I have built and active xover and ran it off batteries as ive posted before. However upon upgrade of the xover the batteries lasted a max of ten hours so a power supply was built (just finished about thirty mins ago). Anyway under battery operation there is ZERO noise added by the xover maybe a bit of hiss but thats because its the preamp to and everything has a bit of hiss. Anyway I wired in the power supply and MAD MAD
hum issue ensued. This is loud.
The powersupply is nothing outrageous a 30VA 12*2 trannie with duel rectifier and 1100uf of smoothing (i plan to add regulators at a later date to see what diff it makes). The input from the mains goes thru a simple shop bought IEC filter stage such as the ones by belden.
More on the hum. This hum is louder then ive ever encountered its at 100hz and 50hz. The 100hz dissapears when u break the ground to the wall socket when you unplug the unit. As the caps drain the 50hz remains however, still very loud. Then the caps drain and the infernal racket ceases.
This hum is clearly audible thru bass heavey material when played at a level which is normally too loud for comfortable listening.
The xover is constructed on vero strip board about 30cm long. Now I dedicated a single track to the earth and everthing wires back to this from the opamp circuits.
Having searched the forum I found a couple of things to try -
One would be a 10ohm resistor from the signal ground to case or to power supply in my situation. (The power supply is wires thru to the case too tho)
and the other was dissconnect the unit from mains ground. This Im not to plussed about, however the unit would still be grounded thru the signal ground form the dac and the power amp so technically there would be a link to the mains.
As the whole xover is on a large piece of vero and none of it has sheilding, (where possible its sheilded like on the input and output wires) its possible that the circuit it picking up the mains hum. I have no objection to sticking the power supply in a seperate box as my poweramps is seperate too and that doesny hum at all, well a little bit but well within acceptable limits.
Im off to try a few things now but would appreciate anyone else input into solving this.
Would seperate wires from each ground location thruout the xover to a common star, not running down a whole track on the vero be worth trying??? if not and the aforementioned doesnt help what else could I try??
Cheers Matt
hum issue ensued. This is loud.The powersupply is nothing outrageous a 30VA 12*2 trannie with duel rectifier and 1100uf of smoothing (i plan to add regulators at a later date to see what diff it makes). The input from the mains goes thru a simple shop bought IEC filter stage such as the ones by belden.
More on the hum. This hum is louder then ive ever encountered its at 100hz and 50hz. The 100hz dissapears when u break the ground to the wall socket when you unplug the unit. As the caps drain the 50hz remains however, still very loud. Then the caps drain and the infernal racket ceases.
This hum is clearly audible thru bass heavey material when played at a level which is normally too loud for comfortable listening.
The xover is constructed on vero strip board about 30cm long. Now I dedicated a single track to the earth and everthing wires back to this from the opamp circuits.
Having searched the forum I found a couple of things to try -
One would be a 10ohm resistor from the signal ground to case or to power supply in my situation. (The power supply is wires thru to the case too tho)
and the other was dissconnect the unit from mains ground. This Im not to plussed about, however the unit would still be grounded thru the signal ground form the dac and the power amp so technically there would be a link to the mains.
As the whole xover is on a large piece of vero and none of it has sheilding, (where possible its sheilded like on the input and output wires) its possible that the circuit it picking up the mains hum. I have no objection to sticking the power supply in a seperate box as my poweramps is seperate too and that doesny hum at all, well a little bit but well within acceptable limits.
Im off to try a few things now but would appreciate anyone else input into solving this.
Would seperate wires from each ground location thruout the xover to a common star, not running down a whole track on the vero be worth trying??? if not and the aforementioned doesnt help what else could I try??
Cheers Matt
5th element I think you should check the ground in your Mains wiring as there might be some corruption there.
I dont know your housing situation but if you own the house try the equipment on another circuit and see if it persists
I typically install a ground rod close to the mains panel and wire a separate circuit for high end installations
Most houses are wired up that the plumbing pipes have a ground attached that has been done for years and over time it becomes corroded
DIRT®
I dont know your housing situation but if you own the house try the equipment on another circuit and see if it persists
I typically install a ground rod close to the mains panel and wire a separate circuit for high end installations
Most houses are wired up that the plumbing pipes have a ground attached that has been done for years and over time it becomes corroded
DIRT®
cheers for your responses ppl much appreciated. This seems the most daft thing to help solve the prob but I guess it makes sense kinda to which my brain says add regs and the prob will vanish.
Here goes.
I added an extra 6800uf of smoothing to each rail this dramitically lowered the hum but not enough for satisfaction. The caps before we bog standard 220uf 5 per rail I dont know abou their smoothing capabilities but to me it wasnt enough. Anyway the next step will be to add regulators im going to do one of those pretracking things (if thats the right name) using lm317's and thier complementaries, hope this helps.
Is it possible that the 5*220uf on each rail has a detrimental effect on the supply? and if so would removing them be a good idea. 6800uf is what 100watt commercials sometimes have in 'em and this xover draws 60mA.
Cheers Matt
Oh and BTW I know ive mentioned my power amp (slones 11.4 design in HPAACM) but with no pics or anything just problems with it. Anyway thats because the case sucks. Anyway Im awaiting the arrival of heatsinks so when the whole shebang is finished ill be sure to share it all with the DIY community. I could even take a pic of its crappyness and then do a before and after makeover of the amp lol.
Here goes.
I added an extra 6800uf of smoothing to each rail this dramitically lowered the hum but not enough for satisfaction. The caps before we bog standard 220uf 5 per rail I dont know abou their smoothing capabilities but to me it wasnt enough. Anyway the next step will be to add regulators im going to do one of those pretracking things (if thats the right name) using lm317's and thier complementaries, hope this helps.
Is it possible that the 5*220uf on each rail has a detrimental effect on the supply? and if so would removing them be a good idea. 6800uf is what 100watt commercials sometimes have in 'em and this xover draws 60mA.
Cheers Matt
Oh and BTW I know ive mentioned my power amp (slones 11.4 design in HPAACM) but with no pics or anything just problems with it. Anyway thats because the case sucks. Anyway Im awaiting the arrival of heatsinks so when the whole shebang is finished ill be sure to share it all with the DIY community. I could even take a pic of its crappyness and then do a before and after makeover of the amp lol.
I have tried removing the ground from the mains plug and all this did was remove the 100hz buzz and leave the 50hz. Initially all I had connected was the power supply earth to the xover earth and that was it with the power supply earth connected to the wall outlet. I then removed the wall outlet earth as recommended as a try it and see. This removed the 100hz buzz but was either replaced of just revealed a 50hz buzz. Having the power supply in different locations didnt help matters one little bit.
So I added the 6800uf and the noise reduced dramatically but not to within acceptable limits.
The power supply design I have build twice for power amps and that has no hum issues at all. I've never encountered a ground loop before. If this is one I dont know what im suppossed to do to combat it. Im not having a go but saying solve this before adding regs is all very well but I dont know where to start to solve it.
Could you please list a few things to try to break the ground loop as ive never had to deal with one before.
Cheers again Matt.
So I added the 6800uf and the noise reduced dramatically but not to within acceptable limits.
The power supply design I have build twice for power amps and that has no hum issues at all. I've never encountered a ground loop before. If this is one I dont know what im suppossed to do to combat it. Im not having a go but saying solve this before adding regs is all very well but I dont know where to start to solve it.
Could you please list a few things to try to break the ground loop as ive never had to deal with one before.
Cheers again Matt.
Can you sketch out and post your grounding scheme? That would help us give some specific suggestions. I have a suspicion that it's how you're tying the PS caps' ground and the analog ground together, but it's a suspicion based on very little data.
Some potential band-aids: run as many grounds separately as possible, all connected at one common point on the chassis. Insulate input and output jacks from the chassis and connect them to the chassis via 10 ohm resistors. Keep power grounds and signal grounds as isolated as possible, connecting together only at the one common point.
Some potential band-aids: run as many grounds separately as possible, all connected at one common point on the chassis. Insulate input and output jacks from the chassis and connect them to the chassis via 10 ohm resistors. Keep power grounds and signal grounds as isolated as possible, connecting together only at the one common point.
Its difficult for me to sketch show as I cant d/l images to net my brother has the know how, how to as he does regularly I just havnt benn bothered to sort anything like that out but I'll have a go at describing what ive got regarding grouding.
Ok PSU 1st. This is done on vero board, duel rectifier, with the centre two tracks soldered together with the caps running the length of the board.
Rec1--------------------------------
==================
Rec2++++++++++++++++++
Kinda like that with caps along the length. So nothing special.
The xover/preamp (alps pot before input) is about 30cm*10cm on vero board again one 30cm track I dedicated to ground/earth with all other points nescessary wired back to this. The power supply board joins this ground via a flying lead. The same point on the PSU joins via another flying lead to the earth pin on the input plug to the mains. And then another wire from the earth pin connects to the case. The input wires are all shielded and earthed to the case. The output are shielded but not earthed.
The input/output was like this on battery operation and I got no noise what-so-ever.
Should I take the ground from the PSU to the case and the case only, then attach the xover ground via the case thru 10 ohm resistors?
The only other thing that I could try but would require some considerable effort is to take individual grounds from the xover to a common star ground on the case. I would only want to do this as a last resort tho.
If you need anymore info I can try and post a pic but that will be tomorrow, I will offer more written help for maybe another hour as Im on the phone too which is gonna last a while.
Anyway cheers for all this I can imagine it could get quite annoying.
Matt
Ok PSU 1st. This is done on vero board, duel rectifier, with the centre two tracks soldered together with the caps running the length of the board.
Rec1--------------------------------
==================
Rec2++++++++++++++++++
Kinda like that with caps along the length. So nothing special.
The xover/preamp (alps pot before input) is about 30cm*10cm on vero board again one 30cm track I dedicated to ground/earth with all other points nescessary wired back to this. The power supply board joins this ground via a flying lead. The same point on the PSU joins via another flying lead to the earth pin on the input plug to the mains. And then another wire from the earth pin connects to the case. The input wires are all shielded and earthed to the case. The output are shielded but not earthed.
The input/output was like this on battery operation and I got no noise what-so-ever.
Should I take the ground from the PSU to the case and the case only, then attach the xover ground via the case thru 10 ohm resistors?
The only other thing that I could try but would require some considerable effort is to take individual grounds from the xover to a common star ground on the case. I would only want to do this as a last resort tho.
If you need anymore info I can try and post a pic but that will be tomorrow, I will offer more written help for maybe another hour as Im on the phone too which is gonna last a while.
Anyway cheers for all this I can imagine it could get quite annoying.
Matt
OK, I think I'm visualizing this. If you're doing what I think you're doing, run the XO board's ground to a point on the case near the input jacks. Run the PS ground there, too. That's the only place they should connect. If the XO board has large electrolytic bypasses, their ground leads should be isolated and returned to the single ground point.
Now, isolate the input and output jack grounds, and return them to that same point via 10 ohm resistors.
This will be a bit of a pain, but should clear up the hum. If you can post any pictures or sketches, that would be cool. You can do that via the Attach File option on the screen you use to write your post.
And it's not annoying at all, except for my rueful thoughts that the 'Net would have been a great thing to have around when I was getting started and making all my easy-to-fix mistakes.
Now, isolate the input and output jack grounds, and return them to that same point via 10 ohm resistors.
This will be a bit of a pain, but should clear up the hum. If you can post any pictures or sketches, that would be cool. You can do that via the Attach File option on the screen you use to write your post.
And it's not annoying at all, except for my rueful thoughts that the 'Net would have been a great thing to have around when I was getting started and making all my easy-to-fix mistakes.
Ok cheers for all your help its way too late to fiddle around with anything my bedroom is adjacent to parents (im 18 going to study acoustics at uni YAY for me) bedroom and I dont wanna risk a thump or click to wake em up.
So I'll try out this all tomorrow well thats really today now but I'll keep you posted with my results.
Over and snore Matt
So I'll try out this all tomorrow well thats really today now but I'll keep you posted with my results.
Over and snore Matt
I've done half of what was prescribed, this involves the single point near the input jacks. Coincidently my input jacks have tags to solder onto for the ground, is it OK to use one of these as the point. I only have two 10 ohm resistors lying around with a few others within two or three ohms of ten are these alright to use also??
The two 10 ohms I had went from the input grounds to the star point. Now I cant turn the volume down to zero, it goes close to it but no complete cut. This wont really be a problem but is there a way to kill it completely. Also doing what I've said I've done, both the inputs and the single star point, not the outputs. It cuts down on the hum a bit but not by much.
With the input and output grounds (shielding on the wire) where should I connect the other end. One is to the case and the other doesnt go anywhere, should this be altered??
Once again Matt
The two 10 ohms I had went from the input grounds to the star point. Now I cant turn the volume down to zero, it goes close to it but no complete cut. This wont really be a problem but is there a way to kill it completely. Also doing what I've said I've done, both the inputs and the single star point, not the outputs. It cuts down on the hum a bit but not by much.
With the input and output grounds (shielding on the wire) where should I connect the other end. One is to the case and the other doesnt go anywhere, should this be altered??
Once again Matt
10 ohms is an arbitrary figure. 15 or 8 would work just as well.
If you isolate your input jacks, the tag won't be at ground; you'll need a separate point for the star ground. Run the 10 ohm (nominal!) resistors from the input jack ground tag to the star point. Your real gain will come, I'm guessing, when you separate the PS and signal grounds and return them to the star ground separately.
If you isolate your input jacks, the tag won't be at ground; you'll need a separate point for the star ground. Run the 10 ohm (nominal!) resistors from the input jack ground tag to the star point. Your real gain will come, I'm guessing, when you separate the PS and signal grounds and return them to the star ground separately.
OK well here goes ive also added a diagram hope this helps.
Now, heres the latest. I have the left channels treb section running silently zero hum, the bass section has some hum which is too loud but much quieter then before.
The right channels trebble section hums but quietly (all this is audible from the listening position) and the right channels bass is the loudest out of the three that hum.
On the vero board which the xvoer is on there is a split in all the tracks mid way along the length that separates the left and the right channel completely in the audio sense. They all link up via the power supply lines and the earth from the star ground on the case.
Each side of the board is identical including the way they are earthed.(doing various things with the earthing doesnt affect anything) So one section is silent (left treb) so that must mean the way everything is attatched is fine, and its corresponding right channel is wired up identical but hums (grounding identical too the right channels grounding wire to star is even shorted then left).
The only thing that isnt totally the same is the PS into the board, it enters close to the left treb pins (see diagram) and then each other section is wired in parallel from this as shown. The only thing left for me to do is power each opamp directly thru hardwireing to the PS. Also regulate. Which I am going to try anyway.
One plus to this tho is that it sounds, hum asside, simply beautiful (OPA627 on treb and buffer then OPA2134 doing bass). Better then when off the batteries, this might be to with the +- rails being identical in magnitude and running at 17.5v, close to max spec.
Cheers again Matt
Now, heres the latest. I have the left channels treb section running silently zero hum, the bass section has some hum which is too loud but much quieter then before.
The right channels trebble section hums but quietly (all this is audible from the listening position) and the right channels bass is the loudest out of the three that hum.
On the vero board which the xvoer is on there is a split in all the tracks mid way along the length that separates the left and the right channel completely in the audio sense. They all link up via the power supply lines and the earth from the star ground on the case.
Each side of the board is identical including the way they are earthed.(doing various things with the earthing doesnt affect anything) So one section is silent (left treb) so that must mean the way everything is attatched is fine, and its corresponding right channel is wired up identical but hums (grounding identical too the right channels grounding wire to star is even shorted then left).
The only thing that isnt totally the same is the PS into the board, it enters close to the left treb pins (see diagram) and then each other section is wired in parallel from this as shown. The only thing left for me to do is power each opamp directly thru hardwireing to the PS. Also regulate. Which I am going to try anyway.
One plus to this tho is that it sounds, hum asside, simply beautiful (OPA627 on treb and buffer then OPA2134 doing bass). Better then when off the batteries, this might be to with the +- rails being identical in magnitude and running at 17.5v, close to max spec.
Cheers again Matt
Attachments
Try removing the jumpers between bass and treble grounds, then running separate wires to the star point.
If that doesn't fix it, keep the separate grounds, then un-daisy chain the supply feeds. Bring the plus supply rails to a point, then run separate wires from each board to that point. And do the same for the minus rails. Note that your quietest channel is the first one in the daisy chain, the noisiest is the last one.
If that doesn't fix it, keep the separate grounds, then un-daisy chain the supply feeds. Bring the plus supply rails to a point, then run separate wires from each board to that point. And do the same for the minus rails. Note that your quietest channel is the first one in the daisy chain, the noisiest is the last one.
This is kinda more complicated then it seems, the trebble is easy as it only has one point to ground but each bass has at least four points to be wired to ground. Still anything is worth a shot.
The hum issue became much less pronounced when I added the big smoothing caps which wouldnt affect the ground, at least I dont think it would. The hum is only at 100hz, which corresponds to the ripple frequency produced by the bridges. So this could be not earth bound but on the supply lines. Its strange tho that the treb left is silent but the rest not. Anyway I am going to regulate Im going to order the components tomorrow and I'll see what that does. I also have another two 6800 caps I could try them aswell giving like 15000uf smoothing for a 60mA current draw. OTT if you ask me, but if it works it works.
Matt
The hum issue became much less pronounced when I added the big smoothing caps which wouldnt affect the ground, at least I dont think it would. The hum is only at 100hz, which corresponds to the ripple frequency produced by the bridges. So this could be not earth bound but on the supply lines. Its strange tho that the treb left is silent but the rest not. Anyway I am going to regulate Im going to order the components tomorrow and I'll see what that does. I also have another two 6800 caps I could try them aswell giving like 15000uf smoothing for a 60mA current draw. OTT if you ask me, but if it works it works.
Matt
Ok here goes second attemp after comp crashed right.
Well I was getting annoyed at nothing working so I decided to have another (like the fifth time) look at the vero board for any irregularities. I found that an un-used track on the right treb opamp was connected up to summit on the bass where it shouldnt be. Anyway I broke the offending article and this made the right treb silent and decreased the right bass hum to the same level as the left.
I am now convinced the two channels are identical (I thought that before right but I hope they are).
Well I started to pull out opamps and see what happens. That way I can isolate specific areas to see where the small amount of hum is being induced. Anyway there are two OPA2134s per bass channel, one does the level set and baffle step, the other does the 24db roll off. Basically on removing the 24db roll off the hum dissapears, and when you put the output from the first opamp directly to the ouput the whole lot runs silently albeit without and xover on the bass.
So Im guessing that maybe isolating the 24bd from the rest and giving it its own star point to earth might solve the problem. I hope that I will get it totally sorted out sometime tomorrow tho.
Anyway cheers again Matt
Well I was getting annoyed at nothing working so I decided to have another (like the fifth time) look at the vero board for any irregularities. I found that an un-used track on the right treb opamp was connected up to summit on the bass where it shouldnt be. Anyway I broke the offending article and this made the right treb silent and decreased the right bass hum to the same level as the left.
I am now convinced the two channels are identical (I thought that before right but I hope they are).
Well I started to pull out opamps and see what happens. That way I can isolate specific areas to see where the small amount of hum is being induced. Anyway there are two OPA2134s per bass channel, one does the level set and baffle step, the other does the 24db roll off. Basically on removing the 24db roll off the hum dissapears, and when you put the output from the first opamp directly to the ouput the whole lot runs silently albeit without and xover on the bass.
So Im guessing that maybe isolating the 24bd from the rest and giving it its own star point to earth might solve the problem. I hope that I will get it totally sorted out sometime tomorrow tho.
Anyway cheers again Matt
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.