Thanks a lot for your replies. Now that I know the terminology will be much easier to find what I have in mind.
Yes I'm planning to do an unregulated snubberized PSU (ala Carlosfm) but with one rectifier bridge and one transformer per channel without common ground like this
http://myweb.tiscali.co.uk/nuukspot/decdun/gc/gcpsu.2t.2r.gif
I will take care to include the separate switching and the ground leads, including an earth safety ground as suggested.
Regards
Antonio
Yes I'm planning to do an unregulated snubberized PSU (ala Carlosfm) but with one rectifier bridge and one transformer per channel without common ground like this
http://myweb.tiscali.co.uk/nuukspot/decdun/gc/gcpsu.2t.2r.gif
I will take care to include the separate switching and the ground leads, including an earth safety ground as suggested.
Regards
Antonio
A Sanchez said:
Or maybe like this? (click photo)
P.S. The metal shell of a commercial 1-piece bridge rectifier, is designed to be grounded, therefore it can go inside the same chassis as commercial broadcast equipment, thereby eliminating the noises caused by excess cable lengths. The noises are from rectifiers, not transformers. Here's some more ways to quiet the rectifier. . .
See also: http://www.diyaudio.com/forums/showthread.php?postid=1500500#post1500500
See photo #4 here: http://diyaudioprojects.com/Chip/Synergy-LM3875-Gainclone/index.htm
And, have a look at other amplifier builds to see that the heatsink is often used as a shield/divider between power and amplifier.
Attachments
Daniel,
I want to have the PSUs in a separate box for practical and WAF reasons.
The WAF reasons is that there is no room so pile up more boxes where I have the hifi system and so I need to have it discretely placed on the floor on one side.
The practical reason is that I want to start trying different chipamp configurations/chips and it is handy to not have to deal with transformers and bridges and electrical cables every time I want to make trials of new amps. Having a separate box with XLR connectors will allow me to plug and play/test new designs quick and easily.
I hope I do not put too much noise doing this and I have seen many pros also do it.
Antonio
I want to have the PSUs in a separate box for practical and WAF reasons.
The WAF reasons is that there is no room so pile up more boxes where I have the hifi system and so I need to have it discretely placed on the floor on one side.
The practical reason is that I want to start trying different chipamp configurations/chips and it is handy to not have to deal with transformers and bridges and electrical cables every time I want to make trials of new amps. Having a separate box with XLR connectors will allow me to plug and play/test new designs quick and easily.
I hope I do not put too much noise doing this and I have seen many pros also do it.
Antonio
A Sanchez said:
Oh sure! Its a very minor audio difference, and nothing to worry about. A difference is a tool, and the thing that matters most is your appreciation of its effect.
I'm so glad that you're doing this for practical reasons. But, I wanted to make it clear, that a single enclosure can perform quite well, and that seperate boxes aren't an audio enhancement.
Seperate boxes are, in fact. . . only boxes.
Its about time to start that power supply thread, isn't it?
Okay. That will be just as soon as I make something really nice to plug into that power supply.
Its been slowly forming here: http://www.diyaudio.com/forums/showthread.php?postid=1514985#post1514985
That's LM1875, now up to a 30w per channel, and it doesn't have the Spike system--so the typical chipamp clipping isn't present.
I've been going slowly so that it looks pretty and do-able.
The input circuit gets put on today. Wish me luck at keeping it good looking.
Its been slowly forming here: http://www.diyaudio.com/forums/showthread.php?postid=1514985#post1514985
That's LM1875, now up to a 30w per channel, and it doesn't have the Spike system--so the typical chipamp clipping isn't present.
I've been going slowly so that it looks pretty and do-able.
The input circuit gets put on today. Wish me luck at keeping it good looking.
Power Supply
Its started here:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=123772
Other progress:
Got a "home run" success on the LM1875 project. It sounds great, deals with difficult speakers, its easy, and it produced a bit more output power than intended.
It was a pleasant experience.
Managed a LM3875, good demonstration; however, it also demonstrated that a printed circuit board, such as the AudioSector product, can contribute to sound and success of the amplifier project. My veroboard project turned out very transparent, but not as "musical" as the AudioSector.
Much research on rectifiers, power supplies, and wiring while observing beginners in the process of building popular designs. From their comments and questions, I do think that a "clearly marked" approach could be beneficial.
Its started here:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=123772
Other progress:
Got a "home run" success on the LM1875 project. It sounds great, deals with difficult speakers, its easy, and it produced a bit more output power than intended.
It was a pleasant experience. Managed a LM3875, good demonstration; however, it also demonstrated that a printed circuit board, such as the AudioSector product, can contribute to sound and success of the amplifier project. My veroboard project turned out very transparent, but not as "musical" as the AudioSector.
Much research on rectifiers, power supplies, and wiring while observing beginners in the process of building popular designs. From their comments and questions, I do think that a "clearly marked" approach could be beneficial.
Just registered here in the first case to see the pictures, but after seeing them I want to do it myself A little bit offtopic, but 2 months ago I ordered at eBay the Xiang Shiang 708B tube headphone/pre amplifier. I thought that it would drive two small speakers, but that was not the case.. So I'm now looking for a DIY amplifier, which I can drive by my Xiang Shiang 708B
Does anybody find the Gainclone a good solution for that?
A little bit offtopic, but 2 months ago I ordered at eBay the Xiang Shiang 708B tube headphone/pre amplifier. I thought that it would drive two small speakers, but that was not the case.. So I'm now looking for a DIY amplifier, which I can drive by my Xiang Shiang 708B Does anybody find the Gainclone a good solution for that?
What i meant to say is that spike "sound" is not a way of clipping but rather an indicator of a design problem. 90% times wrong heatsink, 9.9 times too much voltage 0.001 times people trying to drive ten parallel 2Ohms loudspeakers. Almost all loudspeakers are either 8 or 4 ohms (4 ohms is the load at which the ic shows it's rated power) so this should definitely not be blamed as the most probable cause of spike acting.
As the datasheet states, the system should be designed so as spike does not act during normal operation.
Hi,
take this senario, a massive heatsink on your chipamp, held @ 20degC. Connect a reactive 4ohm load. Allow the supply voltages that National say develops maximum power into 4r0.
Now drive the reactive load. At times the spike protection system will operate into this load. It could be IV limiting or just plain I limiting.
The problem gets worse if the chip temperature is allowed to rise above 25degC (the temp at which most of National's specification/data is published) and becomes very significant at Ts temps exceeding 50degC.
National's advice on heatsinking shows even higher temps in the chip but they conveniently avoid informing us of the complementary specs at this elevated temperature.
That is part of the reason I have been consistently saying make the heatsink twice as big as National's guide suggests.
This is not a National design problem as such, it's more of trying to extract too much power/current from a chipamp feeding low impedance loads i.e. 4ohm speakers. I'll say again that 60W into 8r0 is OK for a domestic music signal where average levels are around -20dB below peak. But that the maximum power target when feeding 6 ohm or 4ohm loads should in my opinion be reduced to somewhere much less than 60W and again only used for domestic listening with typical music signals at average levels of around 20dB below peak.
That is why many report success with chipamps into efficient speakers sensitivity typically >90dB/W/m.
Asking a chipamp to deliver -10dB average level music into a low sensitivity 4ohm reactive load when supplies are around the voltage required for 60W into 4r0 is going to ocassionally/repeatedly exceed the Tc=25degC spec limits and continuously trigger spike when Tc~100degC.
take this senario, a massive heatsink on your chipamp, held @ 20degC. Connect a reactive 4ohm load. Allow the supply voltages that National say develops maximum power into 4r0.
Now drive the reactive load. At times the spike protection system will operate into this load. It could be IV limiting or just plain I limiting.
The problem gets worse if the chip temperature is allowed to rise above 25degC (the temp at which most of National's specification/data is published) and becomes very significant at Ts temps exceeding 50degC.
National's advice on heatsinking shows even higher temps in the chip but they conveniently avoid informing us of the complementary specs at this elevated temperature.
That is part of the reason I have been consistently saying make the heatsink twice as big as National's guide suggests.
This is not a National design problem as such, it's more of trying to extract too much power/current from a chipamp feeding low impedance loads i.e. 4ohm speakers. I'll say again that 60W into 8r0 is OK for a domestic music signal where average levels are around -20dB below peak. But that the maximum power target when feeding 6 ohm or 4ohm loads should in my opinion be reduced to somewhere much less than 60W and again only used for domestic listening with typical music signals at average levels of around 20dB below peak.
That is why many report success with chipamps into efficient speakers sensitivity typically >90dB/W/m.
Asking a chipamp to deliver -10dB average level music into a low sensitivity 4ohm reactive load when supplies are around the voltage required for 60W into 4r0 is going to ocassionally/repeatedly exceed the Tc=25degC spec limits and continuously trigger spike when Tc~100degC.
Rave17 said:Just registered here in the first case to see the pictures, but after seeing them I want to do it myself
Does anybody find the Gainclone a good solution for that?
Well, power op amps (chipamp/gainclone) can work at low impedance. I've never seen one at 127 ohms (headphone).
The best I could suggest is to lower the impedance as far as known specs with a 5k or 10k input impedance for the gainclone. Yet, personally, I don't know how to make up the remaining difference. Perhaps a "buffer" circuit added?
Possible (wild guess):
a "gainclone" with. . .
NFB: 150R vs 10R--with 220uF (cap at CI)
Input impedance: 127R
And, that will match headphones, but I have no idea if it works in practice.
Sure:
I've actually never found it necessary to match impedance when the source is lower than the amp, so perhaps our LM1875 project will do just fine for you.
Rave17 said:. .. So you are saying that a LM1875 amp would be a better idea, in my case, then a LM3875 amp? I didn't had the time yet to read al the topics about those subjects, so maybe it's a to easy question
In my opinion, the LM1875 amps produce fast, easy, fidelity because of its tolerance.
High fidelity, fast, is also possible with the LM3875, especiallly when you buy it from Audiosector.
On the topic of hifi, its all about picking the right size fight, and its okay to call for help when the need arises.
Its also my opinion that smaller amplifiers are easier than larger amplifiers when comparing "apples to apples." That's a very generalized statement.
Of course, I think that you should collect quite a few opinions, and not just mine.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.