FM interference
I have a couple of UCD400s and also some Zappulse. So the other day when I realised that it's impossible to get reception on my FM radio placed in the front room I initially suspected the Zappulse... Turned that amp off and still I could barely get recption... Turned off the UCD400 amp and .... reception back to normal...
Now I have a small baby... So the baby monitor comes into the front room, but I can't place it within several meters of the speaker cables without causing massive interference...
Any thoughts on why the UCD400 is spewing out such large amounts of RFI apparently via the speaker wires? (There are pictures of my amp in previous pages on this thread I think, nothing unconventional). I have a couple more amps to build and I would like to refit the old amps and new ones with whatever it takes to fix this.
Can someone recommend an appropriate ferrite from farnel or RS that should do it? Can I get away with popping the ferrite around the wires inside the amp case? (Do you need a complete loop around the ferite or does it just clamp on?)
Thanks
I have a couple of UCD400s and also some Zappulse. So the other day when I realised that it's impossible to get reception on my FM radio placed in the front room I initially suspected the Zappulse... Turned that amp off and still I could barely get recption... Turned off the UCD400 amp and .... reception back to normal...
Now I have a small baby... So the baby monitor comes into the front room, but I can't place it within several meters of the speaker cables without causing massive interference...
Any thoughts on why the UCD400 is spewing out such large amounts of RFI apparently via the speaker wires? (There are pictures of my amp in previous pages on this thread I think, nothing unconventional). I have a couple more amps to build and I would like to refit the old amps and new ones with whatever it takes to fix this.
Can someone recommend an appropriate ferrite from farnel or RS that should do it? Can I get away with popping the ferrite around the wires inside the amp case? (Do you need a complete loop around the ferite or does it just clamp on?)
Thanks
Ewildgoose,
We have found out that the screws we used did not had always had a good electrically connection with the ground. The reason was the blue anodizing, and this creates a hard non conductive oxide.
We use now a different screw to avoid this problem.
If you like you may ship your modules back and we can check this.
Or you replace the crews by yourself, please send me an email and I will give you the details which screw you have to use.
Farnell has Ferrite Clamps on stock, check theire catalogue.
Regards,
Jan-Peter
We have found out that the screws we used did not had always had a good electrically connection with the ground. The reason was the blue anodizing, and this creates a hard non conductive oxide.
We use now a different screw to avoid this problem.
If you like you may ship your modules back and we can check this.
Or you replace the crews by yourself, please send me an email and I will give you the details which screw you have to use.
Farnell has Ferrite Clamps on stock, check theire catalogue.
Regards,
Jan-Peter
JP, some questions here regarding the UCD400's.
You say you have the blue T piece on the PCB grounded to GND. Now if you connect 2 UCD's don't you have a ground loop? I have my GND at the centre tap of the transformer=0V.
Regarding the caps...they are not low Z caps. Can I replace them later with low Z caps? Specially the input caps. C23 and 24 just look like common caps. Do you use a cap. in the feedback loop also? C25? No problem if I remove the cap. and connect an integrator output there? I never like caps in the audiopath, they wear/dry out. C14 and 26....replacable with low Z caps?
JP, this is in no way critizyzing the design as like you said you wanted to keep the design reasonably priced. But...if there is a possibility to polish it...maybe I am dead wrong now, so please add your comment.
You say you have the blue T piece on the PCB grounded to GND. Now if you connect 2 UCD's don't you have a ground loop? I have my GND at the centre tap of the transformer=0V.
Regarding the caps...they are not low Z caps. Can I replace them later with low Z caps? Specially the input caps. C23 and 24 just look like common caps. Do you use a cap. in the feedback loop also? C25? No problem if I remove the cap. and connect an integrator output there? I never like caps in the audiopath, they wear/dry out. C14 and 26....replacable with low Z caps?
JP, this is in no way critizyzing the design as like you said you wanted to keep the design reasonably priced. But...if there is a possibility to polish it...maybe I am dead wrong now, so please add your comment.
No there is used a small SMD capacitor to connect the heatsink to ground. So you cannot have a groundloop.
The 470uF capacitor are for a special reason no Low Z, when you use Low Z here you can get problems with ringing.
You can remove the inputcaps, but you must be shure that your source does not have any DC. Or you will have problems with Power Supply Pumping Effect (the powerrails are pumped up to an higher voltage). There are no coupling caps in the feedback circuit. And because of the modulator is potted you do not have acces to the integrator capacitor.
Jan-Peter
I just bought a pair of 43v 725VA transformers because I got a great deal on them. After reading the description more carefuly, it looks like I will have to use both of them to get 55vDC +/- that the modules need. I was unaware that transformers were made for AC that didn't supply for both rails +&-, but this is my first project and I am learning as I go.
I was hoping to go dual mono on the amps. Will I need to get
the second pair?
Are there any problems with using seperate transformers for each rail?
Would it be better to scrap that Idea and get two new transformers that would supply both rails?
I am kind of on a budget, so if there are no problems with the two transformer idea, I might just go stereo for now.
Thanks for any help on this
Larry
I was hoping to go dual mono on the amps. Will I need to get
the second pair?
Are there any problems with using seperate transformers for each rail?
Would it be better to scrap that Idea and get two new transformers that would supply both rails?
I am kind of on a budget, so if there are no problems with the two transformer idea, I might just go stereo for now.
Thanks for any help on this
Larry
lne937s said:
Two transformers will work fine. Just make sure the secondaries are in the appropriate phase if you're using a single bridge rectifier.
If you already have the four capacitors for a dual-mono rig, instead of running them in parallel try a CRC filter setup, with maybe a 50 watt 0.2 to 0.3 ohm resistor between the two caps. I picked up the idea from the Pass crowd here; it simulates beautifully so it's going into my UcD rig.
Cheers,
Francois.
Francois,
Thanks for the quick reply. I was planning on using a dual bridge rectifier setup, but how do I check for phase on AC? Put a light bulb in between them and if it lights up they are opposite phase?
Also, I was wondering if the two transformers (one per rail) equal a total of 725VA or 1450VA? 725VA seems to be on the low end of what people are recommending, so I might go ahead and buy the second pair.
I guess it really is necessary to use both transformers for each power supply. I figure I may want to add a shield between them, and probably will just to be on the safe side. They are pretty big (34 lb per pair or ~14kg), so the amp will be well anchored.
I already have four big (30,000 UF, 75v each) caps, but I am unsure what you are recommending. Is it to wire them in series for each rail to ground with small resistors between them?
Thanks in advance for any help you can give me and for being patient with a novice.
Larry
Thanks for the quick reply. I was planning on using a dual bridge rectifier setup, but how do I check for phase on AC? Put a light bulb in between them and if it lights up they are opposite phase?
Also, I was wondering if the two transformers (one per rail) equal a total of 725VA or 1450VA? 725VA seems to be on the low end of what people are recommending, so I might go ahead and buy the second pair.
I guess it really is necessary to use both transformers for each power supply. I figure I may want to add a shield between them, and probably will just to be on the safe side. They are pretty big (34 lb per pair or ~14kg), so the amp will be well anchored.
I already have four big (30,000 UF, 75v each) caps, but I am unsure what you are recommending. Is it to wire them in series for each rail to ground with small resistors between them?
Thanks in advance for any help you can give me and for being patient with a novice.
Larry
Each transformer will give you one half of the required supply. You will end up with a supply capacity of 1450VA if you use both transformers which is rather a lot of overkill, but if you pick them up cheap it won't hurt at all.
For the phase, I would have thought the transformers all have the same colour code for the wires. If you connect a certain colour to one point on the bridge rectifier, simply connect the same colour on the other transformer to the same point on the other bridge rectifier.
You could run 2 UcD400 off this supply great. I think DSP_Geek's idea was to use separate cap banks for each amp. The CRC filter is simply: first cap across rail as per normal, attach one end of a 0.1 ohm resistor to the +ve terminal of the cap and connect the other end of the resistor to the +ve terminal of another cap. Then connect the -ve terminal of this new cap to ground.
For the phase, I would have thought the transformers all have the same colour code for the wires. If you connect a certain colour to one point on the bridge rectifier, simply connect the same colour on the other transformer to the same point on the other bridge rectifier.
You could run 2 UcD400 off this supply great. I think DSP_Geek's idea was to use separate cap banks for each amp. The CRC filter is simply: first cap across rail as per normal, attach one end of a 0.1 ohm resistor to the +ve terminal of the cap and connect the other end of the resistor to the +ve terminal of another cap. Then connect the -ve terminal of this new cap to ground.
Larry, if you're using a double bridge setup (one for +, one for -), then you don't really need to worry about transformer phase. It's only when you use a single bridge with two xformers that you want to make sure one leg goes up while the other goes down. What Richie says about colour codes is valid, although for a single bridge case you want to wire (assuming the secondary CC is red/green) xformer 1 red to one end of the bridge, xformer 1 green with xformer 2 red and call that center tap, and xformer 2 green to the other side of the bridge. Your idea of wiring them together and seeing if a bulb lights is a good one, although at 86 volts I wouldn't expect it to get too bright.
Richie explained the CRC proposal better than I did. You want to take the power from the terminal of the second cap, which acts as a low-pass RC filter after the first cap. It eliminates a fair bit of residual grunge, especially if you slap in a 10 uF film cap (if you don't have those lying around like I do, try a 220 uF electro) across the second electrolytics to compensate for the series inductance the big boys tend to have. You lose a volt or so in the resistors at full power, but that's a small fraction of a dB. They do dissipate some power, so you want to make sure they can take the heat. Something like these are perfect if you mount them on a metal surface:
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=4664&item=3873605006&rd=1
Richie's also right about two 725 VA xformers providing 1450 VA of capacity. You don't need any more for a couple of UcD400s, especially since you'll be running them at +- 55 V.
Richie explained the CRC proposal better than I did. You want to take the power from the terminal of the second cap, which acts as a low-pass RC filter after the first cap. It eliminates a fair bit of residual grunge, especially if you slap in a 10 uF film cap (if you don't have those lying around like I do, try a 220 uF electro) across the second electrolytics to compensate for the series inductance the big boys tend to have. You lose a volt or so in the resistors at full power, but that's a small fraction of a dB. They do dissipate some power, so you want to make sure they can take the heat. Something like these are perfect if you mount them on a metal surface:
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=4664&item=3873605006&rd=1
Richie's also right about two 725 VA xformers providing 1450 VA of capacity. You don't need any more for a couple of UcD400s, especially since you'll be running them at +- 55 V.
If you're just starting out, I'd recommend investing in a good DVM. There are some nice little Chinese jobs which have the same feature set as a Fluke for about one-third the price, good enough for light duty.
You can try simulating your supply with this:
http://www.duncanamps.com/psud2/
which will give you a decent idea of what to expect.
You also have a goodly amount of filter capacitance, so you might want to look into a soft start circuit to reduce your inrush current. Your rectifiers and capacitors will thank you.
Cheers and happy soldering,
Francois.
You can try simulating your supply with this:
http://www.duncanamps.com/psud2/
which will give you a decent idea of what to expect.
You also have a goodly amount of filter capacitance, so you might want to look into a soft start circuit to reduce your inrush current. Your rectifiers and capacitors will thank you.
Cheers and happy soldering,
Francois.
WJMackintosh; said:
This is a very difficult topic for SMPS and Class-D amplifiers. In the second version of the UcD400 we have solved this matter to a very low level of EMI by improved PCB design and a new design of outputcoil.
With a good setup (metal case) you can use your FM tuner in combination with the UcD400.
Regards,
Jan-Peter
Hello Jan-Peter,
is it possible, to share the advanced output coil data with us? So anybody, who bought an early UcD180 or UcD400 could change his coils to prevent from these interferences, if necessary.
Second question: Was it possible to lower the output coil temperature of the UcD400 by the new design?
Thank you in advance,
Regards, Timo
is it possible, to share the advanced output coil data with us? So anybody, who bought an early UcD180 or UcD400 could change his coils to prevent from these interferences, if necessary.
Second question: Was it possible to lower the output coil temperature of the UcD400 by the new design?
Thank you in advance,
Regards, Timo
We have an Advantest R3131A Spectrum Analyzer (http://green.advantest.co.jp/techinfo_e/www_e/product_e/R3131A_e.html) with a Current Clamp.
It's not possible update the outputcoil with this newer coil. The size and connection is completly different, so it will NOT fit in the PCB.
The EMI of the UcD180 and UcD400 is below the official limits. If you still have problems and like to reduce the EMI you can always use a ferrite clamp and ofcourse a good metal case.
Actually the EMI of the new version of the UcD400 is a little bit ABOVE the noise level of the Advantest R3131A.......
Jan-Peter
...hm, the outputcoil temperature is no problem in every UcD version. There are already more as >1.000.000,00 channels produce with the same coil as in the UcD180. When this is not ok, Philips would changed this, don't you think so??
Besides this the new coil on the UcD400 is indeed reduced in heat, hower that was not the reason for this new coil. Philips has designed this new coil for more outputcurrent and lower EMI.
Regards,
Jan-Peter
Besides this the new coil on the UcD400 is indeed reduced in heat, hower that was not the reason for this new coil. Philips has designed this new coil for more outputcurrent and lower EMI.
Regards,
Jan-Peter
Jan-Peter, it's unbelievable!! 100 millions UcD's which have the same coil, and seems there is some quantity with other coils.. Where is this ocean of the UcD's, on the stock? All this UcD's put one by one, would be way from you to me and back again, i'm sure..
100000000 pieces it isn't wrong?