I have a UCD700 amp in my system wich powers my woofers in a three-way active system. The amp is build of two ucd 700 modules, two HG700 power supplys and two potted toroid transformers with one slow start module, all from Hypex.
The amp has never been very quiet. there is some hum and some jet engine noise coming out of my woofers wich reduced pretty much by using a schaffner mains filter in front of the amp and I had to put a lowpass filter with a 200kHZ cutoff in front of the inputs to take the hiss to a lower level. At least it's not audible from the listening position any more. Other amps, all diy are completely dead quiet, this is by far the noisiest amp I have ever had, it's pretty much useless to power my high effiency mids or highs because then you will hear this noise from a two meter distance as well. One time I tried them on my 118db effiecent BMS tweeters and you could hear the hiss even from out side the room. My other amps are completely quiet even with these tweeters.
I have always felt that when I turn on the amps there's something happening in the highs as well, they start to sound edgy. While it is pretty hard to judge what's really happening because I'm comparing the sound with lows or with lows below 100Hz cut off and that alone makes a big difference, even in the highs.
This week I pulled my osciloscoop out of the closet and started doing some measurements. That was pretty shocking. When I turn on the UCD amp there is a pretty strong ringing trace showing up everywhere in the system.
For example when I measure the powerline of my little Trends TA10.1 amps I measure a perfect triangular 1MHz ripple of about 5mV peak to peak. With the UCD's turned on I measure a ringing ripple of 300kHz of about 300mV peak to peak, that's 60 times more than the power supply ripple itself !!!!!! This ripple decreases a bit when I disconnect every connection from the amp but it is still there. I even tried to power the UCD's from completely different mains group but results are the same.
Next step I put a little choke on the scope to see if it would pick up some RF radiation, and well it does, and pretty much actually, shockingly much. I have measured other equipment with this method as well. My computer puts out very little radiation, the Trends Tripath amps I even less and with my other diy amps no trace at all except for a 50hz trace close to the power transformers, but zero RF
I know this is not a very scientific measuring method so I played around with a AM radio as well. With the UCD's turned off I get a little receiving problems and then only when I get close to the equipment, actually the computer monitor being the worst. With the UCD turned on the whole AM band renders useless troughout the entire room. When I get close to the amp it start to overload the receiver completely and it also starts to destroy the complete FM band as well.
I do not have the equipment to take measurement according to the FCC rules but I'm pretty sure it would miserably fail on all tests.
So question is, what's wrong with it ? It's build completely according to the hypex instructions. Anyone else have experienced the same issues ? Why are these modules so troublesome and so very picky to RF (even to there own RF) and foremost, what can be done about it ?
The amp has never been very quiet. there is some hum and some jet engine noise coming out of my woofers wich reduced pretty much by using a schaffner mains filter in front of the amp and I had to put a lowpass filter with a 200kHZ cutoff in front of the inputs to take the hiss to a lower level. At least it's not audible from the listening position any more. Other amps, all diy are completely dead quiet, this is by far the noisiest amp I have ever had, it's pretty much useless to power my high effiency mids or highs because then you will hear this noise from a two meter distance as well. One time I tried them on my 118db effiecent BMS tweeters and you could hear the hiss even from out side the room. My other amps are completely quiet even with these tweeters.
I have always felt that when I turn on the amps there's something happening in the highs as well, they start to sound edgy. While it is pretty hard to judge what's really happening because I'm comparing the sound with lows or with lows below 100Hz cut off and that alone makes a big difference, even in the highs.
This week I pulled my osciloscoop out of the closet and started doing some measurements. That was pretty shocking. When I turn on the UCD amp there is a pretty strong ringing trace showing up everywhere in the system.
For example when I measure the powerline of my little Trends TA10.1 amps I measure a perfect triangular 1MHz ripple of about 5mV peak to peak. With the UCD's turned on I measure a ringing ripple of 300kHz of about 300mV peak to peak, that's 60 times more than the power supply ripple itself !!!!!! This ripple decreases a bit when I disconnect every connection from the amp but it is still there. I even tried to power the UCD's from completely different mains group but results are the same.
Next step I put a little choke on the scope to see if it would pick up some RF radiation, and well it does, and pretty much actually, shockingly much. I have measured other equipment with this method as well. My computer puts out very little radiation, the Trends Tripath amps I even less and with my other diy amps no trace at all except for a 50hz trace close to the power transformers, but zero RF
I know this is not a very scientific measuring method so I played around with a AM radio as well. With the UCD's turned off I get a little receiving problems and then only when I get close to the equipment, actually the computer monitor being the worst. With the UCD turned on the whole AM band renders useless troughout the entire room. When I get close to the amp it start to overload the receiver completely and it also starts to destroy the complete FM band as well.
I do not have the equipment to take measurement according to the FCC rules but I'm pretty sure it would miserably fail on all tests.
So question is, what's wrong with it ? It's build completely according to the hypex instructions. Anyone else have experienced the same issues ? Why are these modules so troublesome and so very picky to RF (even to there own RF) and foremost, what can be done about it ?
The UCD's are running on their linear supplies, they don't have a smps. The little trends amps wich I sometimes use and just put up for testing because it was easy to take measurments on them are runing from a smps. I also measured the supply rails of my Monarchy Audio dip upsampler and it shows the same traces. They are just everywhere even with no connection to the UCD at all, with connections it gets even worse.
Hi Jan-Peter,
I will see if I can make some pictures one of these days.
I do not use symetrical audio lines and I don't think that the problem lies here because it still puts out a lot of rf even with shorted inputs, well a bit less but still. I'm using some jensen input transformers at the moment wich seem to cure the hiss jsut a little bit.
The power transformers are connected together at the softstart module. wiring is done according to the Hypex manuals. That is, all the wiring is connected to the power supply units and from there to the amp modules. It would be hard to see on a picture because the space is a bit cramped and because of the way the modules are mounted to the heatsink.
I do not have an earth connection in my listening but I have tried it with a earth connection with a long extension wire from the washing machine room, nothing chances
I will see if I can make some pictures one of these days.
I do not use symetrical audio lines and I don't think that the problem lies here because it still puts out a lot of rf even with shorted inputs, well a bit less but still. I'm using some jensen input transformers at the moment wich seem to cure the hiss jsut a little bit.
The power transformers are connected together at the softstart module. wiring is done according to the Hypex manuals. That is, all the wiring is connected to the power supply units and from there to the amp modules. It would be hard to see on a picture because the space is a bit cramped and because of the way the modules are mounted to the heatsink.
I do not have an earth connection in my listening but I have tried it with a earth connection with a long extension wire from the washing machine room, nothing chances
I have tried to make a picture of my scoop.
This shows the trace wich is picked up from just a 20cm piece of wire attached to the scope. The wire is lying on top off my shelf wich is about 20cm from the top of the amp enclosure.
scaling is 1uS/div and 20mV/div.
the part of the signal with the high peaks is steady, the other part modulates with the audio signal. This is the same signal formI measure on the supply rails of the tripath amp and the upsampler.
This shows the trace wich is picked up from just a 20cm piece of wire attached to the scope. The wire is lying on top off my shelf wich is about 20cm from the top of the amp enclosure.
scaling is 1uS/div and 20mV/div.
the part of the signal with the high peaks is steady, the other part modulates with the audio signal. This is the same signal formI measure on the supply rails of the tripath amp and the upsampler.
Attachments
Hello,
the UcD(180) performs definitely better:
UcD_Output_20Hz
UcD_Output_100Hz
UcD_Output_1kHz
UcD_Output_20kHz
UcD_Output_Ripple
As one may see, the residual ripple is around 1Vpp or 350mVrms.
Regards, Timo
the UcD(180) performs definitely better:
UcD_Output_20Hz
UcD_Output_100Hz
UcD_Output_1kHz
UcD_Output_20kHz
UcD_Output_Ripple
As one may see, the residual ripple is around 1Vpp or 350mVrms.
Regards, Timo
Hi Jan-Peter,
Yes I know this kind of voltage is normal for a switched mode amp. I was expecting a more perfect sine wave here. Seems there is some corrolation between the riples in the sinewave and the radiated pulseform I measure.
I will take pictures but it's going to take two weeks beacuse I'm away for two weeks but I wil defenatly come back on this, hope to get everything o.k with the wiring.
with best regards,
Sjef
Yes I know this kind of voltage is normal for a switched mode amp. I was expecting a more perfect sine wave here. Seems there is some corrolation between the riples in the sinewave and the radiated pulseform I measure.
I will take pictures but it's going to take two weeks beacuse I'm away for two weeks but I wil defenatly come back on this, hope to get everything o.k with the wiring.
with best regards,
Sjef
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.