I've just about finished assembling my boards. I did not find the SMT aspect of this at all difficult. Just remember NOT to sneeze or the parts might be found some day orbiting the earth!
I am however in a real dilemma! I've also been listening to this Class D T.I. based board I bought from Class D Audio and to be honest it's the finest amp I've ever listened to... not sure I need to finish building this chip amp...? Its very definitely far better than the following amps I have here...
Pass Aleph 2's, Aleph 30, Aleph Mini and both the KSA50 and KSA100 Krell Clones. So the real quagmire is what do I do with all these analog amps and do I even bother to finish this chip amp project??? Perhaps I should do something a bit more exotic like a T.I. based class D amp with vaccum tube front end....
Yes, it IS really that good!
Mark
I think that you are hearing the OPA1632 front end. Despite the 180MHz bandwidth, it is specifically designed for audio. Like a tube, the OPA1632 is super-smooth up till about 10k. And, of course, it does perform best on excellent clean power. If you like that sound, perhaps you'd like to incorporate it into your own amplifier design.
It's funny that you would quote me on the chip amp. There was another thread which I can't rummage up that started a long time ago... probably by Brian GT. I believe that was back about 2002 or so and thats when I first requested a high power chip amp.
It was Star882 that suggested I try some of todays digital stuff out and I did. I'll be the first to admit that digital amplification has certainly come a long long way in this amount of time. If the T.I. chip can pass through the characteristic sound of the OPA1632 analog IC then I feel that is actually another plus for the TAS5630 chip itself because it's probably possible for it to pass through any characteristic one likes. I seriously doubt that I'd be looking at using the OPA1632 in future projects... I probably would consider using a toobe front end on a project like this though...
BTW: The Class D Audio board does have good clean seperate on board regulation just for the OPAs.
Mark
It was Star882 that suggested I try some of todays digital stuff out and I did. I'll be the first to admit that digital amplification has certainly come a long long way in this amount of time. If the T.I. chip can pass through the characteristic sound of the OPA1632 analog IC then I feel that is actually another plus for the TAS5630 chip itself because it's probably possible for it to pass through any characteristic one likes. I seriously doubt that I'd be looking at using the OPA1632 in future projects... I probably would consider using a toobe front end on a project like this though...
BTW: The Class D Audio board does have good clean seperate on board regulation just for the OPAs.
Mark
Last edited:
It seems that we agree in an odd way--from opposite sides. My point is that the power circuit and the preamp are the main contributors to your enjoyment of your new amplifier.
Here's the thing: High quality power circuits and well done preamp/predrive are more often seen packaged with Class D amplifiers because the Class D has very low tolerances and so it enforces using a high quality support. Had you used the same care on a Class A-B amplifier, then its performance would have the same level of excellence. The sound would be quite similar.
The main Class D caveat that concerns me is the nonstop supersonic output. I think that this can be corrected by pre-whitening the signal (treble lift in the preamp/predrive stage) and then have the output dampers compensate (level the frequency response at the output stage). Until this safety measure is in place, I, personally, choose to use Class A-B.
Well, the interesting thing is that the three Pass Labs clones I have built all have the same similar characteristic to them and that is they sound like they add a tiny bit of haze or grunge to the audio. This is very definately true of them while warming up but compared to the Class D board and now to the BPA-300 chip amp(running in stereo mode right now) I have running here there is definitely a bit of the haze or grunge to the Alephs... what ever you want to call it that must just a characteristic of these amps. The Krell Klones I've built also do not have this grunge to them either.
BTW: The chip amp is quite spectaculuar sounding although I think I might go back and add the servos to them!! I can't quite get the offset zeroed on all three chips over time... it does drift around a bit. Probably not as critical when it's bridged. Just need to get the second one running and see how it goes. Believe it or not I am going the Lundahl transformer route to bridge them.
Mark
BTW: The chip amp is quite spectaculuar sounding although I think I might go back and add the servos to them!! I can't quite get the offset zeroed on all three chips over time... it does drift around a bit. Probably not as critical when it's bridged. Just need to get the second one running and see how it goes. Believe it or not I am going the Lundahl transformer route to bridge them.
Mark
If you would like to post a complete copy of the power circuit, I will try to guess the source of the grunge.
It's more or less identical to what Pass used on the production amps. A bit more capacity and almost identical sized toroids. I serously doubt it's power supply related to be honest since all three amps exhibit the same tendency to some degree. Bias on the OP devices is also correct. The only actual difference here is that I used IRFP-240s and I believe Pass used IRFP-244s. That shouldn't really matter though. Others have used the same devices... We can no longer post the Aleph schematics here you'd need to IM me your E-mail...
Mark
Mark
Well, the chip amp wins for sure. Sorry Star882 but it wins by a landslide. The Class D board is still very good and better then the Krells and Alephs that I've built but now I see why Jeff Rowland offers a chip amp! The best way to describe it would be effortless but precise, dynamic but quiet and super fast on transient attack. Seems to not run out of power even when reproducing high levels or extremely wide dynamics from SACD. This amp just doesn't get lost in the music... And this is just the 150 watt per channel stereo version of the BPA-300... I still have to build the other one to use them bridged!
Mark
Mark
Well, the Class D board won't work at 1 ohm so whats your point? I suppose we could parallel in a couple of more 3886 chips and it probably could do 1 ohm.
Star... Considering this is Class AB I'm actually very surprised at how cool the chip amp stays for the amount of power it can put out. Of course this is three chips paralleled not just one. At idle it barely gets warmer than the T.I. board does. I've never had any issues running hot class A amplifiers here during summer... Most houses here use evaporative cooling in the summer!!
So now I have Krell Clone(KSA50 & KSA-100), 3- Pass Clones, BPA-300 Chip amp, and T.I. based Class D board all here in one room... so it's time to invite the listening panel over...
Mark
Last edited:
I saw that the 3 Pass clones you have built are Pass Aleph 2's, Aleph 30, Aleph Mini. The Alephs do have their signature sound - some like it, some don't. I guess the X-version might help in removing some of that haze you notice! Then there is the FirstWatt series which do not exhibit the Aleph Current Source, except for the AlephJ.
Last edited:
the easy solution to arguing over 2ohm or 1ohm speakers is to ban the manufacture of 1ohm and 2ohm drivers/speakers.
Even 4ohm is unnecessary. We can all manage with 8ohm, even for the professionals into high SPL PA equipment.
Now, if we accept that 8ohm is the standard driver impedance then we need the chipamp manufacturers to come clean on the current capability into real 8ohm speakers/crossovers.
A 3886 is limited to >=7Apk when cold. That cannot drive an 8ohm speaker load in real music driving conditions at higher SPLs, if the speaker if less than high sensitivity(>=93dB/W@1m).
The advantage of the parallel chipamp set-ups is the increased current ability.
Chipamps desperately need more current ability into 8ohm speakers. Don't consider using low sensitivity (<=87dB/2.8V@1m) 4ohm speakers.
Even 4ohm is unnecessary. We can all manage with 8ohm, even for the professionals into high SPL PA equipment.
Now, if we accept that 8ohm is the standard driver impedance then we need the chipamp manufacturers to come clean on the current capability into real 8ohm speakers/crossovers.
A 3886 is limited to >=7Apk when cold. That cannot drive an 8ohm speaker load in real music driving conditions at higher SPLs, if the speaker if less than high sensitivity(>=93dB/W@1m).
The advantage of the parallel chipamp set-ups is the increased current ability.
Chipamps desperately need more current ability into 8ohm speakers. Don't consider using low sensitivity (<=87dB/2.8V@1m) 4ohm speakers.
Last edited:
How much did the hybrid cost altogether? What about the analog? At high powers, the hybrid has a significant cost advantage, although it might be obscured somewhat by artificial price premiums.
I have no idea why they even make speakers with an impedance so low. The stray resistance losses are just too significant. Maybe if there are several sets of speakers to be driven from the same source, which can then be simply connected in series. But series/parallel with 8 ohms works as well.
Yes I guess we should all get un-listenable dinky toy speakers so our likkle chip-amps can drive them .....
I've actually built 4 Alephs. Mini, Aleph 3 for a friend, Aleph 30, Aleph 2's... also Both Krell Klones.
Andrew... I think that statement is a little off the wall my friend! Well there are the Apogee ribbon speakers that dip down to 1 ohm... and they are stunning speakers... thats what brought the KSA-100 into existance. In movie theaters we don't like to series-parallel speakers. You have more phase related problems from doing this than load problems from paralleling them. So we usually parallel them in groups that get us to two ohms and this usually means 4 speakers on a channel of a given amp. Many commercial amps are available that will safely drive 2 ohms today. I usually use QSC. I do however consider the Class D board to be high quality to the point that it could be used in Cinemas so I am looking into doing that. I have a myraid of places I can install and test them under all sorts of different conditions.
Star... The LM3886 based amp cost about the same to build as the Class D board and large switching power supply cost to buy. I had a pair of old AB Systems power amps that I stripped out and used the chassis for the chip amps. The Class D board still sits on top of one of the Dynaudios. FYI: If you want to the same chip amp on the cheap you can buy the same chip amp board assembled from China for $40 USD but they do them with through hole parts. I got mine from Per Anders and also bought the SMT parts from him at the same time in that last group buy. The boards and parts were about $110 USD for four boards and sets of parts.
Mark
Andrew... I think that statement is a little off the wall my friend! Well there are the Apogee ribbon speakers that dip down to 1 ohm... and they are stunning speakers... thats what brought the KSA-100 into existance. In movie theaters we don't like to series-parallel speakers. You have more phase related problems from doing this than load problems from paralleling them. So we usually parallel them in groups that get us to two ohms and this usually means 4 speakers on a channel of a given amp. Many commercial amps are available that will safely drive 2 ohms today. I usually use QSC. I do however consider the Class D board to be high quality to the point that it could be used in Cinemas so I am looking into doing that. I have a myraid of places I can install and test them under all sorts of different conditions.
Star... The LM3886 based amp cost about the same to build as the Class D board and large switching power supply cost to buy. I had a pair of old AB Systems power amps that I stripped out and used the chassis for the chip amps. The Class D board still sits on top of one of the Dynaudios. FYI: If you want to the same chip amp on the cheap you can buy the same chip amp board assembled from China for $40 USD but they do them with through hole parts. I got mine from Per Anders and also bought the SMT parts from him at the same time in that last group buy. The boards and parts were about $110 USD for four boards and sets of parts.
Mark
Last edited:
Maybe compare them to a pure digital, then. (You'll obviously need a digital source or it will lose every time.) In theory, pure digital is the very best, even better than hybrid. In practice, I like the smug factor of the hybrid, not to mention that most pure digitals need a microcontroller to start them up.
why put four 8ohm drivers in parallel and then drive them with a 2ohm capable amplifier? It makes no sense other than economy.
A four amplifier system driving four individual 8ohm drivers will be far more resilient if any component fails during a performance. This applies especially to PA, where the sound MUST GO ON.
Mark,
your reasoning may go a long way to explaining why most PA systems, whether disco, or cinema, or live concert (indoor and outdoor) sound pretty near dreadful.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.