MOX builder’s thread

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
some notes on values

...sounds like moral philosophy ;)

The compensation caps on the feedback resistors could use some different values IMO unless I am missing some design consideration.

The input FR blocker LP filter is the R003/C002 combo of 100R and 100 pF by manual. The cutoff (RF filter) is thus about 15 MHz. Consider using a lower LP value, between 150kHz and 300 kHz, say for instance by changing R003 to 5k, and R004 or 47k or 50k.

The inverse situation is found on the LP and HP actual filter sections, for the compensation cap around the feedback R to prevent out of band amplification. C200 and C500 parallel to R200 and R500 22k resistor gives an out-of-band LP of about 80 kHz. This could be set a tad higher, to prevent the cumulative LP filters from starting to affect the audio band, say by making C200 and C500 68pF or 39 pF (LP at 106 kHz or 186 kHz, respectively).

Markus
 
The inverse situation is found on the LP and HP actual filter sections, for the compensation cap around the feedback R to prevent out of band amplification. C200 and C500 parallel to R200 and R500 22k resistor gives an out-of-band LP of about 80 kHz. This could be set a tad higher, to prevent the cumulative LP filters from starting to affect the audio band, say by making C200 and C500 68pF or 39 pF (LP at 106 kHz or 186 kHz, respectively).

In MOAMPS original work up the value was 47pF.

BDP
 
Markus,

While I was testing the prototype I never used the feedback capacitors. This does not mean that they are not needed, but I have not had time to verify the size needed in different configurations.

If anybody would like to simulate or test in real life and post results it would be great.

\Jens
 
Jens,

I'll report as I go along, but my testing is quite limited. I can't test for much more than audio frequency responses. Usually I use best guesses from what I see published. I still have some parts missing but otherwise my 8 boards are 80 % finished, maybe in a week or two I'll have first results. That was quite a marathon btw in soldering.

For input RF filters and for compensation caps, usually I see 150-250 kHz LP's applied. The reasoning behind that is that it should be about 10 times audio range to ensure that there is no phase shift in the audio band even when several of these filters accumulate in a chain, as they usually do.

As you said these caps may not even be necessary. They're just a safeguard for these RFI infested times. It may make a difference for anybody using high speed op amps in a dirty environment. As for instance, having a cell phone transmitter nearby.

Same range applies to decoupling caps at the op amp pins. I use X7R ceramics at 10 nF to get the maximum HF performance. This is the value recommended for high speed op amps such as OPA2134 (as of data sheet). For NE5532 for example, I see anything from 1 uF to 10 uF electrolytics recommended... and not even necessarily at the pins but every 2 inches or so. So, your 100 nF is a good median value, all else depends largely on what op amps ppl will use. High speed types will be most picky, and to be honest, proper decoupling was the single greatest effect I saw in my contraptions from parts alone over the past 4, 5 years of DIYing. For instance I use garden variety resistors because I don't expect major effects here. But that's just me - I will just assume the risk of using my best guesses and I can't guarantee it works, for me or anyone else ;) As for caps here I use PP for the smaller caps and PE for the large ones, but in the past I have used dipped PE's that sounded just sweet. So we'll see. I *do* hope for an improvement over my veroboard messes...

Another thing that struck me, the standoff holes will connect a metal standoff to the ground plane. For ppl like me who will connect 4 boards per channel, it may be a good idea to use plastic standoffs or else get multiple ground loops through the chassis. It may or may not matter but better be safe than sorry... I'll use plastic clip-in types.

Markus
 
Hi Marcus,

I added connection to the ground plane because it’s easier to get rid of than to make later. It is, as you say, a question of using plastic stand offs, instead of metal ones. I have just received a bag of both plastic and metal ones, so it should be possible to find the best solution. I think that I’ll use three plastic and one metal Standoff per board.

I added the ground connection because of the “shielding” effect. Many users (my self included) will use the boards without enclosure and this gives a risk of picking up noise. The shielding may help against this.

\Jens
 
For input RF filters and for compensation caps, usually I see 150-250 kHz LP's applied. The reasoning behind that is that it should be about 10 times audio range to ensure that there is no phase shift in the audio band even when several of these filters accumulate in a chain, as they usually do.
Hi Markus,
You are right.
Also, compensation caps in this design are needed only on input (and eventually output) buffer opamps. IMO

I'll use plastic clip-in types.
This is very good idea. (strongly recommended)
(or cutting cupper foil around holes)

Regards
Milan
 
I'm almost done assembling 10 mox boards. I will be configuring a 4th order capable 3 way xover and a subwoofer xover. Fun, fun, fun.
I'm wondering if someone knows what opamps would be the best sounding for all around use, or are there better opamps for bass, midrange and treble?

Has anyone heard of Sound Odyssey? They are making noises suggesting that the AD8066 is the cat's meow. Expensive little guys, and they would have to be adapted to fit. Maybe just for the midrange???...

Thank you guys for everything, for making this project possible. It is way beyond cool. I'm finally going to build the dream speakers, (and procure a six channel amp.)
 
choosing op-amps

Marsupialx:
For some theory behind choosing op-amps you may try Analog Design Note ADN003, found here: http://ww1.microchip.com/downloads/en/DeviceDoc/adn003.pdf

The website for Microchip Technology at www.microchip.com contains a development tool called "FilterLab" found here:

http://www.microchip.com/stellent/idcplg?IdcService=SS_GET_PAGE&nodeId=1406&dDocName=en010007

for download that may provide some interesting views of what happens to signal with various combinations of R & C.

You may also do a search for "Sallen-Key filter" and find a wealth of info.
 
If powering the mox via batteries, is there reason for concern if the power is slightly different between V+ and V- ?

Any guesstimates on how long a mox board will operate on 4) NiMH 9v cells, two in series for each + and - ? I'll probably be using opa2134's, at least at first.

Eventually I'll build a power supply

Thanks. Sorry if the second question is nebulous. I'm looking for ranges, 30 min. or 4 hours, that kind of thing.
 
SubMox.

Both sides identical.
A few parts left to install.
 

Attachments

  • mox1diy.jpg
    mox1diy.jpg
    71.3 KB · Views: 289
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.