Relais in speaker output
Peter Jan,
I can understand that you want to avoid anything in the signalpath that can be avoided. You never know what it does.
But, on the other hand, would a quality relais contact do more harm to the signal than connectors and cables? You know, some people spend half a fortune on speaker cables and some use cheap UTP network cable to connect their speakers. Hence my question, if you use quality relais, could you hear the difference?
Most DC protections also connect the speaker to the amplifier after a few seconds delay, giving the amplifier the time to set a stable dc biaspoint. I guess this is not needed for UCD since there is no dc biaspoint.
regards,
Ludo
Peter Jan,
I can understand that you want to avoid anything in the signalpath that can be avoided. You never know what it does.
But, on the other hand, would a quality relais contact do more harm to the signal than connectors and cables? You know, some people spend half a fortune on speaker cables and some use cheap UTP network cable to connect their speakers. Hence my question, if you use quality relais, could you hear the difference?
Most DC protections also connect the speaker to the amplifier after a few seconds delay, giving the amplifier the time to set a stable dc biaspoint. I guess this is not needed for UCD since there is no dc biaspoint.
regards,
Ludo
Hi,
I think that Bruno's circuit works well enough, especially if he uses 6V rated coil relay with appropriate series resistor. In my experience relays tipically drop off at approx half the rated coil voltage. But I still think that even in this case there is a good chance of welding the contacts together. Also knowing that lifetime of the amplifier is usually diminished because of the production faults and not engeneering ones, I think that protection circuits operate more oftenly than designer originally assumes. Problem with Bruno's circuit is that welding of the relay contacts will come unnoted after the amplifier is repaired, and also that not all repairs succede at the first time, so it is likely to have more than one operation when amplifier is at fault.
For DIY use it may be perfectly useful provided one leaves a sticker on the relay saying "Check me".
Best regards,
Jaka Racman
I think that Bruno's circuit works well enough, especially if he uses 6V rated coil relay with appropriate series resistor. In my experience relays tipically drop off at approx half the rated coil voltage. But I still think that even in this case there is a good chance of welding the contacts together. Also knowing that lifetime of the amplifier is usually diminished because of the production faults and not engeneering ones, I think that protection circuits operate more oftenly than designer originally assumes. Problem with Bruno's circuit is that welding of the relay contacts will come unnoted after the amplifier is repaired, and also that not all repairs succede at the first time, so it is likely to have more than one operation when amplifier is at fault.
For DIY use it may be perfectly useful provided one leaves a sticker on the relay saying "Check me".
Best regards,
Jaka Racman
Ludo,
In my opinion the best way would be the solution of Bruno, use the ON connection when the relay is off and switch the relay on in a case of DC. The most importend is to have a safe situation, the relaiscontact can degreed over time, so what happen with your outputrelais in the loudspeakerline after a few years?
Ofcourse everybody is free to use an outputrelay in the loudspeakeroutput. On of the nice thing of the UcD is that you don't need them to avoid plops or other small noises during switching on
Cheers,
Jan-Peter
In my opinion the best way would be the solution of Bruno, use the ON connection when the relay is off and switch the relay on in a case of DC. The most importend is to have a safe situation, the relaiscontact can degreed over time, so what happen with your outputrelais in the loudspeakerline after a few years?
Ofcourse everybody is free to use an outputrelay in the loudspeakeroutput. On of the nice thing of the UcD is that you don't need them to avoid plops or other small noises during switching on
Cheers,
Jan-Peter
That is important information. So a time delay to energise the relay is not needed.
There is a difference: Bruno's solution works perfectly but is not fail save: if a failure occurs and for some reason the relais does not break the power, the speakers are unprotected and get the full DC. Any fault in the protection circuitry would remain unnoticed until you need it. And then it is too late. In reality chances that this happens are very little but you never know.
If you put a relais NO in the speaker wires, the relais must operate before you can get anything to your speakers. If something is wrong (relay is not energised, degraded contacts, ...) you would notice it since normal operations of the amplifier is disturbed. This is why I would like to know about the eventual influence of the contacts on sonics. Could you hear it? Is the influence significant compared to the influence of connectors and cables?
regards,
Ludo
Hi,
Best regards,
Jaka Racman
The way I see is this: In medical standards, fuses are supposed to blow only over twice the rated current. If fuse is rated for rated output power, then DC output current will not be twice the rated current because of power supply sagging and fuse could theoretically not blow. Also fuse chosen for 4 ohm load will not blow with 8 ohm load. And besides that fuses come in relatively small number of values, so it is not uncommon to overspecify. I have no idea about sonic impact of fuses.
Best regards,
Jaka Racman
Thank you for that reply.
Fuses are problematic but I thought it might be a simple option.
Considering the nature of this beast, (DIY) we dont' need to worry about selecting a fuse for universal purposes, the builder would be in charge of selecting the right one for their rated load.
It would be a different story for anyone wanting to do OEM with them though.
I can comment on the sonic quality because my speakers themselves are fused.
They degrade over time, continually heating and cooling seems to oxidize them at which point, sonically speaking, they're seriously degrading the sound quality, no longer being a very good conductor. I replace them every so often for that reason, and the quality is brought right back.
Aside from that I've never tried bypassing the fuses so I also cannot attest as to weither or not they are transparent while still in good condition, but they are essentially a wire.
I would think short circuit current caused from a mosfet failure would be far higher than your average operating current so if you selected the fuse just above one that sags at full power it would be sure to blow in a failure condition, but it would take your speaker with it I guess, I understand it doesn't take much DC to cook a voice coil.
So, fuses aren't a good option
Thanks
Fuses are problematic but I thought it might be a simple option.
Considering the nature of this beast, (DIY) we dont' need to worry about selecting a fuse for universal purposes, the builder would be in charge of selecting the right one for their rated load.
It would be a different story for anyone wanting to do OEM with them though.
I can comment on the sonic quality because my speakers themselves are fused.
They degrade over time, continually heating and cooling seems to oxidize them at which point, sonically speaking, they're seriously degrading the sound quality, no longer being a very good conductor. I replace them every so often for that reason, and the quality is brought right back.
Aside from that I've never tried bypassing the fuses so I also cannot attest as to weither or not they are transparent while still in good condition, but they are essentially a wire.
I would think short circuit current caused from a mosfet failure would be far higher than your average operating current so if you selected the fuse just above one that sags at full power it would be sure to blow in a failure condition, but it would take your speaker with it I guess, I understand it doesn't take much DC to cook a voice coil.
So, fuses aren't a good option
Thanks
Bruno, I have a few questions about power output. The Ucd400 generates 400 watts from 60 volt rails, which makes sense to me as 60 V peak = 42 Vrms, and E^2/R implies about 450 W with no losses whatsoever, so 400 W is in the right ballpark.
I don't get the numbers for the Ucd180, however, since 50 V rails imply 35 Vrms, and by E^2/R power output should be close to 300W, yet we see about 60% of that, which implies the losses are on the order of 40%.
Obviously I'm missing something here. Could you please explain what I didn't understand?
Thanks,
Francois.
I don't get the numbers for the Ucd180, however, since 50 V rails imply 35 Vrms, and by E^2/R power output should be close to 300W, yet we see about 60% of that, which implies the losses are on the order of 40%.
Obviously I'm missing something here. Could you please explain what I didn't understand?
Thanks,
Francois.
Power ratings keep into account the fact that a normal power supply doesn't hold up to maximum voltage when fully loaded. To read maximum power output from the maximum supply voltage doesn't work, simply because the power rating actually presumes that the supply will drop!DSP_Geek said:
If you powered the UcD off a stiff +/-50V supply and tried to get 300W from it thus, the protection would act. The output stage is designed to handle about 9.5A peak. It is not unusual for a supply that delivers 50V (max) at idle to come down all the way to 38V when the amp is going full blast.
The fact that the maximum voltage spec of the 400W module happens to yield a figure close to 400W is actually indicative of the voltage spec not being fully fixed yet (should increase).
Efficiency of either 180W or 400W model is on the order of 92% (93 typ).
Bruno Putzeys said:
Ah, now it's clear. Thanks.
I have the best of both worlds - a fairly stiff supply and a speaker comfortably over 4 ohms in the range of interest, so according to the above I should see most of the supply delivered to the load.
In other news, there's an old trick I'm surprised no-one's mentioned trying, and it's easy to do with Ucd modules since they have balanced inputs. Here it is. Stereo signals, especially in the lows, are actually fairly common-mode, which means that one supply rail is getting hit harder than the other at any given time. On the other hand, if one channel is connected in phase (non-inverting) and the other in antiphase (inverting), and subsequently the speaker leads are also switched in phase, then the load is evenly distributed on both rails.
Has anyone tried this recently?
Francois.
Quite impressive modules !!!
Hi Jan-Peter,
I've read most of this thread and the info on your website and must admit I am impressed by the specs of this small 'monster' and the feedback given by those of this forum that have them up and running.
There are a few questions I haven't found answers for, it's about the max continuous current and peak current the UcD180 and UcD400 can deliver. Does bridging allow me to double those numbers?
Furthermore I wonder if you measured the speed of the amps (time to needed to generate the peak current).
Henk
Hi Jan-Peter,
I've read most of this thread and the info on your website and must admit I am impressed by the specs of this small 'monster' and the feedback given by those of this forum that have them up and running.
There are a few questions I haven't found answers for, it's about the max continuous current and peak current the UcD180 and UcD400 can deliver. Does bridging allow me to double those numbers?
Furthermore I wonder if you measured the speed of the amps (time to needed to generate the peak current).
Henk
The Carver cube used the phase trick as well. While it is primarily done with switching amps because of supply pumping it is done with normal amps to get a little more peak output power.
I have done this as well with a conventional power amp about 15 years ago (after having read about it's use within the cube).
Regards
Charles
I have done this as well with a conventional power amp about 15 years ago (after having read about it's use within the cube).
Regards
Charles
Hi guys,
I apologise if this has been discussed earlier in the thread, it's just doing my head in trying to read it all at once!
I've got a few of questions.
-Has anyone actually come to a conclusion on a comparison between this and a Zappulse based amplifier?
-Has anyone compared this to something like a BelCanto Evo amplifier?
-Has anyone compared this to a good Gainclone amplifier?
-What would be the optimum size transformer for best sound quality? I plan to be using a separate transformer per channel/module, and power isn't a huge issue (within reason) as my PHL/Focal drivers are very efficient.
-What would be the recommended amount of capacitance for these modules, the Zappulse amplifier I was thinking of building was going to use 12 4700uF capacitors (6 per channel).
Apologies for all the questions, I'm just starting to really get into this sort of stuff, and my mind is buzzing with all these questions so I can't get any study done!
Thanks,
Nathan
I apologise if this has been discussed earlier in the thread, it's just doing my head in trying to read it all at once!
I've got a few of questions.
-Has anyone actually come to a conclusion on a comparison between this and a Zappulse based amplifier?
-Has anyone compared this to something like a BelCanto Evo amplifier?
-Has anyone compared this to a good Gainclone amplifier?
-What would be the optimum size transformer for best sound quality? I plan to be using a separate transformer per channel/module, and power isn't a huge issue (within reason) as my PHL/Focal drivers are very efficient.
-What would be the recommended amount of capacitance for these modules, the Zappulse amplifier I was thinking of building was going to use 12 4700uF capacitors (6 per channel).
Apologies for all the questions, I'm just starting to really get into this sort of stuff, and my mind is buzzing with all these questions so I can't get any study done!
Thanks,
Nathan