[Reactcore]

I can't help wondering whether Mod A with triple capacitors is worth trying instead with just a single capacitor...

Mod A is what gave Dave's amp the energy speed and explosive impact.. but higlighted also existing background problems which in stock form are less evident.

The other mods are kind of needed to solve those and to harness the energy better.

I'm actually sad few get to hear what Dave is really capable of.. via line out and specially driving headphones directly.

 

[Jawed]

With a line out configured system (e.g. speakers via a power amplifier or headphones with a headphone amplifier) these mods should make little difference, shouldn't they? Current draw is effectively constant across audio frequencies in this situation, isn't it? A few milliamps into a typical 50Kohm or higher load?

The primary problem is to stop mains RF noise from upsetting the output circuit, when the system is configured for line out. With good RF isolation you get "better bass, increased liquidity, no harshness".

In theory it should be possible to calculate the low pass filter function for the original circuit LC filter and the effect of the various mods. And also how these interact with the regulator. e.g. stability and ESR interaction:

https://electronics.stackexchange.c...ay-from-the-lm317-liable-to-cause-instability

The sound quality problems you've heard seem likely to be about more than "energy storage". The variety of sound quality problems (varying bass, varying midrange, varying congestion) also hints at there being multiple factors at play.

I also found this interesting:

https://www.edn.com/simple-circuits-reduce-regulator-noise-floor/

As usual my typical disclaimer: I know practically nothing about electronics.

 

[Reactcore]

With a line out configured system (e.g. speakers via a power amplifier or headphones with a headphone amplifier) these mods should make little difference

True the improvement isnt as vast and obvious as with driving headphones.
But LO surely improves too.. i hear it using my speakers and amps.
Remember Dave's amp draws 100mA+ bias current in idle state which seem to get modulated with small signal related currents.

The primary problem is to stop mains RF noise from upsetting the output circuit, when the system is configured for line out. With good RF isolation you get "better bass, increased liquidity, no harshness".

I battled RF with my optic link and this mostly smoothed the treble response.. taking the 'digital glare' away.. specially with mscaler..
But the impact of the cap mods is the biggest step imo

https://electronics.stackexchange.c...ay-from-the-lm317-liable-to-cause-instability

This article mentions a too slow or instable startup time when output caps are chosen too big in capacitance.. something im not bothered with.

The sound quality problems you've heard seem likely to be about more than "energy storage". The variety of sound quality problems (varying bass, varying midrange, varying congestion) also hints at there being multiple factors at play.

These are my careful observations while playing around with caps.. where parallel setups -with longer wires/tracks in between- had these strange retractions on varying frequencies.

I placed the TDK's extremely close to each other with short thick wires between them to make them as much as one cap as possible.

I also found this interesting:

https://www.edn.com/simple-circuits-reduce-regulator-noise-floor/

Yes ive seen this.. it seems the LM307/317 are noisy regulators.. still Rob chose to implement them, i wonder if he explored alternatives or just wasnt aware of that..

 

[573531]

Wouldn’t this mod void your house insurance?

 

[Reactcore]

Wouldn’t this mod void your house insurance?

Yes ..but the audio nirvana is worth it

 

[573531]

Why did chord skimp on powersupply?

 

[TheAttorney]

Wouldn’t this mod void your house insurance?

I'm not sure that it would void a typical house insurance. If it did, then the same would apply to the many DIY enthusiats who build their own amp etc.
There would also need to be some proof that the mod was the actual cause of a fire - which in this case seems unlikely, as the mods are just increasing capacitance and resistance to an existing, relatively low powered, circuit.

More likely causes of fire would be when people overload a single socket with multiple adapters and then connect a room heater. Or run a class A tube amp with insufficient ventilation. Would the insurance be void due to the person's stupidity or ignorance? Again, not sure.

With every action there is a risk vs benefit equation. Every time I get out of bed in the morning I risk getting an injury at some point. But the benefits or getting out of bed each morning are huge and completely outweigh the risks.

IMO, the main risks of this mod are:

1. You do all the mods, and DAVE stops working properly. You undo the mods and DAVE still doesn't work properly.
2. The mods work, but later you want to sell your DAVE. Will the mods hurt or enhance its salability?
3. Your DAVE needs to return to Chord for any reason. You then have to put up with Chord chastising you.
4. As amatuer solderer, you do the mods yourself. It works fine, but looks like a dog's dinner.
5. You pay a soldering expert to do the mods neatly. You're not satisfied with the result. Who is to blame?

Against the above, the insurance risk is small IMO, but yes it is still there.
Of course, any warranty will be totally invalidated, but nobody should be trying this within the warranty period.

 

[Reactcore]

Now after playing and trying different resistor values for a time i learned that my Dave's sound signature is not consistent with every value..

To give it a description the higher the resistance in series with +/- 15 lines.. the more body and fullness it gets.. (sub) bass staying about the same, until it kind of gets too full sounding where it gives the feeling it overshadows the highs and transparency..

Its just weird because the output voltage of the regulators remains the same at 12v but the 15v drops a few tens of mV.

Using some parallel resistors i found a sweetspot to my liking at ~0,6 ohm in front of the LC filter and 0,47 ohm after per line.

So this gave me the idea to equip my external original boxed SMPS with two 1 ohm potmeters to replace the 0,6 ohm front resistors.

Now i can finetune Dave's tonality on the fly while playing.. to suit different recordings giving stunning results.

I just Luv to DIY around and get rewarded

 

[Reactcore]

And for the ones using a Farad, Sean Jacobs or having the original PSU in Dave..

The potmeters can also be placed as a removable module in Dave

 

[marcmccalmont]

Has anyone tried a Sean Jacobs power supply with this mod?

 

[GryphonGuy]

Do you mean the cap board in the space previously occupied by the SMPS shown in the above picture?

 

[marcmccalmont]

Do you mean the cap board in the space previously occupied by the SMPS shown in the above picture?

No I mean a SJ linear power supply

 

[Reactcore]

No I mean a SJ linear power supply

The mods will be tested with the DC ARC6 by @motu81 maybe he can post his impressions

 

[SteveHulk]

The mods will be tested with the DC ARC6 by @motu81 maybe he can post his impressions

Wow. Did he do the mod himself, or did he send his Dave to you all the way from Beijing?

 

[marcmccalmont]

The mods will be tested with the DC ARC6 by @motu81 maybe he can post his impressions

That would be very helpful for me!