Yamaha M512 Mods

Help Support GroupDIY:

collinTHEbrewer

Active member
Joined
Jul 29, 2013
Messages
35
Here goes another Yamaha mixer thread...

I got a pretty clean Yamaha M512 from a local shop, and I thought I'd consider using it as a vibey preamp option. 12-channels of cheap-ish input transformers sounded like a fun deal to me especially for drums. I had an old RAMSA mixer years ago that I used for a similar thing, but the size was a little ridiculous. This is a lot more compact and could easily be put away when not in use.

So here's what I'm thinking:
-Add balanced outputs
-Swap Preamp TA7322P opamps (IC1, IC2)
-Recap
-Tweak EQ frequencies

Balanced Outputs
I think I've got two main options here. Either way, I would be diverting the pre-fader FB send to a switching TRS jack. The question is, to transform or not to transform? I can either do some psuedo impedance balancing using a resistor and capacitor as described by Jensen (https://www.jhbrandt.net/wp-content/uploads/2014/11/Interconnection_of_Balanced_and-Unbalanced-Equipment.pdf), or I could add in some OEP A262A2E transformers. One is obviously way cheaper than the other, but I did find a good deal on a dozen OEP transformers.

Opamp Swap
The TA7322P is a real pain. For starters, it's a single opamp in a 9-pin SIP package that runs on a +-25VDC supply.

Problem #1 - SIP Package
There are some solutions/adapters that will fit the SIP pinout and adapt to something like a OPA604 or similar. The most common solutions are fairly pricey (~$4 per adapter x 2 x 12 = $$$).

Problem #2 - Supply Voltage
Most modern opamps won't tolerate more that +-18VDC. I was thinking I could throw in a pair of 7.5V zener diodes to each opamp to lower it down closer to 18V.

Problem #3 - Inventory
I have about 20 LME49720 dual opamps laying around, and I'm thinking that sounds cheaper than 24 new single opamps. However, I would need to figure out an adapter solutions that could potentially utilize both channels of the LME49720.

Recap
I only count about half a dozen electrolytic caps touching the the signal path between the input and fader. Any reason to replace any others?

EQ Frequencies
Without getting to use this in a session yet, I have no idea what to expect from the EQs, but the word on the web is that they could use some love. (https://groupdiy.com/index.php?topic=51214.0)

Anyone have any thoughts or see any red flags?
 

Attachments

  • M512 Balanced Output.png
    M512 Balanced Output.png
    173.2 KB · Views: 88

collinTHEbrewer

Active member
Joined
Jul 29, 2013
Messages
35
I've got some parts in, so I'm prepping to install the TRS jacks along with the OEP output transformers. I got them for $6 each, so I think it's probably worth the mounting headache verses the RC impedance balancing.

I've got to work out if there's an easy way to still use the switching jacks along with the transformer. I don't really care to preserve the FB output, but maybe it's a good thing to attempt.
 

collinTHEbrewer

Active member
Joined
Jul 29, 2013
Messages
35
Does this make sense for connecting the output transformer? If I grab the signal directly from the opamp output while maintaining the existing connections, am I causing any issues?
 

Attachments

  • M512 Xformer Balanced Output.png
    M512 Xformer Balanced Output.png
    41.5 KB · Views: 49

dogears

Well-known member
Joined
Nov 15, 2017
Messages
477
Location
Texas
I made a version of our M2 discrete op amp in that SIP footprint / pinout for a client. If you want to try some out let me know.
 

Whoops

Well-known member
Joined
May 9, 2008
Messages
5,925
Location
Portugal
collinTHEbrewer said:
So here's what I'm thinking:
-Add balanced outputs
-Swap Preamp TA7322P opamps (IC1, IC2)
-Recap
-Tweak EQ frequencies

For the Balanced outputs why not just use That 1646 opamps?
They sound really good and are easy to implement
OEP transformers are a bit on the lo-fi side, so they will have an impact on the sound that you might enjoy or not, they will be more expensive also


Swap Preamp TA7322P opamps (IC1, IC2)

Whats the problem with the TA7322P? why do you want to replace it?


As for the Recap, I think you should do it. Use Panasonic or Nichicon caps and you are sorted

As for the EQ change, that's a matter of personal taste, but you should use it first and see if you really need to do it, maybe the freq are fine as is


 

abbey road d enfer

Well-known member
Moderator
Joined
Jan 22, 2008
Messages
13,456
Location
Marcelland
collinTHEbrewer said:
Does this make sense for connecting the output transformer? If I grab the signal directly from the opamp output while maintaining the existing connections, am I causing any issues?
Yes. Whatever output offset will magnetize the xfmr and cause distortion on both the pre and post-xfmr signal. You should insert a large capacitor (>470uF) in parallels with a 47r resistor.
Beware that the opamp may struggle to drive the xfmr at low frequencies. You may have to ensure no infrasonics go into the signal.
 

abbey road d enfer

Well-known member
Moderator
Joined
Jan 22, 2008
Messages
13,456
Location
Marcelland
collinTHEbrewer said:
Recap
I only count about half a dozen electrolytic caps touching the the signal path between the input and fader. Any reason to replace any others?
Everything is in the signal path, except the pilot light.
Decoupling (by-pass) capacitors deserve the same attention than coupling caps.
 

Newmarket

Well-known member
Joined
Oct 10, 2016
Messages
830
Location
Brighton Sussex UK
abbey road d enfer said:
Yes. Whatever output offset will magnetize the xfmr and cause distortion on both the pre and post-xfmr signal. You should insert a large capacitor (>470uF) in parallels with a 47r resistor.
Beware that the opamp may struggle to drive the xfmr at low frequencies. You may have to ensure no infrasonics go into the signal.

Agreed (esp wrt the op amp drive capability) except that I'm struggling with why the 47R resistor in parallel with the capacitor ?
(I'll readily admit that I do still get 'transformer confusion' when it gets down to the detail.)
 

abbey road d enfer

Well-known member
Moderator
Joined
Jan 22, 2008
Messages
13,456
Location
Marcelland
Newmarket said:
Agreed (esp wrt the op amp drive capability) except that I'm struggling with why the 47R resistor in parallel with the capacitor ?
(I'll readily admit that I do still get 'transformer confusion' when it gets down to the detail.)
The combination of the coupling cap and the primary inductance make a resonant circuit, which results in a subsonic hump. Adding this resistor damps the resonance.
Actually, for best performance, this resistor should be adjusted to the xfmr's characteristics. 47r is a ballpark figure.
 

Newmarket

Well-known member
Joined
Oct 10, 2016
Messages
830
Location
Brighton Sussex UK
abbey road d enfer said:
The combination of the coupling cap and the primary inductance make a resonant circuit, which results in a subsonic hump. Adding this resistor damps the resonance.
Actually, for best performance, this resistor should be adjusted to the xfmr's characteristics. 47r is a ballpark figure.

Thanks. Yes - I can see there is an 'LC thing' going on without a resistance in there. But intuitively I'm having trouble with having the resistor in parallel with the cap' and forming a dc path ?
 

abbey road d enfer

Well-known member
Moderator
Joined
Jan 22, 2008
Messages
13,456
Location
Marcelland
Newmarket said:
Thanks. Yes - I can see there is an 'LC thing' going on without a resistance in there. But intuitively I'm having trouble with having the resistor in parallel with the cap' and forming a dc path ?
The only alternative for damping te resonance is to put the resistor in parallels with the primary, which is bound to overload the driver stage.
There is a indeed a possible DC component, but it is generally perfectly admissible.
 

Newmarket

Well-known member
Joined
Oct 10, 2016
Messages
830
Location
Brighton Sussex UK
abbey road d enfer said:
The only alternative for damping te resonance is to put the resistor in parallels with the primary, which is bound to overload the driver stage.
There is a indeed a possible DC component, but it is generally perfectly admissible.

Thanks. Helpful as always.
I'll try and work out a transformer sim for LTSpice or Simetrix when I get the chance.
 

collinTHEbrewer

Active member
Joined
Jul 29, 2013
Messages
35
Man this thread really lit up when I wasn't looking. I've been on a hiatus from my project, but I should be restarting it soon.

I got some DOAs courtesy of dogears from Iron Age Audioworks. He's got a DOA that's compatible with the 9-pin SIP package. The initial swaps have been promising. THD dropped overall, but when driven, the THD looks more evenly distributed. I'll have to post some of my captures if I still have them.

PXL_20210509_235746070.jpg

For the Balanced outputs why not just use That 1646 opamps?
They sound really good and are easy to implement
OEP transformers are a bit on the lo-fi side, so they will have an impact on the sound that you might enjoy or not, they will be more expensive also

Man, I wish I would have seen this before I ordered the OEP transformers... I found a killer deal on the 'bay that really made it a cost effective solution. The OEPs are nice addition to the sound I think. When I ran some mono drum tracks through it, you could feel the transformer in the circuit. Definitely not HiFi, but that's not really what I want from this mixer. I'm looking for a little bit of color while retaining the transients required for drums which I think will be my primary use.

Plus, transformers on input and (now) output makes the marketing sheet look a lot better. :p

Everything is in the signal path, except the pilot light.
Decoupling (by-pass) capacitors deserve the same attention than coupling caps.

Fair point. I was mostly excluding the caps on the +/-25VDC rails to the op-amps.

Yes. Whatever output offset will magnetize the xfmr and cause distortion on both the pre and post-xfmr signal. You should insert a large capacitor (>470uF) in parallels with a 47r resistor.
Beware that the opamp may struggle to drive the xfmr at low frequencies. You may have to ensure no infrasonics go into the signal.

Is the recommendation to add the capacitor/resistor in series with the primary?
 
Top