final tests for repairs

[danmagruder]

This is my first post here and I would like to thank everyone for their wonderful input and willingness to help others. I have been repairing guitar amps and such for a while now and am having a tough streak where the amps are coming back. What tests does everyone do to ensure they have gotten the problem fixed?

[Chuck H]

When I first started I had that happen a couple of times. The real problem was that I was in over my head. I gave up trying to repair amps for a while. Now I don't have that problem but in the course of things I realized that many times simply repairing the amp by replacing broken parts "fixes" nothing. You need to find the underlying cause for the failure. Sometimes it's user error. The owner is using the unit in some inappropriate way that causes a specific failure. Experience will also teach you what those things are. Typical bench test would be all voltages and current conditions in the unit are correct and the unit "works" and allow it to sit on for several hours before testing again. Experience will also teach you when you don't need to do this. That is, there will be some repairs where the cause is obvious and simply using the unit will tell you if it's fixed. The most important thing is to understand why something broke rather than simply identifying whats broken inside.

[danmagruder]

Chuck,

Thanks for the reply and the advice. One of the problems I am having isn't necessarily making sure it works but the "does it sound right". I work for the U.S. Navy Music Program and all the repairs come in via shippers. I have little or no interaction with the players. I am a sound tech and have good ears but do you have any advice as to how to determine if a particular amp sounds like it should?

[jrfrond]

What is the nature of your bounceback failures?

Guitar amps, especially combos, can be funny animals. A static burn-in under load is a good start for final QC, but doesn't suss out microphonic tubes, or other defects affected by vibration and/or transients. Also, ALL guitar amps need to have routine maintenance performed before buttoning them up and testing. This involves cleaning pots, jacks (don't forget the EFX loop jacks, which are often the cause of intermittent audio) and switches, cleaning and/or retensioning tube sockets, board vibration and flex testing for cold solder joints, grounds tightened, especially where they are completed through chassis-mounted jacks or pots. Once all of this is done, then give it the torture test with a guitar and nail it! I don't care what amp it is, blast it on 10 and see what happens. Microphonic tubes are easily weeded-out like this. You will sometimes turn up cabinet or chassis rattles, which can then be isolated by running a low-frequency sine wave from your audio generator, because the rattles will be heard easily over the pure tone. This sometimes will turn up speaker issues as well. When all passes muster, only then should it be released, and in our case, we also do a cosmetic cleaning or the panel and covering (Armor All works great on Tolex) and dust blowout as well, especially important for fan-cooled amps, which accumulated the "dust bunnies". In amps with reverb, we always check the reverb on 10 while playing, because you will sometimes get feedback from the tank, caused by either the mount being too tight, or standing waves in the pan. A trick I learned from my good friend Andrew Barta at Tech 21 is that this can often be quelled by bending out the sides of the pan slightly so that the corners do not touch. Neat trick!

The idea is to abuse the amp and make sure it doesn't fail BEFORE it goes back to the customer, which is NEVER a good thing. Even then, sometimes you will get returns, as customers find new ways to abuse their gear. Nobody's perfect.

[Chuck H]

In the case of sound reinforcement it should simply be that the unit works properly. In the case of instrument amps it can become subjective. Especially with tube amps where so many parameters can be +/- by a large margin. It's been said before that often guitar amps sound best right before they blow up. In that case if you fix whatever is causing the amp to fail then of course it will sound different. Often to the dissapointment of the owner.

Since you don't have the benefit of the users at hand all you can do it see that the unit tests correctly for voltages and current and seems to function properly. As in amplifies the intended signal to the correct wattage without any problems. In the case of tube guitar amps you may also want to test the unit in overdriven conditions to see that the square wave is somewhat symmetrical without excessive crossover distortion (in AB type circuits) and no squealing or other signs of oscillation. What can you possibly do beyond that??? Unless the Navy is willing to designate an office for amp modification just to do "tweaks" for the musicians

[jrfrond]

In the case of vintage tube amps, the repairs may also incorporate some electronic restoration. What this means is that you need to replace resistors and caps with components that are as close to the original as possible in terms of construction, BUT electrolyitc caps need to be down-valued, while carbon comp resistors need up-valued to be close to the component value prior to failure. In the case of noisy plate resistors, they can be measured prior to replacement. This "pre-aging" of components will further ensure sonic satisfaction for the customer.

The old adage that amps sound best right before cooking is usually the case with output transformers and power tubes. It all sounds better when it runs hot!

In the case of sound reinforcement it should simply be that the unit works properly. In the case of instrument amps it can become subjective. Especially with tube amps where so many parameters can be +/- by a large margin. It's been said before that often guitar amps sound best right before they blow up. In that case if you fix whatever is causing the amp to fail then of course it will sound different. Often to the dissapointment of the owner.

Since you don't have the benefit of the users at hand all you can do it see that the unit tests correctly for voltages and current and seems to function properly. As in amplifies the intended signal to the correct wattage without any problems. In the case of tube guitar amps you may also want to test the unit in overdriven conditions to see that the square wave is somewhat symmetrical without excessive crossover distortion (in AB type circuits) and no squealing or other signs of oscillation. What can you possibly do beyond that??? Unless the Navy is willing to designate an office for amp modification just to do "tweaks" for the musicians

[danmagruder]

I pretty much just make it work. My two most hated words are "intermittent problem". Especially when the user is in Japan or on a ship somewhere and can't be easily reached. I also get "It just doesn't sound right". What a joy those are. But now I'm just complaining.

[gbono]

Do you look at the output with a scope? I've noticed that some people are able to pick up on crossover distortion. Does the limiting circuit work properly?

Make sure the amp is putting out its rated output power and then run it at 1/8 the RMS rated output for 15mins - I'll check for overheating components during this test with a cheap IR "gun". This test can also be run with program (I have some guitar tracks, gated sine waves, etc). monitor the power supply during these tests.

Do a rattle and buzz on the speakers, by themselves, with a known clean signal generator. are speakers out of phase?

Try and get a good description of the issue(s) so you have something to work with.

Chop sticks are great for tapping boards looking for bad solder joints/loose connectors/etc.

Check all ground connections

Be patient and write down what you are doing so you have a record of what works and what doesn't.

Sounds like a fun job.

[Euthymia]

For final QA on repaired amps, while it's on the bench, I crank it up to clipping into a dummy load and make sure it's making full power.

Then after it's all put back together in its cabinet, I play through it for a while, just working out whatever song I'm working on currently for a few minutes.

A very handy thing I do, although this is not possible in your case, is have the person play through it before they leave if they and I have time. Sometimes they set the controls in a way that does not seem logical, but brings out a failure mode in the amp. Also, I would guess that if your sailors are playing in places all over the world, they may encounter electrically noisy environments with transformers, dimmer switches, missing ground connections, etc.

Sometimes the problem isn't even the amp. You might print up a troubleshooting cheat sheet: "try it with a different cable and try it with a different instrument before you decide it's the amp's problem."

[km6xz]

The "don't sound right" problem is often one of user memory of what it most recently sounded like, which might be very different from when the unit was working properly some time in the past. What they are remembering it their impression of the sound, not the sound itself. Humans have very short memory of sound characteristics but long memory of our impressions or internal dialog we used to describe it.

The return rate is in direct inverse relationship to experience. The longer you track down problems that are less than the usual the more mental checks you account for in diagnosis. A proper diagnosis is one that explains ALL of the symptoms as the expected, reproducible result of the identified mechanism of fault. MoF is the process and progression of the fault, why it happened and what happened. The vast majority of "tech" do not understand materials, electronics or mechanical properties to ever be able to form quick accurate diagnosis, but instead treat symptoms and replaced parts, and when asked why the problem occurred can only say"they are always doing that" or "its a cheap piece of junk". The reason there really are no good techs in pro audio or M.I. less than 45 years old is part due to the experience and part because there was more of an appreciation for science and math when the old timers were first involved. There are many other issues but in decades I have not found a tech in the field who was good at diagnosis who was young. 30 years ago there were lots of young people who really knew electronics. You had to to get anything done in the hobby because there were not many alternatives than doing it yourself in designing gadgets or systems.

Any report of gradual deterioration of sound requires questions about anything they use in the front end, any peddles, effect loop units etc. Batteries are often last suspected.

John had a good list of routine work to be done after the unit was returned to original operating parameters.
One thing that helps both you and the owner...measure the good as well as the bad. No unit leaves my bench without base line measurements of power consumption, gain, noise floor, a capture of spectral distribution, distortion, frequency response, power out into all the impedances that it is rated for etc.
These help you prove your work returned it to factory specs. They also tell you precisely if a unit fails to meet one of those prior made measurements.
Just plugging the amp into the metered variac reading idle current from the wall compared to past units of the same version is a good head start in diagnosis.
To make these measurements not only as insurance proving your work to yourself and to the customer, it saves time. When set up correctly all these tests can be done in 5 minutes. Doing so can save hours or even end any debate with a customer of it "not having the power it used to have" or "its not clean anymore".

What to work on, if left up to the customer, is often the wrong item, now more than ever a unit is part of a system. It was quite common for someone to bring in a piece of equipment and describing the problem that could have been caused by many parts of the system, not just that one unit in the system, I would ask why they pulled that item out of the rack to bring in. They would usually say "the other items were too hard to get to" or "the others were too heavy". Ask right upfront how the customer proved to himself that the unit he sent in was the source and victim of the fault. You will be amazed by the lack of logic and reasoning that went into the determination of where the problem is.
You will get lots of people who are very specific in what they want done.."change the tubes and bias". Don't take it for granted that is what they really want. Ask why. Usually the answer is "some guy on the internet said if it blows fuses it's the tubes".

After a unit is repaired and meets spec, time for burn-in.
Set up a burn-in rack. A signal generator, load bank and scope. Mine had a large switch on it to switch which amp the scope was looking at when 10 amps were burning in at a time. Each power line had its own breaker. Run the amp at slight into clipping on peaks if solid state for 3 hours with a music signal, radio due to its compression and peak limiting is perfect. Periodically switch the generator in to look at a sine-wave at full power. That room can get pretty warm but no one has to sit there with the gear. Tube amps are pushed harder into clipping on the music. But after the first hour of hot running, re-bias, again, the amp if it got new tubes.
Some amps require a more labor intensive burn-in, like old SVTs. They would be burned in with only a pair of tubes at a time(change load Z to reflect the change in plate Z) until they have settled, after all have been burned in, put the all the tubes in, re-bias and burn in normally. The current tubes available are highly variable in their ability to handle the circuit conditions so burning-in in pairs saves a lot of time if one fails out of the box. When balanced that way they stay together for a long time, a couple years before it is needed again. If not, they will make someone poor very quickly.

Those are just some random thoughts on the subject of minimizing returns.

[danmagruder]

It is especially being active duty. I spent 10 years out in the bands as a trumpet player/ sound tech and got to do and see a lot of cool places. I also advise on what musical electronics the Navy Music Program buys. I do everything from tubes amps to digital mixers, reconing and road case design and repair. My job is all over the map and always something different. I really enjoy the variety.

[Prairie Dawg]

I tend to agree with John. When I'm finished with something, I try and burn it in for a day or two, play through it and and thump it hard with a balled up fist. Whenever possible I like to have the owner of an amp show me what's wrong with it and road test it before they take it home. I have a very light touch and some players really beat the crap out of their guitars and have very hot pickups to boot. Ron, you know who you are.

[danmagruder]

What does the burn-in process do for the amp? I have never done it before and have retubed many amps. Maybe that's part of my issue.

I have taken a few EE classes and the Navy's electronics correspondence courses but mostly I have taught myself through research and asking a lot of questions to factory techs who I am fortunately able to contact. Most of them have been amazing and have taught me tons. I do feel I lack some of the basics but I appreciate everyone's ideas here.