p3a seperate fuses for i/p, VAS o/p ??

[Mark25]

All,
As it says in the subject tittle really, i want to seperatly fuse the 3 stages of my ESP p3a amp. So that the input stage and VAS fuses are not subject to the current drawn by the output stage. My audio experimentation thus far has indicated that this will improve sound quality. But the question is:

Are there any disadvantages in terms of loudspeaker life/protection?

i.e. would i be any worse of than the current situation, which is; if one supply rail fuse blows, the other rail will put X amount of DC on my expensive bass drive units.

Thanks Mark

 

[richie00boy]

What experiments have you condcuted to make you think this will improve sound quality?

It's a bad idea as if the output stage fuse blows, the driving circuitry will still be trying to drive the dead stage and you will get the output stuck at one of the rails (very nasty to your speakers). Purely theoretical this bit but, if you used MOSFETs as output devices you would also destroy them as their gates would be forced up to the rail voltage.

 

[Mark25]

What experiments have you condcuted to make you think this will improve sound quality?

going from stereo psu to dual mono, going from one bridge per split rail psu to one bridge per rail, separating or enlarging shared grounds in pre amp psu's, etc, etc. Basically anything that limits the sharing of resources, or wire, seems to have a positive effect on the sound.

 

[kASD]

Hi Mark,

I completely agree with RITCHIE about the drastic effects of seperate fuses and I dont think it will do any sort of improvement in sonic quality.

regards,
Kanwar

 

[pinkmouse]

Fuses are not good... To even attempt to protect the semis, they have to run at very close to their overload limit. This leads to heating with current demands, and as the fuse gets hot, it's resistance increases, lowering current flow...

If you have to fuse anything, just fuse the outputs. Run the rest direct from the PSU. Local decoupling for each stage, (Thanks Pavel), is a good idea as well.

 

[Mark25]

..........
If you have to fuse anything, just fuse the outputs. Run the rest direct from the PSU.............

doesn't that increase the risk of fire?

 

[pinkmouse]

Fires in electronic kit usually are caused by long term overheating of components, and a fuse won't protect you from this. Most other failure modes are explosive, and are not sustained enough to cause the heating that ignites other flamables. If the psu inputs are properly fused, and all connections are tight, then an amp is as safe as you can make it, (providing of course it's not filled with paper, thin bits of wood, or a large quantity of dust).

 

[richie00boy]

At risk of contradicting pinkmouse I will say just leave the amp alone Fuses should not be run close to their limits for the reason mentioned by pinkmouse, however, a fault usually results in a short of some kind so the fuse will blow OK anyway. The ripple caused by a power amp will far exceed any current ripple caused by a fuse impedance.

 

[Mark25]

........... The ripple caused by a power amp will far exceed any current ripple caused by a fuse impedance.

Yes, but the current flow and hence volt drop across the fuse is also modulated by the music. Which although lower in magnitude than the psu ripple, will be of a higher order.

 

[richie00boy]

By PSU ripple I meant the modulating of the PSU by the music demands, not the intrinsic repetitive/consistent charging ripple voltage, thus I believe that my original statement still stands.

 

[Mark25]

............... thus I believe that my original statement still stands.

Then maybe you'd be kind enough to re-word your original statement in a manner that my less than perfect English can grasp. As I understand it; fuses have resistance and don't cause "current ripple".
Thanks Mark

 

[richie00boy]

Sorry. Basically that if you ran the i/p and VAS from the main PSU instead of through the same fuses as the o/p stage, the varying load presented by the amp (which is also connected to that same PSU) will cause the rails to sag and be modulated (what I was originally terming ripple) such that the drop across the fuses is completely negligible by comparison.

And, your right, I should have written voltage ripple in the original post.

 

[Mark25]

I see what you are saying; The amount the i/p and VAS rail fulctuates will be reduced, by the ommision of the Rfuse, but this will be a relatively small change, due to Rfuse being low wrt to the overall PSU impeadance.

I accept that, but find quite large listening differences when shared R and silicon are reduced. For instance; in the context of my setup, going from one brigde reftifier per psu, to one brigde reftifier voltage per rail, provides a substantial positive effect on the sound quality.

However, the real reason for the thread is to establish how much danger i am putting my LS driver units in, by testing this upgrade to 3 fuses per amp rail, from one.

 

[richie00boy]

I think the dual bridge rectifier is a valid upgrade, as it allows the 0V currents to circulate independently in their own path. This may well be audible. I think running the i/p and VAS off a 'fuse bypassed' point may not be such an improvement. Try it by all means.

The real point as you mention is the problems this will cause your speakers. It is essential that good DC protection is employed and also be prepared for more expensive amp repairs should something ever go wrong.

 

[AndrewT]

Hi,
Borbely amps are all fet outputs and he runs separate rails to driver and output. Fuses all 4 rails appropriate to current demand.
His simple layout isshown using one dual pole PSU with double fuses to the -+ rails.
I second the dual wire/dual fuse route to help separate low current from high current demands. Go one step farther & suggest reg. psu to driver and brute force to output.