Ah, your design is an inverting op-amp, mine is noninverting.
I've not built an inverting amp yet, but if I remember well, there's no connection from input to ground via a resistor like that R1 in my circuit...
But - do you already use that amp??? And R2 is 110k? So the gain would be 55, that's huge. I think you could only set a very low output volume in your mp3 player, can you?
Cheers,
Dominique
EDIT: Forgot to mention: If you got dual 25V transformers, you have to assure the final DC voltage at the amp will be +-30V at max (look at the datasheet)
'Cause 25V AC x 1,41 = 35,25V
Then it's a question of how big your diode losses are and what voltage your transormer really delivers. Because with low load, the transformer should output even more than 25V.
So I guess those 25V transformers are too much without "destroying" some of the voltage via a regulator (that would have to sit on a rather big heatsink then or a high power resistor...
Hmmm.... better use a different transformer or another LM chip?
I've not built an inverting amp yet, but if I remember well, there's no connection from input to ground via a resistor like that R1 in my circuit...
But - do you already use that amp??? And R2 is 110k? So the gain would be 55, that's huge. I think you could only set a very low output volume in your mp3 player, can you?
Cheers,
Dominique
EDIT: Forgot to mention: If you got dual 25V transformers, you have to assure the final DC voltage at the amp will be +-30V at max (look at the datasheet)
'Cause 25V AC x 1,41 = 35,25V
Then it's a question of how big your diode losses are and what voltage your transormer really delivers. Because with low load, the transformer should output even more than 25V.
So I guess those 25V transformers are too much without "destroying" some of the voltage via a regulator (that would have to sit on a rather big heatsink then or a high power resistor...
Hmmm.... better use a different transformer or another LM chip?
Yeah...it looks a gain of 55 but in an inverting design, you need to count all impedances including the two little 1K summing resistors. I'm combining the right and left channels with these. I wasn't 100% sure if these two little resistors should be added to the calculation as parallel or if one could be ignored. If you think of them as parallel, then you get
110K / (2K + parallel value of 1K and 1K)
110K / (2K + 500)
110K / 2.5K = 44
Note that R3 is 2.49K to approximate the parallel value of R2 and R1 (it's actually 2.44K but 2.49K s the closest value resistor I had)
C1 as a capacitor also adds some resistance but it's hard to measure this. I used a nice little non-polar electrolytic here since the film capaictors of the same value were huge and expensive. I still ordered some film caps but wound up going with eletrolytics...yeah...I know....a no-no...putting electrolytics in the input path but it sounds OK
110K / (2K + parallel value of 1K and 1K)
110K / (2K + 500)
110K / 2.5K = 44
Note that R3 is 2.49K to approximate the parallel value of R2 and R1 (it's actually 2.44K but 2.49K s the closest value resistor I had)
C1 as a capacitor also adds some resistance but it's hard to measure this. I used a nice little non-polar electrolytic here since the film capaictors of the same value were huge and expensive. I still ordered some film caps but wound up going with eletrolytics...yeah...I know....a no-no...putting electrolytics in the input path but it sounds OK
Ah ok, if you get those 25V at the amp it's fine.
Maybe you could add a resistor that consumes some additional current when the amp isn't working to drop the voltage below those 31V.
I guess 31V should still work ok, but it's a risk that your will live shorter...
For the gain calculation, now you confused me
Anyway, you should remark if your gain is ok.
Polar caps are better in general. At least when I used a cheap nonpolar electrolytic cap on an amp's input I could easily hear a loss of high frequencies, while even a cheap polar electrolytic sounded fine to me.
If you didn't compare, then I suggest you try a (good quality and bigger) polar electrolytic against that nonpolar. Well, just connect it in parallel when music is playing and you may hear a difference if there's some. Desoldering and replacing the cap makes you forget the previous sound anyway...
Cheers,
Dominique
Maybe you could add a resistor that consumes some additional current when the amp isn't working to drop the voltage below those 31V.
I guess 31V should still work ok, but it's a risk that your will live shorter...
For the gain calculation, now you confused me
Anyway, you should remark if your gain is ok.
Polar caps are better in general. At least when I used a cheap nonpolar electrolytic cap on an amp's input I could easily hear a loss of high frequencies, while even a cheap polar electrolytic sounded fine to me.
If you didn't compare, then I suggest you try a (good quality and bigger) polar electrolytic against that nonpolar. Well, just connect it in parallel when music is playing and you may hear a difference if there's some. Desoldering and replacing the cap makes you forget the previous sound anyway...
Cheers,
Dominique
" ... what do you mean with that voltage divider? I use a dual voltage transformer and this gives me two rails and the center which I connected to the ground of the amp's pcb. ..."
Yes, a + / 0 / - , dual coil secondary with center tap for establishing equal potential difference "above and below" the 0 VDC ground plane ... and then you don't need to establish any other ground reference to the Vcc+ and Vdd- (V+ & V-), no need for a voltage divider ...
= no problem found.
Yes, a + / 0 / - , dual coil secondary with center tap for establishing equal potential difference "above and below" the 0 VDC ground plane ... and then you don't need to establish any other ground reference to the Vcc+ and Vdd- (V+ & V-), no need for a voltage divider ...
= no problem found.
Dominique:
You cannot go polar on the input cap since the audio signal is inherently alternating.
Yeah...the beauty (curse) of electrolytics is that they "break in" or more accurately the electrolytic juice starts to make bonds with the film and changes everything as it does so. See how many post about Black Gates taking an inordinate amount of time to "break in." The Nichicon website has some really cool info on this process. Likewise, the enemy of electrolytics is idle time and aging...like good wine
You cannot go polar on the input cap since the audio signal is inherently alternating.
Yeah...the beauty (curse) of electrolytics is that they "break in" or more accurately the electrolytic juice starts to make bonds with the film and changes everything as it does so. See how many post about Black Gates taking an inordinate amount of time to "break in." The Nichicon website has some really cool info on this process. Likewise, the enemy of electrolytics is idle time and aging...like good wine
hey dominique,
i would like to build a lm1875 amp based on your design is that okay with you?
how did you design the pcb? do you have gerbers or a pdf that you can send me, something i can give the board house for them to etch it. i will be getting very basic boards done so actually i just need the copper layer and the position of the component holes...
can you send me these files?
love, light and peace
mymindinside
i would like to build a lm1875 amp based on your design is that okay with you?
how did you design the pcb? do you have gerbers or a pdf that you can send me, something i can give the board house for them to etch it. i will be getting very basic boards done so actually i just need the copper layer and the position of the component holes...
can you send me these files?
love, light and peace
mymindinside
sticking my nose in ...
Looking at http://www.national.com/mpf/LM/LM1875.html ... It seems to be almost an exact fit ... (IMOP: " The LM1875 delivers 20 watts into a 4 or 8 load on ±25V supplies " so C6 and C7 could easily be larger = to 470 uF or 1000 uF (50 volt, working), but nothing wrong with keeping to National's suggestions.)
The notes above in this thread about that low pass filter usage on the output (R5 & C5) would also apply = probably not needed, but could be used with high confidence.
DO use a heat sink, of course.
Its a good little board design. FWIW: I don't see anything I would change ...
(If you can print the image ( http://www.diyaudio.com/forums/attachment.php?s=&postid=1085737&stamp=1166558658 ) to ~~1/2 size (~50%), that would probably work ... I would size the image so that the pin holes for the LM1875 were approximately correct spacing, then the rest would also be good. >> Note that the National artists have the LM1875 locations on their own drawings WRONG on page 8 of the data sheets at http://www.national.com/ds/LM/LM1875.pdf <<
)
Looking at http://www.national.com/mpf/LM/LM1875.html ... It seems to be almost an exact fit ... (IMOP: " The LM1875 delivers 20 watts into a 4 or 8 load on ±25V supplies " so C6 and C7 could easily be larger = to 470 uF or 1000 uF (50 volt, working), but nothing wrong with keeping to National's suggestions.)
The notes above in this thread about that low pass filter usage on the output (R5 & C5) would also apply = probably not needed, but could be used with high confidence.
DO use a heat sink, of course.
Its a good little board design. FWIW: I don't see anything I would change ...
(If you can print the image ( http://www.diyaudio.com/forums/attachment.php?s=&postid=1085737&stamp=1166558658 ) to ~~1/2 size (~50%), that would probably work ... I would size the image so that the pin holes for the LM1875 were approximately correct spacing, then the rest would also be good. >> Note that the National artists have the LM1875 locations on their own drawings WRONG on page 8 of the data sheets at http://www.national.com/ds/LM/LM1875.pdf <<
)- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.