MJL21193 said:
That's such a nice looking project.
Oh, no hard feelings. Thanks for the compliment, and for the project photos.
I do have a big question. . .
Some folks will want to use their PC, Mac, or MP3, or TV as the source.
A needed gain of 38 to 45. . .
For increased gain, I need (and lack) the figures on how to minimize DC offset through NFB resistor value selection. (page 16, lm3875.pdf)
Without those figures, I could suggest an AC coupled NFB (adding a cap); however, that is quickly off into the "deep end" of component selection. Or, I could use the "factory standard" guidelines (gain 21) and add a preamp (gain 15).
So, you see my plan in using a preamp. . . This postpones the topic of "voicing" and eq, onto the preamp project (traditional), instead of doing that to a power amp.
The esoteric, hifi genre voicing of power amps can work well, with 25w (lm1875) and smaller amplifiers; but, since that's not what we're doing, I'd like to skip the esoterics. However, the preamp produces less than 25w (grin), and its "fair game" for voicing efforts. . . later.
We could skip that too, if we could pump up the LM3875TF's gain up to about 38 (barely sufficient), and also have a dc coupled (no cap) NFB, and also have very low dc offset, as not explained on page 16 of LM3875.pdf.
So, how do we do it?
danielwritesbac said:
Typical line output of all of the above sources is 1.2Vrms (3.4Vpp). Consulting the chipamp .xls, this is an optimal input for a chip amp with gain of 21.
More of the "gray" area...
MJL21193 said:
There's a slight problem with variable output devices, like mp3, pc, tv. . . At full output, their dynamics are hampered.
MJL21193 said:
danielwritesbac said:
Line output is not the headphone output of an mp3 player. You DEFINITELY don't need a preamp for headphone out, as this will already be amplified. If you need to use headphone out for your source input, the volume control on the device has to be turned down to approximate line level (1.2Vrms) or you will drive the chipamp into clipping.
This is the basic education I was referring to earlier (and an example of inaccuracy and misinformation).
MJL21193 said:
Actual application and live testing is involved.
The education was sufficient, as in the Edison method of a 1000 ways not to make a light bulb. This isn't theory.
I've tried a selection of mp3 players, and I've tried a selection of computer sound, both onboard chip and add-in cards.
The majority of these are notoriously insufficient (well documented) and unable to drive a chipamp to full power, while retaining expected fidelity. With the modern "ear protection" variable outputs, the output power is lacking on these devices. They are often able to output "tones" at their stated power; however, there is much distortion of mp3 format music whenever headroom is zero.
I'm currently listening to a chipamp sans preamp, simply using a pot.
This works quite well with my CD player, but leaves room for improvement with my MP3 player.
I think that a simple implementation with a pot can be a valuable learning tool and it may well be all that some people need... not everybody uses an MP3 player in their home system 🙂
7/10
This works quite well with my CD player, but leaves room for improvement with my MP3 player.
I think that a simple implementation with a pot can be a valuable learning tool and it may well be all that some people need... not everybody uses an MP3 player in their home system 🙂
7/10
seventenths said:
Oh thank you!
Here I was wondering if I made a mistake of following page 16 (LM3875.pdf) too closely, and used the 100k example because that roughly matches headphones. It seems that everyone else is using 22k (or so). . . which is a stronger load for the source, and possibly could encourage a bit more power out of the source's op amp? Is that true?
danielwritesbac said:
A full shift into the realm of fantasy again.
I can't speak for anything you have tried to do, but can only comment on what works in the real world.
My opinion: Headphone output from a TV, mp3 player or soundcard are not ideal sources for any amplifier. I wouldn't bother building anything if that's the only thing I had to drive it.
Line output from my soundcard drives my 21V/V gain chipamps just fine.
Help Please!
I am going to build this version:
http://dogbreath.de/Chipamps/GainCardCopy/GainCardCopy.html
I am ordering the parts from Farnell and I am having difficulties locating the right ones due to unknown terminology and other aspects, for example:
1- They do not have the 4.7uf polyester capacitor in 50v version but only in 63v version. Does it matter? would the 63v work the same?
2- The schematics refer to carbon resistors, but there are three types: ceramic, compound and film.
3- I cannot find 22k and 680R 1/4 W resistors. They only have 1W ones. Again, does the difference in Watts matter?
4- Cannot find 10k .35W metal film resistors. Only 10k .33w metal film "fusible" ones. Is this OK?
5- Is 2R7 the same as 2.7R?
Many thanks for your help.
Regards
Antonio
I am going to build this version:
http://dogbreath.de/Chipamps/GainCardCopy/GainCardCopy.html
I am ordering the parts from Farnell and I am having difficulties locating the right ones due to unknown terminology and other aspects, for example:
1- They do not have the 4.7uf polyester capacitor in 50v version but only in 63v version. Does it matter? would the 63v work the same?
2- The schematics refer to carbon resistors, but there are three types: ceramic, compound and film.
3- I cannot find 22k and 680R 1/4 W resistors. They only have 1W ones. Again, does the difference in Watts matter?
4- Cannot find 10k .35W metal film resistors. Only 10k .33w metal film "fusible" ones. Is this OK?
5- Is 2R7 the same as 2.7R?
Many thanks for your help.
Regards
Antonio
MJL21193 said:
Thanks for the info. It is very helpful.
Since our results differ, the probable meaning is that the ouput capabilities of sources differ.
This has me thinking that I should do a standardized power amp project and propose a preamp as an add-on project.
danielwritesbac said:
??
Typical headphone impedance is quite a bit lower than 100k.
danielwritesbac said:
With each successive post you sink further. 🙂
I'm done here.
1- They do not have the 4.7uf polyester capacitor in 50v version but only in 63v version. Does it matter? would the 63v work the same?
Yes, it will work fine. You can always go higher in voltage rating (if physical size isn't an issue) but not lower.
2- The schematics refer to carbon resistors, but there are three types: ceramic, compound and film.
Carbon film.
3- I cannot find 22k and 680R 1/4 W resistors. They only have 1W ones. Again, does the difference in Watts matter?
They have 1/4w metal films in those values.
4- Cannot find 10k .35W metal film resistors. Only 10k .33w metal film "fusible" ones. Is this OK?
Use 0.5w metal film.
5- Is 2R7 the same as 2.7R?
Yes.
Yes, it will work fine. You can always go higher in voltage rating (if physical size isn't an issue) but not lower.
2- The schematics refer to carbon resistors, but there are three types: ceramic, compound and film.
Carbon film.
3- I cannot find 22k and 680R 1/4 W resistors. They only have 1W ones. Again, does the difference in Watts matter?
They have 1/4w metal films in those values.
4- Cannot find 10k .35W metal film resistors. Only 10k .33w metal film "fusible" ones. Is this OK?
Use 0.5w metal film.
5- Is 2R7 the same as 2.7R?
Yes.
>This has me thinking that I should do a standardized power amp project and propose a preamp as an add-on project.
Good idea. Keep it simple and add the preamp later if you feel it is required. You can use the pot plus gainclone as a standalone integrated for the time being.
Good idea. Keep it simple and add the preamp later if you feel it is required. You can use the pot plus gainclone as a standalone integrated for the time being.
Re: Help Please!
1. That is a component specific WIMA brand 4.7uF 50v capacitor, and the design appears reliant on this choice (IMO).
2. Carbon film resistors are a tan/cream color, and they are the most common sort.
3. You can find a "variety pack" of 1/4w resistors at your local electronics hobby shop.
4. Use an ordinary metal film resistor.
5. Yes, and a speaker output zobel needs at least 1/2w if not a 1w resistor.
For reference, the amplifier design from the link above will play 1 decibel louder than lm3875. Given that there's no noticable difference in power output, there must have been something else that you find attractive about that (above link) project. Personally, I would inquire about component specific information, and if the amplifier from 2005 is still in use or if it was replaced/updated. Here's the email address associated with that link above: mick 'at' feuerbacher 'dot' net
A Sanchez said:
1. That is a component specific WIMA brand 4.7uF 50v capacitor, and the design appears reliant on this choice (IMO).
2. Carbon film resistors are a tan/cream color, and they are the most common sort.
3. You can find a "variety pack" of 1/4w resistors at your local electronics hobby shop.
4. Use an ordinary metal film resistor.
5. Yes, and a speaker output zobel needs at least 1/2w if not a 1w resistor.
For reference, the amplifier design from the link above will play 1 decibel louder than lm3875. Given that there's no noticable difference in power output, there must have been something else that you find attractive about that (above link) project. Personally, I would inquire about component specific information, and if the amplifier from 2005 is still in use or if it was replaced/updated. Here's the email address associated with that link above: mick 'at' feuerbacher 'dot' net
Correction.
Thank you for pointing out the error.
Insulting others isn't an appropriate response for a request of information.
Typical headphone impedance is, in fact, 120 ohms.
http://sound.westhost.com/project100.htm
And there's a link, in case someone wants a headphone jack.
MJL21193 said:
Thank you for pointing out the error.
Insulting others isn't an appropriate response for a request of information.
Typical headphone impedance is, in fact, 120 ohms.
http://sound.westhost.com/project100.htm
And there's a link, in case someone wants a headphone jack.
Daniel.......
I built my first Gainclone some years ago and it sounded fantastic. May be I just got very lucky because because I didn't know anything about the likes of "component specific information", or any of the other wonderful issues that you come up with.
You see, the real beauty of the chip amp is the very low component count and the simplicity of the circuit. So it is quite possible to build a working GC, enjoy it, and then think about tuning a bit by substituting other parts and hearing the differences that they make.
I guess that you are trying to be helpful but if you read between the lines of the posters on this thread, they are complete novices and looking for ABC guidance, rather than being educated to degree level on the finer points of tuning a chip amp. As such you may be doing them a dis-service rather than moving them forward! 😉 But perhaps my signature says this all a bit more clearly!
I built my first Gainclone some years ago and it sounded fantastic. May be I just got very lucky because because I didn't know anything about the likes of "component specific information", or any of the other wonderful issues that you come up with.
You see, the real beauty of the chip amp is the very low component count and the simplicity of the circuit. So it is quite possible to build a working GC, enjoy it, and then think about tuning a bit by substituting other parts and hearing the differences that they make.
I guess that you are trying to be helpful but if you read between the lines of the posters on this thread, they are complete novices and looking for ABC guidance, rather than being educated to degree level on the finer points of tuning a chip amp. As such you may be doing them a dis-service rather than moving them forward! 😉 But perhaps my signature says this all a bit more clearly!
Re: Correction.
Hi Daniel,
You actually do insult to yourself when you steadfastly agrue with established fact.
Read more of Rod Elliot's site, he has lots of sound information to share. Short on mumbo-jumbo but long on fact. 🙂
danielwritesbac said:
Hi Daniel,
You actually do insult to yourself when you steadfastly agrue with established fact.
Read more of Rod Elliot's site, he has lots of sound information to share. Short on mumbo-jumbo but long on fact. 🙂
Re: Re: Correction.
I like to find the "high" and "low" parameters rather than assume, so this is most of the esoteric content that you'll find in other threads.
As for this thread, I have reported post#68 (my own post) to the moderator and asked for its removal on the grounds that its confusing. . .because I was really confused when I asked the question.
Rod Eliot's site is fascinating. Check out the corner frequencies topic on his subwoofer equalizer boost technology. Short on fact, but rattles the house nicely. 😀
EDIT: got to get on topic. . . Rod Elliot also has some amplifier and preamplifier information, and most is a good read.
MJL21193 said:
I like to find the "high" and "low" parameters rather than assume, so this is most of the esoteric content that you'll find in other threads.
As for this thread, I have reported post#68 (my own post) to the moderator and asked for its removal on the grounds that its confusing. . .because I was really confused when I asked the question.
Rod Eliot's site is fascinating. Check out the corner frequencies topic on his subwoofer equalizer boost technology. Short on fact, but rattles the house nicely. 😀
EDIT: got to get on topic. . . Rod Elliot also has some amplifier and preamplifier information, and most is a good read.
Nuuk said:
I'm actually requesting help in . . . avoiding. . . voicing the amplifier.
EDIT: We're doing the NFB next, its decision time, and I "hit the brakes"--time to ponder. . . and ask for help.
I think we got a solution. It was "Plan A" anyway.
That's a "factory standard" chip amp implemention with component values from LM3875.pdf.
and later. . .
a preamplifier (a more traditional item to "tinker" is the preamplifier).
Suggestions needed! 😀
- Status
- Not open for further replies.