Drop testing your new builds? / Percussive maintenance...

Member
Joined 2011
Paid Member
Does anyone else drop test their builds?

When I build a tube amp (using point to point) and it's up and working nicely, I'll "test" it by tapping the tubes, the chassis, and eventually picking up the amp and dropping it up to 12 inches onto the desk while it's running to check for bad connections (and make sure it would survive being shipped).

Who else does this?

How many of you still use "percussive maintenance"? It's the art of fixing something by hitting it for those who haven't heard of the term.
 
After I finish building it, I turn it upside down and shake it vigorously to get stray solder / cutting out of the chassis.
If it's built right, it should survive a fall from a table (30 inches average, right). Without the tube, anyway.

I'd guess most tube amps would have a bent chassis after dropping that far, due to the weight.
Just from shipping, I've seen transformer mounting flanges well bent from dropping the carton.
 
Mine don't bend, but that's part of the design, right? I've yet to ship one though. I think I would ship it
with "some assembly required" to eliminate that problem (for me).

Later on, I found you can use a steel washer on top of the flange (large enough to cover most of it),
and prevent the flange bending from carton dropping. Double boxing, with foam in between, helps a lot also.
At work I've had equipment on a pallet (a $50k prototype EV motor) during shipping, dropped 20' to a
concrete floor by a rookie fork lift operator. We got the video, it wasn't pretty.
 
Last edited:
Almost 30 years ago, I traded a McIntosh monoblock for a pair of Dynaco MKIIIs. I packed the Mac using custom cut foam block, while the other guy just threw the MKIIIs into a box with few styrofoam peanuts. Needless to say there was a lot of glass dust, the chassis was bent between the two transformers on both amps. After straightening them out and getting new tubes, they worked. I later sold them to a guy in Qatar, shipping via USPS. Cant remember how I packed them, but no complaints.

The last two tube amps I bought off ebay suffered physical damage from negligent packing. Got a lot of spare paper towels out of one package. One arrived with a couple output binding posts sheared clean off, but the tubes survived - go figure. Most dont know how to pack such a beast -

I recently sold a 35 lb - apparently the sweet spot these days for most possible damage - Phillips receiver on ebay. I thought I did a very reasonable packing job, using sequenced foam of different density - softer on the inside, harder against the box edge. It arrived (Texas) extensively damaged, cracked PCB; I immediately refunded the full payment. Buyer and I then tried to figure out how it could have possibly sustained that; concluded it had actually ejected from the package. What the &^%$! I put in a claim and got a check for the full sale amount.

Working at Intel, I got to create a LabVIEW program to operate the product drop table. The table had two feet that fell into two pressurized piston affairs, which determined the deceleration by the foot type and applied pressure. (Of course the drop height was involved as a variable) They monitored this with an accelerometer, to be sure the impact didnt exceed spec. I'd say you wont know much about what your product is actually sustaining, just from a raw 1ft drop onto a solid table surface.

When I worked an Amazon, I got to see first hand how min wage workers handle packages. Imagine a 40 X 40 X 4" box, labeled "do not lay flat". Ka-Blammo! sound echos across the floor as it's slammed flat on the conveyor belt. I thought "these poor manufacturers"...

When I say 35 lbs is the sweet spot, I'm imagining someone picking it up, cursing it cause it's pretty heavy and giving it a good chuck. They dont really give a .... so you better pack your heavy item using cut foam blocks, with zero movement within the package when taped up.
 
Last edited:
I sometimes have trouble with SMD pcb's.
The pads are so small its hard to see if they have soldered well.
If they dont work I twist the pcb a little to see if it makes contact.
Any bad joints show up pretty quickly.
I also always put a line of vias along the pins so I can check for shorts.

On a TH pcb I once found a fault by twisting the pcb.
I had completely missed a transistor solder joint.
 
Member
Joined 2011
Paid Member
When I worked an Amazon, I got to see first hand how min wage workers handle packages. Imagine a 40 X 40 X 4" box, labeled "do not lay flat". Ka-Blammo! sound echos across the floor as it's slammed flat on the conveyor belt. I thought "these poor manufacturers"...

When I say 35 lbs is the sweet spot, I'm imagining someone picking it up, cursing it cause it's pretty heavy and giving it a good chuck. They dont really give a .... so you better pack your heavy item using cut foam blocks, with zero movement within the package when taped up.

Like "phonograph/photos, DO NOT BEND"... Or "Fragile"... I worked in the shipping industry for a time, too. "Fragile - Handle with care" means "kick me!".

If I pack something for shipping, before it even hit the mail, I throw it on the floor as a test (gross generalisation). If it can't do that, it'll get there broken and I've wasted my money and everyones time.
 
Always assume a drop from 2m high for your shipment...


About tubes durability: once I dropped a ~10kgr monoblock from the bench to the floor:eek:. While testing... i.e. alive:eek:. It landed on the tubes. I got two Piza tower EL34 but the glass remained intact. I discarded them but the small signal tubes are still in use!
 
Last edited:
How many of you still use "percussive maintenance"? It's the art of fixing something by hitting it

Ram Tuning, "calibration", or attitude adjustment, is practically useless on modern SMD boards in thin plastic boxes. Well placed match books and tape can effect a temporary fix for the jumbled screen on a TV or computer display when one of those conductive elastomer strips fails to connect.

It used to be my first reaction when I was a "factory floor support tech" at Motorola in the late 70's. The factory test technicians were subject to reprimand, including dismissal for whacking the test equipment.....I however, was not subject to such a restriction. I just had to fix whetever I broke, since I worled in the Cal Lab.

I carried my Motorola HT220 walkie talkie in my back pocket so That I could be allerted to such a malfunction. The first response was to take out the walkie talkie and whack said equipment with the battery end of the radio. Much of the 5 to 10 year old stuff in that factory would respond favorably to this, since it was dirty, old and worn. As the old junk got replaced by "modern" computer controlled equipment with fewer interconnects and switches, my first action became "reboot the damn thing."

I found you can use a steel washer on top of the flange

My personal SSE amp (about 20 pounds) got dropped when I was moving out of Florida. No tubes were installed at the time. It landed on its side but the 6 pound power transformer got one of it's mounting flanges ripped off and the end bells were bent. The internals were OK. The big motor run cap wound up with a gash down it's side and an oil leak. It got trashed. The chassis was originally glued together with JB weld, which held up surprisingly well, but some of the vertical seams with very minimal overlap split. Six years later it's still crying out "rebuild me." It looks ugly but works, so why mess with it.

I subject most of my stuff to plenty of physical and electrical abuse, but I'm reasonable as to expectations. I would not expect a HiFi design with exposed tubes and transformers on a base constructed of mostly wood to eat the same physical abuse as a guitar amp head small enough to be used as a football by some roadies fed by beer and caffeine.

I just do it for piece of mind. I've never build at 50k$ prototype before.

I remember the really ugly looks I got when I blew up the first of of 10 prototype boards at work. There had been over $1M invested in the design so far, and there were only 10 proto boards in existence (1985). I thought my engineering career was over before it started, but I lasted another 30 years.

Fortunately I wasn't the guy that dropped the only wafer in existence during an IC design project (1994). Over $100K had been spent to take one wafer from the que and walk it through the fab process, bypassing it's spot in the normal process line to get a test wafer about one month early. Yes, it's possible to pick up the pieces, hand cut them and package them in order to recover about 30% of the parts expected from the wafer.

The full cost of developing a custom RF IC chip is well over $1M, and time IS money. The usual processing time from sending out data to receiving packaged IC chips is 4 months. It takes 2 to 5 of these cycles for an IC design, assuming most of the circuits on the chip are known to work. There are dozens of small test chips known as "pizza masks" sent out to test individual circuits long before the "big picture" comes together.
 
Jeez, my 1625 amp usually suffers some sort of issues just from transporting it in the back of a vehicle for an hour on Wisconsin roads.

Usually its one of the power tubes that has an issue, though I did have a speaker impedance switch fail (my own stupid fault for using a knock-off part that was in my junk bin).

For being made as modulator tubes in the ARC 5 sets, these 1625s have not been holding up to vibrations nearly as well as I would like. They sound great, but I'm unimpressed by their reliability. Granted, they are about 75 years old now. The Tung-Sol 6L6GCs that are in one of my other amps have seen twice the abuse with (so far) a 0% failure rate.

Most of them need reactivation to avoid arcing as well, but that is to be expected with that age. 715s will hold off less than 1500V until you reactivate them, at which point they can take 20kV. The 715 was designed as a pulse modulator for radar magnetrons. Just about the biggest cathode you'll ever see on a 60W tube. Not really a tube I'd use for audio due to the power requirements (and because they are still quite valuable for their original use in hard-tube pulsers, usually in a scientific setting now), but an interesting case none the less.
 
I once got into an argument with a mechanical engineer about a connector that I had chosen for a cell phone design that I was responsible for.

The connector was originally specified by Toshiba for use in a laptop computer to connect the flex circuit from the display to the motherboard. I proposed using it to connect a flex circuit from the front panel board (buttons, LED's and LCD) to the main board.

To verify the constant connectivity under test, I devised a simple latch circuit with two LED's. Battery power flowed back and forth through every connection in the flex, winding up in the front panel board where it lit one of the two LED's. The green LED would remain lit as long as there was no interruption in power, but a glitch of 1 microsecond or more would extinguish the green and light the red LED, and of course a loss of connection would extinguish both LED's.

We made 5 of these "phones" and subjected them to the usual drop, stress, temperature, and humidity testing, and all passed. Then we "George tested" them. First test, toss them down one flight of concrete stairs. Then two..... one of the guys on the in house softball team threw them up in the air to let them hit the asphalt. The football guys threw passes with them, not all were caught. In the end all 5 were destroyed.

Autopsy revealed that 4 were due to battery failures (not my problem), and one had the front panel board penetrated by a foreign object (again not my problem). No connector or flex circuit failures were seen.

I devised all sorts of electrical "George tests" too. One of the usual tests for a two way radio or phone involve subjecting it to random reflections of it's own RF signal. This is usually done by moving the device around inside a metal box or cavity while observing the signal on a spectrum analyzer. This does not catch all issues. George test, stand the device next to a large fan with a metal blade. Use a remote microphone to avoid wind noise and actually talk over the radio / phone. The fan blade may cause low frequency modulation of the voice which impairs intelligibility, or in bad cases, up set the low frequency signaling used to control the device.

Other "George tests,"

Take the device into the "transformer room." This room receives an underground high voltage feed and converts it to the voltages needed to power a 1 MILLION square foot facility. There are several transformers inside, and some are about the size of a cargo shipping container. I always felt funny for an hour or two after being in that room.

The "roof test," Being a Motorola plant that made two way radio equipment, there were a zillion antennas on the roof. Most concentrated on the top of the tallest spot, so it's a "high RF environment."

The "spark plug wire test," one of my hot rod friends came up with this one since there was this blue Dodge Demon with a hot rodded 340 engine that wiped out AM and sometimes FM radio when it drove by. We would hole the radio to be tested under the hood near the distributor when the owner fired it up.

There was some "unsanctioned" lithium battery and phone testing that might have occurred at the local gun range.....

Not to be outdone, the mechanical engineers behind the early versions of the APX radio product made a video of an unaltered APX5000 being driven over with a Ford Explorer, than picked up and used. OK, I handed over my crown of destructive radio testing to him as I left the company 6 years ago.

I went looking for that video on the web, but couldn't find it. I did find the product announcement for what was still secret the last time I visited the plant. This is the culmination of nearly 10 years of research and development by myself and many others. Let's just call it the offspring of a traditional police radio and an Android smart phone. I'm sure there will be several cops that want to banish ViQi in the same manner that I would Siri or Cortana.

APX NEXT Overview | Two-way Smart Radio - Motorola Solutions