夏季 DAC 原理和製作班 (AD1955) 5-SEP

Hi KK sir,

I am using default plan since I want to look for my dream component set myself. However it is really extremely hard to find all the components I desired in the market. Therefore I asked you and william for the upgrade plan late on yesterday. But William replied both upgrade plan A and B were not complete right now (some missing parts in both plans) because I was too late to request.

"Both heads cannot reach the shores", "兩頭唔到岸", 好慘! 頭髮白晒!

I think I cannot complete soldering my dac before Monday. 仲撲緊零件

BUT I would come on SEPT 15 to look at you testing for other c chings. I think I have no problem testing mine at home since in fact I have built several dac before: (3623, 8412, 8420, 5843, pcm63pk, 1704.......) If not many issues, I could handle the test at home.

話晒都係紅褲仔出身麻 !

If possible I would bring a small bottle to take some cleaning fluid home to wash the PCB myself. I know the washing procedure.

KK sir,

I've checked 1645 and 10100 spec and found that the reverse leakage current of schottky is surpprissingly high (4mA @ 20V reverse voltage), though it has favorable low (0.57V) forward voltage drop at large 16A foward current. That means even when the diode is switch-off, there is still 4mA current flowing from the 4700uF to the power transformer.

10100 is even worse

MUR420 seems normal on reverse current (in uA microamp range)

From the diagrams above, in order to use Schottky with least reverse current, the Schottky may use heatsink to keep it cool in operation.

Maybe the forward characteristic of rectifying diode is more important than reverse characteristics. Letting more peak current to charge up the 4700uF capacitor compensates the minor 4mA loss in diode off.

The diagram shows the current waveform of Schottky diode (grey color).

The diode is only conducting for a relatively small period in the cycle. But the forward current is huge in magnitude (10A peak can be expected).

Therefore the diode forward characteristics is critical. Lower voltage drop of Schottky not only means lower diode power loss but also ensures a practical low impedance path and fast switching response for the high current spike charging up the 4700uF capacitor (Close to idea diode).

When reverse bias, the 4mA leakage can be considered as a 5kohm load across the 4700uF capacitor (R=20V/4mA). Not a great issue in fact.

Having recongnized the result above, if we incorporating Schottky in our design, we should also employ higher current rating transformer so that to provide the high peak current in diode conduction. Otherwise, transformer becomes the limiting factor in the system.

Is this your machine?

KK sir,

I am using default plan in DAC1955, may I know the default transformer spec now? how many secondary windings?

And also what is the size of the default transformer? Is it toroid?

Then what amp, what speaker? do you play LP?

我果塊DAC1955底板有少少彎, 我至今未焊through hole料, 辛苦咁耐撲完零件, 今晚焊SMD料時, 塊PCB平放玻璃上睇, 而家發現成塊板向一邊角慢慢彎起有2mm高, 我相信冇問題, 但係大家有冇注意到? 之前我塊PCB放書架上.

一遍amp海 !  有冇位坐