Same thoughts here, see http://www.diyaudio.com/forums/soli...-old-ideas-1970s-ips-ops-189.html#post5039736
Best regards!
Well, in some cases I connect the trimmer (or some servo) in parallel with NFB, in some cases - just keep it at the input. Does not really matter. Input stage like this reacts on differential signal, regardless from what input it comes from. Do you see any advantage of connecting the trimmer particularly in parallel with NFB?
CFA designs - a kind of a different story, as we have to use much lower impedance feedback network there. That means - if we connect a trimmer (or a servo) in parallel with NFB, those have to be rather low-impedance as well.
That's why, I like to connect them to high-impedance input in CFA designs.
Over the time, I started connecting them to the input regardless of CFA / VFA topology - gust for unification.
Then I'd prefer another suggestion of you: Moving the first 56k resistor to the non-inverting bases, thus paralleling it to the one coming from the offset pot and approximating the NFB resistor's value. The very first resistor mustn't be omitted, but has to be replaced instead by one of somewhat bigger value (100k or more) in order to have defined potential over the input cap.
Best regards!
Hi Valery,
Lichtstark, another nice design.
Is the ecap in the feedback loop necessary, if you have a DC offset trim? I realize that the ecap makes the DC gain unity but you know that they add some degree of thd. The simulations do not show it. We know Bob Cordell does not like them and prefers a DC servo instead. I guess the design could accommodate both types and make it a user/stuffing option.
If you need that ecap, what type do you recommend, a bipolar, such as Nichicon Muse UES type, regular polarized ecap such as a Nichicon low ESR, UPW type or I have used one of those organic polymer types with very low ESR.
Do you have a layout for Lichtstark, I see pcbway has now lowered its cost to $5 for 10 qty, for a 100x100mm 2 layer pcb, what a deal for us hobbyists.
Thanks
Rick
Lichtstark, another nice design.
Is the ecap in the feedback loop necessary, if you have a DC offset trim? I realize that the ecap makes the DC gain unity but you know that they add some degree of thd. The simulations do not show it. We know Bob Cordell does not like them and prefers a DC servo instead. I guess the design could accommodate both types and make it a user/stuffing option.
If you need that ecap, what type do you recommend, a bipolar, such as Nichicon Muse UES type, regular polarized ecap such as a Nichicon low ESR, UPW type or I have used one of those organic polymer types with very low ESR.
Do you have a layout for Lichtstark, I see pcbway has now lowered its cost to $5 for 10 qty, for a 100x100mm 2 layer pcb, what a deal for us hobbyists.
Thanks
Rick
Last edited:
Hi Rick!
You're right - the cap can be dropped (as well as servo can be implemented).
I think, what I will do - I will leave a placeholder for the cap, and then if you or somebody else wants to jumper it - just solder a small piece of wire, if not - use the cap. This kind of symmetric design is rather thermally stable, however... it would actually be cool to compare the performance with the cap and without it.
Cheers,
Valery
Hey Rick, I think we're in a regime where a well-sized capacitor will not be much of an issue: http://www.diyaudio.com/forums/part...ipolar-caps-measured-120db-thd-140db-imd.html
Yes, my practical experience don't show any noticeable quality degradation, caused by electrolytic, working in this kind of low voltage conditions - even if it's a polarized one.
Anyway, it would be interesting to see the difference.
Anyway, it would be interesting to see the difference.
Hi!
The prototype is still on its way - I can only show the sim results for now.
They look particularly good - see the main parameters here (note different frequencies and power levels):
Lichtstark-X preliminary performance
I plan to design three layouts:
1) Standalone IPS within 100x100mm (TH), etching-friendly;
2) IPS module for NS Modular hi-end series (SMD, standard size);
3) Full power amp module with dimensions similar to VHex+ (SMD) - that's Lichtstark-x, the one with IRFP240/9240 at the output, mentioned above.
I will post the updates as soon as the layouts are ready 😉
Cheers,
Valery
Those Specs look better then the Honey Badger!
But for listening most of the time .1w and 1 watt distortion figures are important too.
But for listening most of the time .1w and 1 watt distortion figures are important too.
That sounds very promising, Valery
!Would you consider to provide three pairs of output devices and design the board according to UMS? Thanks a lot!
Best regards!
That sounds very promising, Valery
Would you consider to provide three pairs of output devices and design the board according to UMS? Thanks a lot!
Best regards!
Hi Kay,
What is UMS? 🙂
Hi,
Late comment but I like what I see. Great layout on great design.
Obvious is interleaved P-N outputs and +/- rails on top of each other. Minimal inductance and stray fields.
Regards
