Crown XLS802 shuts down

[fredcapo]

I have something on the bench that I reallly dont want to work on, because it will be very time consuming and likely not worth repairing. But maybe someone has seen a similar problem and can help me pinpoint.

A Crown XLS802 is shutting down in one second after powerup. Nothing looks obviously burned. I would assume there is a massive short - likely the output transistors. Is this a good guess?

In a case like this, is the relay shutting down saving the board from massive failure? Is it possible that only one or two devices have failed at this point? Would all need to be replaced?

It seems to me that there is no way to test the output transistors without removing them, first because they are in parallel and second because there is no way to clip to them... will have to be unsoldered and removed to test.

Has anyone had such a problem with this beast? Thanks to any takers.

[The Dude]

For starters, you can check for shorts across output transistors in circuit. It won't tell you which ones are shorted in a group of paralleled transistors, but you can narrow it down to:

A) Is shorted outputs even my problem?
B) If so, which channel has the shorted components?

Before you start taking out parts, try to narrow down where the possible short is by doing a quick check of parts in circuit. Once you have it narrowed down to which channel, then you can remove the paralleled transistors in the shorted group to determine which ones are shorted.

Possibly a bit confusing, but I hope this makes sense to you.

[fredcapo]

Makes good sense. Actually I clipped the in circuit transistor tester across one (bank) of the (i assume) mosfets and it showed leakage...
will continue the process and then maybe remove a few to test individually. Thanks.

[The Dude]

Sounds like you've got a good start. FYI- I always start with a DVM on diode check to check out junctions. It's a little quicker if your just looking for a short. Though, whatever your choice of tool, my main point is, "start with the forest and then narrow it down to a particular tree".

[fredcapo]

I dug into it yesterday with meter... there are 4 banks of 5 transistors in parallel... 10 for each channel in push pull. I am reading a hard short across one bank. Now to remove those 5 and test. Wondering now what other devices went with them, though initial ohm readings across like-devices are the same across the banks.

These are silicon Transistors, not FET. How critical is matching? pair to pair or all in the channel?

[The Dude]

There are varying schools of thought on this. I've found in the past that, as long as you use original parts- no substitutes, you can just replace the shorted parts. That said, many suggest replacing all of them. Also, you'll want to check all of the emitter resistors. They often open up when the transistors short.

[fredcapo]

Thanks. I pulled all of the power transistors...wasnt that big of a job...after desoldering they came out with the heat sink as a unit. Only one transistor is hard shorted collector to prob emitter. Is there a good out of circuit test for the transistirs while I have them out?

Also no other devices so far test bad in circuit. Using meter and BK502.

Can I power this up with a single pair of output transistors just to test? or with no output transistors for same? thanks

[Jazz P Bass]

As long as one pair is installed, the amp is o/k to test.
Assuming that the transistors are attached to the heatsink.
If only one transistor is bad, replace the one.

[fredcapo]

I powered it up with 16 of the 20 total transistors - still waiting for the new transistor to arrive. Looks and sounds (and smells) OK but the B+ rails are symmetrical at + and - 109VDC. However the schematic I received from Crown, which only identifies XLS602, indicates voltage should be as much as +/- 95VDC. Crown insisted that this is the correct schematic though it only shows 4 banks of OT pairs instead of the 5 in the 802 and the nomenclature does not match what is etched... So I dont want to keep it powered up until I can determine if this is OK. May try to contact Crown again Monday.

In general, what types of tests should be done on an amp of this class to determine it is operating within specs? Thanks.

[Jazz P Bass]

Attached is the XLS Checkout file. Page 8 & 9 are the test spec's.
Do you need to do all that.
Probably not.
An output transistor blew.
Replace it. Make sure the all of the transistors are well greased, the output section soldering is sound & the bias is set cool.
A half power test into a 4 ohm dummy load should prove out the amp.

[fredcapo]

Nice document. Thanks. Reminds me of my days (years) as a tester at Bell Labs. Interestingly it too only references up to XLS602. Sounds like Crown or whoever owns them fired most of their engineering staff when the 802 was released...