Fram, whats the story with your PCB boards? Are they working well?
Having never built a PCB before can someone suggest a cheap place where I might get his boards built, with shipping to the UK?
Fram, I'm curious as to whether you have a list of the other parts you used to populate the board
Good luck
Having never built a PCB before can someone suggest a cheap place where I might get his boards built, with shipping to the UK?
Fram, I'm curious as to whether you have a list of the other parts you used to populate the board
Good luck
Hi ewildgoose! 😉
Just look at this: (4 ways with UCD180)
The same amp. with 4 UCD400 is in progress (just missing two 8 caps PSU : this design is in progress just like my last release of 4 caps PSU)
I received UCD last week and the last parts this friday (XLR, powercon and molex connectors) except class X2 caps for the soft start.... 🙁
I'm actually in progress to build cable with XLR and molex connector.Tomorrow i'll finish this project, and hope it will run this next week (i'm going to find X2 caps quicky!!
)
i'll let you know some news very very soon.
😉
Just look at this: (4 ways with UCD180)
An externally hosted image should be here but it was not working when we last tested it.
The same amp. with 4 UCD400 is in progress (just missing two 8 caps PSU : this design is in progress just like my last release of 4 caps PSU)
I received UCD last week and the last parts this friday (XLR, powercon and molex connectors) except class X2 caps for the soft start.... 🙁
I'm actually in progress to build cable with XLR and molex connector.Tomorrow i'll finish this project, and hope it will run this next week (i'm going to find X2 caps quicky!!
)i'll let you know some news very very soon.
😉
Cool. Why go for an 8 cap board rather than two 4 cap boards? Space seems to be the main reason?
...Hmm, actually in retrospect also the number of torroids would be an issue...
I'm still interested in building a couple of your 4 cap boards for my own use. Suggestions on where to get them made (for uk delivery) would be appreciated.
Anyone else want to build some? I can't believe everyone else is happy hardwiring this stuff together? A decent PCB must be better from a safety point of view...? (wires coming unsoldered and touching where they ought not...)
How many people leave their home built stuff on and unattended for reasonable periods of time?
...Hmm, actually in retrospect also the number of torroids would be an issue...
I'm still interested in building a couple of your 4 cap boards for my own use. Suggestions on where to get them made (for uk delivery) would be appreciated.
Anyone else want to build some? I can't believe everyone else is happy hardwiring this stuff together? A decent PCB must be better from a safety point of view...? (wires coming unsoldered and touching where they ought not...)
How many people leave their home built stuff on and unattended for reasonable periods of time?
Hi,
One problem with big fat caps on a PCB in a high SPL environment, the shake rattle and roll can in time cause the solder joint to fail, certainly if poorly soldered, a common problem in cheaply assembled OEM stuff.
This is why they're glued down to the PCB, in my amp that had this problem, each cap only had about three drops of glue! and only 1 of those drops actually made it to the cap, the other "drops" were more like smears, and 1/4" off the mark, must have been a friday.
In a real good case it'll even lift the traces off the board, I guess if they aren't properly seated tight to the board, glued, traces too small, caps too big.
Now if you bolt the cap down with a nice clamp, and use screw terminals on a nice computer grade cap..... what can go wrong?
Anyway, something to consider.
One problem with big fat caps on a PCB in a high SPL environment, the shake rattle and roll can in time cause the solder joint to fail, certainly if poorly soldered, a common problem in cheaply assembled OEM stuff.
This is why they're glued down to the PCB, in my amp that had this problem, each cap only had about three drops of glue! and only 1 of those drops actually made it to the cap, the other "drops" were more like smears, and 1/4" off the mark, must have been a friday.
In a real good case it'll even lift the traces off the board, I guess if they aren't properly seated tight to the board, glued, traces too small, caps too big.
Now if you bolt the cap down with a nice clamp, and use screw terminals on a nice computer grade cap..... what can go wrong?
Anyway, something to consider.
I can see that you use a kind of CPU cooler on the self.
What is the model?
Can this be a problem to place a metallic cooler on the self?
Hi steph
This is Zalman ZM-RHS1 kit aluminium heatsink which is in fact a kit for graphic card chip.😀
in this kit you have 2 kinds of heatsink (around 7 euros for 1 kit)
Zalman heatsink
😉
This is Zalman ZM-RHS1 kit aluminium heatsink which is in fact a kit for graphic card chip.😀
in this kit you have 2 kinds of heatsink (around 7 euros for 1 kit)
Zalman heatsink
😉
Are those little sinks enough for Ucd180? wow 
Another question that comes to mind when designing power amps is the chassis. I know where to get all my components now but build it into/onto what?
Are there any good 43cm/17" audio chassis to be found out there?
Another question that comes to mind when designing power amps is the chassis. I know where to get all my components now but build it into/onto what?
Are there any good 43cm/17" audio chassis to be found out there?
Audiokit.it has some nice ones from Hifi 2000
http://www.audiokit.it/ITAENG/Cabinet/HI-FI2000/HI-FI2000.htm
I have quite a few of these and am very happy. The postage can be a little steep, but the cases themselves are pretty good for the price (base is a little flimsy and made of steel though).
There is a UK place which does a similar kind of case, but with alu top and bottom. Forget the URL now.
Lots of other places of course, and you can get them built to order fairly cheaply as well.
http://www.audiokit.it/ITAENG/Cabinet/HI-FI2000/HI-FI2000.htm
I have quite a few of these and am very happy. The postage can be a little steep, but the cases themselves are pretty good for the price (base is a little flimsy and made of steel though).
There is a UK place which does a similar kind of case, but with alu top and bottom. Forget the URL now.
Lots of other places of course, and you can get them built to order fairly cheaply as well.
Capacitor choice
In response to a post elsewhere by Bruno where he states he could not find a correlation between (psu) capacitor sound quality and specifications as opposed to a correlation between sound quality and size.
In light of that remark, how much of an improvement may one expect when using two capacitors in parallel instead on a single (identical) one?
The reason I'm asking is I had a look at Siemens Sikerol (B41550)and wet my pants while calculating the total amount needed for 6 ucd180 stable into <4ohm. Suffice to say, it was a 'bit' more than what the price of the modules... I had a look at loads of caps and found one (EPCOS B41456, also Siemens, right?) that was reasonable in price and had better specs than a single sikerol when using two in par. So there you go...
Thanks,
Hans.
In response to a post elsewhere by Bruno where he states he could not find a correlation between (psu) capacitor sound quality and specifications as opposed to a correlation between sound quality and size.
In light of that remark, how much of an improvement may one expect when using two capacitors in parallel instead on a single (identical) one?
The reason I'm asking is I had a look at Siemens Sikerol (B41550)and wet my pants while calculating the total amount needed for 6 ucd180 stable into <4ohm. Suffice to say, it was a 'bit' more than what the price of the modules... I had a look at loads of caps and found one (EPCOS B41456, also Siemens, right?) that was reasonable in price and had better specs than a single sikerol when using two in par. So there you go...
Thanks,
Hans.
And I think the size of those things is larger than 6 modules yeah? They are the really huge chunky things right?
The Sikorel, 63V, 10000uF is 107x51mm, so it's big, but not as huge as 6 modules 😉. It's EUR 37,35 overhere.
ESRmax = 18 mOhm and Zmax = 14mH
The Epcos I'm look is also 63V, 10000uF, but 'only' 81x36mm and costs EUR 11,78.
ESRmax = 30 mOhm and Zmax = 30mH
I was planning 2 per rail... Anyone have an opinion on my original question on using two caps parallel?
ESRmax = 18 mOhm and Zmax = 14mH
The Epcos I'm look is also 63V, 10000uF, but 'only' 81x36mm and costs EUR 11,78.
ESRmax = 30 mOhm and Zmax = 30mH
I was planning 2 per rail... Anyone have an opinion on my original question on using two caps parallel?
Look at this supply
Hello all,
From the UcD180 thread, I found this company (White Noise Audio), They seem to have a very interesting active PSU that looks nice for an UcD power supply. It has a shutdown input, so likely you can connect that to some DC detection circuit to shutdown the power supply in case of DC at the amp output.
Thinking seriously of trying this one.
http://www.wnaudio.com/active.html
Anybody tried this supply?
Gertjan
Hello all,
From the UcD180 thread, I found this company (White Noise Audio), They seem to have a very interesting active PSU that looks nice for an UcD power supply. It has a shutdown input, so likely you can connect that to some DC detection circuit to shutdown the power supply in case of DC at the amp output.
Thinking seriously of trying this one.
http://www.wnaudio.com/active.html
Anybody tried this supply?
Gertjan
Hello,
All I don't like about it at this point is that I dont' see anything there regarding the price, if you find out what it is let us know please.
It kind of sounds like it might be using a capacitance multiplier.
Anyone interested can read about them here:
http://sound.westhost.com/project15.htm
Regards,
Chris
All I don't like about it at this point is that I dont' see anything there regarding the price, if you find out what it is let us know please.
It kind of sounds like it might be using a capacitance multiplier.
Anyone interested can read about them here:
http://sound.westhost.com/project15.htm
Regards,
Chris
classd4sure said:
Yves Smolders said:
You can download a catalog on their web-site. The catalog has additional information and a pricelist as well. They state that the supply is high efficient, not like a shunt regulator, so they must be doing some switching.
About the 10 amps, well, if you have a symmetrical supply with +/- 60V, so 120V in total, then that 10 amps is a power of 1200Watt in total that it can deliver constantly. I was thinking of adding some extra decoupling caps after that supply (assuming that power supply is stable with extra caps at the output) so that short current peaks during the top of the sinewave that are bigger than 10 amp can be delivered by those caps. If you stay under 40V peak voltage, then you will not reach the 10amp. 40V peak in 4 Ohm is still a respectable 200W. For the woofers, I'm thinking of bridging two UcD400 modules. In 4 Ohm load, I then exactly should be OK with 10 amp per supply rail, since the same amount of power will be drawn from both rails. Then you can get probably close to 800W in 4 Ohm with those supplies, should be OK.
Best regards
GErtjan
Gertjan,
Build the beast and let me know 🙂
I'm going to build my stereo model with a classic transformer 😀
Build the beast and let me know 🙂
I'm going to build my stereo model with a classic transformer 😀
ghemink said:
A shunt regulator is basically a class-a regulator, so very inefficient. A series regulator is what most regulators are and they are much more efficient.
The Active PSU does not do any switching. Is is simply a little more efficient than a series regulator because the voltage differential requirement is not as great. However, as I said previously, IMO the benefit is minimal as you still need heatsinks etc. and the only real improvement is ripple and impedance. You do not get more charge storage.
ghemink said:
Yes, but that is peak to peak power, which is only 300W RMS. It also assumes no losses in the amp.
ghemink said:
Adding caps will not increase the current capacity of the supply, merely reduce the ripple and impedance. I get the impression that the Active PSU current limit is by current sensing, thus the limit is fixed.
ghemink said:
When you bridge an amp each amp sees half the impedance. Thus if each module is to supply 400W into 2R this demands a current of 14.1A (just under the 15A UcD400 limit) Limiting yourself to a 15A supply this is just achievable. Having said that, as the average current waveform of class-d is less than the average load current, you may have some headroom.
Hi Gertjan,
It doesn't use any type of switching, it just makes a small capacitance "appear" like a much bigger capacitance (X the gain of the transistor).
This has the effect of significantly reducing the ripple, just like a 100 000uF cap might. However, unlike a 100 000uF cap, it just can't store the same level of energy, and will readily sag more under load, further reducing power.
The reason it sags and is also so efficient is because it isn't regulating anything at all, and has less reserve.
If you think you have a hum problem or poor psrr it might be an interesting option to play with, but you have to be aware of what you're giving up, a nice stiff supply that can really deliver on demand, which is why unregulated raw DC supplies are good, and why they're still so commonly used in high end audio.
Maybe if you never plan on turning it up at all, it would be just the thing.
As far as the price of them goes, well, they closely match or slightly exceed the cost of the class-d modules themselves, and for what? I'm not really impressed with that. Although as far as what it is, it seems decent.
Also, the shut down feature is said to operate 5ms after triggered, that'll never protect anything at that speed, it seems like a useless feature to me.
Best Regards,
Chris
It doesn't use any type of switching, it just makes a small capacitance "appear" like a much bigger capacitance (X the gain of the transistor).
This has the effect of significantly reducing the ripple, just like a 100 000uF cap might. However, unlike a 100 000uF cap, it just can't store the same level of energy, and will readily sag more under load, further reducing power.
The reason it sags and is also so efficient is because it isn't regulating anything at all, and has less reserve.
If you think you have a hum problem or poor psrr it might be an interesting option to play with, but you have to be aware of what you're giving up, a nice stiff supply that can really deliver on demand, which is why unregulated raw DC supplies are good, and why they're still so commonly used in high end audio.
Maybe if you never plan on turning it up at all, it would be just the thing.
As far as the price of them goes, well, they closely match or slightly exceed the cost of the class-d modules themselves, and for what? I'm not really impressed with that. Although as far as what it is, it seems decent.
Also, the shut down feature is said to operate 5ms after triggered, that'll never protect anything at that speed, it seems like a useless feature to me.
Best Regards,
Chris
Hello,
Sorry didn't mean to be at all repetitive, think we were writing at the same time.
Regarding David's email reply in the 180 thread,
For what it is, it really seems like a robust little supply, but it would have to be. It may not work as well when expecting it to work at 400W or more.
Maybe you can email David and ask how much the rails will sag at the kind of a load you plan on presenting it with. Seems like a reasonable question to ask before you drop the buck on it.
Keep us posted.
Regards
Chris
Sorry didn't mean to be at all repetitive, think we were writing at the same time.
Regarding David's email reply in the 180 thread,
For what it is, it really seems like a robust little supply, but it would have to be. It may not work as well when expecting it to work at 400W or more.
Maybe you can email David and ask how much the rails will sag at the kind of a load you plan on presenting it with. Seems like a reasonable question to ask before you drop the buck on it.
Keep us posted.
Regards
Chris
- Status
- Not open for further replies.