I'm putting together a jam-session suitcase with an active speaker, mixing and power supply for some synths. I still haven't received all the parts I've ordered yet, so I can't try out the full setup yet. Below is a simple schematic of the signal path and power.
I've got most of it sorted out, but a few questions remain.
1. The 4 channel line mixers are two MX400 which are really small and neat, but maximum level adjustment lies at roughly 60% before it starts to distort. Should I just learn to live with it or can I install some resistors maybe before the inputs to take down the signal? Or should I just lower the output from the synths? The synths that are connected are set to about 80% of maximum output, at this level the waveforms looks good without clipping.
2. It feels like the output from the mixer is a bit low, I think it might not be enough as a clean input for the amp. So I started looking at a cheap DIY preamp kit, it has pots for gain and output so I should be able to adjust the level pretty good. I'm thinking about buying a hand-held scope, but for now I've used my audio interface and Visual Analyzer, but I'm not sure I can trust the readings since it depends on the gain I set on my interface etc. How would I decide if I need the preamp? Just set it up without it first and listen? (I will not drive the speaker especially hard so it might be enough?)
Maybe a bit off topic, but someone maybe knows.
3. The way I routed the power, I'm thinking just a big bus from the 24V PSU and connect the other parts directly to it. Is this a viable connection or could there be cross-talk? Most of the components have a regulator between the actual part and the 24V line, except the amp and pre-amp.
4. Ground loops.. I've read about them, but I'm not quite there yet. How do I minimize them? Should all the ground connectors be connected to the same point? Should it be connected to- or separated from protective earth? I'd say the +/-12V reg and X/O earth should be kept isolated right? Since I guess the regulator creates a virtual ground?
5. What is a good size and type for a protective cap for the tweeter? It's a ribbon if that matters. Someone recommended a PTC resistor as well to protect the tweeter, but I'm not sure if I need it.
Hopefully someone can help me out so I can get started during the holidays
A bit delayed though, since the 24V 120W PSU I ordered from Aliexpress turned up at my doorstep yesterday as three 5V 2A PSU's...? Gaah.
An externally hosted image should be here but it was not working when we last tested it.
I've got most of it sorted out, but a few questions remain.
1. The 4 channel line mixers are two MX400 which are really small and neat, but maximum level adjustment lies at roughly 60% before it starts to distort. Should I just learn to live with it or can I install some resistors maybe before the inputs to take down the signal? Or should I just lower the output from the synths? The synths that are connected are set to about 80% of maximum output, at this level the waveforms looks good without clipping.
2. It feels like the output from the mixer is a bit low, I think it might not be enough as a clean input for the amp. So I started looking at a cheap DIY preamp kit, it has pots for gain and output so I should be able to adjust the level pretty good. I'm thinking about buying a hand-held scope, but for now I've used my audio interface and Visual Analyzer, but I'm not sure I can trust the readings since it depends on the gain I set on my interface etc. How would I decide if I need the preamp? Just set it up without it first and listen? (I will not drive the speaker especially hard so it might be enough?)
Maybe a bit off topic, but someone maybe knows.
3. The way I routed the power, I'm thinking just a big bus from the 24V PSU and connect the other parts directly to it. Is this a viable connection or could there be cross-talk? Most of the components have a regulator between the actual part and the 24V line, except the amp and pre-amp.
4. Ground loops.. I've read about them, but I'm not quite there yet. How do I minimize them? Should all the ground connectors be connected to the same point? Should it be connected to- or separated from protective earth? I'd say the +/-12V reg and X/O earth should be kept isolated right? Since I guess the regulator creates a virtual ground?
5. What is a good size and type for a protective cap for the tweeter? It's a ribbon if that matters. Someone recommended a PTC resistor as well to protect the tweeter, but I'm not sure if I need it.
Hopefully someone can help me out so I can get started during the holidays
A bit delayed though, since the 24V 120W PSU I ordered from Aliexpress turned up at my doorstep yesterday as three 5V 2A PSU's...? Gaah.
Keeping the cables as short as possible is the best way to cut down hum from ground loops. While standard RCA plug cables are 6', you can sometimes find 3' ones. Even better you can cut one end off and use solder RCA plugs to make them shorter still, if your enclosure allows less cable length.
As far as the distortion that happens when you turn up the mixer, apparently your active XO has some sort of input limit. Try to look at the spec sheet and see what it is. Cutting the input level to the mixer is a bad idea, as that is a formula to increase broadband noise. Limiting the output of the mixer with a fixed potentiometer or fixed resistor divider is a better idea.
Incongruently, you say the output of the mixer is too low. Which is it? too low or too high? 1.6 vac is the max input some of my 1960's amps use, 2 vac is the modern standard for line level signals.
I think rather than a *****y scope which probably has a max probe voltage too low to repair amplifiers (60 v max probe voltage is common among ****ese bargain scopes, whereas I own transistor amps with 85, 75 & 70 v rails. My tube amps use 450 V plate voltage ) you should buy an analog VOM. My Simpson 260XLPM has 20 vac & 2 vac scales which are handy for measuring line level signals, especially in the input stage of an amp. Even a $20 GC analog meter from the hardware store with a 50 VAC scale would tell you some facts about your levels.
As a dimyselfer, I monkey with input stages of amps to change the gain. If you want more loudness, you want more gain out of the amp. If BJ transistor input stage, you increase the collector resistor. If op amp input stage inverting, you increase the feedback resistor. Of course if you hit the clipping limit, all sound becomes garbage. That is why the high rail voltages in my amps are so nice, on LP's or CD's I don't actually use 200 W/ch for more than a few tenths of a second when the cannon goes off in 1812 overture. If you're PA amping a synth, then the sound if more constant value and your amp better be PA quality with the rated watts available for hours at a time. I melted solder in a home hifi amp with a keyboard at a 4 hour rehearsal one time. It had a 1 hour watt rating whereas Peavey, Crown (except macrotech) QSC & Yamaha PA amps are 24/7 rated for wattage. Make sure the speaker drivers are rated at least 1/2 the watt amp rating, or put fuses in the speaker lines.
The value of a blocking capacitor on the tweeter depends on the impedance of the tweeter. 4 ohm, 8 ohm, some high value? 1/(2*pi*R*C) where C is in farads (microfarad is one millionth of a farad) is the inflection frequency of a RC filter. A simple RC filter will keep bass out of your tweeter. A series inductor will keep high frequency out of your woofer. My 2 way SP2-XT speaker system uses a crossover frequency of 2 khz, although others use 1.2 khz. These high current speaker crossover components are best purchased in the US from partsexpress. Europe or Asia, I have no idea. If you are Biamping, that is your active crossover separates high and low into two channels in the amp, then the crossover components don't need to be high current out in the speaker wiring. PTC resistors on tweeter is something good package speakers have for protection, but I don't know where you buy them naked, newark certainly doesn't have any. Perhaps PE again. If bi-amping, your main source of tweeter damage would be big bangs from the input cable pulling loose or RF interference from flourescent lamps turned off or similar pop sources.
While you're waiting you can get 24 v 48 W (2 amps) from a doorbell transformer from the hardware store, a couple of MR856 rectifiers, (look for the wires about 22 gauge on a rectifier) and a couple of 1000 uf >35 v capacitors. Put in a nice grounded box with a fuse. All old Projection or LED TV's and switcher supplies found on garbage day will have the rectifiers & filter capacitors.
Best of luck, I'm playing acoustic piano instead of electronic this holiday because the reliability is so much higher, if you keep it out of the rain. I just fixed a 1998 PA amp (CS800s) yesterday I bought used for $200, now blowing the main breaker. The last guy in there replaced a main 200 v filter cap with a 1981 date version. He had the gall to initial the board. Caps age whether used, or sitting on the shelf, aged rubber is dirt. These people call themselves "repairmen"!!!!
As far as the distortion that happens when you turn up the mixer, apparently your active XO has some sort of input limit. Try to look at the spec sheet and see what it is. Cutting the input level to the mixer is a bad idea, as that is a formula to increase broadband noise. Limiting the output of the mixer with a fixed potentiometer or fixed resistor divider is a better idea.
Incongruently, you say the output of the mixer is too low. Which is it? too low or too high? 1.6 vac is the max input some of my 1960's amps use, 2 vac is the modern standard for line level signals.
I think rather than a *****y scope which probably has a max probe voltage too low to repair amplifiers (60 v max probe voltage is common among ****ese bargain scopes, whereas I own transistor amps with 85, 75 & 70 v rails. My tube amps use 450 V plate voltage ) you should buy an analog VOM. My Simpson 260XLPM has 20 vac & 2 vac scales which are handy for measuring line level signals, especially in the input stage of an amp. Even a $20 GC analog meter from the hardware store with a 50 VAC scale would tell you some facts about your levels.
As a dimyselfer, I monkey with input stages of amps to change the gain. If you want more loudness, you want more gain out of the amp. If BJ transistor input stage, you increase the collector resistor. If op amp input stage inverting, you increase the feedback resistor. Of course if you hit the clipping limit, all sound becomes garbage. That is why the high rail voltages in my amps are so nice, on LP's or CD's I don't actually use 200 W/ch for more than a few tenths of a second when the cannon goes off in 1812 overture. If you're PA amping a synth, then the sound if more constant value and your amp better be PA quality with the rated watts available for hours at a time. I melted solder in a home hifi amp with a keyboard at a 4 hour rehearsal one time. It had a 1 hour watt rating whereas Peavey, Crown (except macrotech) QSC & Yamaha PA amps are 24/7 rated for wattage. Make sure the speaker drivers are rated at least 1/2 the watt amp rating, or put fuses in the speaker lines.
The value of a blocking capacitor on the tweeter depends on the impedance of the tweeter. 4 ohm, 8 ohm, some high value? 1/(2*pi*R*C) where C is in farads (microfarad is one millionth of a farad) is the inflection frequency of a RC filter. A simple RC filter will keep bass out of your tweeter. A series inductor will keep high frequency out of your woofer. My 2 way SP2-XT speaker system uses a crossover frequency of 2 khz, although others use 1.2 khz. These high current speaker crossover components are best purchased in the US from partsexpress. Europe or Asia, I have no idea. If you are Biamping, that is your active crossover separates high and low into two channels in the amp, then the crossover components don't need to be high current out in the speaker wiring. PTC resistors on tweeter is something good package speakers have for protection, but I don't know where you buy them naked, newark certainly doesn't have any. Perhaps PE again. If bi-amping, your main source of tweeter damage would be big bangs from the input cable pulling loose or RF interference from flourescent lamps turned off or similar pop sources.
While you're waiting you can get 24 v 48 W (2 amps) from a doorbell transformer from the hardware store, a couple of MR856 rectifiers, (look for the wires about 22 gauge on a rectifier) and a couple of 1000 uf >35 v capacitors. Put in a nice grounded box with a fuse. All old Projection or LED TV's and switcher supplies found on garbage day will have the rectifiers & filter capacitors.
Best of luck, I'm playing acoustic piano instead of electronic this holiday because the reliability is so much higher, if you keep it out of the rain. I just fixed a 1998 PA amp (CS800s) yesterday I bought used for $200, now blowing the main breaker. The last guy in there replaced a main 200 v filter cap with a 1981 date version. He had the gall to initial the board. Caps age whether used, or sitting on the shelf, aged rubber is dirt. These people call themselves "repairmen"!!!!
Last edited:
A setup as proposed is just begging for trouble, and I would thus consider it a Bad Idea.
If you galvanically connect the power supplies of all of these units, you must have (properly implemented) balanced connections between them to stay out of trouble. Otherwise it'll be a maze of ground loops. (While these do not magically disappear when going balanced, their currents will flow past any signal-carrying circuitry.) That would be a whole bunch of isolation transformers.
Is there any particular reason you want to supply everything from the +24V supply instead of using individual wall-warts?
BTW, I would only ever be ordering power supplies directly from China if I knew they'd be from a trustworthy brand and had all the necessary approvals. "China Export" and "electrical safety" have some overlap but it's definitely not 100%.
If the MX400 starts distorting, it's probably running into clipping with the input level provided. It may have more gain than its 12 V single supply will support under these circumstances, which if you do the math is not generally going to allow much more than about 9 Vpp or around 3 Vrms of unbalanced output. (Same goes for the XO.) I'd just reduce input level. But if 3 Vrms of output level doesn't cut it, there clearly is something wrong with the rest. Power amp input level might be turned down, or the active XO might have a lot of loss for some reason.
If you galvanically connect the power supplies of all of these units, you must have (properly implemented) balanced connections between them to stay out of trouble. Otherwise it'll be a maze of ground loops. (While these do not magically disappear when going balanced, their currents will flow past any signal-carrying circuitry.) That would be a whole bunch of isolation transformers.
Is there any particular reason you want to supply everything from the +24V supply instead of using individual wall-warts?
BTW, I would only ever be ordering power supplies directly from China if I knew they'd be from a trustworthy brand and had all the necessary approvals. "China Export" and "electrical safety" have some overlap but it's definitely not 100%.
If the MX400 starts distorting, it's probably running into clipping with the input level provided. It may have more gain than its 12 V single supply will support under these circumstances, which if you do the math is not generally going to allow much more than about 9 Vpp or around 3 Vrms of unbalanced output. (Same goes for the XO.) I'd just reduce input level. But if 3 Vrms of output level doesn't cut it, there clearly is something wrong with the rest. Power amp input level might be turned down, or the active XO might have a lot of loss for some reason.
Many thanks for an extensive answer!
All cables will be soldered or put into terminals, except from the synts, will have an I/O panel built in to the flightcase. So the cables will be kept as short as possible. Any guidelines of the thickness of the cables for the different applications? I bought 2 x 1.5mm cables for the speakers, 30W RMS + 15 W RMS, both 8 Ohms, is that enough?
In the instructions for the crossover (kit from KMtech) it only says "line-level", so I'm not sure really.
What I'm trying to say with the mixer is that it feels like it's disting when i go higher than ~60% och the level. Using my computer as a scope (Visual Analyzer) i see that there is significant noise coming from the output if i dial up the level adjuster of a channel, even if it is not connected to a source. So I guess it might just be a shitty mixer and I'll have to keep from turning the knobs all the way up.. Made a recording yesterday with the levels set about 50-60% and it sounds pretty good.
About the low level, normally when I have my monitors connected, I have them at maybe half volume, and that's pretty loud. With the mixer in between, it's a fair bit lower. But I think it's just due to the low output from the synths then.
About the scope, I'm mostly interested in buying one to study waveforms coming from the synths. Thinking about starting to put one together myself. So it's all low level stuff, for the higher loads I'll use my multimeter.
I have oversized that part of the build pretty good I think. I've got a 2x100W amp Bi-Amping a 30W+15W system, fed by roughly 100W if I subtract the power draw from other components from the 120W PSU. So I don't think clipping will be a problem..
I'll just go with a 10uF cap, X/O is at 4.5kHz so that should be safe. I'm guessing it should be non-polar, any special kind you recommend?
All cables will be soldered or put into terminals, except from the synts, will have an I/O panel built in to the flightcase. So the cables will be kept as short as possible. Any guidelines of the thickness of the cables for the different applications? I bought 2 x 1.5mm cables for the speakers, 30W RMS + 15 W RMS, both 8 Ohms, is that enough?
In the instructions for the crossover (kit from KMtech) it only says "line-level", so I'm not sure really.
What I'm trying to say with the mixer is that it feels like it's disting when i go higher than ~60% och the level. Using my computer as a scope (Visual Analyzer) i see that there is significant noise coming from the output if i dial up the level adjuster of a channel, even if it is not connected to a source. So I guess it might just be a shitty mixer and I'll have to keep from turning the knobs all the way up.. Made a recording yesterday with the levels set about 50-60% and it sounds pretty good.
About the low level, normally when I have my monitors connected, I have them at maybe half volume, and that's pretty loud. With the mixer in between, it's a fair bit lower. But I think it's just due to the low output from the synths then.
About the scope, I'm mostly interested in buying one to study waveforms coming from the synths. Thinking about starting to put one together myself. So it's all low level stuff, for the higher loads I'll use my multimeter.
I have oversized that part of the build pretty good I think. I've got a 2x100W amp Bi-Amping a 30W+15W system, fed by roughly 100W if I subtract the power draw from other components from the 120W PSU. So I don't think clipping will be a problem..
I'll just go with a 10uF cap, X/O is at 4.5kHz so that should be safe. I'm guessing it should be non-polar, any special kind you recommend?
Speaker cables are not where ground loop noise will be picked up. 1.5 mm should be fine at this wattage. Inter-component cables are where noise will be picked up, both RF and hum. So those, short as possible in the rack.
Inter-line level component cables are either RCA connectors on coax, or 1/4 phone plug connectors on twisted pair. The latter is set up for differential op amp inputs and line driver outputs, which avoids the ground loop problem. Every power-DC supply-power-interline connector return makes an area that mains power frequency will drive current around. Remember magnetic fields make current flow in a circle. AC driven components have a power transformer in each device to break up the loop. Wall transformers are cheaper than the isolation transformers sgrossklass was talking about. Jensen's the ones with actual specifications are about $50 each.
I have no knowledge of what the impedance of the tweeter is so I have no opinion on the crossover. Don't know what continent you are on so no opinion on the parts brands. PartsExpress has a house brand, there the biggest source of crossover parts here unless you buy 10000 quantity of something from the orient.
DVM produce random numbers on music IMHO, except the $180 RMS Flukemeter. And the RMS fluke averages over some time, missing peak events, and cuts out at 7000 hz, missing RF oscillation if you have any. Analog VOM can do 20 hz to 20 mhz with a capacitor in the negative lead on AC scale. analog VOM can pick up events as fast as 1/60 of a second. I had a Radio shack analog VOM that the pointer vibrated full scale at 60 hz when I tested the wall plug with 300 VDC scale. RS used to sell a good analog VOM, but that was long ago.
Inter-line level component cables are either RCA connectors on coax, or 1/4 phone plug connectors on twisted pair. The latter is set up for differential op amp inputs and line driver outputs, which avoids the ground loop problem. Every power-DC supply-power-interline connector return makes an area that mains power frequency will drive current around. Remember magnetic fields make current flow in a circle. AC driven components have a power transformer in each device to break up the loop. Wall transformers are cheaper than the isolation transformers sgrossklass was talking about. Jensen's the ones with actual specifications are about $50 each.
I have no knowledge of what the impedance of the tweeter is so I have no opinion on the crossover. Don't know what continent you are on so no opinion on the parts brands. PartsExpress has a house brand, there the biggest source of crossover parts here unless you buy 10000 quantity of something from the orient.
DVM produce random numbers on music IMHO, except the $180 RMS Flukemeter. And the RMS fluke averages over some time, missing peak events, and cuts out at 7000 hz, missing RF oscillation if you have any. Analog VOM can do 20 hz to 20 mhz with a capacitor in the negative lead on AC scale. analog VOM can pick up events as fast as 1/60 of a second. I had a Radio shack analog VOM that the pointer vibrated full scale at 60 hz when I tested the wall plug with 300 VDC scale. RS used to sell a good analog VOM, but that was long ago.
"you must have (properly implemented) balanced connections between them" - Not quite sure what you mean by that? Are you talking about the signal path from the synths to the amp? The leads are very short, maybe 10-15 cm max.
The reason is that everything will be housed in the flight case, with just a fused c14 connector. I'm trying to minimize the crap I have to drag around if I just want to go for a quick jam.
The reason is that everything will be housed in the flight case, with just a fused c14 connector. I'm trying to minimize the crap I have to drag around if I just want to go for a quick jam.
With leads that short, you might get away with it without a lot of hum.
Synth most likely won't have a balenced output, a 3 pin XLR. Mostly the have 1/4" phone receptacles.
Preamps, PA amps, pro audio equipment sometimes have XLR connectors that come from balenced line driver IC's and go to two inputs of an op amp in differential receive mode. Used with twisted pair cables to avoid hum.
Synth most likely won't have a balenced output, a 3 pin XLR. Mostly the have 1/4" phone receptacles.
Preamps, PA amps, pro audio equipment sometimes have XLR connectors that come from balenced line driver IC's and go to two inputs of an op amp in differential receive mode. Used with twisted pair cables to avoid hum.
That Belden 8721 has a twisted pair, covered by a foil shield.
Great for XLR cable. The 3 pin kind. Great for pro audio equipment having XLR connectors and differential op inputs and split line driver outputs.
If you're not using XLR connectors, but instead 1/4 phone plugs, there is nothing to connect the shield to. Merging the shield and one wire on the jack shaft buys IMHO nothing.
1/4 phone plugs, one tends to use unshielded twisted pair in the PA market. Think snake cables from mixer to stage power amp.
Amateur line level equipent usually uses RCA plugs & jacks. The cable tends to have one straight wire inside, and a mesh shield enclosing. Impedance about 110 ohms in the days when people made these cables for themselves (late 50's early sixties). The best bargains these days are salvage HDTV converter cables, and triple RCA plug video game cables. I would envision shortening 6' and 8' versions of those I pick up for $1 to $2 each at resale shops, with custom solder on RCA plugs. I get my solder on RCA connectors from tubesandmore.com and triodeelectronics.com in the US, neither a major supply house. More vacuum tube and obsolete amp specialists.
Great for XLR cable. The 3 pin kind. Great for pro audio equipment having XLR connectors and differential op inputs and split line driver outputs.
If you're not using XLR connectors, but instead 1/4 phone plugs, there is nothing to connect the shield to. Merging the shield and one wire on the jack shaft buys IMHO nothing.
1/4 phone plugs, one tends to use unshielded twisted pair in the PA market. Think snake cables from mixer to stage power amp.
Amateur line level equipent usually uses RCA plugs & jacks. The cable tends to have one straight wire inside, and a mesh shield enclosing. Impedance about 110 ohms in the days when people made these cables for themselves (late 50's early sixties). The best bargains these days are salvage HDTV converter cables, and triple RCA plug video game cables. I would envision shortening 6' and 8' versions of those I pick up for $1 to $2 each at resale shops, with custom solder on RCA plugs. I get my solder on RCA connectors from tubesandmore.com and triodeelectronics.com in the US, neither a major supply house. More vacuum tube and obsolete amp specialists.
Pin1 direct to Chassis via the lowest impedance connection you can manage. Don't use a 20mm long pig tail, keep the connection under 5mm and a wide tape/strap is better than a narrow wire..
As above:
a] Connect the signal interconnect shields to the chassis at the connector.
b] Connect the Protective Earth/Safety Ground to the chassis near where the AC power enters the chassis.
c] Connect the audio circuit & power supply common to the chassis at one point near the signal input connectors. (maybe with a RC network)
a] Connect the signal interconnect shields to the chassis at the connector.
b] Connect the Protective Earth/Safety Ground to the chassis near where the AC power enters the chassis.
c] Connect the audio circuit & power supply common to the chassis at one point near the signal input connectors. (maybe with a RC network)
Hi again, I need some help from you guys again. I've got this PSU and I'm about to set up my project. But I'm getting a bit confused about the grounding, when I try to measure between V+ and PE, i get 0V. So what I found trying to read up on it, the V- is separated from PE.
I'm trying to follow the principle in a guide posted by a user here, but I'm unsure what happens if I connect PE to ground. I'm worried that there will be interference if I plug in a source with different ground potential? Probably misunderstood something, but my question is, should PE be connected to ground or not?
I'm trying to follow the principle in a guide posted by a user here, but I'm unsure what happens if I connect PE to ground. I'm worried that there will be interference if I plug in a source with different ground potential? Probably misunderstood something, but my question is, should PE be connected to ground or not?
There are two separate yet related issues here. One is shock hazard safety, and the other is audio quality. Connecting signal ground to safety/protective/chassis ground ensures that that a product's electrical fault protection will activate should high voltage appear on the signal ground. If signal and safety grounds are disconnected from each other, a high voltage voltage fault connecting to the internal signal ground would then also appear on any exposed signal connector ground surfaces, making for a shock hazard apart from any on the chassis surface. Unfortunately, from an audio quality perspective, such connection is also a recipe for power mains generated ground loop hum. National electrical codes are written to ensure safety and are not concerned with audio quality.
Regarding ground loop hum, the points of entry for it in your wiring diagram are the mixer inputs when connected to other mains powered devices. The most effective way to suppress ground loop noise is also the most costly, which is by utilizing transformer coupled inputs to the mixers. The next best solution is to utilize one of the ground isolation devices mentioned in that Rane note, such as employing an diode string, or an thermistor. These devices insert an conditional isolating impedance between signal and safety grounds. Should the voltage difference between the two grounds exceed a low threshed, then the devices become a low impedance, locking signal ground to safety ground and activating the power mains protection circuit.
BTW, so long as any devices connected via the mixer inputs are ALL floating, such as with electric guitar pick-ups, then there shoudn't be any power mains generated ground loop hum coupling. Although, it is still possible for hum to couple (be induced) into the interconnecting cables via stray external magnetic hum field.
One other tip. Cascade only linear voltage regulators after that primary 24V switchmode power supply module. Multiple SMPS units can easily generate intermodulated noise and distortion products off of each other's non-synchronous switching frequencies.
Last edited:
Big thanks Ken, you cleared a few questions there. The "next best solution" is what I'm going with. Tried out a circuit with 2 diodes in opposite directions, a 10 Ohm 5W wirewound resistor and a 100nF ceramic cap, it did a big difference. Even though I tested the circuit with a PSU which is audibly noisy (got damaged when my first amp got shorted, looks like the heatsink shorted a smd cap...).
Still, I'm wondering what happens if I connect the PE to V-? Because I'm not sure how this works really. I mean if I measure between V+ and V- I get 24V difference, but it's just a potential. I mean, now I can't be sure if V- is -12V and V+ is +12V, or any other combination.. Will tying V- to PE, set it to 0V, and thus, V+ to 24V? I just want to be sure I don't break anything else, I'm tired of waiting for parts
Still, I'm wondering what happens if I connect the PE to V-? Because I'm not sure how this works really. I mean if I measure between V+ and V- I get 24V difference, but it's just a potential. I mean, now I can't be sure if V- is -12V and V+ is +12V, or any other combination.. Will tying V- to PE, set it to 0V, and thus, V+ to 24V? I just want to be sure I don't break anything else, I'm tired of waiting for parts
Big thanks Ken, you cleared a few questions there. The "next best solution" is what I'm going with. Tried out a circuit with 2 diodes in opposite directions, a 10 Ohm 5W wirewound resistor and a 100nF ceramic cap, it did a big difference. Even though I tested the circuit with a PSU which is audibly noisy (got damaged when my first amp got shorted, looks like the heatsink shorted a smd cap...).
Still, I'm wondering what happens if I connect the PE to V-? Because I'm not sure how this works really. I mean if I measure between V+ and V- I get 24V difference, but it's just a potential. I mean, now I can't be sure if V- is -12V and V+ is +12V, or any other combination.. Will tying V- to PE, set it to 0V, and thus, V+ to 24V? I just want to be sure I don't break anything else, I'm tired of waiting for parts
This depends on how that 24V supply is configured internally, and how it's being specified. Normally, any power supply voltage should be specified relative to the signal/power ground (to the supply common tap). So, -24V should be what one would measure between the power ground tap and that specified output tap. If the supply is specified as being +12V and -12V, then those outputs should each be as measured relative to the power ground/common.
Safe practice is to tie power ground to PE in a way that does not permit any significant voltage to appear on the power supply ground relative to PE, in case of some fault condition. Most common is to simply hard wire one to the other. However, as has been stated, this is a recipe for allowing ground loop noise.
Once the supply power ground and the PE are tied, connecting any of the supply output taps (-24V, +12V, -12V, etc.) to the PE will short circuit the supply and SHOULD trip it's overload protection circuit.
Last edited:
I think maybe I explained it the wrong way. I'm saying it is a single polarity supply, so the only thing coming out of the supply is V+ and V-, no ground/common. If I measure between V+/- and protective earth, I get 0V. So that's what's bothering me, I can't relate the 24V difference in potential to anything.. But supposedly, V+=24V and V-=0V, then tying V- to PE wouldn't be a problem.
It seems the output is indeed not referenced to PE. It seems to be floating. So, you likely can assign either the positive or the negative output terminal as power/signal ground. That allows you to easily make the supply +24V, or alternatively -24V, simply by designating the opposite output terminal as ground. Refer to the supply's data sheet for application details.
ASSUMING the supply output is in fact effectively floating (verify!) you can then connect the terminal you have designated as ground to the PE terminal. Such connection could be made via the isolation devices you had mentioned utilizing earlier, so as to safely impede a potential power mains generated ground loop.
ASSUMING the supply output is in fact effectively floating (verify!) you can then connect the terminal you have designated as ground to the PE terminal. Such connection could be made via the isolation devices you had mentioned utilizing earlier, so as to safely impede a potential power mains generated ground loop.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.