Well I guess it just goes to show you, you can’t keep a good diyer down. It’s nice to see the high end class-d market opening up with another hat being tossed in the ring. Although a worry that Lars may be pricing himself out of the market.
NewClassD NCD1 400W = 248.99Euro
UCD400 = 100Euro
This of course all depends; if it performs better even by a small margin (in listening tests) the price could be justified. Lets see what the free market forces do.
The socket allowing one to replace the IC seems a good idea. As it allows one to try many options, although it may be susceptible to chip creep over time (many many years). A method for securing it down may be something to think about in the next version.
Nice to see you back in the business Lars,
Leve
NewClassD NCD1 400W = 248.99Euro
UCD400 = 100Euro
This of course all depends; if it performs better even by a small margin (in listening tests) the price could be justified. Lets see what the free market forces do.
The socket allowing one to replace the IC seems a good idea. As it allows one to try many options, although it may be susceptible to chip creep over time (many many years). A method for securing it down may be something to think about in the next version.
Nice to see you back in the business Lars,
Leve
Sassen: Right.....
Here is with 8 Ohms, 4 Ohms, and for reference 4 Ohms with Local Feedbacl loop.
20 - 20.000 Hz is within +/- 0,25 dB
Here is with 8 Ohms, 4 Ohms, and for reference 4 Ohms with Local Feedbacl loop.
20 - 20.000 Hz is within +/- 0,25 dB
Re: Re: Patents
That..... is a bold statement. I suppose I should go and read the patent but you are suggesting in the above that Every Amplifier with voltage feedback violates the patent.
No prior art?
I think you might be overstepping the mark there. Unless there is additional information associated with that claim it doesn't seem likely that it would stand on its own.
DNA
Bruno Putzeys said:
That..... is a bold statement. I suppose I should go and read the patent but you are suggesting in the above that Every Amplifier with voltage feedback violates the patent.
No prior art?
I think you might be overstepping the mark there. Unless there is additional information associated with that claim it doesn't seem likely that it would stand on its own.
DNA
LBHajdu said:
Hi Leve, and thank You for the nice wishes 🙂
I don't mind being the most expensive choice, in this comparison. However it should be said, that if you take out the VAT from the price of NewClassD NCD1, then you can compare more directly.
Because also UcD400 price it is without VAT.
NewClassD NCD1 400W = 199 Euro. (excl VAT).
I forgot to put that information in the price field on newclassd.com , sorry ...
http://www.diyaudio.com/forums/showthread.php?postid=491322
It looks like the post filter feedback technology has been really well known for a while. The schematics of the feedback used here also looks really identical to the one described in the Philips fact sheet. The one in the patent is also well known, and as i said before, used back in 1970 in a class D construction based on bipolar transistors.
I think it is very bad if this makes you compromise the circuit Clausen. When reading the Philips patent one really ends up with being unable to do any kind of post filter feedback.
In sum I think Philips, Sony, Mitsubishi, Samsung and Matsushita together has patented every traditional solution known to the world, at least 3 times.
It looks like the post filter feedback technology has been really well known for a while. The schematics of the feedback used here also looks really identical to the one described in the Philips fact sheet. The one in the patent is also well known, and as i said before, used back in 1970 in a class D construction based on bipolar transistors.
I think it is very bad if this makes you compromise the circuit Clausen. When reading the Philips patent one really ends up with being unable to do any kind of post filter feedback.
In sum I think Philips, Sony, Mitsubishi, Samsung and Matsushita together has patented every traditional solution known to the world, at least 3 times.
Snickers: Thanks, but it would appear that the new solution is actually better than the standard setup, as used in the patent.
The new solution has lower distortion, and higher bandwidth. And so far no negative effects.
The new solution has lower distortion, and higher bandwidth. And so far no negative effects.
Lars Clausen said:
The UCD400AD is € 130,- so yours is actually priced quit well.
The only problem I have with your amp. is the input imp.
I use a passive preamp. Is there any option to go 100k?
BTW nice to see you'r busy with a new class D amp.
Bert
Regarding the UcD patent: I think it is just a matter of long breath/money to challenge the details of post-filter feedback or the feedback path using a PD function since they are prior-art IMO.
The novel part of UcD (and the enviously admitted really clever one) is the topology as a whole and it is definitely worth of getting a patent.
Regards
Charles
The novel part of UcD (and the enviously admitted really clever one) is the topology as a whole and it is definitely worth of getting a patent.
Regards
Charles
High impedance might be O.K. since it is paralleled with the pot. But switching amps with single ended inputs are definitely a NO-NO.
Regards
Charles
A high Z input would have to be bandwidth limited to 20-20.000 for general stability, and i'm not sure this is without cost of sound quality.
On the other hand if you can live with using a 10k pot, it works just fine as is. Is this an acceptable solution?
On the other hand if you can live with using a 10k pot, it works just fine as is. Is this an acceptable solution?
How does the input look like? Is it a discrete input stage with BJT's? Maybe a input buffer would be appropriate?
I have an input buffer in both my "gainclones" and by that I eliminate all interfacing trouble. As you may have noticed a common problem with at least LM3875, 3886 designs are problems with pots and the signal source impedance.
Some tube preamps would probably benefit from this.
I have an input buffer in both my "gainclones" and by that I eliminate all interfacing trouble. As you may have noticed a common problem with at least LM3875, 3886 designs are problems with pots and the signal source impedance.
Some tube preamps would probably benefit from this.
Any special reason why you chose an inverting input? Is LM6172 (special kind of current feedback) a good choice compared to "conventional" opamps?
Only 5 kohms as input impedance may be a drawback for the DIY market at least.
AD8620 is pretty nice and specified for +- 5 volt operation also. According to datasheet you will also have better phase margin at unity gain. 60-70 degrees vs. only 40 deg for LM6172
Only 5 kohms as input impedance may be a drawback for the DIY market at least.
AD8620 is pretty nice and specified for +- 5 volt operation also. According to datasheet you will also have better phase margin at unity gain. 60-70 degrees vs. only 40 deg for LM6172
5k is not only related to choice of op-amp, even if a higher input impedance with most bipolar op-amps (like this one) will lead to higher background noise.
A more pressing problem is stability, as higher input impedance like 22k or 47k will lead to more critical wiring, and potential oscillation problems. I have experienced quite a few of those in the past. Of course one solution is to limit the bandwidth of the amplifier, but i think that is a poor solution, as it will affect sound quality. So in this sense a higher impedance will also become a drawback in the DIY area.
Even if the new safety processor will dodge breakdowns on this account, i think it's better to start with high stability, and then add safety on top of that.
A more pressing problem is stability, as higher input impedance like 22k or 47k will lead to more critical wiring, and potential oscillation problems. I have experienced quite a few of those in the past. Of course one solution is to limit the bandwidth of the amplifier, but i think that is a poor solution, as it will affect sound quality. So in this sense a higher impedance will also become a drawback in the DIY area.
Even if the new safety processor will dodge breakdowns on this account, i think it's better to start with high stability, and then add safety on top of that.
What will happen if you desided to have a non-inverting input?
Do you need a full speed (>> 10 MHz) input bandwidth?
I mean ultra high speed input, no input filters and an environment with DVD players, hometheatre gear and lot's of digital noise. It's maybe better to filtering out this instead?
Do you need a full speed (>> 10 MHz) input bandwidth?
I mean ultra high speed input, no input filters and an environment with DVD players, hometheatre gear and lot's of digital noise. It's maybe better to filtering out this instead?
The inverting input is good because it's easy to protect agains static overload, without it affecting the audio signal, like otherwise a Zener diode would.
- Status
- Not open for further replies.