Fender Passport 250 in protect mode once again...Help Enzo

[DaCentaur]

I only had to replace Q232 last time this time it is Q212, 215 completely shorted and I get a bleep from 214 and 216 with 460 ohms on their legs as well.
Wondering if their are some other transistors that are better or what.
Thanks

editing this post ...I went ahead and ordered replacements for all those Q's from MCM...I was really wondering if there was a cure to prevent this from happening again.

[52 Bill]

I only had to replace Q232 last time this time it is Q212, 215 completely shorted and I get a bleep from 214 and 216 with 460 ohms on their legs as well.
Wondering if their are some other transistors that are better or what.
Thanks

editing this post ...I went ahead and ordered replacements for all those Q's from MCM...I was really wondering if there was a cure to prevent this from happening again.

Without looking up the schematic, all I can tell is that the same part Q232 did not short, so it may have been totally unrelated to the original repair.

Sometimes when an output shorts, it will stress the other transistors. Not enough to cause an immediate failure, but enough to cause an eventual failure. This may or may not be the case here, but the recommended procedure is to replace all of the transistors that share loads with the dead one, like paralleled outputs and the drivers.

How are you mounting the new transistors? Are you reusing the gray silpads and applying heatsink compound? Have you checked the rest of the system, like the speakers and the speaker cables? How is this unit being used?

[DaCentaur]

this time the shorted one was Q212...yes this time I plan on shotgunning it and replacing all of them. Should I not reuse the silpads? Never heard that was a problem..I do it on other test equipment. and yes I do apply heatsink as well. Use is fairly light right now and everything else checks fine. I was surprised though that the manual says you can turn the unit on without a load but that is exactly what happened...I inadvertantly unplugged the unit-moved it outside and plugged it back in without realizing the pwr sw was still on and without the speakers hooked up...maybe a coincidence?? IDK

[52 Bill]

I've run them on the bench with no load without any problems. Did it just fail when it was plugged back in or did it fail when you set it up and tried to use it?

The silpads are easily torn and can be compressed so thin that they can cause shorts, so check them before reinstalling.

[Enzo]

When an amp blows up, something has failed. So far so good. When you fix a blown amp and it keeps blowing up, it is not a matter of "how do you prevent this happening." It is a matter of COMPLETELY repairing the amp. For example, we may do the basic repair and restore function to the amp, but we may not have restored all the protective circuits, so it cannot fend off some sort of situation that arises. Or we may not have excised all the parts involved with thermal tracking, so it works OK, but cannot adjust itself to compensate rising temperatures. That sort of thing.

So the key to preventing it blowing up repeatedly is to complete the repair. Really.

[km6xz]

To amplify Enzo's comment, the most important part of a repair is the diagnosis. A diagnosis is also the most difficult part and requires the most test instruments and electronics knowledge. What is a diagnosis? It is the gathering of evidence and determining a model of the failure, why it happened, what set the failure into motion, and what would be risk factors in it happening again. If the model developed does not fit ALL the evidence, it is wrong or incomplete.
Shotgunning is ill-advised because it introduces uncontrolled variables. The two most likely times of component failure is when its design specs are exceeded and when it is brand new.
What, would be the components with elevated risk from the failure mode the evidence lead you to? Any parts that were in the current path of the failure path are likely to have been stressed. It does not mean it is was driving into a failure mode itself but it means you need to verify to yourself that it is not a problem for future reliability.

Simply replacing parts that are obviously defective, like shorted transistors, does not deal with the most important part, the diagnosis, so future reliability can't be predicted. Replacing parts without determining why, is further increase the risk of incomplete repairs that are harder to diagnosis the second time. Get the circuit out and trace what would happen if your assumptions are right, what would the current be through some transistor or diode if another transistor was found to be shorted. If the calculated peak current would have exceeded the spec of the surrounding parts, it is a good reason to inspect it carefully. One thing you find quickly is that a shorted transistor does not cause parallel devices to have excess current through it very often, if anything, they are starved of current, not overloaded. But tracing the circuit would tell you which parts would need special attention. Shotgunning increases the costs to the customer by replacing perfectly good parts, takes more of your time so further cheats the customer, and does not do anything to attack the original problem so risk of later failures increase, which even further disadvantages the customer. Good diagnostics results in cheaper, more accurate and reliable repairs, and take less time for the tech. It is good business and good ethics to do it right.

[52 Bill]

My bad on the paralleled devices. Gotta pay more attention when I'm posting.

[km6xz]

Parallel devices might face higher overall circuit risk if one goes bad, even if the parallel devices are themselves fine. Some circuits, by nature, are critical for current or parameter matching. Replacing one with an odd man out transfer function can lead to that one running with heavier load. But most circuits are using modern transistors or IC that are not as critical as it was back in the old days of SS when gain matching was essential because there was such a wild variation in transistor performance from the same production line. Up until about 1975, tubes had tighter tolerances than power transistor. Now, the transistors are great and tubes vary over a wide range. I can pull any power transistor from my stock and measure family curves on my curve tracer and see almost identical traces to any other of the same type. You used to he able to assume that with new tubes 40 years ago.
So common sense comes into any rule of thumb, but mostly, in general rules of thumb work out pretty good.
One of the things that makes a profitable shop profitable is efficiency which goes to the bottom line faster and more surely than higher prices. Having few units return for incomplete repairs is one of the most efficiency improvements a shop owner can introduce. Parts costs for stocking goes down, advertising costs go down because good word of mouth is free but overcoming bad relations with even one client can be very expensive, backlog remains manageable, time other shops spend in trying to resolve customer complaints with redo's is spent making new customers happy, and the basis of good reputations. The root of all this is good diagnostic methods and skills.
It is something that can be taught to a degree because elements of it are habit but really good diagnosticians are rare in any field. One way I got techs to think more logically was to require a brief description of the reasons why a particular diagnosis was reached. I would pick several during the day from work orders and go sit at their bench and have them defend the conclusion and how it fit ALL the evidence. Some hated that at first but it really increased their game. If there was a rare principle involved, the whole repair would go to the blackboard and all techs could walk through the evidence. At first I walked around all day to tech benches to help wth one diagnosis after another but I soon figured my time would be better spent helping techs learn a new approach that would not require so much of my time. With 15 techs and 24 workbenches full of gear, it had grown to be a major task for me. I love the evidence gathering phase and nailing an complex diagnosis, it makes repair fun, not work. Having good repeatable measurements by keeping test gear calibrated and predictable in performance really helps in making the evidence be less ambiguous. As a result of that need, every shop should have a couple basic calibration references. A saturated cell voltage standard or a precision resistor and a high stability time base can be the tools needed to keep every other device calibrated because voltage, resistance, capacitance, inductance, time, frequency, current, can all be derived from a couple knows. Resolution of some of the newer low cost digital meters are pretty good but for anything dealing with engineering, other than time, 3 significant figures is all that is needed. We can do a SPICE model run and get lots of significant figures but no matter how carefully your prototype construction is done, the best you could ever expect is 3 SF. Almost nothing in analog practice is as the well understood theories predict to any more than those 3 pesky significant figures. So any test instruments can be off a bit in precision and still be effective tools but no in stability where repeatable readings are needed. The same stimulus needs to read consistently , even if it is consistently off in accuracy by 1/1000...except in time measurements. I have an oven stabilized time base used to sync everything. It stays on all the time and has a drift within 5 x 10^10 per day after a 3 day warm up.

[DaCentaur]

Thanks for all the great replies and I do try to always do quality work but this time I was going for the quick fix because this has become my unit which I dont use or value that much. I am an electronic tech for the Kentucky State Police which keeps me busy and like a car mechanic who usually drives a jolopy. One of my big mistakes with this was not biasing the amp after I replaced the transistor the first time and I do always try to find out why it blew in the first place. But I was under the impression that this was an ill designed product by Fender and its circuit and components were inherently bad. That is why I was asking if there was some mod or other components I should replace with beefier ones. I have 27 years in electronics but unfortunately most of my positions were not component level and therefore I have lost alot of knowledge.
Thanks again and this time I will remember to bias the amp after replacing the transistors. Is there something else I am missing in this circuit besides checking the resistors/transistors. I remember from the other thread some other component others were having difficulty with...I will go back to the original thread and see if I can find it as well.
Thanks!!!

[Enzo]

The internet is a wonderful thing, but problems are WAY over-represented on it. Imagine if you went to the local Best Buy store and hung out at the service counter. You would only meet people with problems and defective products. SHould we then assume all products are defective? No.

When amps blow up, we can usually expect certain things to fail. But there is a great huge difference between "These parts usually fail," and "These are the parts that are usually bad WHEN the amp fails."

The Passport amps work just fine, and the vast majority of them never have a problem. The components in them are common parts used in many amps from many makers. To claim these are ill designed and the parts inherently bad is unwarranted. It simply is not true. Unfortunately you have one that has failed, but I would not write that off to them being poorly designed. Too many people take their one example of something then decide that therefore all of them must have some systematic problem.

[km6xz]

A manufacturer does not like in-warranty returns but will tolerate a certain percentage of failures because reducing failures by a any significant amount will cost well over the the price of fixing a typical number of returns, and drive up the selling prices as to be noncompetitive. It is a balancing act between making it within the buying range of the largest group of potential buyers versus reliability secured by using more expensive manufacturing techniques and components. Currently, people are getting the greatest range of features for the lowest prices in history because manufacturers have gotten really good at that balancing calculation. The Asian manufacturers have raised that balancing to an art form. We can afford a far wider range of products than ever before due to that great value
How much is the acceptable failure rate? Industry average is well below 5% and the worst units for failure are around 5% of units sold. Some, internally brag about 2% or less across their whole product line. Behringer accepts a higher rate, around 5%, for the ability to sell for such incredibly low prices. Still, even that relatively high rate is better than just about any company 30 years ago. The higher the end, and lower production levels, and higher the price, that failure rate can be quite a bit higher. Specialized boutique gear is not more reliable, in fact is often finicky, and less carefully engineered than general purpose high volume products because small producers do not have the engineering staff to handle optimizing that ratio of engineering value to the price point. But that higher potential problem rate is accepted by buyers for the more specialized specs or features that buyers seek.
In this case, Fender does not sell the Passport to advanced users who are looking for extremes in roadworthiness or performance. The typical customer walks into a store and asks the salesman for a portable easy to use sound system that does not weigh much. For that type of customer, it is a very good match of modest needs and value so it sells well. Very well, so if you see a few, it is an indication that it is not a defective product with defective design or defective components. They sell by the thousands yet few techs have worked on more than a few dozen. There are more stores than techs, many smaller towns have no techs and even some large cities have no qualified techs in pro audio so you would expect those failed units would be filling the limited number of repair shops shelves. We worked on them but never saw that many, maybe 30 total for a shop that handled more volume than any shop in a region with 20,000,000 people. Did our region just not buy them? Go to small cafes, service clubs, churches, AV departments of schools or non-technical businesses/office buildings and you will see a great many of them. I attended a seminar for 30-35 attendees recently in a conference room of a 5 star hotel and they wheeled in a small AV cart with the pa system was a Passport. I was not surprised.
The idea that some product can be modified to make it better by a non-engineer is not well supported by actual results. In fact I see few mods that did not lower some other aspect of expected performance or reliability. There is no free lunch, if one specific parameter is bettered by a mod, the customer does not think about the other parameters which are made worse. When someone comes to me asking for mods I usually refuse to waste my time with them but I ask what they are willing to give up to get their goal. They can't have both

[DaCentaur]

ok..so I finally got around to repairing this again and now I am ready to bias this beast. Never done that before. Can you please possibly post step by step instructions to do that? I have access to just about any electronic test equipment. Thanks!!