[R-390] OT: Pass Transistor Question

2002tii bmw2002tii at nerdshack.com
Fri Oct 17 01:55:57 EDT 2008


Jon wrote:

>This is an old beast that I just happened upon that has one bad 
>channel.  Right now I'm resisting working on it just for the 
>mechanical reasons you describe.  Sounds like I don't need to worry 
>about matched/selected parts and your comments do explain the 
>parasitic suppression components.  Also sounds like the older 
>transistors may have had a lot of phase shift at their upper limits 
>with the potential for creating havoc with the amp's feedback loop.
>
>The mechanical packaging really is pretty bad.  I was unpleasantly 
>surprised.  Documentation's nothing to write home about 
>either.  Digging in to the docs still leaves ???  Anyway I'll be 
>sure to sweep the output for any signs of life after I get it back 
>together.  Any special conditions that's apt to make it go into 
>flight; level, freq etc?

In all honesty, unless you are just obstinately curious, if you 
haven't put the effort into it yet -- don't bother.  After hassling 
with the worst mechanical design you've ever experienced in an 
electronic product, you will have a thoroughly mediocre amplifier (at 
best).  Sell it on eBay and buy an Ashly FET-200/2000 or a Bryston 4B.

If you insist on repairing it, use MJ15003 output transistors (real 
Motorola ones) instead of the original 2N5631s.  If the TO-66 drivers 
are broken, good luck -- very few manufacturers make TO-66 packages 
these days.  You can make a TO-220 transistor fit, but being 
plastic-cased transistors, they don't last.  Whatever you do 
(including using stock parts), be prepared to adjust the compensation 
to maintain stability.  (Note that Crown provides two schematics, one 
for each brand of output transistors they used.)

Run it up to clipping without a load, with an 8 ohm load, and with a 
4 ohm load, using any convenient sine wave in the 500 Hz to 2 kHz 
range.  Look for bursts of oscillation as you do this (it will appear 
at the same spot each cycle -- check particularly as the amp enters 
and leaves clipping).  Don't push it very far into clipping without a load.

Do not try to reach full power above 10 kHz unless you are willing to 
fix it again.  I generally ask power amplifiers to drive 2 kHz square 
waves into an 8 ohm load resistor paralleled by a variety of 
capacitors from 0.001 uF to 1 uF, but Crowns will not tolerate this, either.

Good luck!

Don 




More information about the R-390 mailing list