What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

Do it yourself audio panel repair (?)

Ed_Wischmeyer

Well Known Member
So my otherwise wonderful PS Engineering 5000EX died this afternoon, and since it's out of production, support is pricey -- $200 plus time and materials. I've purchased a plug-compatible Garmin intercom, but the question is, what do I do with the old one? Sell it? Try to fix it? Use it as a boat anchor for a very small boat?

The symptoms observed this afternoon were:
1. Moderate frequency buzzing in pilots right earphone if the headphones weren't on tight. (Lightspeed Sierra headset) (That symptom might be a red herring);
2. Passenger microphone had difficulty breaking squelch;
3. Passenger microphone would not break squelch;
4. Unit died completely -- no intercom, passenger side PTT inop, no lights on the front of the unit.

Any thoughts? Suggestions? Offer$?

Thanks!

Ed
 
+1
Typically electrolytic capacitors degrade over time. Sometimes you'll see the telltale sign of the bottom (top) bulging out.
If a voltmeter won't do for troubleshooting and you don't have an oscilloscope to look at the power supply output, you can use the brute force approach of simply replacing the electrolytic capacitors. Trying that sure beats $200+.

Finn
 
What FinnFlyer said. Electrolytic caps are cheap, and you can probably put in better parts than original. Also look for telltale color changes on the PC board, areas where the green is a little darker or lighter. This can indicate overheating components, often a voltage regulator or amplifier.
 
I haven't opened up that unit, but if it's less than about 15 years old I'm willing to bet is SMT. Not easily fix with out some special equipment on the bench.

Bob burns
Rv-4 n82rb
 
I would tend to disagree about the SMT caps. The electrolytic caps can be replaced by hand pretty easily depending on your soldering skills.

I had a coworker that was collecting bad PC monitors at our work. He investigated and it turned out that most had bad caps. He bought some caps and told others how to fix them if they wanted.

Given that your failure was not a "hard" failure I would tend to agree with the assessment. When the caps dry out you lose their filtering capability. This results in the power supply having more AC noise on the DC supply. This is probably what you are hearing.
 
Thanks, all. This was a gradual degradation to a total failure -- now the lights won't even come on.

I've got the unit apart, emboldened by the knowledge that a replacement is on the way and that if I do irreversible damage to this unit while I'm screwing around with it, no big loss.

It has two surface mount boards with a zillion tiny components, a few custom ICs, and one umpty-gazillion pin connector running down the middle. One board is only half populated, indicating that that board was designed to work on other units as well.

There are no immediate signs of any distressed components -- no obvious discoloration, no holes where the smoke escaped etc. I'll get out the magnifiers and look again, but at this point, no smoking guns... or anything else smoking.

I'll get any number of hundred dollar hamburgers out of this one and won't even have to drive to the airport.
 
Last edited:
Start at the +12 volt input and see if you can track the traces from there.
I do agree that trying to repair surface mounted components is a lot harder than old stuff. I actually pretty much stopped doing electronics repairs around the time they became prevalent. (Yes, I'm old)

Finn
 
I have found that a 10x lighted bench magnifier, a pair of curved sharp pointed tweezers, and a decent soldering station with very fine tips for the iron make SMT work not that big of a deal. Doesn't take anything really fancy, my Weller WES50 was under $100. I've hand soldered a number of fairly dense prototypes. I use solder paste and a heat gun when able, but using the iron is not that tough.

And speaking of the heat gun... Ed, another thing that will cause intermittent problems degrading to total failures, as you have experienced, is bad reflow solder joints. They can be almost impossible to spot. SMT boards like that get a solder/flux paste applied with a mask (similar to silkscreening), then the components are placed by robot. After that they go through an oven that melts the solder and everything gets soldered at once. Problem is, sometimes that reflow process is not perfect. You can get bad solder joints that don't show up for weeks, months or even years. Trust me, I've had it happen on my own products. If you have nothing left to lose, a heat gun or toaster oven can be used to gradually heat the board until the solder flows again. That may fix it. Gotta be careful with a heat gun, though -- it's possible to melt the solder and blow the components out of place.
 
Dale's not kidding about the toaster oven. One of the products quite a few of you guys have in your panel has its circuit board soldered in a literal toaster oven. You do need to know the correct temperature & timing, though.

Charlie
 
Dale's not kidding about the toaster oven. One of the products quite a few of you guys have in your panel has its circuit board soldered in a literal toaster oven. You do need to know the correct temperature & timing, though.

Charlie
Yep. I had a ham radio product that I sold for a while -- USB to serial interface for some ham rigs. I used a mask and squeegee to apply solder paste, placed the parts with tweezers, and into the toaster oven they went. THAT product had a 0% failure rate in service. Another one that I had done by a well equipped, US based professional production shop didn't do quite as well... of course it had some pretty fine pitch SMT packages on it. My hot air gun repairs all worked fine though.
 
Still haven't seen anything suspicious, other than the markings on the MOSFET power transistor are very hard to read. F9540N IszR 225P are the markings, but I've not found the exact part nor do I know what the fine print means, nor do I know if faded markings are significant or not...

Replacement parts, quantity 1, are about $1.25 for a similar part. Maybe I'll call the manufacturer and ask them.

I'm a software guy, not a hardware, but I did once point out to the hardware guys that they had no pull-up resistors and that's why the board didn't work. A well-remembered one of the few times I knew something about hardware.
 
Hi Ed,

What about posting a few detailed pictures of the boards? At worst it would add to the education part of EAB and at best maybe we could help you figure it out...

(Picture posting help available if needed.)
 
Decided I'd better put the audio panel back together before I forgot how or lost a part. Got that done with no real problems, but I bet the folks who know what they're doing do it a whole lot faster.

If I make it out to the airport tomorrow, I'll check and see how much functionality I have. If I have failure mode functionality (can still use Comm 1), that's really all I need. If the audio panel works, by some chance, I'll probably sell it at a good discount.

Taking it apart and comparing surface mount to older computers or older radios, where you could recognize the components, reminded me of looking under the hood of a modern car -- there's lotsa stuff there, but mostly I have no idea what it is. And I used to work on my own cars...
 
And speaking of the heat gun... Ed, another thing that will cause intermittent problems degrading to total failures, as you have experienced, is bad reflow solder joints. They can be almost impossible to spot. SMT boards like that get a solder/flux paste applied with a mask (similar to silkscreening), then the components are placed by robot. After that they go through an oven that melts the solder and everything gets soldered at once. Problem is, sometimes that reflow process is not perfect. You can get bad solder joints that don't show up for weeks, months or even years. Trust me, I've had it happen on my own products. If you have nothing left to lose, a heat gun or toaster oven can be used to gradually heat the board until the solder flows again. That may fix it. Gotta be careful with a heat gun, though -- it's possible to melt the solder and blow the components out of place.

Can all components handle the heat? I think I've seen instances where some pin-though components (for example bigger electrolytic caps) were added after the oven treatment.

Finn
 
Can all components handle the heat? I think I've seen instances where some pin-though components (for example bigger electrolytic caps) were added after the oven treatment.

Finn

Depends on the part, of course. But if they are SMT parts, they're designed to handle reflow soldering. And most thru-hole parts are the same as their SMT counterparts, other than the leads.
 
It works again!

It works! at least, it does right now.

So here's what I did. I brought the audio panel home, disassembled it mechanically, looked for any distressed electrical parts, said the correct incantation ("Wow, look at all those tiny little components!"), put it back together before I either forgot how or lost a part, put it in the back of the SUV along with a 12 pack of soda, went to Sunday school and church (don't think that had anything to do with it), and went out to the airport.

When I reinstalled the audio panel, it was in backup mode, but at least Clearance Delivery responded that the comm was loud and clear. But no lights, hence, no power... at least the plane is still flyable.

So I decided to see if there was power to the unit. The installation manual said which pins were power, but didn't give a picture of the connector with pin numbers. Found that on the web, but was unable to reliably reach the correct pin, no matter what Rube Goldberg assortment of improvisations I tried.

So I put the audio panel back in and then, in a stroke of inspiration, wondered if the unit was turned on or not. I pushed the power knob, and it came to life!!

I was wondering if the problems might have been the unit coming loose from the connector in flight, and that seems plausible except that my headphone and mike continued to work after the audio panel went into fail safe mode. If it had been loose connectors, I would not have been able to use the radio.

Anyway, late last week I went looking for a good deal on a pin-compatible replacement (rewiring the plane would require lots of time and bucks to have a shop do it, so pin compatible was a major consideration), and found a new in the box Garmin 245 at a very good price. Differences are that the Garmin has 3D audio (great for side by side and with two radios), but the present PS unit has a MONitor button which gives comparably useful two radio capability. But I only have one radio. The Garmin has Bluetooth and a USB port on the front, the PS has a 3.5 mm audio plug. Audio cables are more flexible but don't provide power to the unit. So the units are different but comparable.

The Garmin does have a clearance recorder, but the manual doesn't say what it records or really how to use it. It also says that if power is removed from the unit (low voltage during engine start, maybe?), all the recordings are lost. Very few vendors these days have manuals that really tell you what you need to know...

Anyway, I'll probably sell the PS unit at a good discount (About $1200 new, IIRC, although this one is out of production) and go with the Garmin unit, just for the sake of reliability. PS doesn't sell a separate install kit, but the Garmin install kit is about $125. And the Garmin should arrive this week.

Anybody want a PS Engineering audio panel that works fine?
 
Back
Top