Trouble with Crown Power Base 1 repair.
John Rosenberg
I am having a challenge with the repair of an older Crown Power Base 1
amplifier. The amp was brought in with a completely clogged dust
filter and transistors on both output boards open and or shorted. I
made the repair; returned unit to customer and it came back with one
side gone. I have since repaired the damaged channel and had it fail
again under testing. I know I have spent more time than this amp is
worth but I am determined to find the problem.
First: Crown specs the quiescent power draw at 90 watts but this unit
draws 168 watts with the output modules disconnected and even with
open circuits on both secondaries. Crown's tech department said they
have never seen a partially shorted transformer. Could this be normal
production variation?
Second: With an 8 ohm load and a 40V p-p 10 khz square wave the amp
rings on the low-side waveform. I also noticed oscillation at low
power levels...also on low-side. Have I missed an open or damaged
capacitor? Could the inverting 0p amp that feeds the low side be bad.
Does the use of non-crown matched transistors really invite this kind
of instability as the Crown techs have inferred. (I am using
MJ15016/MJ15015 output devices. Is this correct or would the
MJ15022/23 be better in this particular amp.)
Thanks in advance for your input,
John Rosenberg
Yeoman Service
Asimov
--- on the topic of "Trouble with Crown Power Base 1 repair."
JR> From: jlr...@ix.netcom.com (John Rosenberg)
JR> Hello,
JR> I am having a challenge with the repair of an older Crown Power Base 1
JR> amplifier. The amp was brought in with a completely clogged dust
JR> filter and transistors on both output boards open and or shorted. I
JR> made the repair; returned unit to customer and it came back with one
JR> side gone. I have since repaired the damaged channel and had it fail
JR> again under testing. I know I have spent more time than this amp is
JR> worth but I am determined to find the problem.
JR> First: Crown specs the quiescent power draw at 90 watts but this unit
JR> draws 168 watts with the output modules disconnected and even with
JR> open circuits on both secondaries. Crown's tech department said they
JR> have never seen a partially shorted transformer. Could this be normal
JR> production variation?
Remember that larger transformers have a high reactive current flowing.
Were you using a watt-meter or simply measuring the current and
multiplying by voltage? This latter method will give a wild
interpretation of power consumption. Like when one turns on an arc
welder it will draw a lot of reactive current but little real power!
JR> Second: With an 8 ohm load and a 40V p-p 10 khz square wave the amp
JR> rings on the low-side waveform. I also noticed oscillation at low
JR> power levels...also on low-side. Have I missed an open or damaged
JR> capacitor? Could the inverting 0p amp that feeds the low side be bad.
If you were outputing 40V p-p 10KHz square waves then it is good reason
why capacitors and resistors are being damaged in the ouput Zobel
network. That would be a good cause for ringing and oscillation!
If you're getting a signal then the opamp is fine.
JR> Does the use of non-crown matched transistors really invite this kind
JR> of instability as the Crown techs have inferred. (I am using
JR> MJ15016/MJ15015 output devices. Is this correct or would the
JR> MJ15022/23 be better in this particular amp.)
Matching helps to control idle bias currents better but isn't absolutely
required as long as the transistors have sufficient gain to allow the
drivers to work properly. The huge gain in an opamp will more than make
up for ordinary gain variations in the output transistors. It doesn't
hurt to use transistors with the same gain and from the same batch.
However, the transistors you are using may perhaps have too much gain
and making the feedback loop critical. You will have to modify the
feedback loop frequency compensation if that's the case, not so easy.
JR> Thanks in advance for your input,
JR> John Rosenberg
JR> Yeoman Service
... If all else fails, hurl it across the room a few times!
to...@nospamprepair.com
.22/5W (or thereabout) emitter resistors if you haven't already done so. I
think your choice of outputs are probably OK though I usually use MJ
15024/15025. check the bandwidth difference between the 015/016 and the
024/025 pairs. You could possibly have selected a type that will be prone to
oscillate if they have considerably wider BW than the original. Are you
getting them from a good source?...I have heard of shoddy bootleg Motorola
transistors being sold. What does Crown say to use? I don't think you need
to use their matched transistors - they do match their stock for switching
times as well as gain but they still fail here and there. It is a good idea
to check beta if you can and match what you have.
My best guess is you have a driver or small signal device leaking or bad -
or maybe a diode in the drive section has gone bad. I would check every last
component of your failed channel...fun, I know! What I usually do with a
catastrophic failure is quick check all the components in the driver. pre
drive, etc. sections, including caps. I have an "octopus" hooked up to a
scope which makes this go pretty quick. Don't forget to check your low
voltage supply's as well.
When you are pretty damn sure you have replaced all bad components, go ahead
and install as few of the output devices as you can get away with and still
have a complete circuit (all feedback hooked up). Look at your output
without a load - if you still see your oscillation or ringing then you
haven't fixed it yet. This method is a little tricky but if you go at it
patiently you should be able to test the amp out without a dozen outputs
installed, and under low current conditions and -troubleshoot without
risking $100 worth of output transistors. A (current limiting) light bulb in
series with one side of your AC line is also a tremendous aid in doing this.
Usually I will apply a load of 16 ohms and have a quick look at a very low
level signal across the 16 ohms load (maybe 10 watts output power max) once
I am confident that all the driver stages are working properly. Then and
only then do I install the rest of the paralleled output devices and bias up
& test at full power into rated load.
I have found Crown's quiescent current spec's to be way off many times -
perhaps I don't understand the method by which to test.
Hope this info helps. Good luck!
to...@amprepair.com
(615) 591-7556 10am - 6pm Mon thru Fri
please see our website:
www.amprepair.com
"John Rosenberg" <jlr...@ix.netcom.com> wrote in message
news:60474107.03022...@posting.google.com...
John Rosenberg
the other. I dropped in these types and viola no oscillation and no
ring. Somewhere down the line I got the idea that crown used the
15/16 pair in the PB1 amp. (I never came out and asked them, sure
would have saved me some headache) It is also possible that these
15/16 transistors were substandard parts how you say bootleg? As it
turns out the gain bandwidth of the 15/16 is less than that of the
22/23, 24/25. I have heard of the current limiting lamp trick before
but never used it. I prefer the combination of variac and clamp meter
and a slow power up.
Thanks again and nice amprepair website!
John Rosenberg
Yeoman Service
609-345-3800
<to...@nospamprepair.com> wrote in message news:<uyOdnVA6WMR...@comcast.com>...
John Rosenberg
Different transistors worked great...no other changes. Must be that
the 15/16 transistors had too much gain as you said.
Thanks for the input,
John Rosenberg
Yeoman Service
"Asimov" <warpcastgate@-removethis-bbs.juxtaposition.dynip.com> wrote in message news:<MSGID_1:167/133.0_3...@fidonet.org>...