Sure man, in what post is the BV mod ? there are so many changes that I lost track
Regards
Juan
Post#2861 .Sure man, in what post is the BV mod ? there are so many changes that I lost track
Regards
Juan
PDF contains all new modifications .
Just wait for BV says that is ok.
You can work on OS PCB ,I'sure you can do it.
Post#2861 .
PDF contains all new modifications .
Just wait for BV says that is ok.
You can work on OS PCB ,I'sure you can do it.
Ok I will wait for BV no problem
Regards
Juan
I have tried the first modification,post #2922 .
I'm absolutely sure that is a big difference from unmodified ver.
I'm not sure that i like the second mod.with more flying parts.
Last edited:
The second modification looks like the more interesting of the two and is the one I'm currently interested in trying. What's slowing me down is the prospect of it changing again So I figured I'd wait a little just to see if any refinements are made. Also, as people currently have IPS boards they can chop up and modify, it would be nice to see something in action before I go ahead and etc/make a board.
Having seen Terry get the first mod working properly without too much difficulty, it would seem that the second mod should also work fairly well.
So I figured I'd wait a little just to see if any refinements are made. Also, as people currently have IPS boards they can chop up and modify, it would be nice to see something in action before I go ahead and etc/make a board.Having seen Terry get the first mod working properly without too much difficulty, it would seem that the second mod should also work fairly well.
Thanks Thimios,
I think it will take a redesign of the circuit layout to implement correctly, or at least optimally the second modification that BV has proposed. Perhaps when someone who has the skills can take the time to do that we will find out how far we can go with the CFA-XH BV version.
I think it will take a redesign of the circuit layout to implement correctly, or at least optimally the second modification that BV has proposed. Perhaps when someone who has the skills can take the time to do that we will find out how far we can go with the CFA-XH BV version.
The second modification looks like the more interesting of the two and is the one I'm currently interested in trying. What's slowing me down is the prospect of it changing again
Having seen Terry get the first mod working properly without too much difficulty, it would seem that the second mod should also work fairly well.
If we all waiting for the this last modification, we wouldn't have any working amplif up to now.
I have the last BV mod built and am trouble shooting it right now. Should have it working by tonight. I have to go and deliver a wedding cake so it will be couple of hours before I can get back to it. It is working but I have a .5V offset. Hunting down the problem. I do have a question for BV.
Why are R4 and R17 different? It causes the - rail to run higher. Not questioning your design, just trying to learn.
I am attaching the mods I did to the stock board. I'll let you know if anything has to be changed.
Why are R4 and R17 different? It causes the - rail to run higher. Not questioning your design, just trying to learn.
I am attaching the mods I did to the stock board. I'll let you know if anything has to be changed.
Attachments
Well I have just finished the first pass of the PCB design for this mod like 1 minute ago. I just need to check it to make sure that all is well, then I will probably give it a go.
Having just altered my preliminary CFA-XH design into the CFA-XHmod2 from BV, I can say that it does change a reasonable amount. Not so much as to be impossible to mod, but you are going to end up with a fair few cut traces and the like.
Having just altered my preliminary CFA-XH design into the CFA-XHmod2 from BV, I can say that it does change a reasonable amount. Not so much as to be impossible to mod, but you are going to end up with a fair few cut traces and the like.
Jason it sound like Terry will have something preliminary by tonight. It does look like with all the cutting and jumpers and deletion of parts that a new layout would be in-order if it is a significant improvement. I think the final arbitrator will be a listening comparison with an earlier unmodified board with a good set of speakers. If we can't here a difference then things just become an academic exercise and then why keep chasing our tails?
For certain people the advancement of a design, if only in measured specification, is also of interest. I think on the whole this is of interest to those who like to design circuits and understand how they work.
From my point of view, the second modded CFA-XH is of interest because it helps to overcome some of the limitations of the original design and simplifies parts of it. Sure you add in a couple of other parts, but what you gain is significant.
Better compensation, better distortion performance, better PSSR, no need for zener regulation and it also removes a lot of connections back to the sensitive ground. These things should make the design more consistent across different platforms, ie power supplies, internal case layout and how the amp is connected up.
From my point of view, the second modded CFA-XH is of interest because it helps to overcome some of the limitations of the original design and simplifies parts of it. Sure you add in a couple of other parts, but what you gain is significant.
Better compensation, better distortion performance, better PSSR, no need for zener regulation and it also removes a lot of connections back to the sensitive ground. These things should make the design more consistent across different platforms, ie power supplies, internal case layout and how the amp is connected up.
5TH Element,
I agree and for those reasons I would want to use those modifications. On the other hand if it was just to chase from 5ppm to 3ppm then I wouldn't care much. The fact that it lowers the psrr so dramatically and brings it into the vfa range is something to be happy about and any time you can get better specifications with fewer parts that always seems to be the way to go. I really am looking forward to building this amp with that input section and trying it with my own speakers, that will tell me all I need to know.
I agree and for those reasons I would want to use those modifications. On the other hand if it was just to chase from 5ppm to 3ppm then I wouldn't care much. The fact that it lowers the psrr so dramatically and brings it into the vfa range is something to be happy about and any time you can get better specifications with fewer parts that always seems to be the way to go. I really am looking forward to building this amp with that input section and trying it with my own speakers, that will tell me all I need to know.
I think it depends on what perspective you look at this from. Going from 5 to 3ppm might not seem like much, but within that reduction there is design refinement. Before this there might have been another design refinement that carried the design from 9ppm down to 5 and before it one that carried the performance from 20 down to 9.
I understand that it might not seem like a lot in isolation, but this kind of fine tuning is what leads, slowly over time, from one generation of products into the next.
It also depends where this distortion is. If we're talking mid band distortion, say 1kHz, then going from 0.0005% down to 0.0003% isn't really of any use to anyone you could say. It's vanishingly low and you're certainly not going to hear it. Usually tweaks like this come from component value refinement. But, if that performance increase also lowered the distortion at 20kHz, from 0.005% down to 0.003% I would most certainly want to take it.
Low, high frequency distortion, is something of a holy grail and to my way of thinking, getting the real life, measured, performance from 10k+ down to the 0.001% and lower bracket is just awesome. Not only does this require the design to be extremely well executed (thank you OS!), but if you're able to achieve it, it also means that your implementation is top notch too.
I understand that it might not seem like a lot in isolation, but this kind of fine tuning is what leads, slowly over time, from one generation of products into the next.
It also depends where this distortion is. If we're talking mid band distortion, say 1kHz, then going from 0.0005% down to 0.0003% isn't really of any use to anyone you could say. It's vanishingly low and you're certainly not going to hear it. Usually tweaks like this come from component value refinement. But, if that performance increase also lowered the distortion at 20kHz, from 0.005% down to 0.003% I would most certainly want to take it.
Low, high frequency distortion, is something of a holy grail and to my way of thinking, getting the real life, measured, performance from 10k+ down to the 0.001% and lower bracket is just awesome. Not only does this require the design to be extremely well executed (thank you OS!), but if you're able to achieve it, it also means that your implementation is top notch too.
I think that kind of amplifier probably suits most users more than monster power amplifiers. Still a two pair slew is going to out perform single pair output stages, so even for low power designs it still makes sense to go with two pairs. Of course if space is a real concern then you'd have no choice but to go single pair.
From what I understand though, the CFA-XH only works properly when driving the benign load of the EF3 output stage. I know that standard complimentary feedback output stages are easier to drive than EF2s, I wonder how the CFA would work coupled to a single pair CFB?
From what I understand though, the CFA-XH only works properly when driving the benign load of the EF3 output stage. I know that standard complimentary feedback output stages are easier to drive than EF2s, I wonder how the CFA would work coupled to a single pair CFB?
I think of single transistor on VAS and EF2 with two pair. I will choose driver transistor with high hFE.