Amplifier using MJL21193/94

Jump to Latest
#82

Attachments

  • IMG-20211126-WA0000-1.jpg
    IMG-20211126-WA0000-1.jpg
    117.9 KB · Views: 241
  • FB_IMG_1638263085197-1.jpg
    FB_IMG_1638263085197-1.jpg
    250.2 KB · Views: 236
Last edited:
#91
Discrete amps have a few parts to learn about.
Long tailed pair at front end which acts like a differential amplifier.
VAS stage which amplifies voltage. VAS bias constant current generator.
Vbe multiplier which sets bias across bases of output transistors.
Driver transistors which amplify current to output transistor bases.
Output transistors amplifiy current to drive the speaker.

Power supplies where you need to provide enough voltage and current for amplifiers needs into 4 and 8 ohms.
 
#92
@john_ellis
As you know i was in search of KSC3503/A1381 but different hfe group was not acceptable for me that's why i bought Sanyo bjt's; spec is also very good.
Now i've everything to build some excellent discrete amp! ;-)

@nigelwright7557
That's why my hunt for bipolar transistors ends here. Now i want to give some time in reading & experimenting. For me it's not much difficult to build a SS amp but that's not my aim. I want to dive deep ;-)
 
#94
As you wish 🙂 But you didn't reply privately, can i ask you why? You replied 8days later after your last comment & you asked me for 4pair. I said you privately that you're late but if you want i can try to arrange for you. In this situation price is very unpredictable. I paid more for mje15034/35 pairs than 2sc5200/1943 (similar quantity). Are you going to pay for $2.8 usd/pair for high power to3? Anyway i'm not a component seller & don't do this in my every thread.

Best regards!
 
Last edited:
  • Like
Reactions: VIPINSAINI20
#95
Well, I have recommended using matched devices in the output stages. But even for one lot of devices purchased with the same gain group, I found most of the NPNs at the low end of spec and the PNP's at the top.
So now I am using the standard triple EF output stage, with loads of current gain in higher power amps, although matching BD139/BD140's and 2N3055/MJ2955 is still doable (just, within perhaps 1.25 for the drivers and 1.5 for the otuput) for moderate powers.
What this does do as a downside not often mentioned is that the effective Early voltage of the triple is poor, as it multiplies the Early voltages of the individual stages.
Which is why we need a lot of NFB (read open loop gain) which will offset the asymmetry as well as countering the Early effect and thermal effects.
 
#98
+/-72V is starting to come down out of the stratosphere of ridiculousness for an EF2. Also makes your existing drivers more suitable. On +/-81 you probably would have driven them a little outside the SOA since the s/b breakpoint doesn’t quite make 40 volts. Less average dissipation too - I’d still upgrade any stock heat sink they were using (unless they are sharing the output heat sink which is ok). They are the full pack version, which puts them at a thermal disadvantage already compared to the bare tab.

I’d like to know where those KECs came from too. It would be really nice if one could get legit second source Jap drivers (obsolete outputs might be too much to ask) from regular distributors like Mouser. Other than the stuff that was taken over by Fairchild/ON, which seem to be in the business of trimming down their offerings. At some point there won’t be any left. We can still get legit 2N copies but those are really for repair and maintenance of old equipment (where procedures and documentation requirements do not allow for a substitute and the company is willing to pay three prices for it) - limited application in high performance audio. Those 25W driver types with 100-200 MHz fT and flat beta to half an amp are just plain critical to making a really good amplifier. Even Mr. Mile (Apex) recognized this with the “retro” AX6.
 
You must log in or register to reply here.
Top Bottom