Author |
Message |
hedefalk
Joined: Aug 29, 2017 Posts: 51 Location: Stockholm, Sweden
|
Posted: Thu Sep 28, 2017 7:36 am Post subject:
Debugging MFOS panning mixer… |
|
|
Newb alert here.
I'm just starting out building my diy eurorack thingie.
Second module I'm building is a simplified version of Ray Wilson's panning mixer. I've just taken two channels and stripped the entire FX bus. That leaves me with just 4 TL07 opamps and two 386n-4's so I figured I would just use a TL074CN as opposed to using 4 TL071's as schematic says. So first question, is this a bad idea in any way? Leakage?
Second question, is there something that I'm missing in believing I could just strip it to use only the circled parts here in the schematic:
After breadboarding it half decently I built it on a breadboard-like proto board and I'm currently debugging it. Seems like first stage buffers are fine but something weird happens with the signal at the pan stage (pans seems to not work correctly)
I've checked everything I can think of offline with a digital meter and I'm currently debugging with a cheap scope hooked up, but I'm really out of ideas what to check for.
Any pointers are highly appreciated, like a standard checklist to go through or anything…
…while writing this, I was gonna do list of my observations and doing that I found that there's most likely just something wrong with the pan pots and I had turned the panel wiring for the main phones level the wrong way.
Ok, having fixed the phones level I now just have to fix the pan's. I guess I have a little bit of distortion due to that, signal's being too strong since the grounding via pan wiper is gone, but my biggest concern is a white noise bright hizzing I have in the phones that is entirely unrelated to the turn of any pots, both input gain and also master headphone level.
I would think that this is coming from the LM386's and since the main volume pot is actually only doing something with the input signal to these power amplifiers this is probably something I'll have to live with? If I want true quietness on volume 0, then I'd have to have the pot attenuate after the last power amp too, right?
Description: |
|
Filesize: |
1.08 MB |
Viewed: |
501 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
Description: |
|
Filesize: |
1.26 MB |
Viewed: |
474 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
Description: |
|
Filesize: |
2.31 MB |
Viewed: |
473 Time(s) |
This image has been reduced to fit the page. Click on it to enlarge. |
|
|
|
Back to top
|
|
|
hedefalk
Joined: Aug 29, 2017 Posts: 51 Location: Stockholm, Sweden
|
Posted: Fri Sep 29, 2017 3:05 am Post subject:
|
|
|
I've googled a bit and tried adding a 10k resistor and a cap in series over output pin 5 and gain pin 8 or 1. Like here https://dmohankumar.wordpress.com/2015/03/15/reducing-hiss-in-lm-386-amplifier-design-note-12/#more-10895
This removes the hizz but replaces it with an even more pronounced humming. I think this method probably isn't applicable unless the gain pins are used, but the MFOS one leave them floating to use the internal 20 gain factor I think. I'll revisit the data sheet |
|
Back to top
|
|
|
PHOBoS
Joined: Jan 14, 2010 Posts: 5591 Location: Moon Base
Audio files: 705
|
Posted: Fri Sep 29, 2017 7:54 am Post subject:
|
|
|
Adding components between pin 1 & 8 increases the gain so yeah you're better of to leave them floating. You can indeed use it
for filtering and I don't think you would need to use the gain pins for it, but I'd have to check the internal structure of the chip.
What might help with the hum is adding a large capacitor (100uf or something like that) directly over the power pins, I am actually
a bit surprised it only uses 0.1uF in the schematic. Of course a regulated powersupply is also very important. I've had circuits that
work fine with a simple unregulated wallwart but others produce a lot of hum without any regulation. Personally I am not a big fan
of the LM386 especially not when used for an headphone amp. They are ok for adding a small speaker to a circuit or if you need to
boost a signal for processing that isn't audible. _________________ "My perf, it's full of holes!"
http://phobos.000space.com/
SoundCloud BandCamp MixCloud Stickney Synthyards Captain Collider Twitch YouTube |
|
Back to top
|
|
|
hedefalk
Joined: Aug 29, 2017 Posts: 51 Location: Stockholm, Sweden
|
Posted: Fri Sep 29, 2017 8:12 am Post subject:
|
|
|
Thanks!
At this point I fell I'm stuck with this choice though since I don't think there are any pin compatible and I just want to finish this one quickly.
But out of curiosity, what power amp would you recommend to use instead of lm386? For my next amp build
I'll try adding larger caps next to the IC. There are larger electrolyts on the power connection (isn't shown here)
However, the hum was just something I got when doing that feedback to the gain pins. I think I'll leave that idea aside.
I just read this once again: http://www.transkommunikation.ch/dateien/schaltungen/space_station_receiver/informationen/08_-_reducing_hiss_in_an_lm386_audio_amplifier.pdf
I don't really understand why this works only when using the gain pins because for me it gets really hummy. |
|
Back to top
|
|
|
PHOBoS
Joined: Jan 14, 2010 Posts: 5591 Location: Moon Base
Audio files: 705
|
|
Back to top
|
|
|
AlanP
Joined: Mar 11, 2014 Posts: 746 Location: New Zealand
Audio files: 41
|
Posted: Fri Sep 29, 2017 1:40 pm Post subject:
|
|
|
Another option is the headphone driver in the Yusynth Output Module -- works very well. Mind the 1/2W resistors, though, where indicated. |
|
Back to top
|
|
|
hedefalk
Joined: Aug 29, 2017 Posts: 51 Location: Stockholm, Sweden
|
Posted: Mon Oct 02, 2017 3:36 am Post subject:
|
|
|
Thanks for all the tips!
I think for this module, I'm stuck with my choice and have tried all "band aid" things I've found:
1) Doing the feedback from output to gain:
1.1) If no connection between 1 and 8 and connecting the feedback cap+resistor to pin 8 I get a maxed out beep, maybe 5000Hz that almost killed my ears a few times.
1.2) Connecting to pin 1 I instead get a hum that is more pronounced than the original hizz.
2) No decoupling caps seem to make any difference at all for this hiss - it doesn't seem to be related to power flucts.
3) Modding my circuit to actually use something like a 1k + 10mF series between pin 1 and 8 I CAN connect the feedback from output to pin 8 as suggested in places like here https://dmohankumar.wordpress.com/2015/03/15/reducing-hiss-in-lm-386-amplifier-design-note-12/#more-10895
Again though, that gives me a humming that is simply worse than original hizz.
I think I'm out of options. I have only one batch of lm386-n4's that I bought from aliexpress and I'm thinking maybe they're a bit defective or something. I have tried a few. I have a few standar lm386 (lower voltage ones) on the way and I'll try those to compare but for now I'll have to live with the hizz. |
|
Back to top
|
|
|
hedefalk
Joined: Aug 29, 2017 Posts: 51 Location: Stockholm, Sweden
|
|
Back to top
|
|
|
hedefalk
Joined: Aug 29, 2017 Posts: 51 Location: Stockholm, Sweden
|
|
Back to top
|
|
|
hedefalk
Joined: Aug 29, 2017 Posts: 51 Location: Stockholm, Sweden
|
|
Back to top
|
|
|
hedefalk
Joined: Aug 29, 2017 Posts: 51 Location: Stockholm, Sweden
|
|
Back to top
|
|
|
hedefalk
Joined: Aug 29, 2017 Posts: 51 Location: Stockholm, Sweden
|
Posted: Tue Mar 20, 2018 5:23 am Post subject:
|
|
|
Just an update:
Tried this earphone output with my old Beyer Dynamic DT-100 which is at 400 ohm. Sounds waaay better and the noise is hardly audible. I would just want to state the to some probably obvious. The LM358 is waaay to noisy for low impedance earphones, but on this one it seems like a perfect match. I have tried to use these earphones in more traditional equipment before and they sound really bad there. I guess they need a lot of power and therefore also cover up a lot of weak noise. |
|
Back to top
|
|
|
|