done this way, since only the V- has much of a trace length.
moving_electron said:
When I said "widely separated PSU caps" I was talking about the ground.
Miserable.
Anyway, I have one of those boards here looking at me (didn't pay for it, someone offered it to me😀 ) and please can you tell me how to solder a cap on the bottom of the board?
The PG and SG are on the top layer, masked.
If you solder a cap on the bottom from a PSU pin on the chip to one of the PSU caps' legs, it's too far away.
Complete disgrace, don't make me tell anything more, I won't.
I'm not ever gonna use this PCB, maby put it on the wall?
carlosfm said:
Anyway, I have one of those boards here looking at me (didn't pay for it, someone offered it to me😀 ) and please can you tell me how to solder a cap on the bottom of the board?
Again, it shows how close minded, lacking imagination and bad mannered you really are. I will not tell you how to do it, as industrial spies are everywhere, but looking at this picture it obviously shows it can be done.
An externally hosted image should be here but it was not working when we last tested it.
One more, if you still didn't figure it out. You might also get some clues, why the board is shaped the way it is.
An externally hosted image should be here but it was not working when we last tested it.
Nuuk,
Yes.
The cap will see whatever voltage that is present at the ADJ pin with respect to ground. Don't forget ye ol' protection diodes. $.10 to protect a few bucks worth of IC (or just the hassle of opening the little %$#@ up) is worth it IMB.
Glenn
Yes.
The cap will see whatever voltage that is present at the ADJ pin with respect to ground. Don't forget ye ol' protection diodes. $.10 to protect a few bucks worth of IC (or just the hassle of opening the little %$#@ up) is worth it IMB.
Glenn
The only thing decent I see there is a good box, you know someone with machinery to do that.
Peter, do you know how to bypass an op-amp?
What are you showing me here?
BTW, I didn't wanna say more, but people are asking...
Peter, do you know how to bypass an op-amp?
What are you showing me here?
BTW, I didn't wanna say more, but people are asking...
er...
Your schematic has two C5's - as such, to clarify which caps should be as close to the regs as possible - those connected to pin3 (IN) of the regs I gather are two for sure - the other two that should be as close as possible would be across the output?
C
Your schematic has two C5's - as such, to clarify which caps should be as close to the regs as possible - those connected to pin3 (IN) of the regs I gather are two for sure - the other two that should be as close as possible would be across the output?
C
carlosfm said:The only thing decent I see there is a good box, you know someone with machinery to do that.
Peter, do you know how to bypass an op-amp?
What are you showing me here?
BTW, I didn't wanna say more, but people are asking...
With the caps I'm using here, I don't need any bypasses. I tried them all.
I'm showing what you were asking for: installing the caps on bottom side.
I could add also that the space between your ears might not be wide enough for you being able to comprehend what I'm showing here.
Peter Daniel said:
I'm looking at the PCB now.
If you call that bypassing, I'm not surprized why you can't put an OPA627 playin' good music.
Peter Daniel said:
What you're showing me here doesn't tell me anything.
Well, it tells, but we would go directly to Texas. 😀
- Status
- Not open for further replies.