New in solid-state, which Amp to build?

Status
Not open for further replies.
Well, new as in this is the second time I'm building a discrete amplifier. The first time was ages ago, me and a friend of mine started doing a Quad405-2 clone from scratch. I've also worked on a Gainclone that behaves very well and a PIMETA headphone amp that I'm especially happy with.

Now I want to make something discrete. All the amazing and fun designs here make my head spin however. My goal - build a stereo amp with output power between 100W and 200W in 8ohm speakers, but with enough safety margin so it doesn't blow if I attach anything with lower impedance. I'd like to have a nice PCB ready, but I've no problems tweaking things myself.

I was thinking of trying the Project101 or the Lynx boards, but I'm really not sure what to prefer. I'd like to try an all-FET design, I think I saw on those forums, but I don't know if there's anything with boards available..

Any suggestions?
 
Hi,
your target of 100W to 200W into 8r0 determines your transformer voltage at 35+35 or 40+40 or 45+45Vac. Expect the VA rating ~ 1times to 2times the maximum total output power.

For a ClassAB amplifier expect the device dissipation capability ~ 4 to 6times the maximum output power. i.e. 2pair of 150W devices to generate ~100W to 150W into 8r0 and safe into any 8ohm reactive load at any sensible domestic duty operating temperature. 1pair of 200W devices for 100W into 8ohm from 35+35Vac.

If you require damage limitation at lower than specification impedances then extra circuitry/hardware will need to be added, that is not strictly audio related.
 
I haven't listened to one or built one but Paul Ball's SKA might fit the bill, apparently able to cope with 3 ohm loads (though with lower supply rails) I recently found out about it from another diyaudio member "stuey" who has one. http://www.ska-audio.com/ everything I've read about them since has been positive.

Another with an almost cult following would be the AKSA 100 (again not fortunate enough to have heard one). Hugh is a very nice guy extremely helpful (even if you don't have one of his amps) and certainly from the rave reviews his amp gets, should probably be on your shortlist. http://www.aksaonline.com/products/products_aksa100.html should be ok into 4 ohms, not sure about lower but I'm sure Hugh can answer that question should you ask.

Both of these I have read a lot of reviews/listening impressions on and anyone who has compared them to the p101 has said that they are in another league. Of course budget also comes into it, so perhaps you need to mention that as well.

Me I have a 100W mosfet amp I built in the late 80's I've been very happy with it over the years, though once I complete my speakers maybe I'll be searching for something better. Certainly at the moment my speakers seem to be the weakest link in the chain, though I must admit that when my daughter took one channel out of action the other day and I temporarily put my gainclone in it's place I was presently surprised by the quality of the gainclone... The mosfet will be becomming the sub amp when I finish my speaker project, whether the gainclone then takes over main duties or I go on a quest for amplifier nirvana waits to be seen 😉 but I guess the fact I have been looking at alternatives tends to hint at the direction I'm likely to take.

Tony.
 
The SKA has good looking modules indeed, I've seen the site before and it looks quite suitable. I think the AKSA goes over what I'd like to spend, unless I can get only the boards instead of a kit.

I'm doing the PSU myself and actually it's ready, except on deciding on my rail voltage. It needs to be under 63V so maybe a 40V transformer will be the perfect choice with a decent safety margin on the cap ratings.

I don't think I'd ever want to plug in anything lower than 4ohm into any AMP I've got around, but a 6ohm speaker would be quite likely. I just want to be on the safe side, I'm not looking for a crazy amp that can handle whatever speaker I decide to throw at it. It's all fun and games until one power stage evaporates 🙂
 
63V is a sensible economy target for the PSU caps.
That does indeed limit you to ~<=240:41Vac, depending on mains tolerance and transformer regulation.
The margin is actually quite tight.

eg.
if you buy a 230:40Vac transformer and you input 254Vac, the specified maximum for a UK supply, then the peak output voltage at full VA loading is ~62.46Vpk less the diode drop.
 
Last edited:
If you are willing to delve a bit deeper into the DIY aspects you could look at something like the DX and it's brethren. The sound quality has been attested to by a number of builders and Carlos (the designer) lurks on this Forum and will be glad to help

Link:http://users.tpg.com.au/users/gerskine/dxamp/default.htm


My preference would the DX Precision, with CFP front end (I tried a similar topology in my TGM2 amplifier and it sounds wonderful)
 
Just by looking at the schematic, it seems that The DX precision is a really nice amp. The good part is I actually understand what almost all components do, which is always a plus 🙂

I'll look around for details of this amp. Bigun, how did you make yours, do you have a PCB design somewhere ?
 
Well I wanted to try everything myself, this is DIY afterall, so I went the whole-hog and made my own pcb in the basement sink. My TGM2 isn't exactly the same as DX Precision - it's actually got fewer components, see details here....

http://www.diyaudio.com/forums/showthread.php?t=145652&highlight=TGM2


But the DX amplifiers are designed to allow others to make them in that there is the web site I pointed to and then there is Carlos who designed them and frequents this forum who will advise and help you if you need. I think there are pcb designs for the regular (non-precision) DX amplifier you can use.

If you want to try making everything yourself I can send you a copy of my pcb artwork as a .pdf
 
The CFP in the LTP is a jewel.

It does many things for you very well, most of them are inter-related of course, but It reduces inter-modulation distortion which cleans up the sound nicely. This is less important if using high supply rail voltages but I believe at lower rail voltages the regular LTP's performance in this regard degrades if not biassed with a CCS. The improved linearity and accuracy of the front end has a direct impact on the effectiveness of the global negative feedback. I found a big improvement in bass and in clarity at the high end. I also think it improves the ability of the LTP to drive the next stage, the VAS, since a lot more current is now flowing.

In my view, you don't need a CCS supply for the LTP if using CFP, or at least it's benefits are reduced.
 
I was looking at this thread as well, but it's easy to miss something important when going to a few hundred page thread.

Nevermind, just looked at the last page, I'll send a mail and see if I can get a few of those etched for a test 🙂
 
Last edited:
Status
Not open for further replies.