Crown XLS 402 Troubleshooting

Anything about hardware musical instruments.
Post Reply New Topic
RELATED
PRODUCTS

Post

Hi All!

I have a Crown XLS 402 that has stopped functioning.

It will power on for a second or two, all the lights will come on, the fans will start and then it shuts off.
The amp is then unresponsive and pressing the power button or the breaker switch does nothing.
If you let it set for a while (a couple of minutes) it will turn on again for another second or two.
I have found that it if you only wait a shorter period of time (like 30 seconds) the lights will flash for a moment but the fans don't turn on.

I have opened it up and it sounds like a maybe a relay is tripping on power supply when it shuts off.

I have the equipment and soldering capability to replace a component or two but I don't know where to start.

And I know this amp can kill me if I don't properly respect the capacitors and I don't feel like dying so I will be mindful of them.

Any ideas on where how I can start to find and fix the issue?

Post

aciddose wrote:
BertKoor wrote:
maybe these 2 can help

Post

https://www.crownaudio.com/en/support_d ... ts/xls-402

The schematics downloadable there. Later xls are digital amps but this appears a fairly ordinary old fashioned class AB bipolar design using complentary NPN and PNP output transistors.

Given a schematic it should be easy to fix if you know how. If you don't know how then it will be very difficult to fix because it contains numerous parts and any one part or combination of more than one parts could be bad and result in the symptoms.

Because of closed loop negative feedback it is hard to find bad sections by signal tracing because a problem anywhere in the circuit tends to show up everywhere else in the circuit both ahead and behind the problem location.

If more than one part is fried and you don't find/fix all of them, then that lonely remaining bad part you did not find will probably burn out the shiny new parts you just got thru installing. It will probably fry all the new-installed parts "instant" as soon as you power up the amp first time after finishing the incomplete repair.

If you can read the schematics and understand the operating principles thus shown then probably you can fix the amp.

If you do not easily understand the schematics then it is less likely the amp will ever get fixed, unless you make it a "science project" and an opportunity to learn how to understand and debug that kind of analog amp. In which case the final repair completion may lie many moons in the future. :)

Post

nix808 wrote: Fri Apr 12, 2019 11:37 pm
aciddose wrote:
BertKoor wrote:
maybe these 2 can help
Thanks for showing apparently I have a reputation, but me not likely...

Some years ago a rather nice hifi amp of mine had some simple water damage (don't ask). I have a multimeter and a solder iron, but no scope or means to test semiconductors. I can read schemas but do not fully understand why/how it works if it gets beyond simple.

So I let a "professional" deal with it, which are harder and harder to find. He replaced the main end transistors as these were blown. Powered up, and poof! Found something else, fixed it, powered up: no direct poof, but pcb tracks were melting and poof!
Apparently he also did not really know what he was doing.

So everything JCJR said above is spot on.
We are the KVR collective. Resistance is futile. You will be assimilated. Image
My MusicCalc is served over https!!

Post

Me and u Bert-
we are mortal
heh

---I'm still going to ask u to fix my stuff heh

JCJR, - U rock

Post

With a discrete linear unregulated supply like that you're probably best to walk through the circuit from start to finish. Basically all it does is filter a tap off the input transformer using a diode ring and capacitors.

All the other portions of the circuit are a simple timer that checks things have charged up before it switches on the main supply relays. Depending on the age the biggest things to check would be:
  • Look over the PCB to ensure there is no heat damage
  • Any ICs (tl074) or other actives (transistors, diodes)
  • The electrolytic capacitors, start with the smaller ones first after you measure the expected voltage on the nodes with the larger ones
  • The relays themselves
Basically you want to inspect for aging and heat damage. De-soldered pins on components that heat up or can wiggle and components that are past their expected service life. You need either a simulation or print-out of the expected operating voltages at each node.

The PDF schematics don't provide node voltages so you'll either need to input that into a spice simulator yourself or measure them from a known working version of the circuit. A scope to measure proper operation of RC filters (the capacitors) can help a lot too if you see unfiltered waveforms where they're unexpected.

Generally speaking working with any high-voltage and high-current circuit is extremely dangerous and should not be done without extreme caution, most of all by an expert. ("I'm an expert, I don't need to be cautious" = no, experts would say the opposite so saying this proves you better keep your hands off that circuit!)

Relays and electrolytic are known to fail over time. That would be my main suspicion.
Free plug-ins for Windows, MacOS and Linux. Xhip Synthesizer v8.0 and Xhip Effects Bundle v6.7.
The coder's credo: We believe our work is neither clever nor difficult; it is done because we thought it would be easy.
Work less; get more done.

Post

I'm speculating here, but based upon glancing at the schematic I might guess that there is an issue with unbalanced input from the supply capacitors after the input rectifiers.

This would slowly charge the averaged voltage (in the center, input to the opamps) in one direction or the other away from ground and that would switch the relays off. This would be a protection feature for unbalanced current on one side and could possibly point to an issue with the amplifier circuit itself drawing power in unexpected ways.

After switching off the input power you'd need to wait at least several minutes for the 4.7 uF or 100 uF "DC Sense" capacitor to discharge back to ground on its own.

So that would explain the symptom and point to the cause. (Either a problem with the sensing circuit, the supply or the destination circuit demanding too much current.)
Free plug-ins for Windows, MacOS and Linux. Xhip Synthesizer v8.0 and Xhip Effects Bundle v6.7.
The coder's credo: We believe our work is neither clever nor difficult; it is done because we thought it would be easy.
Work less; get more done.

Post

Yes, "DC Sense" is a filter averaging both A+B channels via 47k into 100uF. The voltage then passes via diode into the opamp which controls the power and audio output protection relays.

So if the average offset voltage of both channels is offset more than a certain amount determined by the temperature sensor (TH1 = LM35DZ) feedback circuit the relays switch off. The temperature sensor should normally only trigger the fans, so it's possible there is a problem either with the temperature being measured or the sensor itself especially since you report the fans turn on immediately.

That points to a problem with the temperature measurement... so again you'll need to walk through the circuit and compare against a spice simulation or a working version of the circuit.

Is the amber "Protect" lamp turning on?

Figuring out exactly what the problem may be is a fairly complex task so you're likely best to just take it to an expert and expect to pay for at minimum a day (possible near 1k?)
Free plug-ins for Windows, MacOS and Linux. Xhip Synthesizer v8.0 and Xhip Effects Bundle v6.7.
The coder's credo: We believe our work is neither clever nor difficult; it is done because we thought it would be easy.
Work less; get more done.

Post

Yes there are so many things that will cause the protection circuitry to trip, including faults in the protection circuitry itself.

Such amps generally get zero volt output offset (and low distortion) with high open loop gain and heavy global negative feedback. When power comes up it is unstable for a short time and the output wiggles around then as the power supply voltages settle to normal the feedback takes effect and the output "snaps to zero".

The speaker protect relays don't connect the speakers until after the output has stopped wiggling around, avoiding speaker thumps and possible damage. That same DC sense will protect the speakers and limit catastrophic amp damage if the amp fries. If the amp fries DC will appear at the output and the protection relays open to avoid frying speakers and also limit current draw out of the wounded amp.

This crown apparently also has power supply protection relays that should further limit catastrophic damage by turning off the DC when the amp gets DC problems.

So maybe the issue is with the power supply protection. But if the issue is dc output then it could be ANYTHING causing it. Bad front end parts, bad driver parts, bad output transistors. Anything bad in that big power opamp feedback path can mess up the DC output tracking.

Post

I don't mean to be over pessimistic. It isn't rocket surgery but was just trying to describe the general shape of the problem.

Pro amps tend to be easier to fix than consumer amps and not every repair is a tough dog. Tough dogs are those that take so much bench time to fix that a pro repairman loses his shirt fixing it but he has to fix it anyway because of pride and trying to maintain a reputation for competence.

It is bad form to return to a customer an even worse smoking melted ruin than was delivered by the customer for repair and explain "I'm not smart enough to fix it." Also had form to present a repair bill bigger than the price of a brand new amp, at least unless the customer has been warned in advance and for some reason the repair has sentimental value worth a ridiculous fee.

Sometimes problems are real minor and easily found and fixed.

Just an example about "finding all the bad parts" for something intermediate tricky-- Maybe you open it up and find fried driver transistor, some fried output transistors, and maybe a couple of well-carbonized resistors.

Those bad parts purt easy to find. So you replace them all but unfortunately maybe what caused all the damage was a tiny little open diode in the driver circuit, or an insignificant looking small shorted capacitor in the driver circuit. Or just a bad solder connection for such small part.

If the tiny little faulty part that caused the meltdown isn't located and fixed, then first time you power on after repair, that tiny little remaining problem will near instantly burn out the new parts the same way the old parts got burned out.

There are ways to minimize those risks if you have the right repair gear and a little experience. Unless you want to professionally fix amps, it is probably not worth acquiring the tools and equipment for a one-off job. I found amp repair to be kinda boring when I would do it fairly regular. Fun once in awhile but not fun day in day out. A person who likes the work could probably still make a good living at it even nowadays.

Post

Thank you All for your input, suggestions and analysis!

Your comments, along with a fair amount of research, have shown me that I have a A LOT of things to learn and understand before I will be able to diagnose and fix this amp.

I will not be spending any money sending this out to a repair tech as I can probably pick up a functioning one for around $100 USD so this will definitely fall into "science project" territory. I am good with that though as I have been diving pretty hard into DIY hardware the last year or so and have already learned quite a bit. The past couple days researching this have actually been more informative because I am able to see how several components actually work in real devices. I had a basic understanding of what transistors and transformers do but this exercise is forcing me to dig in deeper and REALLY understand how they work. I am not a total stranger to schematics but most of my exposure has been with really small circuits like simple effects pedals and this is far beyond my experience level. It is going to take me quite a bit of time to figure it all out.

I have an identical working amp that I can compare against so that should be very helpful.

The failing amp is actually in really good condition and I see no physical damage on the board from the top. I have partially dissembled it and haven't found anything that looks burnt or damaged but hopefully will be able to see something amiss when I get the main board out.

So first real question...

I started going through the power supply schematic as it is much smaller than the amp side and started reading about the main input transformer and amplifier transformers in general. I was trying to find out out I can test one and I found some info and several videos showing how wires can get hot burn up causing either a short or just a full break in one or several of the coils. I have tested for continuity and all things look good on the input side (that is if a continuity test is good enough). On the output side though I have 5 wires that should have 2 outputs (red/black/red = 65V and yellow/black/yellow = 1-18.1V). All 5 wires test positive for continuity in all combinations. Does that make sense? I thought that the yellow and red wires would not be connected in anyway.

Sorry for the noob questions... :-?

If I get this done and get good at it then maybe I can help others fix their stuff too!
Thanks for your help!

Post

Measuring for continuity doesn't make a lot of sense in a transformer, so just avoid doing that at all. Generally what you want to do to test a transformer is run a low voltage low current (via resistors) clean sine into the input and measure the outputs. Instead of mains AC you can just connect a 1v sine from a function generator for example and measure off the expected ratios from the output without even producing enough voltage to run any current through the rectifiers.

A function generator will generally have built in protection so you don't even need resistors, just connect the signal and probe the outputs.

To test the supply on its own what you could do is disconnect the amplifier circuit entirely from the supply. The supply should run fine with only the ground connected to the amp and the power rails disconnected completely. Then you can power up the supply from the mains input and ensure it works on its own... which will identify whether the issue is with the supply itself without loading and let you check that off your list. (All the capacitor, transformers, transistors, heat damage, wiring stuff can be partially eliminated with that one test.)

After the PSU is verified to work you can do a load test. Identify the usual current load in a working version of the PSU/amp combo and load the PSU rails to approximately match that with large (high wattage) resistors. That will then verify 100% that the PSU is not the problem if it functions correctly under the approximate load.

At that point you can move on to trying to figure out what is tripping the temperature sensor and DC safety circuit... if the PSU has been verified to work you're left with narrowing down which current path is blown in the amp which would lead to potentially both high temperature readings and a DC offset: one of the transistors in the output path is conducting when it shouldn't be.

Ideally finding the issue should only take a few hours at most but you need to be very careful as it's very easy to do damage while probing around in an active circuit by accidental shorts.
Free plug-ins for Windows, MacOS and Linux. Xhip Synthesizer v8.0 and Xhip Effects Bundle v6.7.
The coder's credo: We believe our work is neither clever nor difficult; it is done because we thought it would be easy.
Work less; get more done.

Post

Hi Dusty. What kind of digital multimeter do you have? Some of the points inside this kind of solid state analog amp can be "sensitively biased" and if you probe some of the points with a low-impedance measuring device (power-on testing) then the load from the probe can be enough to mess up the biasing and cause the amp to self-destruct (or destroy itself worse than it was beforehand, presenting even more parts to fix).

Not all points are so sensitive, but generally you don't want to probe inside a solid state amp (measuring for voltage) with an instrument lower than about 1 megohm impedance. Preferably much higher than 1 megohm impedance.

Maybe even the cheapest chinese DMM has > 1 meg input impedance nowadays, dunno. Long ago it wasn't hard to find cheap meters with impedance too low to be safe for this usage but maybe that is a problem of yesteryear. Just check to make sure your meter has high impedance in voltmeter mode. Not some cheap old passive radio shack meter with a swing-arm needle or whatever, which will almost certainly be too low-impedance for the purpose. Maybe they still sell such thangs for auto electrical repair and such, dunno. Just wanted to mention that.

It is good to buy a few clip test cables. Maybe with alligator clips on one end and the push-button "grabbie" precision probes on the other end. Nice to have an assortment of sizes. As aciddose said, if you accidentally probe touching together two parts, or your hand slips and accidentally shorts some random points on the board when powered up, it can do heavy damage very quickly. I was always a little shaky and worse as I got older so I would usually use the grabby pushbutton probes, or alligator clips a lot especially in crowded locations. Power off, connect clips to where I want to test, then power on and make the measurement, then power off again to disconnect the probes. Also helps in keeping hands away from high voltages.

====

Maybe won't help fixing the amp, but generic comments about the power supply. In my experience power supply issues are low probability in failed solid state power amps but statistics happens one event at a time so you always want to verify that it is working OK.

You don't want to take anything for granted, don't want to assume you know any answer before it is tested and verified.

That diamond-shape configuration of diodes in the power supply is a full wave rectifier. You can google full wave rectifier and get lots of good info. Down in the +/- 24 volt control circuit power supply is another full wave rectifier but they just didn't draw it in the shape of a diamond this time.

The 5 wire transformer secondary is a tapped secondary. One long loop of wire with three taps soldered in to points in the middle of the wire loop. If you measure the AC volts between the two outside ends of the coil maybe you measure "up 130 volts". But the center tap is in the middle of the coil, so if you measure from the center tap to either end you get "up to 65 volts RMS". Using the center tap as ground reference, when the top coil end swings positive, the bottom coil end is swinging negative. And vice-versa.

So one end is always "pulling" electrons when the other end is "pushing" electrons. The bridge rectifier of four diodes routes both half-waves of both output ends into the + and - DC smoothing filter capacitors. When the top coil end is negative the diodes steer that coil end to charge the negative power supply rail. At the same time the bottom coil end is positive and the diodes steer the bottom coil end to charge the positive power supply rail. And then when the AC reverses direction, the full-wave bridge steers the Top coil end to charge the positive power rail while the bottom coil end is steered to charge the negative power rail.

That way, regardless whether the AC input is positive or negative, both supply rails are getting charged "all the time" rather than "half the time" as with a half-wave rectifier circuit. Which delivers more/smoother power.

The lower voltage +/- 24 volts come from two taps "closer to the center tap" that measure about 18 V RMS relative to the center tap.

I suspect this power supply schematic "generically" applies to several amp models. Maybe the expected main power supply rails would really be +/- 95 volts DC in your amp but I would expect that kind of voltage to apply to higher-power models in that year's xls amp lineup.

OK, typically these bipolar analog amps can get within a volt or two of the supply rail before clipping.

You can calculate RMS power-- RMSPowerWatts = (RMSVolts ^ 2) / LoadImpedanceOhms

So if the amp is designed to deliver 300 watts into 8 ohms, RMSVolts = SQRT(RMSPowerWatts * LoadImpedanceOhms). It needs to deliver about 49 volts RMS. Peak = 1.414 * RMS, so the power supply rails need to be at least a few volts higher than +/- 69 volts so that it can deliver a +/- 69 volt peak sine wave without clipping. Maybe at idle the power supply rails really would get as big as +/- 95 volts DC but I'm guessing maybe that high a voltage would be for a more powerful amp in the same model lineup.

You see the peak vs rms thing in the control power supply as well. They claim 18 volt AC for and show +/- 24 volt DC power supply after the full wave rectifier and smoothing capacitors. 18 VRMS * 1.414 = about 25.5 volts peak. The diodes have about a 0.6 volt drop and you get some drop from loading. The heavier the load on this secondary power supply, the lower the voltage and the bigger the ripple, because it is apparently unregulated.

If you measure in there and it is a volt or two high or low from +/- 24 volts it may be entirely normal. Odd things you can notice if you decide to build some little gadget with a power supply-- Maybe you buy a transformer spec'd for 18 volts RMS at 300 mA. Ok if you lightly load that transformer, maybe only draw a few mA out of it, then the voltage will usually be a LOT BIGGER than 18 volts RMS. If you load it down to 300 mA then the voltage would drop down to 18 volts, but if honestly spec'd it will probably have much higher output voltage if lightly loaded.

Post

Thanks again for you help!

I'm limited on time during the week so I only focus on this for some hours on the weekends... so "many moons" in the future is looking to be a correct assessment on the timeline for this project.

Your answers, thus far, have mostly sent me deeper into the research rabbit hole and have shown me how much I don't know!

My multimeter is a PDI DM62 http://pdimeters.com/dm62.html
I have not been confident enough to do any power on testing yet but will get to that soon enough but thanks for the heads up on the "sensitively biased" issue. I have alligator clips so those will definitely be used.

I now have the amp fully disassembled but the power supply is still connected to the amp and there is no visible damage on the bottom side of the board that I can see. I will test the power supply side while disconnected from the amp and check to see if that works on it's own. I will also do some research on rectifiers and transformers because I think I get the gist of what you are saying but need to solidify it.

I am lacking a variable power supply, oscilloscope and function generator.
Do you have any recommendations on these
And are PC based USB oscilloscopes/generators a viable choice?

Post

"PC-based", the whole point of the scopes and generators is accuracy. You can synthesize almost anything with a DAC driven by a PC, but you'll never get the ideal DC-accurate output (or input) and sync capability that dedicated hardware provides. Another big reason to use proper hardware is the interface. There is simply no comparison between multi-turn pots and a GUI.

I can't recommend anything specifically as I believe this is another thing you need to experience to understand clearly on your own. I can't make a recommendation for you since anything I'd recommend would be based on my own preferences rather than yours.

A power supply is the least essential element and you can reasonably trivially build your own low current (<1 amp) supplies. You won't be needing a heavy variable supply for 99% of tasks and most of the time it makes a lot more sense to build a supply for your projects or whatever you're working on directly. That's one thing I'd never buy unless I found I actually needed it on a regular basis. My own linear bench supplies work fine while I'm testing things (down-regulating from +25v to +/-15, 12, 10, 9, 5, 3.3, ... works great) and for anything real I've simply built the appropriate proper supply into the finished devices.
Free plug-ins for Windows, MacOS and Linux. Xhip Synthesizer v8.0 and Xhip Effects Bundle v6.7.
The coder's credo: We believe our work is neither clever nor difficult; it is done because we thought it would be easy.
Work less; get more done.

Post Reply

Return to “Hardware (Instruments and Effects)”