-
Notifications
You must be signed in to change notification settings - Fork 3
/
Copy pathaudio-cheap.txt
111 lines (101 loc) · 4.8 KB
/
audio-cheap.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
Summary, PAM8301 "worked pretty good!" (good enough that emeb went and bought some)
Make sure to go to the bottom for the updates
13:48 < dongs> emeb: found PAM8301 for $~0.20
13:49 < dongs> whats your higher highs assessment of that one
13:49 < emeb> looking...
13:50 < emeb> that's cute
13:50 < emeb> tiny, no big caps needed...
13:58 < emeb> dongs: looks like a nice part. You can make a simple RC LPF for
reconstruction and gain
control but it's hard for me to tell right off what the ideal
values would be.
13:59 < emeb> so that plus 3 passive parts hooked to the F4xx DAC output would
probably work fine.
13:59 < dongs> emeb: on where? input or output
14:01 < emeb> F4xx DAC pin - DC blocking cap (0.1-1.0uf) - LPF & Gain resistor
- PAM input
14:01 < emeb> then an LPF cap from PAM input to GND
14:01 < dongs> hm thats adding more parts than i really care about
14:01 < dongs> i can understand rc lpf, what about rest of shit?
14:01 < dongs> wahts gain for
14:02 < emeb> DC blocking cap?
14:02 < dongs> no like
14:02 < dongs> i ahve dc cap now
14:02 < dongs> dac -> cap -> pam input
14:02 < dongs> i can add a rc lpf after cap (?or before cap?)
14:02 < dongs> but whats the 'gain resistor' thing?
14:03 < emeb> gain resistor is to control how hard the DAC drives the amp
14:03 < emeb> if DAC full swing overdrives the amp then you lose resolution
14:03 < emeb> so resistor is there to limit gain
14:04 < emeb> DAC gives ~3.3Vpp
14:04 < dongs> but if you turn on buffer
14:04 < dongs> its less
14:04 < emeb> only slightly less
14:04 < emeb> 3.1Vpp maybe
14:05 < dongs> so you think its necesary
14:05 < emeb> you could always just reduce the signal range in DSP
14:05 < emeb> but that cuts resolution and raises noise floor
14:05 < emeb> so it's better to use analog gain -> resistor
14:06 < emeb> But - for minimum parts, just use DC blocking cap and make up for
it in software.
14:06 < dongs> yay
14:07 < emeb> If it were me I'd put two more passives in the signal path just
for pro / hifi. :)
14:08 < dongs> so R in series and C to gnd right? but thats just LPF?
14:08 < emeb> yes
14:08 < dongs> or do you mean another R after that for gain?
14:08 < emeb> no
14:08 < dongs> just thast?
14:08 < emeb> yes
14:08 < emeb> the PAM input has 10k impedance
14:08 < emeb> so the value of R you use affects gain too
14:08 < dongs> ah right
14:08 < dongs> im looking at some tarduino shit
14:08 < dongs> 1.5k/10nF for 11khz lpf
14:09 < Laurenceb> lol
14:09 < Laurenceb> dongs is a perman00b
14:09 < dongs> yuep
14:09 < dongs> and proud of it
14:09 < dongs> i leave analol'g shit to pros
14:10 < dongs> emeb: wait, that RC is before or after ac bias cap
14:10 < emeb> dongs: yes - those values give ~10.6kHz corner freq
14:10 < Laurenceb> Quinetiq etc
14:10 < emeb> and it doesn't matter what order you put them in
14:11 < dongs> mmkay ill do them before
14:11 < emeb> but I'd put DC cap @ F4 output, then R, then C
14:11 < dongs> o okay
14:11 < dongs> then after.
14:11 < emeb> the 1.5k resistor won't affect gain much
14:11 < dongs> i could always replace it wiht 0R
14:11 < dongs> and NP the cap
14:12 < dongs> for max pro
14:12 < emeb> yes
14:12 < dongs> if something isnt working
14:13 < emeb> if you need to cut gain then just multiply R and divide C by same
factor to preserve corner freq
14:13 < emeb> (so double R, half C, etc)
--- "later" ---
00:29 < emeb> dongs: PAM8301 working - such loud, very gain, wow.
00:29 < dongs> lol
00:29 < dongs> indeed
00:30 < dongs> i had to like 1/4 the dac outputs
00:30 < dongs> and it was still almost clipping
00:30 < emeb> I put a 100k resistor in the path to the amp input
00:30 < emeb> still very loud
--- "later still" ---
22:00 <emeb> karlp: heh - that PAM8301 mentioned in the zypsnips is kinda weaksauce. I've tried to use it to drive cheap 8ohm speakers and it distorts badly at any usable volume.
----- separately ----
18:47 <BrainDamage> yes, you need a 12ax7 tube driven by 120V
18:47 <kakimir> how many watts?
18:47 <BrainDamage> encased in a 10kg block, for vibration dampening
18:48 <BrainDamage> 3-6 for the filament, depending on the drive
18:48 <BrainDamage> output will be few douzens mW, a hundred at best
18:50 <kakimir> not enough
18:51 <Steffanx> Just ask what crt would do. He's into amplification
18:54 <BrainDamage> if you want super pro solution that no audiophile would ever approve, get an ADSL line driver
18:54 <BrainDamage> their current capability, slew rate and distortion figures beat all headphone shit hands down
18:54 <BrainDamage> and they cost 2$
-- testing audio --
20:20 <emeb> Chasing down low level audio noise in my bench-top development setup. Discovered that the STLink V3 mini that I was using generates a lot of broadband audio noise.
20:20 <emeb> Just hooking the GND pins of the SWD connector causes loud "shhhhhhhhhh".
20:21 <emeb> but using the STLink V2.1 on an old Nucleo board is quiet.