Hafler XL-280 : PC-40D circuit design

[keith721]

Up-front disclaimer: I'm not an EE, and I'm definitely NOT an electronics technician. I'm a computer software developer with an average understanding of basic electronics and average PCB soldering skills (no SMT work).

What I'd like to get is a better understanding of the original circuit design within my favorite vintage amplifier. I've been searching and reading all of the old threads and gathered plenty of troubleshooting and modding information (thanks!!) Could someone here provide a summary of the various stages in the XL-280 PC-40D circuit, and which specific components comprise each stage? With this information and the schematic from Hafler's website, I could probably learn a lot more than I understand now.

Many thanks in advance to Dick_West, mlloyd1, anatech, fab, nw_avphile, djk, amp-guy, and carlmart for all their previous contributions and comments on the DH-200 and DH-220 and occasional XL-280 threads.

[mlloyd1]

go to borbely's web site and look for downloads on the servo 50 and dc102 amplifiers: http://www.borbelyaudio.com/special_articles.asp there is a lot of similarity between the xl280 and servo 50 and you WILL get a lot from from the articles.

best,
mlloyd1

[keith721]

mlloyd1: Thanks, last week I read both of his 1999 articles on JFET: The New Frontiers, but not being an EE, I can't fully appreciate everything in them. Matching some of the circuits described in the articles to the XL-280 schematic isn't something I've attempted (but I will! ) I wasn't aware his Servo-50 circuit was so similar. I'll take a look at it for comparison.

I'd guess the unobtainium JFET pairs making up the input cascode are Q1/Q2 (2SK147) and Q3/Q4 (2SJ72). So what circuit topology best describes the remaining transistor groups on the board:

Q5/Q6 (2N5550) and Q7/Q8 (2N5401) - ?????
Q9 (2N5550) / Q11 (2N5415) - ?????
Q10 (2N5401) / Q12 (2N5415) - ?????
Q13 (NP2222A) - ?????
Q14 (2N3440) and Q15 (2N5401) - ?? power to the output MOSFETS ??

[mlloyd1]

Q5/Q6 (2N5550) and Q7/Q8 (2N5401) - cascodes for the unobtaniums
Q9 (2N5550) / Q11 (2N5415) - VAS stage (aka second voltage gain stage). first device is emitter follower. second device, well, a combo of folded cascode and common emitter. others may have a different opinion on what to call it. whatever ...
Q10 (2N5401) / Q12 (2N5415) - ditto for the other polarity (we'll ignore your device number error )
Q13 (NP2222A) - sets the output stage bias - not thermally coupled to outputs since output mosfets are set close to zero tempco bias point
Q14 (2N3440) and Q15 (2N5401) - yep, drivers for the output mosfets. borbely uses mosfets for these in the servo 50 and dc102

mlloyd1

[keith721]

You sir are a gentleman and a generous one at that.

Now that I know the appropriate names, I'll have an easier time searching for (and reading, of course) the relevant EE-type articles discussing this stuff.

As for my error, I'll have to look it up when I'm at work tomorrow to see which it was. Thanks for telling me.

[keith721]

. . . and I see I used the wrong part numbers for Q12 and Q15.
They should have been Q12 - 2N3440 and Q15 - 2N5415.

[keith721]

If the bias transistor (Q13, NP2222A) were bad, would MPS2222A be an acceptable replacement? Although it has a lower Ic of 600 mA, it otherwise matches regarding hfe, Vceo, Vcbo, and Vebo.

[keith721]

Rather than starting a new thread for substitution advice on the bias transistor, I'm bumping this thread. No, I haven't determined this specific component has failed, I'm just trying to determine a suitable replacement if one of these eventually fails.

[Dick West]

Almost any transistor will suffice in the Vbe multiplier circuit, as long as it is in the same voltage and hFe range and proper polarity. Replacements available even at Radio Shack.

[keith721]

I wondered if the RadioShack MPS2222A would be a reasonable substitution, and I tried searching the web for suggested substitutes. I eventually resorted to comparing datasheets. Since the bias on this amp is generally set no higher than 250 mA, I thought the MPS2222A should be acceptable.