Adding Direct outs on Yamaha m1516

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

bmaughan

Active member
Joined
Jul 15, 2011
Messages
39
Hi there, I apologize in advance for my ignorance and silly questions.. I’m very much new at this haha. But I’m attempting to add direct outs on my Yamaha m1516 and had some questions for anyone willing to help.

I was planning on wiring in a CA2623 transformer right off of the echo pre/post switch. Many questions

1. Is this the right transformer to use? The spec sheet says 75ohm primary. Seems low compared to the 600:600 I’m used to seeing but I’m trying to wrap my head around transformers not having inherent impedance rather they reflect impedance.

2. Will having the transformer on the channels output opamp at all times negatively affect the signal down other paths? For instance if I try to rout to another bus simultaneously. Or come mixdown time when I will only be using stereo bus outs.

3. It was recommended to me to use a dc blocking output cap. Necessary?
What size would be recommended? 470uf? What voltage rating?

4. Would it be best to wire the xfrmr 1:1 or 1:2? Maybe it’s subjective but I didn’t know if this would affect opamp loading.

Sorry for the barrage of questions, I’m just trying to get this thing done and quickly realized I was a little over my head haha. Can’t thank you all enough!
 

Attachments

  • M1516 input .pdf
    648.9 KB · Views: 2
  • M1516 input layout.pdf
    1.9 MB · Views: 1
Hi there, I apologize in advance for my ignorance and silly questions.. I’m very much new at this haha. But I’m attempting to add direct outs on my Yamaha m1516 and had some questions for anyone willing to help.

I was planning on wiring in a CA2623 transformer right off of the echo pre/post switch. Many questions

1. Is this the right transformer to use? The spec sheet says 75ohm primary. Seems low compared to the 600:600 I’m used to seeing but I’m trying to wrap my head around transformers not having inherent impedance rather they reflect impedance.
The transformer is correct for a line output. It's inductance is a little low so ideally it should be driven from a low impedance. But a typical line load is 10K so the reflected impedance should make driving it with a higher impedance ok. The thing to watch out for would be peaking or loss of bandwidth. You could pretty easily test this by just connecting it to the EQ out jack and looking at the spectrum with different loads like 10K and 600.
2. Will having the transformer on the channels output opamp at all times negatively affect the signal down other paths? For instance if I try to rout to another bus simultaneously. Or come mixdown time when I will only be using stereo bus outs.
It might. Depending on the echo switch pos, you could see a shift in EQ frequencies, loss in bandwidth, levels being off, etc. Again, ideally that transformer should be driven from a low impedance. But if you have a good way to see a spectrum and test, you could just try it and see. If it works, it works.
3. It was recommended to me to use a dc blocking output cap. Necessary?
What size would be recommended? 470uf? What voltage rating?
I would say probably no since both the EQ amp and the fader amp already have blocking caps. The exception is the patch input. If something with a DC offset plugged into that, that might not be good. But that would be a strange occurrence.
4. Would it be best to wire the xfrmr 1:1 or 1:2? Maybe it’s subjective but I didn’t know if this would affect opamp loading.
Neither. It seems the CA2623 has split secondaries. So wire the secondaries in parallel and get 2:1 step down. That will reduce the chance of negative effects on connected circuits.

Actually another possibly better way would be to wire it in reverse. Use both secondaries in series to get a 2:1 step down but with double the primary characteristic impedance.

The step down will attenuate 6dB (12dB compared to 1:2 step up) but unless you're trying to drive some pro gear to a high level, you don't need it. If you're connecting to a digital audio recording device, you definitely don't need it. Most modern gear just doesn't need +24dB levels. The amps in that mixer probably can't drive anything to that level anyway.

Another transformer that is not super expensive but has higher inductance and therefore less likely to cause negative effects connected directly to existing amps would be the Jensen JT-123. But wired 2:1 I think the CA2623 would work.

You will need to get spectrums to verify that it actually works correctly however. To do that you need a USB audio interface (preferrably one that goes over 20KHz by a bit) and some FFT software or at least maybe something in a DAW that shows you the spectrum / fft / bode plot / frequency response / whatev. Take baseline measurements of all the channels and save screenshots of their performance, EQ frequencies and so on. Then try adding a CA2623 to an EQ out with a 600 ohm dummy load and measure again at that load and compare screenshots. Look for incorrect changes in level, loss of highs / lows, shifting in EQ frequency, etc.
 
Interesting about wiring the xfrmr backwards. Didn’t realize it was that simple.
I would like to drive the inputs of a mci jh-110 1/2” 4 trk machine. Probably not ideal I would assume?

So when you say, “if it works,it works” does that mean if it’s not negatively affecting the channel when connected, it shouldn’t cause problems when other buses are routed etc?

I guess what I’m trying to figure out is (if it’s working correctly obviously)
will the opamp not “see” the transformer if I have nothing plugged into the direct out? Or does the mere presence of the transformer in the circuit present a load at all times? (In which case a switch would be the only option I would think..?..)

In regards to getting measurements..what is the proper way to create a dummy load?

Can’t thank you enough for you help!!!
 
Also, just for more detail. I attached the data sheet for the TA7322p which is the opamp driving the signal. I have some boards made up to swap it out with a OPA604.
I’m regards to this scenario..Better,worse?

I’m also making up some adaptor boards to fit the 2520 opamps into the sip format. Would any particular discrete opamp be able to handle these Xfrmrs no problem?
 
Interesting about wiring the xfrmr backwards. Didn’t realize it was that simple.
I would like to drive the inputs of a mci jh-110 1/2” 4 trk machine. Probably not ideal I would assume?
No idea. What are the impedance of the inputs? Are they balanced?

If you know that you will always run into the tape machine and the tape machine inputs are balanced, then I would not add transformer outs to your mixer at all. I would make impedance balanced outputs like this:

1660172560851.png

Impedance balanced outputs provide good ground noise isolation and are trivial to implement.
So when you say, “if it works,it works” does that mean if it’s not negatively affecting the channel when connected, it shouldn’t cause problems when other buses are routed etc?
If you measure things as described previously and there are no negative effects, then it works.
I guess what I’m trying to figure out is (if it’s working correctly obviously)
will the opamp not “see” the transformer if I have nothing plugged into the direct out? Or does the mere presence of the transformer in the circuit present a load at all times? (In which case a switch would be the only option I would think..?..)
You don't need to worry about disconnecting the echo 1 / 2 pot loads and inserting the transformer with no load on the secondary. You only need to worry about loading it too much (with an impedance that is too low) or resonance created by the transformer which should not be a problem but needs to be confirmed by said measurements.
In regards to getting measurements..what is the proper way to create a dummy load?
A resistor on a breadboard or with some jumper alligator clips would do. But the cable from the load to the USB interface should be shielded. And preferably the USB interface should have a balanced input so that you can use pin 2 for signal and pin 3 for 0V at the load. This will minimize ground noise which is a huge problem when trying to connect outboard gear to USB things and PCs.
 
Thanks for getting back to me. I haven’t done the spectrum analyzing tests yet. Plan on doing it tonight.

Yes the 4trk is balanced. Both the 4 trk and my interface are 10k balanced.
I probably should have just gone impedance balanced or used balanced line amps in hindsight but I can’t return the transformers at this point so I’d like to try and use them.

So I hooked it up to my interface wired 1:2 just to see and I’ve got a few things going on. (I’m going to experiment with the wiring you suggested here shortly)

1. wired 1:2 and I notice I have about 12db less level coming out of the direct out vs. the pgm bus out.
I thought 1:2 would give me 6db more than the other buses. What am I missing?

2. I notice a parasitic loss when monitoring the pgm bus when switching the direct out transformer into the channel. I lose a db or so. And another when I switch the direct out send to pre-fader.

3. I’m getting some noise/hum. Pretty low level, but a good bit more than I have on the pgm bus outs.

I’ll post again once I do the tests you said I should try. If anything I just listed is a red flag please let me know.

Thanks again!
 
I should add one more weird thing. I’m getting a lot of microphonics in the cable that’s connected to the direct output. Not sure if that’s a clue as to what might be going on.
 
1. wired 1:2 and I notice I have about 12db less level coming out of the direct out vs. the pgm bus out.
I thought 1:2 would give me 6db more than the other buses. What am I missing?
Sounds like those amps cant drive the transformers. Try 2:1 with the secondaries in series. Should be only 6dB loss. Confirm perf. with measurements.

Cheap cables with foil shielding can make crackling noises when moved. A proper installation would use snake cable with a tight stranded shield like this:

Redco Shielded Snake Cable
 
So I did a few sweeps (using REW so pardon the "SPL" graph) on a channel with the transformer on a switch so I could test with it in/out.
A lot of stuff going on..not sure I understand. Seems like some serious resonance going on. I'm way over my head at this point. let me know what you make of this.
I attached 6 screenshots of the notable tests.


1. channel going out the PGM bus out with no transformer connected in any way.
2. channel through the PGM bus out with the transformer connected but no load.
3. channel through the PGM bus with the transformer connected to interface 10k input.
4. Is the direct out through the transformer into the 10k interface input.
5. PGM bus with transformer on channel wired 2:1 w/ no load
6. PGM bus with transformer on channel wired 2:1 to 10k interface input


Obviously the transformer is causing some high passing in the other busses when engaged and the direct out signal is high passed as well.
To add more confusion..I tried wiring it 2:1 and things were better on the PGM but I couldn't get signal from the direct out. ???

Eq points/gain seems fine when tested.

One additional note. When the transformer out was set to prefader, while not fixed.. things improved slightly across the board.

I know this is a lot.. sorry to bombard with so many questions/issues...but are there any fixes I can try before ditching the xfrmrs all together ?
I'd hate to waste all this time/money :\

thanks again!!
 

Attachments

  • 1. PGM bus no transformer .png
    1. PGM bus no transformer .png
    1 MB · Views: 0
  • 2. PGM bus transformer in but no load.png
    2. PGM bus transformer in but no load.png
    1.1 MB · Views: 0
  • 3. channel through PGM bus transformer in w: load.png
    3. channel through PGM bus transformer in w: load.png
    1.1 MB · Views: 0
  • 4. channel through direct out.png
    4. channel through direct out.png
    1.1 MB · Views: 0
  • 5. PGM bus xfrmr 2-1 w: no load.png
    5. PGM bus xfrmr 2-1 w: no load.png
    1.1 MB · Views: 0
  • 6. PGM bus xfrmr 2-1 w: load.png
    6. PGM bus xfrmr 2-1 w: load.png
    1 MB · Views: 0
1. channel going out the PGM bus out with no transformer connected in any way.
Ok. So nominal level of PGM bus out is 99dB in the graph. So I'll describe all levels relative to that. Meaning 99dB -> 0dB ref.

This assumes you didn't change any levels on the mixer, interface or in REW. If that's not the case, my comments will be much less meaningful.
2. channel through the PGM bus out with the transformer connected but no load.
+11dB resonance at 10Hz. If that signal is isolated by an amp, it's probably fine. If you're going to use the direct outs and not the PGM bus, then you could just put the faders down and no signal will even make it to unloaded transformers.

What do you mean by "going out the PGM bus out". Your schematic only shows the inputs. Is there an amp before the transformer? Where exactly on the schematic did you attach this? For the moment, I will assume you mean the regular mixer output which has an amp.
3. channel through the PGM bus with the transformer connected to interface 10k input.
Looks good. 0.5dB attenuation with a +2dB resonance at 10Hz. However, if you went through the transformer 2:1, I would expect a 6dB atten. Maybe you tried 1:2 and it coincidentally lost 6.5dB? Explain how the transformer is wired exactly.
4. Is the direct out through the transformer into the 10k interface input.
Looks good. 6dB attenuation with a +2dB resonance at 10Hz. I guess this is with the transformer using two windings in series and one as secondary for 2:1. That would explain 6dB atten. Explain how transformer was wired.
5. PGM bus with transformer on channel wired 2:1 w/ no load
+2dB boost with +15dB resonance at 7Hz. Again, just turn the faders down. But strange that 2:1 created boost and not atten. Are you should you're not changing levels between measurements?
6. PGM bus with transformer on channel wired 2:1 to 10k interface input
Looks good except again +2dB instead of expected 6dB atten is strange.
Obviously the transformer is causing some high passing in the other busses when engaged and the direct out signal is high passed as well.
Where do you see that? If you're talking about roll off down at 10Hz, that's super low. Personally, I would engage low cut way above that.
 
So I think I realized a mistake/miscommunication on my part. The first sweep was a signal going through a channels line input and going out through the pgm bus +4 out. Not simply the channels unbalanced insert out. I should probably redo the tests with that as the baseline.. my bad.
So levels relative to the first test are skewed I would assume. Hopefully that explains why it seemed that the levels were all over the place. When I said a test was “PGM out” I was speaking of the signal from master “PGM” bus out with its additional amplifiers.

- I included the PGM out schematic-

That said the tests still show some questionable things to me. I plan on using the pgm outs for stereo mixdown as well as some summing while tracking so it’s important for the signal to the PGMs to be unaffected by the direct outs if at all possible.
So as it is now, having the transformer always wired in on each channel but potentially not connected to anything, doesn't seem like it’s going to work.

Is there anything that can be done to mitigate the channels DO transformers effect on the PGM bus sends? Reduce the resonance?


Looks good. 0.5dB attenuation with a +2dB resonance at 10Hz. However, if you went through the transformer 2:1, I would expect a 6dB atten. Maybe you tried 1:2 and it coincidentally lost 6.5dB? Explain how the transformer is wired exactly.
I had the transformer wired 1:2.
Maybe this is obvious, but just to be clear.. this sweep was only measuring the PGM out, not the channels direct out. I wanted to see how the DO transformer was affecting my bus outputs. So as you can see it attenuated and created a 10hz resonance. Is this normal for it to affect the signal being sent to the buses like this?

Looks good. 6dB attenuation with a +2dB resonance at 10Hz. I guess this is with the transformer using two windings in series and one as secondary for 2:1. That would explain 6dB atten. Explain how transformer was wired.
So this sweep was the signal from the channels direct output with the xfrmr wired 1:2. 6db attenuated from the PGM bus out. Does that calculate out correctly?
Where do you see that? If you're talking about roll off down at 10Hz, that's super low. Personally, I would engage low cut way above that.
Yeah I’m talking about the roll off. If it was just in the direct outs I wouldn’t care because who needs 10hz haha. But because that same high passing effect is showing up in the PGM outs it makes me feel that just having the direct outs in circuit will negatively affect the sonics come mixdown time.

One additional question: if I were to put the transformer on a switch. Should both the signal and common be switched on/off or just the signal.
Wasn’t sure if having the extra shielded cable connected could affect capacitance or something.

I’m sure I’m missing some info/details but hopefully that clears a few things up. Sorry for the mistake!

Thank you again for all the help. I obviously need it. Please keep it coming if you can! :)
 

Attachments

  • M1516 pgm .pdf
    572.7 KB · Views: 0
Sorry. I just can't follow what you're doing well enough.
no worries, I really appreciate all the help!
one last question if you don't mind. I took a few screenshots of the distortion measurements.
1. distortion on a factory bus output
2. distortion on my recently added direct output
3. distortion on the factory bus output while the direct output is also being driven.

does anything look suspect? It seems to sound ok from what can tell.
Distortion is obviously higher on the direct output than the factory bus output...
but does it look to be within acceptable limits?

your help has been invaluable...thanks a ton!
 

Attachments

  • factory bus output.png
    factory bus output.png
    1.2 MB · Views: 0
  • recently added direct output.png
    recently added direct output.png
    1.2 MB · Views: 0
  • factory bus output while the direct output is also being driven.png
    factory bus output while the direct output is also being driven.png
    1.2 MB · Views: 0
Back
Top