DIY Passive Line Mixer?

  • Thread starter Thread starter FALKEN
  • Start date Start date
FALKEN said:
thanks!! I got a box for it. in your ps schematic there is a 120v 1A fuse. all i could find is 250V 1A. is that ok or am I screwed? also, is it ok to omit the switch, or put it on the +/- 15V side ?

Yes I think it's the current that makes a fuse pop, so that should be OK. However I wouldn't put the switch on the secondary, something tells me that's a bad thing, although I am pretty weak on transformer theory.
 
FALKEN said:
thanks!! I got a box for it. in your ps schematic there is a 120v 1A fuse. all i could find is 250V 1A. is that ok or am I screwed? also, is it ok to omit the switch, or put it on the +/- 15V side ?

The current rating is what truly matters. for 120V use a 125V 1A and a 250V 1A fuse serves the same purpose.

Generally power switches are put between the primary transformer coils and the mains, so when you open the switch the power transformer drops its magnetic field and as the capacitors on the secondary side drain from the load, the entire power supply becomes unpowered.

*I* prefer to use a DPDT switch and cut both the hot and neutral lines, so the transformer isn't attached to either leg of the outlet - just in case it's plugged in when I take the cover off and poke around. Even though plugs and outlets are keyed, once in a while you may have the misfortune of plugging it into an outlet that's miswired, so when the SPST switch breaks the hot lead, the white lead, i.e. neutral, might actually be hot. Hit that and the metal case with a screwdriver...

The last +/-15V power supply I made also had a 120V coiled DPDT relay in parallel with the primary transformer coil. When the power switch was on, the relay would energize too, and using the normally closed contacts, disconnect a 32V 1A light bulb that was inside the case. When I powered off the power supply, the relay immediately de-energized, switching the light bulb onto the secondary windings, draining the filter capacitors almost instantly. This is a monkey way, i.e. cheap way, to make the secondary side of the power supply "die" faster, if the load is smaller than the filtering capability of the output capacitors (which is what you want with audio circuits). This way, the "thump" upon power down is significantly reduced.

Power up thump is a whole 'nother story.
 
ok cool. the reason I ask is that I have never seen a power supply with a switch on it. Not that I mind being different.. :cool: its just that the switch always seems to be on the unit and the PS on the floor behind the racks...

that is very interesting about the thumps!! how long does it take the secondary to drain out usually (sans lightbulb)? Also, where can I find info on how to wire the transformer? there are 2 primary wires and 3 secondary, and they're not labeled or anything...? I'm not really sure how to tell which one is supposed to go to ground, and which is the + and - ..?
 
FALKEN said:
Also, where can I find info on how to wire the transformer? there are 2 primary wires and 3 secondary, and they're not labeled or anything...? I'm not really sure how to tell which one is supposed to go to ground, and which is the + and - ..?

On the primary, it doesn't matter, wire neutral to one wire and hot to the other. The ground wire goes to the chassis, not the transformer. On the secondary, measure the resistance between the leads to figure out which one is the center tap. + and - DC voltage is a function of the rectifier, not the tranny.
 
ok let me rephrase. (i sound like such an idiot its unreal) i know which is the center tap. your diagram only has two leads leaving the transformer. what to do with the third?
 
FALKEN said:
ok let me rephrase. (i sound like such an idiot its unreal) i know which is the center tap. your diagram only has two leads leaving the transformer. what to do with the third?

Oh, sorry. The center tap should be grounded :o
 
so I've been using this thing passive and it sounds pretty damn good! I haven't tried it with the aux yet but if it works ok I think I'm just going to keep it this way and call it a day. Although I still might attempt a separate "wired" project.

I do have one question though, before I go any further with this thing. Since I am using mic pres for gain makeup; what would be the simplest way to prevent against accidental application of phantom power and the frying of all my gear?
 
FALKEN said:
so I've been using this thing passive and it sounds pretty damn good! I haven't tried it with the aux yet but if it works ok I think I'm just going to keep it this way and call it a day. Although I still might attempt a separate "wired" project.

I do have one question though, before I go any further with this thing. Since I am using mic pres for gain makeup; what would be the simplest way to prevent against accidental application of phantom power and the frying of all my gear?

Keeping in the spirit of the project, use an output transformer.
 
cool; that's what I was thinking. I'm not sure which route to go...a 150:150 after the shunt, or a 600:600 with no shunt, or perhaps a 600:150 with no shunt? Doesn't the primary have to be loaded somehow?
 
FALKEN said:
cool; that's what I was thinking. I'm not sure which route to go...a 150:150 after the shunt, or a 600:600 with no shunt, or perhaps a 600:150 with no shunt? Doesn't the primary have to be loaded somehow?

Hmmm I have to refresh my memory on this project . . . this is strictly a passive mixer, or did it have an output buffer? I can't remember. Do have a final schematic?

If there is an output buffer, I would use 600:600. If not, then you might want to use a high impedance input tranny. Something like a 10K:600, that would drop the level some more though, but so would having a 600 ohm primary. I would think 150:150 would kill your level entirely.

I'm not a transformer guy, I usually avoid them because they are expensive :o This might be a question for the Lab.
 
right now it is a passive mixer. There are 8 channels. each channel gets split between two 10k pots. one goes directly to aux mix bus. the other pot is the volume pot and from there goes to a 4-pole 3-position pan switch (only half the poles are used since it is unbalanced). The center position goes through a pair of 8.66k resistors. from there to a mute switch and the mix bus which is 20k resistors shunted with probably a 250ohm resistor. The aux bus is also 20k mix resistors shunted.

This is a very simple setup. This weekend I got all of the kinks worked out, tightened up all the pots and got it racked up. I have to tell you, it sounds very good. the effect that the aux has on the stereo pair is virtually unnoticeable. The required makeup gain is ideal (around 40 db probably). It really doesn't even need transformers. But I think it might be a good idea. There is a very small amount of hum. I can draw the schemo if you would like.
 
Last edited:
Sounds pretty good. For the hum, just have another look at your grounding scheme and see if there is room for improvement. Try to isolate the hum that is in the box and the hum that is in the cable by using a really really short cable and see if that helps.

Then run out to Radio Shack and pick up a cheap 600:600 audio transformer. If that helps the hum issue, then spend the money on better quality transformers.
 
hmm...all of the cables are 2' patch cables so I doubt the hum is there. I could tie the signal ground to the box, do you think this is a good idea? right now it is floating. still not sure which transformers to use. and I still have to check for crosstalk because I realized I haven't done that yet.
 
dude this mixer has been kicking serious ass. I've been using with with a vintech 1272 and it sounds really sweet. I plan on trying it with a pair of vintage api's today. it really is incredible.

But I really have to find the best way to protect all of my gear from accidental phantom power! I was thinking transformers on the outputs but I really want the output to be as transparent as possible, so maybe high quality blocking capacitors would be the way to go? I'm not really sure on this one...also maybe its better to use blocking on each individual input rather than the outputs....any suggestions are appreciated.
 
Back
Top