I'll send you the circuit drawing this weekend - could come in handy to DIY your DC protection as well I guess?
Yves Smolders said:
Thanks, I already am aware exactly of what it consists of ....if I wanted to make my own. However my speakers are fused already. I'm not concerned with DC protection for my amp 🙂
I'll have a look at it for you though, easier would be Jan-Peter could just let you know.
Regards,
Chris
nudge: can someone confirm all this chat about it being safe to not tie the center of the powersupply to safety/chassis ground and instead leave it floating?
I don't want to risk blowing anything up, but I would like to try this on my dual mono amps to see what effect it might have
Cheers
I don't want to risk blowing anything up, but I would like to try this on my dual mono amps to see what effect it might have
Cheers
ewildgoose,
Maybe the following will be of some use regarding earthing techniques which is taken from an article by the late Les Sage which I have found to give the best results with linear amps and UCD.
The golden rule is that true earth only exists at the signal source,from there on the further down the chain one goes towards the loudspeaker the further away from true earth one gets.
At the extreme end of the chain i.e. at the loudspeaker and junction of the PS capacitors in a power amp,this is furthest away from earth and so the loudspeaker 0v return line or PSU caps must NEVER join to earth or chassis,the result is at best, increased harmonic distortion and most often instability or even mysterious output stage failure [as a result of instability].
True earth only exists at the signal source and it is this that should be earthed. However,many amps with metal cases need to be earthed for safety reasons. If this is so, then the case should be connected directly to mains earth.
Then the chassis/case must be isolated from the electronics via an isolation bridge, typically a 100ohm res.[ at the input sockets; signal earth to chassis ] which is low enough to reference the circuit to earth, thus avoiding any potential build up between them, but high enough to severely attenuate any earth loop current. A small 22nf cap in parallel with the res. will short circuit any RFI but not 50/60Hz mains frequencies.
Bob Lewis
Maybe the following will be of some use regarding earthing techniques which is taken from an article by the late Les Sage which I have found to give the best results with linear amps and UCD.
The golden rule is that true earth only exists at the signal source,from there on the further down the chain one goes towards the loudspeaker the further away from true earth one gets.
At the extreme end of the chain i.e. at the loudspeaker and junction of the PS capacitors in a power amp,this is furthest away from earth and so the loudspeaker 0v return line or PSU caps must NEVER join to earth or chassis,the result is at best, increased harmonic distortion and most often instability or even mysterious output stage failure [as a result of instability].
True earth only exists at the signal source and it is this that should be earthed. However,many amps with metal cases need to be earthed for safety reasons. If this is so, then the case should be connected directly to mains earth.
Then the chassis/case must be isolated from the electronics via an isolation bridge, typically a 100ohm res.[ at the input sockets; signal earth to chassis ] which is low enough to reference the circuit to earth, thus avoiding any potential build up between them, but high enough to severely attenuate any earth loop current. A small 22nf cap in parallel with the res. will short circuit any RFI but not 50/60Hz mains frequencies.
Bob Lewis
Quick question for Bruno or Jan-Peter: I'll be using UcDs to drive a subwoofer, which has a seriously rising impedance (Le = 3 mH). Would a Zobel, even a partial one (I'm NOT getting a 300 uF cap for this thing) like 12 uF - 15 ohms, help at all?
BOB LEWIS said:
I rebuilt my Zappulse amp as described above (yeah I know it's not a UCD180..), there was immediate and VERY loud hum though.
Question: the mains input socket has it's metal shell connected to safety earth and hence the whole case is effectively grounded through that socket. Does this mean that a supplemental connection from my chassis star point back to the safety earth constitute another ground loop? Since the case is all anodised I initially felt better about the direct connection because I wasnt sure how good the connection was through the case to the back plane (actually I am not sure there is one...)
I suspect that the hum is actually caused by chassis earth being connected to the input XLR pin 1. I think the noise on the earth is then injected into the audio inputs? I am not sure to what extent the UCD is different, but I would like a second opinion that safety earth should effectively be connected (via the chassis) to pin 1 on the XLR input for a UCD amp?
Thanks
Ewildegoose,
I`m not familiar with the Zappulse.Try testing for continuity/resistance between pin1 of the XLR and earth pin of the mains input socket, if wired as suggested this should read 100ohms, if the reading is zero that would obviously indicate some other direct connection somewhere on the module perhaps?
Bob Lewis
I`m not familiar with the Zappulse.Try testing for continuity/resistance between pin1 of the XLR and earth pin of the mains input socket, if wired as suggested this should read 100ohms, if the reading is zero that would obviously indicate some other direct connection somewhere on the module perhaps?
Bob Lewis
Hello, just a sidenote... Found this baby on the net:
Integrated CD player & amp:
http://www.moremusic.nl/primareproducten/cdi10.html
Primare has always been nice quality high-end and their latest offering has class-d... I'm wondering if if might be UCD180's, because there is talk of going balanced just before the class-d input...
Guess we will never know 😀
If this is actually UCD, it would be nice for my living room, as my HT system is moving towards a different room.
Integrated CD player & amp:
http://www.moremusic.nl/primareproducten/cdi10.html
Primare has always been nice quality high-end and their latest offering has class-d... I'm wondering if if might be UCD180's, because there is talk of going balanced just before the class-d input...
Guess we will never know 😀
If this is actually UCD, it would be nice for my living room, as my HT system is moving towards a different room.
ewildgoose said:
Yeah it's connected exactly like that, pin 1 to earth via the chassis, earth to chassis nearest to point of entry, output stage power grounds connected to cap center point which are floating from the chassis ensuring there's no loop with the earlier and small signal stages.
The shield in my case also connects to the chassis of the source, the 100 ohm resistor can be used from pin 1 to chassis to lift the signal ground and minimize ground loop current that forms with an earthed source. Altnernatively you can float the the earth of the source, which for me meant my entire computer, or break the shield at one end. Originally I had lowest background noise by breaking the shield at one end, but noticed my soundcard being connected at earth is very suscetible to feeding any ground noise through to the outputs, so I opted to connect the shield to chassis of both units giving them a clean and low impedance path as a point of common reference, and float the earth of the source/computer, which leaves me with one true earth at the input of the amp, and no way heavy output stage currents can interract with it.
Double isolation should be used for the circuitry which is left to float and it seems a static shield groundstrap tied to an earthed chassis qualifies from what I can make of it anyway.
Rather than float the ground of the computer I could have used an isolation transformer to isolate it, but found that far too expensive to be an option. It depends on the earthing of your source and what your requirements to drive it are, but its' best to avoid a ground loop of any kind with the output stage /input stage and also from your input stage to the source. If you ignore how the source is connected to earth and form any type of loop you're apt to get hum, signal transformers might be required here if you're unwilling to float the source or break the shield at one end.
I also took pin 1/signal ground from both amp modules, soldered them together with a wire and then took that one wire to a single point on the chassis to earth it, ensuring no chassis earth ground loops from one input to the other. You can experiment where the point with the lowest noise is to connect the wire, it doesn't seem to really matter though, usually it's closest to the inputs as possible.
Good luck applying that to your zap.
Regards,
Chris
Chris that's really helpful. To try and keep things on topic: I assume this all also applies to the UCD as much as to the Zaps?
Can I just clarify a couple of points.
- How did you measure the noise levels on your inputs/outputs in order to determine which grounding arrangements worked best? I assume a scope on the speaker outputs and watch what happens? I don't have a scope at the moment, any issues to watch out for buying one of these cheap things from Maplins? What kind of freq range do I need it to run to in order to just check this stuff out? Presumably just audio band? Do you have a scope which does FFT for analysing broadband changes? Is this useful?
- What does double isolation construction mean? (As in practical points) I worry about a wire coming unsoldered and touching the case. Since I am using balanced connections everywhere it seems likely to me that I can just join all the chassis's to mains earth, float the 0V wires from chassis earth and then I am not sure about the small signal grounds, but it seems to me that they could float as well and then only the (computer) source is grounded?
- You said you tied pin1s together, etc. Can you clarify if you do this as described by Bruno by taking amp board to chassis ground, then chassis ground to pin 1, ie rather than tying XLR1 to amp ground and then taking XLR1 to chassis ground? The implication from Bruno that the XLR grounds should go to chassis first, THEN to the amp module ground wires?
TO be honest, I am not far off the ideal arrangement with my Zap so I am a little confused why it is humming quite so badly? The only thing I might be doing wrong is having a loop within the chassis, ie the IEC inlet is wired internally so that it earths the backplate and then I take another wire from safety earth to a second point on the chassis. Also I wired my XLR pin 1 to the XLR chassis earth tab and then took XLR pin1 to the amp ground. The 0V line is disconnected. My 2 power supplies appear not to be symetrical though and this may be causing problems (ie their 0V levels have a PD difference between them.)
Can I just clarify a couple of points.
- How did you measure the noise levels on your inputs/outputs in order to determine which grounding arrangements worked best? I assume a scope on the speaker outputs and watch what happens? I don't have a scope at the moment, any issues to watch out for buying one of these cheap things from Maplins? What kind of freq range do I need it to run to in order to just check this stuff out? Presumably just audio band? Do you have a scope which does FFT for analysing broadband changes? Is this useful?
- What does double isolation construction mean? (As in practical points) I worry about a wire coming unsoldered and touching the case. Since I am using balanced connections everywhere it seems likely to me that I can just join all the chassis's to mains earth, float the 0V wires from chassis earth and then I am not sure about the small signal grounds, but it seems to me that they could float as well and then only the (computer) source is grounded?
- You said you tied pin1s together, etc. Can you clarify if you do this as described by Bruno by taking amp board to chassis ground, then chassis ground to pin 1, ie rather than tying XLR1 to amp ground and then taking XLR1 to chassis ground? The implication from Bruno that the XLR grounds should go to chassis first, THEN to the amp module ground wires?
TO be honest, I am not far off the ideal arrangement with my Zap so I am a little confused why it is humming quite so badly? The only thing I might be doing wrong is having a loop within the chassis, ie the IEC inlet is wired internally so that it earths the backplate and then I take another wire from safety earth to a second point on the chassis. Also I wired my XLR pin 1 to the XLR chassis earth tab and then took XLR pin1 to the amp ground. The 0V line is disconnected. My 2 power supplies appear not to be symetrical though and this may be causing problems (ie their 0V levels have a PD difference between them.)
Hi,
No I didn't do it like Bruno said by wiring the shield to the chassis first and then to the module if I understand you there? I haven't gone into the finner detail of putting it in a case yet, this is evolving as I go and grounding is something I've recently been working at. Right now the signal wires are just soldered directly to the pins.
Perhaps the point of Bruno's method is to have the shields connect at a single point though to keep chassis currents interfering between modules, which is effectively what I've done.
I haven't bothered to try taking noise measurements with my soundcard yet, it would slow me down more than anything. I also dont' have any test/measurement equipment. I just plug it in and put my ear next to the tweeter (if there's no obvious hum), the cleanlyness and level of hiss tells me worlds, breaking the shield at one end was the only way to remove switching noise from the computer's SMPS from it, evident now with the shields connected.
I had opted to break the loop at one end previously because my soundcard was earthed through the computer and by earthing the chassis of the amp it formed a loop. By floating the computer that loop was broken and so was the commong point of reference which brought on very loud hum, that's when I connected the shield at both ends. I still hear the dirty SMPS ground noise this way but there is no ground loop induced noise/hum.
I haven't found much on double isolation standards online that's of much value but what I did find let me to believe there's a number of ways of going about it. A static shield tied to earth offers one form of protection and the earthed case another. I'm sure there's more to be told on that story.
The advantages of doing it this way is it is closest to the ideal with no transformers in the signal path, no sharing of high power currents with that of the signal's ground since the supply is left to float, and the input of the amp, as well as the output, is correctly referenced to the exact same earth as the source at a low impedance point with minimal noise current flow.
For dual supplies I don't see anything needing to be different since each supply is left to float and each module connects to a different floating supply they're even better isolated. The rest of the wiring is probably sound theory for all amps.
With the soundcard earthed through the computer I had to break the loop at one end, I never tried the lift circuit, breaking it at one end was the lowest noise option out of what I tried. I'd assum if your shields connect to the chassis of the computer now and your amplifier chassis is earthed that's the source of your hum as the card's tend to be earthed through computer case as well. Are you using any form of isolation there?
My problems were kind of solved for me though as per what option I went with because the card itself passes through the smallest bit of ground noise, it had to be isolated from earth, which meant I had to connect the shields at both ends. Whether or not you float the power supply doesn't really make much difference in all these loops, its benefit is additional.
I tried to be real specific in all this as to what ground I'm referring to because it becomes confusing real fast. I can imagine if you're taking safety earth from the backplate to a second point in the chassis that would be a loop, if you're connecting the XLR shields and the input signal grounds to a point in this loop it could induce noise. Really there's no use for that wire though, but I can't see that being as bad as the loop which is likely forming through your computer case. Is pin 1 of your card tied to earth through the case?
Regards,
Chris
No I didn't do it like Bruno said by wiring the shield to the chassis first and then to the module if I understand you there? I haven't gone into the finner detail of putting it in a case yet, this is evolving as I go and grounding is something I've recently been working at. Right now the signal wires are just soldered directly to the pins.
Perhaps the point of Bruno's method is to have the shields connect at a single point though to keep chassis currents interfering between modules, which is effectively what I've done.
I haven't bothered to try taking noise measurements with my soundcard yet, it would slow me down more than anything. I also dont' have any test/measurement equipment. I just plug it in and put my ear next to the tweeter (if there's no obvious hum), the cleanlyness and level of hiss tells me worlds, breaking the shield at one end was the only way to remove switching noise from the computer's SMPS from it, evident now with the shields connected.
I had opted to break the loop at one end previously because my soundcard was earthed through the computer and by earthing the chassis of the amp it formed a loop. By floating the computer that loop was broken and so was the commong point of reference which brought on very loud hum, that's when I connected the shield at both ends. I still hear the dirty SMPS ground noise this way but there is no ground loop induced noise/hum.
I haven't found much on double isolation standards online that's of much value but what I did find let me to believe there's a number of ways of going about it. A static shield tied to earth offers one form of protection and the earthed case another. I'm sure there's more to be told on that story.
The advantages of doing it this way is it is closest to the ideal with no transformers in the signal path, no sharing of high power currents with that of the signal's ground since the supply is left to float, and the input of the amp, as well as the output, is correctly referenced to the exact same earth as the source at a low impedance point with minimal noise current flow.
For dual supplies I don't see anything needing to be different since each supply is left to float and each module connects to a different floating supply they're even better isolated. The rest of the wiring is probably sound theory for all amps.
With the soundcard earthed through the computer I had to break the loop at one end, I never tried the lift circuit, breaking it at one end was the lowest noise option out of what I tried. I'd assum if your shields connect to the chassis of the computer now and your amplifier chassis is earthed that's the source of your hum as the card's tend to be earthed through computer case as well. Are you using any form of isolation there?
My problems were kind of solved for me though as per what option I went with because the card itself passes through the smallest bit of ground noise, it had to be isolated from earth, which meant I had to connect the shields at both ends. Whether or not you float the power supply doesn't really make much difference in all these loops, its benefit is additional.
I tried to be real specific in all this as to what ground I'm referring to because it becomes confusing real fast. I can imagine if you're taking safety earth from the backplate to a second point in the chassis that would be a loop, if you're connecting the XLR shields and the input signal grounds to a point in this loop it could induce noise. Really there's no use for that wire though, but I can't see that being as bad as the loop which is likely forming through your computer case. Is pin 1 of your card tied to earth through the case?
Regards,
Chris
Hi
Just done with my UCD180. Pretty nice!🙂
Haven't mod anything yet.
cheers
Coffin
Just done with my UCD180. Pretty nice!🙂
An externally hosted image should be here but it was not working when we last tested it.
Haven't mod anything yet.
cheers
Coffin
That looks very nice! What is the copper foil for exactly? Where did you get the softstart circuit and what else does it do (can't read it on the pic)? And what's your opinion on the sound?
Cheers,
Hans.
Cheers,
Hans.
Hans L said:
Hi
The copper foil is grounded to prevent any interference( don't know exactly if any).
Yes it is a soft start with speaker protector. The circuit is from Taiwan.
The sound of UCD is very transparent, balanced with punch. It can reproduce the sound of your Preamp.
cheers
Coffin 🙂
coffin said:
What you've got there is an electrostatic shield. I think what you were going for was a magnetic/gauss shield.
Usually to wrap the transformer you'd use mu-metal or grain oriented silicon steel for the high permiability.
Magnetic permeability of air m = 1; Magnetic permeability of copper m = 1
Perhaps this link is of some interest to you:
http://www.mushield.com/design-guide.html
You may get more of an advantage in shielding your mains input wires which look to be running in close quarters to both of your modules.
If/when you mod it, you might like to try it without the speaker protection, see how it affects the sound.
Best regards,
Chris
copper shield
Did anyone ever try glue a small piece of copper foil on integrated circuits? Especially the digital ones. The copper absorbs somewhat the radiation from those components. There's no need to connect these shields with ground.
Did anyone ever try glue a small piece of copper foil on integrated circuits? Especially the digital ones. The copper absorbs somewhat the radiation from those components. There's no need to connect these shields with ground.
johnrtd said:
Magnetic permeability of air m = 1; Magnetic permeability of copper m = 1 ?
Do you think if I tried it, it would interfere with my Shatki stone?
stones
That stone does a lot, it weighs somewhat.
Next time I go to Britain I might try out my UcD amp at Stonehenge. Who knows what will happen?
That stone does a lot, it weighs somewhat.
Next time I go to Britain I might try out my UcD amp at Stonehenge. Who knows what will happen?