HiFi Subjectivist ¦ Audio Forum ¦ NVA User Group

valvesRus wrote: ↑Fri Jul 05, 2019 12:33 pm

savvypaul wrote: ↑Fri Jul 05, 2019 12:17 pm

Do you have anything to post that is relevant to the OP?

Does the same apply to DQ ?

Just asking.

DQ hasn't gone off topic. He posted a comment on empirical evidence versus electrical theory - entirely on topic with the preceding posts.

You posted a comment that was defensive about Nick and unrelated to the preceding posts.

DQ replied that "Your desire to provide retorts to my posts, alas means you ignore what i posted and talk little sense".

The above comment from DQ seems correct, to me, based on what you posted and have subsequently posted.

Time to stop playing games. If you are keen to debate with DQ then do so on the genuine content of the thread.

Been for my hair cut and a pint . Takes forever if I'm walking.

I think there is a balance to be found with the optimum amount of smoothing for a given transformer and load. Cost and amount of space in the case also plays a part.

My NVA based amps (per channel) are running on 60,000uf for the output transistors and 40,800 uf for the front end and driver transistors.

Ok a simple rule of thumb, 1,000uf at 1A constant gives 1V pk-pk ripple, so 10,000uf is 100mV this means the dc voltage will. At 50Hz Main thus full bridge 100Hz Ripple.

A) have less ripple and will give a higher pk voltage out.
B) CMRR 0r Common Mode Rejection Ratio is the amp ability to ignore hum/ripple and noise on the PSU so the bigger the better not always so.

Peak output from the caps is again rule of thumb 0.5CVV = A/S so the shorter the current demand the more current you need to replace and can be available to the load.
The problem is that cap have internal resistance so this will limit current in and out it is normally in the mOhm range, this will also have a reflection on the resistance and saturation current of the transformer so bigger here is better, but this will now effect the pk current in the diode the IIT rating.
For example a 6A diode may have a peak rating of let say 60A but for only 1mS.
So now you can work back and decide the ESR of the cap and track resistance.
Complex no fun oh yes I love it.
Now if you use a Post Regulator linear PSU you can if designed and use correctly ignore the ripple but with a voltage loss across the REG, often in power amps this type of cheaper reg is to slow and can add HF harmonic to the op stage and make the amp sound hard and crap.

A rule thumb again if the amp has a BW of 10Hz - 100KHz the responce time of the reg will need to be faster than 1uS (1MHz) then it will sound better.

So who game to design a proper PSU ??? IS IT YOU

As short not I like PUSH PULL PSU design complex and they can be bloody fast 10-50MHz fast.
This has ripple and over and under voltage regulations, and the PSU Impedance is lowered like a damping factor so more control, so go for it lads.

SMPSU in power amps.
Great for Class A as long as there are full bridge resonate mode, never you Flyback in Audio it is shit, lamps and battery charger OK but audio, just buy a gun and end it.
Forward are OK but need a constant load like flyback but with less noise normally.
CUK converter are good but I am not sure Dr. Slobadan Cuk still owns the Patents so not for commercial use only MIL spec stuff. And complex maths,

Interesting Colin.

Are low ESR capacitors always beneficial in power supplies ?

Yes but it become more critical not for ripple reject but peak current demand, unfortunately the specs given are normally set above 100Hz as the ESR peaks. So it is good practice to use let say 100nF poly of film cap in parallel as the peak high speed current can be to fast for the large electrolytic.

Have a look at these and they are some of the best I have ever used.

https://paktron.com/wp-content/uploads/ ... pstick.pdf

This part no is very good for audio 106K050CS4

_D_S_J_R_ wrote: ↑Fri Jul 05, 2019 11:46 am
Tom, I don't think that's right. The larger amps (AP50 upwards) had double value supply caps over the smaller models and the Phono 2 has double that again (one lot in the PSU and the other set in the head unit).

RD told me early on that getting a larger reservoir was a good move, but maybe four 4700 rather than two 10000, but this may depend on the caps used?

Putting phono stages aside as we were discussing power amps

A20 has a 50VA transformer and 2x4700 caps (total of 9400)
A60 has a 160VA transformer and 2x10000 caps (total of 20k)

160/50 = 3,2
3,2*9400 = 30k

So I think A20 has more smoothing caps than A60 if you take size of transformer into account

Niave question but is there any mileage in using voltage regulators in NVA amps ? I assume the boards are designed to work best with raw 38v.