Langevin AM1A Mixer Direct Outs

[groselicain]

Several years ago I came across one of these units that someone had given direct outs on four of the channels. They have some sort of homebrew buffer on the output of the potentiometer that controls preamp output gain to the master section booster amps. Impedance balanced at the output jack. Fast forward, I purchased a second unit that had no direct outs.

I've often wondered if I could add direct outs to the second unit on my own, and wanted to know if a buffer is necessary—or could I just do as the previous owner had done and get signal from after the potentiometer (the 220k gain controls in the attached block schematic) but work without the buffer.

 

[groselicain]

Here's the schematic for the AM4100 amp card itself.

 

[mjrippe]

I have an AM1A that I started modding but have not completed due to other problems I ran into.  I built clones of the 4700 output amps to drive the direct outs through 1:1 transformers just like the main out.  They are a simple circuit that fits on a little piece of protoboard.  Taking the signal off the pot directly means you are taking it from a high impedance point and it will not be able to drive any sort of load.  The 4700 board give it some oomph and lowers the output impedance.

The main issue I ran into was bleed from the direct out channels into the mix bus.  I had installed switches after the pot to select bus or direct, but switched to direct out I could hear it faintly in the mix.  Probably a grounding or decoupling issue, but I set it aside for other things.

Hope this helps,
Mike

 

[groselicain]

I have an AM1A that I started modding but have not completed due to other problems I ran into.  I built clones of the 4700 output amps to drive the direct outs through 1:1 transformers just like the main out.  They are a simple circuit that fits on a little piece of protoboard.  Taking the signal off the pot directly means you are taking it from a high impedance point and it will not be able to drive any sort of load.  The 4700 board give it some oomph and lowers the output impedance.

The main issue I ran into was bleed from the direct out channels into the mix bus.  I had installed switches after the pot to select bus or direct, but switched to direct out I could hear it faintly in the mix.  Probably a grounding or decoupling issue, but I set it aside for other things.

Hope this helps,
Mike

Thanks, that does help! I may try and copy the buffer the original owner created, otherwise I'll look into cloning the 4700s as you have. If you ever come up with any other tips or trick please feel free to share. I've always really liked these amps. Have you experimented with changing the value strapped between pins 4 and 5 for input gain on the 4100s? I've wondered if there's any disadvantage to using a pot there, or if it's better to use resistors on a switch. With the amps set as is, they're great for a lot except for certain high transient sources like close miced drums, and I'd love to remove the guesswork of sessions where I may or may not have to run out and add an inline pad.

 

[mjrippe]

It may be a while before I revisit them, but I'll post here when I do.  I also have a box of 4100B cards that I may rack in sets of four with 4700 clones and xfmrs.  Someday...

Mike

 

[groselicain]

Mike,

      I've been working on this and decided since the 4700s will get me down to Low-Z, I'm going to go with your advice.  I've been building a Mouser cart for components and came across one issue. The 100uF capacitor at the output of pin 1 doesn't appear as polarized in the schematic I have, but I've seen recapped 4700s with a polar electrolytic there. Any reason in your mind to go one way or the other? Thanks!

 

[mjrippe]

My schematic (attached) shows polarization.

 

[groselicain]

Thanks Mike!

 

[groselicain]

Mike, did you make a new PSU or use the original Langevin PS 4900? I'm looking at the data sheet here for both the 4900 and the 4700s and wondering if I'm going to be drawing too much current when I add 6 4700s for direct outs. Literature states a maximum output current of .25A for the 4900 and that the 4700s draw 40mA. Thanks for any thoughts you've got!

 

[mjrippe]

I only built two 4700s and was using the Langevin psu (recapped of course).  You could very easily drop in a Meanwell 48v switching supply to replace it.  They can be adjusted to 45v, which is also within spec for phantom power.  The RS-25-48 gives you 0.57A and costs about $10.  You might want to add some filtering to the output, see other posts around here.

 

[groselicain]

So the scope of this project has changed a bit—I’m hoping to get some clarity, though, to see if I’m being unrealistic.

I’ve thought about building a full fledged mixer out of what I’ve gathered, but the 4100 cards would never work as is for accepting line level. Would the 4700s accept line level? My plan has been to build a switch in each channel to bypass the 4100 mic amp and 4700 direct out, and send signal directly to the mix buss—4700 cards. This means taking a balanced, line level signal and dropping pin 3, directly into a 4700 without any impedance matching.

I’m just so green at this sort of thing, you’ll have to forgive me for being gun-shy about level and impedance. Perhaps the 4000 series just isn’t a good format for this sort of thing, so I really need someone like Mike to let me know if I’m down a rabbit hole—I’ve just had a lot of great experiences tracking through the 4100s over the years and now that I’ve got 15 4700s at my disposal I’m thinking of any and all possibilities to try.

 

[gridcurrent]

I’ve thought about building a full fledged mixer out of what I’ve gathered, but the 4100 cards would never work as is for accepting line level. Would the 4700s accept line level?

fairly sure the Beyer input transformer normally fitted to the 4100 is a 1:15 ratio.
the voltage gain specified for the similar but transformer-less model 4300 is 20 dB.
you could reverse the transformer to achieve a floating/balanced line input.
a high ratio mic input transformer backwards can normally take significant level,
and if not, easy to add a resistive pad.

 

[groselicain]

fairly sure the Beyer input transformer normally fitted to the 4100 is a 1:15 ratio.
the voltage gain specified for the similar but transformer-less model 4300 is 20 dB.
you could reverse the transformer to achieve a floating/balanced line input.
a high ratio mic input transformer backwards can normally take significant level,
and if not, easy to add a resistive pad.

So, cut the traces on the 4100 from the secondaries, and wire a 4PDT switch that sends either
        A. Input to transformer primaries and out secondaries, or
        B. Input to transformer secondaries  and out primaries.

Am I following? Thanks gridcurrent.

 

[groselicain]

Well, here’s something worth mentioning that maybe Mike can chime in on.

I had PCBs made from the 4700 cards I have. Only later—once they’re in my hand and I’m populating—do I realize that they’re 4700Bs, not the 4700 in the published schematics. In fact, I don’t own anything but 4700Bs. Now granted, I know they both do the same job—but my Mouser order was taken from the 4700 schematic. I wonder if there are any significant tradeoffs by using the 4700B. I can’t find any documentation on them anywhere. I plan to wire up a 4100 and 4700B this weekend, so I’ll report back with results.

 

[mjrippe]

I am not sure what you are concerned about.  The difference between a 4700 and the "B" version may be as simple as a different layout or a solder mask revision.  If the components match up there will be no difference in performance.