Need schematic review. motherboard for Expat's Eden /8ch Micpre frontend

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

jwhmca

Well-known member
Joined
Jan 28, 2009
Messages
931
Location
USA
Hi Guys,

Hoping for some more sets of eyes on a new schematic design...  It is going to be a motherboard of sorts for Expat's Eden... 8 ch's in one box. Really my first time playing with relays and multiple power rails and subsequent separate grounds.

I'm thinking of daisy chaining the 48vdc & relay/LED power rails between the 8 cards and doing discreet "home runs" for the audio power/ground.

 

Attachments

  • EdenMB - Schematic.pdf
    59.6 KB · Views: 37
I would be a little worried about noise switching tx/caps. For that I would leave connected the caps before RL4 all the time and R18/R19 before RL5, duplicated, one set for the Tx, one for the caps, so the caps are always charged as they should. That load shouldn't be a problem for the mic.

In the other hand, I understand the idea of having the option, but usually a transformerless mic pre is not optimized to work with a transformer 600:3k would make a level difference in the preamp, but I guess you still didn't chose the Tx.

XLR input shouldn't have the ground to the board, just to the chasis as close as the connector as possible. Use that same point for C1  ground, not the ground coming from the supply. If you have extra poles in a switch you could use them paralleled with the other to get better reliability.

Some mics prefer to see more than 2k loads, you would have to tweak the pad to archive that, is not wrong how it's now, depending on the use. This way is good for noise, higher values are good for distortion and freq response, using a transformer less preamp won't suffer from having slightly higher input impedance.

This are all details, in general terms it looks good. Regards.

JS
 
joaquins said:
I would be a little worried about noise switching tx/caps. For that I would leave connected the caps before RL4 all the time and R18/R19 before RL5, duplicated, one set for the Tx, one for the caps, so the caps are always charged as they should. That load shouldn't be a problem for the mic.

Would this be an issue just with the Phantom power turned on?

joaquins said:
In the other hand, I understand the idea of having the option, but usually a transformerless mic pre is not optimized to work with a transformer 600:3k would make a level difference in the preamp, but I guess you still didn't chose the Tx.

The transformer idea is an experiment... I did choose it.

joaquins said:
XLR input shouldn't have the ground to the board, just to the chasis as close as the connector as possible. Use that same point for C1  ground, not the ground coming from the supply. If you have extra poles in a switch you could use them paralleled with the other to get better reliability.

Curious as to why not use the 48vdc supply common for 48vdc ground?

joaquins said:
Some mics prefer to see more than 2k loads, you would have to tweak the pad to archive that, is not wrong how it's now, depending on the use. This way is good for noise, higher values are good for distortion and freq response, using a transformer less preamp won't suffer from having slightly higher input impedance.


This pad circuit was taking from Jensen's application schematics... I'm sure it was for a transformer coupled micpre... what values are more normal for transformerless designs?

Thanks for the help!
 
The phantom power should (or must) be conected to the chasis, since pin 1 of the XLR MUST be connected to the chasis directly and is where the mic takes it's reference. In order for the mic to see clean phantom power it must be referenced at the pin 1 of it's XLR, that is done with R3 and C1, R3 for decoupling from the source and C1 for cleaning it referenced to pin1 from XLR. This could be done in the board, XLR pin1 coming only for the cap, then the 3 resistors connected as they are. The other way (which I use most for my designs) I connect the phantom powering directly to the XLR connector, using the same pin from the cap (or other short wire) to connect to the chasis at the screw holding the connector.

As I decribed, the cap switch shouldn't have problems with phantom, caps are always charged as they suppose to be, 100k may be too high as bleeding resistors, in a 2k mic preamp, using 10k or 20k at that point should be good. But they are only connected to the output when they are needed. Transformer is out completely which is important. When switching the caps already have the positive side connected, so no strange things happening there, and the other side connected to ground through the bleeding resistors, so they are charged just as they need to show 0V at the input of the gain stage.  What I describe is to avoid switching noise when phantom is on, when phantom is off it shouldn't be much of that anyway. It's always preferable to have caps connected to somewhere all the time, not a big deal with leaky electrolytics but a cap may maintain a charge quite some time, and if a charge is builded there for some reason and you then connect the cap you may damage something. The resistors are there to avoid that.

About the attenuator, I would use 2x1k resistor and a 220Ω resistor. This is a bit higher than 150Ω expected but I always assume that if you are using the pad is because the signal from the mic is too high to use it at the minimum gain of the amp, likely to bee too high, then I'm more concerned for the overload of the mic driving low impedance sources than the noise floor for it. Then I go for at least the most common rated impedance of 2k than lower.

JS
 
Back
Top