digitally controlled stepped attenuators..??

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Thanks.  So there are 64 variations of combining 6 different K sections with relatively 1dB between each one?


 
Thanks,I got it harpo.  I just didn't believe you could get those possibilities with 6 numbers.  Sometimes
math is very cool and powerful.  I had to see for myself and manually add down to -48 based on Igor's numbers. ::)

what is the name for that list of possibilities? binary array?

Time to buy one of these pcb's since I am sick of running through my console as a glorified output select and
level control.  With horrible stereo tracking  :'(
 
Don't know if anyone linked to this page before. Here it is anyway, good info! http://www.vaneijndhoven.net/jos/switchr/index.html
 
Time to buy one of these pcb's since I am sick of running through my console as a glorified output select and
level control.  With horrible stereo tracking

Yes, totally agree. Those CRM sections on console's master are unbearable, after you try once good mastering quality level control.
First time we implemented stepped passive relay level control on La-Vi console:
http://www.groupdiy.com/index.php?topic=23907.0
Than, master section of La-Vi console became our commercial product, as part of mastering console.

We have a plans to add 3xCRM audio boards in DB studios, to replace weak SSL G6000 deck's control room section,
and have surround level control as well. CRM is finished and built many times thing...you can check it on my ordering thread,
as well as use other dig. controled relay attenuator, from MCS, ebay, etc., one MUST thing imho: replace the CRM level pot! :)
 
Guys, I'm using the MCS's digipot in one of my units for more than a year but recently it started to loose some steps. I tried to change encoder, cables, measured the voltages but no success . Here is the video link of behavior https://www.dropbox.com/s/qgaz3x898kkarcs/CONTROLL%20ERROR.mov
Any ideas ? :-\
 
Can be capacitors at encoder's inputs of uP. Can be pcb contacts near encoder, etc.
Possible to use optic encoder, HRPG series etc, they always give perfect behaviour, however, not cheap ($20-60 per encoder).
 
Thanks Igor. I doubt it's encoder since it worked great for about a year, then , when "drops" started I replaced with new encoder but without any improvement. Caps at encoder inputs? Interesting, I will try that too :)
 
Igor said:
Can be capacitors at encoder's inputs of uP. Can be pcb contacts near encoder

Nothing, still the same behavior after recapping and resoldering the pcb. Can it be the oscillator???
 
No response from MCS. I still can't figure out the problem. here is the schemo maybe someone can figure out what's going on with this "guy"  :(
 

Attachments

  • VolCon2_sch-neve volume.gif
    VolCon2_sch-neve volume.gif
    36.6 KB · Views: 91
Huh... I don't see any caps on the encoder...

Maybe put a scope probe on the two lines going into the micro (pins 6 and 9) and look for clean edges as you rotate the encoder.

JR

 
JohnRoberts said:
Huh... I don't see any caps on the encoder...

Maybe put a scope probe on the two lines going into the micro (pins 6 and 9) and look for clean edges as you rotate the encoder.

JR
Thanks John, yes, at first I didn't checked the schemo, just replaced all the caps at the board. Now, when I saw the sch. I figure out the "missing" caps. I will check  ASAP with scope, but adding two caps from 6 and 9 to GND cant make it worse… maybe I can go with 15nf ?
 
I am guessing the micro is providing a weak pull-up current and the encoder alternately shorts one line then the other to ground. The order of that sequence determines the Up/Down direction.

Since this looks like a simple one step at a time up/down encoder, not an absolute value encoder, missing a step actually means it is reading multiple hits or 2 or 3 at a time when you only want it to get one.

Again my suggestion to look at the encoder lines with a scope probe.

You said you replaced the encoder, so it "should" be working correctly.

Caps might help but doesn't explain why it worked perfectly for years without them.

JR
 
Thanks John. Unfortunately, I'm dumb for this kind of electronics but I will check  the signal on the oscilloscope as you suggest.
You said you replaced the encoder, so it "should" be working correctly.

Caps might help but doesn't explain why it worked perfectly for years without them.

Yes, strange indeed  :(
 
I tried to measure with scope but it's really hard without digital scope. Anyway, I tried to put 15n from both lines to ground and it smoothed the drops. Now it works but sometimes don't react (encoder clicks but micro  don't count it)
 
The clicks are just mechanical so do not affect the micro other than tell you it should have registered.

Before it was getting too many counts,

Now not enough.

Maybe use smaller caps.

JR
 
JohnRoberts said:
The clicks are just mechanical so do not affect the micro other than tell you it should have registered.

Before it was getting too many counts,

Now not enough.

Maybe use smaller caps.

JR
Clear :) Yes, I will experiment a bit. Still don't understand how it's possible that it worked withouth caps smoothly . Maybe I have to change few encoders to find accurate one.  :(
 
Moby said:
JohnRoberts said:
The clicks are just mechanical so do not affect the micro other than tell you it should have registered.

Before it was getting too many counts,

Now not enough.

Maybe use smaller caps.

JR
Clear :) Yes, I will experiment a bit. Still don't understand how it's possible that it worked withouth caps smoothly . Maybe I have to change few encoders to find accurate one.  :(
I thought you replaced the encoder... is it a new one? If not, dirty intermittent switch contacts can cause issues.

While I am mostly speculating about the details, a simple encoder has two switch closures and they close first one then the other based on which direction the shaft is turned.

Logic inside the micro can denounce to prevent multiple counts (jumped levels) by only allowing one step per pair of transitions. A simpler decoder might trigger multiple times if there is noise on the line. Note: Slowing it down with a cap, could have the unintended effect of introducing ground noise to that node depending on where the cap is grounded.

These are simple and robust interfaces. If transition edges look very sharp on the scope (they should) it should work well.

A bad ground connection on the encoder common could cause problems, but that shouldn't work at all.

Some micro's have built in firmware for decoding the encoder output, some require rolling your own code.

JR
 
Code:
I thought you replaced the encoder... is it a new one? If not, dirty intermittent switch contacts can cause issues.
Yes, but now days (China manufacturing) everything is possible. I tried firs one from a bag ;) . I'm not familiar with firmware details, unfortunately MCS who wrote software is not available.  :-[
 

Latest posts

Back
Top