https://www.digikey.dk/product-detail/en/fairchild-semiconductor/BC556CTA/BC556CTA-ND/974186
BC556B will do the job too.
BR, Toni
BC556B will do the job too.
BR, Toni
...fully agree. And many thanks for supporting all the questions in the thread! Also thanks to jlester.
@ fushfush:
The design is reasonably forgiving regarding tolerance of the filter choke. +/-20% will not mess the amp.
5% is intended to achieve most identical results like I published.
Expected impact in case your particular sample really shows a huge tolerance is a slight imperfection in step response (higher inductance will be more visible than lower inductance) and THD changes in the area of 0.00x% THD.
In case you have something on hand with +/-20% which works satisfying for you in terms of heat+sound, then it is fine.
The spec of the Coilcraft ED0006 is sort of incomplete (their ESR graph pretty sure is not the ESR but the total Z), nevertheless from overall description I agree with jlester87 that it should work basically.
For bridged operation you will either need to customize your own non inverted & inverted signals or use the gain board.
Astx also has customized his signals and published his board here in the thread. Just let us know which path you want to follow and I can settle a small drawing which explains how to wire (...erhm, well sometimes astx or jlester are faster
, but for sure someone here will give you support then).Please note, the liteamp main board acc. original BOM needs a large input signal. Driving it directly from standard line level signals will not push it to full power.
Last edited:
...fully agree. And many thanks for supporting all the questions in the thread! Also thanks to jlester.
@ fushfush:
The design is reasonably forgiving regarding tolerance of the filter choke. +/-20% will not mess the amp.
5% is intended to achieve most identical results like I published.
Expected impact in case your particular sample really shows a huge tolerance is a slight imperfection in step response (higher inductance will be more visible than lower inductance) and THD changes in the area of 0.00x% THD.
In case you have something on hand with +/-20% which works satisfying for you in terms of heat+sound, then it is fine.
The spec of the Coilcraft ED0006 is sort of incomplete (their ESR graph pretty sure is not the ESR but the total Z), nevertheless from overall description I agree with jlester87 that it should work basically.
For bridged operation you will either need to customize your own non inverted & inverted signals or use the gain board.
Astx also has customized his signals and published his board here in the thread. Just let us know which path you want to follow and I can settle a small drawing which explains how to wire (...erhm, well sometimes astx or jlester are faster
Please note, the liteamp main board acc. original BOM needs a large input signal. Driving it directly from standard line level signals will not push it to full power.
I know, then you don't know anything, dont obey the designers demands
I dont know where to make short-cuts without ruin the design...
First ill run it from the my NAD amplifier, just for starters, before i decides to run it with or without the gain board. But i got them and the parts for it.
About the running it bridged..
I don't need the power and my speakers is rather low-impedance about 3.2 ohm, so i don't wont it to run to close to the limits, because of distortion and so on.
Astx and Jlester has been an enormous help, and you for that sake
I just need to get the last details with Reichelt.com in order, then ill order the wima-caps and the parts for the inductor...
Now i just have to wait for the digikey-parts
That's Ok for starting, just be prepared mentally that you will
need to turn the volume knob pretty high in order to hear your music.
Which NAD are you going to use and which of it's outputs?
C326BEE, ill just skip the amplifier section and drive it from the pre directly.
I have to admit its only a temporary solution
The spec of the Pre Out is looking pretty fine.
It offers a low output impedance and has approx 10db voltage gain vs the line inputs.
In case the chain plays loud enough for your needs, then there is no real need for anything else.
P.S.
What's wrong with the power amp section of the NAD?
Sounding harsh and somehow messing details?
...some years ago I owned a NAD which had the idle current of the
power devices running just at the limit to class B.
IMHO they messed this without any need, because the amp had
a good thermal head room and the temperature compensation was simply perfect. So I increased the idle current of the amp and turned it into a proper class AB amp and immediately it was sounding fine.
First of all, its a project.. Its intended in the long run to drive a 2.1 set for my desktop. For now it should be tested on my speakers (These are mkI, i got mkII with a different and better crossover)
Second, ill want to try Class D, i wanted the clear sound, and the huge power they promise. The NAD is fine, but a bit warm in the sound and a bit closed.
These speakers can be driven from a 50WPC amp, but ill like to see how they are with a lot more power.
The main designer drive them with a UCD400.
+/-10% for all your listed caps will work without headache. Lite amp is forgiving if you relax just one or two tolerances. However if you double up all tolerances then with some bad luck trouble might start.
No, they are limiting the integrator and by this they ensure blameless clipping performance of the amp.
Thank you for your answer. Obviously, the solution could be to buy a few more caps and match these myself to a satisfactory range. It's the kind of thing one forgets while obsessing over a BOM. I'll share the Reichelt's cart later on.
If I may ask about some details :
- the wire links to gnd on the mosfet heatsink can be replaced by small caps... which value would you suggest ? And btw, why two links and not simply one ?
- what standoffs' height do you use for attaching the pcb to the heatsink ?
Btw, I was surprised by how big the elna caps for C20, C22 were. I'm rather planning on using pana FR which are only 12.5*20mm for 470µ/50Vdc, with better current rating and lower ESR.
Sorry for all the questions...
I've just opened the gerber files with a gerber viewer (usual routine before ordering) and strangely the drl file doesn't match the rest of the board (the drills appear out of the outline). I've generated a new drl file from Kicad (thanks for providing the files btw !) and then it matches.
Any reason to worry ?
Edit: the attached drl file works correctly at oshpark and at Online Gerber Viewer
I've just opened the gerber files with a gerber viewer (usual routine before ordering) and strangely the drl file doesn't match the rest of the board (the drills appear out of the outline). I've generated a new drl file from Kicad (thanks for providing the files btw !) and then it matches.
Any reason to worry ?
Edit: the attached drl file works correctly at oshpark and at Online Gerber Viewer
Last edited:
I would suggest 10nF ceramics (..it simply has to be much larger than the parasitic capacitances of the MosFets vs heat sink).
Two links in order to ensure low HF impedance (halving the inductance).
5mm
The panasonics FR should be fine.
The Elna types simply were in my assortment - waiting to be used.
Furtheron the I enabled the PCB for even larger caps in order to serve the needs of the +/-80V version.
Thanks for highlighting and publishing a correction.
Really strange. I published the same file bundle that I had used for my second PCB order and received no questions from the vendor, but correct PCBs.
Nice to see your customized gain board.
One remark:
Your gain stage is non iverting. The main board of the LiteAmp is inverting.
So in total you will have an inverting behavior, while the connector which are using will be labeled as In+.
Please note, in my gain stage this connector was routed to the inverting input of the instrumentation amplifier, ending up to be in total the non inverting input of the full assembly.
It really makes one realize how nicely tight the original is
..yes, I had to scratch my head once or twice until everything was fitting and keeping correct distances (up to +/- 80V on the board) and achieving reasonable signal wiring at the same time..
Thank you for all the detailed answers. That really helps.
For drl files, it's probably just a matter of mirrored coordinates or something like that; the vendor probably automatically corrected it.
For the gainboard, I realize there will be a mismatch between the markings on the pcb and the actual phase. Not a big deal for one off production but it's worth noting indeed. It's also assuming only +/-33v rails as it is what I'll be using. Its far less complete than the original.
Still, if it interests someone, the eagle's .brd is attached. Down to 28.27x40.49mm
One more question if you don't mind. The iraudamp7 app note states that in between +/-28 and +/-45Vdc, one IRFI4212H should be enough into 8r and 4r. In post 223, you said that dropping one IRFI4212H would forfeit the ability to drive 2r load. Is there anything else lost (increased thd ?) by going for one dual mosfet rather than two in 4 to 8r speakers ?
For drl files, it's probably just a matter of mirrored coordinates or something like that; the vendor probably automatically corrected it.
For the gainboard, I realize there will be a mismatch between the markings on the pcb and the actual phase. Not a big deal for one off production but it's worth noting indeed. It's also assuming only +/-33v rails as it is what I'll be using. Its far less complete than the original.
Still, if it interests someone, the eagle's .brd is attached. Down to 28.27x40.49mm
One more question if you don't mind. The iraudamp7 app note states that in between +/-28 and +/-45Vdc, one IRFI4212H should be enough into 8r and 4r. In post 223, you said that dropping one IRFI4212H would forfeit the ability to drive 2r load. Is there anything else lost (increased thd ?) by going for one dual mosfet rather than two in 4 to 8r speakers ?
Attachments
Nice to see that IR has the same opinion.
There should not be any other noticeable impact.
However as you already have pointed to THD, please be aware I tuned the THD for two devices in parallel. And after tuned all harmonics down to the sub -100db range (at low and medium power), I have to state...
...at such low levels of distortion most likely any mod will become visible in the spectrum (better or worse). I am convinced using just one IRFI4212H will not cause any dramatic worsening.