What did you last repair?

In the 1980's I remember working on an analogue modem.
It was picking up 50Hz.
Power supply was fine so wasnt from there.
Then my boss came in and asked why I had the sig gen on top of the scope.
I said because its handy there.
He said there is a mains transformer in the sig gen radiating into the scope.

Funny and it reminds me of when, in the 90's, I was going to solder a couple of XLR connectors at the back of an equipment rack at an FM radio station. I'd parked the Weller soldering station on top of an audio processor and before I could get started I was called away to attend to something else. Then later on the program director of the station came to tell me that they were hearing a strange noise in the background sometimes. I listened and heard a pulsing sound that would go for a while and stop and come back again later. It turned out to be the Weller's power supply coupling into the audio processor. When the iron's tip cooled enough the supply would turn on to warm it back up.
 
Of course the above trick only works with a battery-operated voltmeter; anything with a 3-wire power cord, such as a scope, would require a differential measurement.

Does anyone else out there have a problem "seeing" these grounded-emitter circuits, or is it just me? 🙁

Does make simple things like scoping the rails more difficult, especially checking the operation of class H switches. You have to resort to an isolation transformer for power, and another one for the SIGNAL, since both the source and the scope have grounds.

As simple as these amps are, I don’t build them that way at all because monitoring during development and tuning is such a pain in the ***.
 
I was chasing down a 50 kHZ “oscillation” in a power amp for about a week. The oscillations only occurred with the volume pot turned all the way down, and followed that one amplifier around. Another being built at the same time was clean so it had to be the amp, right? Wrong. Turn off the lights and the “oscillation” was *gone*.

My LED/SMPS 4 foot linear shop light fixture is 3 1/2 feet above my service bench.
Nice and bright like sunlight with the GE LED tubes instead of the original floro tubes.
But that nasty SMPS in it... interferes with me doing alignment on radio/tuners particularly AM, near the bottom of the band. (55-700kHz).
So I have to turn off the fixture and use my 100w halogen swing-arm lamp.
 
My LED linear bulb is Phillips and is running from the same magnetic ballast as the fluorescent lights were.

My main LED "worklight" (it's more for lighting up a parking lot) flickers at 60Hz even if it's switching...

The only light I ever used in here that threw off HF / RF garbage was a cheap HID H7 car light. The light was fine, the ballast though? They are noise bombs! I can get FM radio to go away in the parking garage if I turn on my HID fogs.

Other than that, I've had good results and luck using SMPS. Frankly, I can't imagine what the PSU in this computer would look like if it was linear, but it would probably be the size of a welder.

And in a low power stand by situation (12V supply to light a LED and power a power button relay circuit) I built two versions. The linear supply used 3 watts, but the SMPS only used 600mW. Over a period of years, and times X number of devices, that's a giant power savings potentially.

That and I can buy 5 of those little SMPS for the cost of a single transformer, never mind the caps, regulators, etc. and it's a no brainer.
 
Frankly, I can't imagine what the PSU in this computer would look like if it was linear, but it would probably be the size of a welder.

In 1975 I bought the SWTPC (Tiger Amps) MC6800 computer system. It had 2 Kilobytes of static RAM and ran at a blazingly fast 921 KHz. You can see the fat cap and power transformer in the linear supply in this picture:

SWTPC

By the late 70's I had upgraded (2 MHz 6809 CPU) and expanded (192 KB of paged memory) this system with DIY color graphics (MC6847 chip), cassette, reel to reel tape storage and even an 8 inch floppy disk system. By then it took up a whole workbench and had 4 or 5 power transformers. The room lights blinked when you turned it on. The combined 5 volts supplies totaled about 60 amps, most of which fed the many DIY 16 KB static memory (2114 chips) boards which occupied several DIY SS-50 buss backplanes.

It took 4 of us with similar systems (Motorola computer club) several months to get a static image of the Starship Enterprise on the TV screen at a 256 X 192 resolution.

Today a modern $2 PIC chip could seriously kick it's A$$. I donated it all to a museum in the 1990's.
 
I always found it more enjoyable to write assembly code for the 6809 than for the 6502. But the companion graphics processor in the C64 was *way* better than that 6847. I tended to do things that didn’t involve a lot of graphics. I had a buddy that got pretty good at coding for the 6502 - his games were better than anything you could buy.

When everything went to high level scripting on MS based systems I just lost interest in coding.
 
And in a low power stand by situation (12V supply to light a LED and power a power button relay circuit) I built two versions. The linear supply used 3 watts, but the SMPS only used 600mW. Over a period of years, and times X number of devices, that's a giant power savings potentially.


Energy savings are good, yes.
However, "too much of a good thing" can also lead to problems, and in some cases offset reliability, durability, and actually cost more due to needing repairs.


In my shop through the years, I repaired thousands of VCR's for customers.
The older models used a linear supply, eventually going over to SMPS.
The linear supplies rarely ever needed service, instead, these VCR's had the usual mechanical and consumer-related issues.
Once the SMPS machines started coming through the door, the increase to repair the SMPS was amazing. - we techs saw this dramatic change.
And the troubleshooting time/labor involved also increased repair prices for the customer.
All attributed to saving a few measily stand-by watts and reducing the size and weight of these machines.
........So much for evolution/technology. 😉
 
That's not necessarily because of SMPS, but more likely from cheap parts. Almost every SMPS I've fixed had a bad cap, or was underbuilt. Some had shorted FETs.


If you were one of the techs at my shop between 1991 and 2015, you'd know where I was coming from.
Most SMPS VCR's had well-designed supplies, and yes, suffered from cap issues. The one common suspect was a 1uF 50v cap that kept the supply from starting up.
But depending on the machine, some of these supplies were a beast to get to and service - resulting in more time to repair, thus costing more shop time.


One good thing though... in 1995....

A woman brought in a 1988 GE vcr that she only used as a clock on top of her tv set. She said her son bought it for her as a Christmas gift and she never put a tape in it!
One day the clock wasn't lit.... thus she brought it to me.
Diagnosis = that 1uF cap, and a few others, costing $80 a reasonable standard repair rate. (we check/replace the belts, clean/lube as standard procedure)

She turned the service down, told me to keep the machine, she was going to walmart for a new one.


Upon some investigation, this machine was a Panasonic "high end" model made for GE, loaded with features, including a digital freeze frame, video editing, and..... quadrial picture-in-picture along with dual tuners!
40+ controls and switches!
And a viscous damped drop down access door!
Illuminated tape window!
LED bar graph stereo audio/video meters, the works!

MSRP was about..... $960.00!


I wound up fixing it, brought it home, I still have it, its still flawless for tapes!
The drop-down door's got a ton of touch pad controls too, as well as the top of the case.
 

Attachments

  • ge vcr front 9-7885.jpg
    ge vcr front 9-7885.jpg
    73.8 KB · Views: 136
Last edited:
But when it comes to audio and digital circuits the high frequency garbage from the marginal SMPS takes its toll. Put a good analog scope AC coupled and trigger off of the the residual high frequency noise on the SMPS power supply in the circuit and see the nastiness for yourself. The circuit designer expects perfect DC power and the SMPS designer is working to make a price point product. The accountants choose the cheap SMPS and divergence happens.
Loads of examples of things that fail after a while.
The AT&T 3G cellular home hotspot for people with poor cellular phone coverage had wall plug SMPS that had a short life. People would have increasing problems to the point of complete failure. Users would test the power supply with a multimeter and say that the power supply is fine because the DC voltage is correct. They would then throw them away.
 
I assume you gave her some cash, a working VCR, or at least a clock for it, right?
Probably unknown at the time. Once a customer abandons a repair, they are escaping the refused estimate fee. So in effect they are being paid. It may seem heartless, but they made the choice and don't deserve anything.

Most shops lose enough on abandoned equipment. It's nice to get something useful now and again. Its even worse for an abandoned completed repair. However, if I know something is well worth repair, I will often give the person something for it. That's if they are nice about it.

-Chris
 
I assume you gave her some cash, a working VCR, or at least a clock for it, right?


Regarding Chris's post right after yours, he's got it right.
We've had plenty of abandoned stuff left at the shop - people think we were a garbage dump I suppose, and with the "throw away" mentality they simply bought something new.

And no, the woman spoke with me over the phone when I called to give her the estimate.
She never came back.


I was lucky plenty of times to own some nice stuff to repair, and some already serviced, abandoned due to customer ignoring it.
Who am I to complain about some cosmetically great stereo equipment left behind?


Once, a guy brought in some "high end" equipment that his brother who passed away owned.
He didn't want any money for the stuff, just told me to give them a good home.

The cream of the crop was an absolutely stunning "like new" Denon DP-59L turntable, not a blemish on it, dust cover still wrapped in plastic, and w/owners manual..... with a Grace F9E cartridge that had a custom re-tipping by Soundsmith - a ruby cantilever too! - cost for that new needle was $450 I was told.
So... I put the turntable on Craigs List asking $800 firm for it.
And a guy grabbed it up less than 2 weeks after I listed it.
800 tax free smackers ain't too shabby.
 
But when it comes to audio and digital circuits the high frequency garbage from the marginal SMPS takes its toll. Put a good analog scope AC coupled and trigger off of the the residual high frequency noise on the SMPS power supply in the circuit and see the nastiness for yourself.

Additionally, you can even put the scope on a piece of gear further down the line in the audio chain, and find out the latter gets messed up as well because of that SMPS, even if it has a LPS itself.

Even audio captures out of a DAC showed this.

All timbres are affected. It also robs a lot of low frequency energy. High Freqs take on a very unnatural tone, and the overall sound is 'excited' to the point of being fatiguing - sure signs of digititis...

Some of the noise is inherent to how SMPSes are built.
 
Hi YashN. I've been dealing with USB isolation myself. I have a DAC that I love ... The differences in listening between the external supply and the USB were very significant. Also having the laptop charger connected to or off of the MacBook was a noticeable difference when using USB power (off was better).

That experience got me on the path to finding the best solution for USB as a source. I bought Analog Devices' ADUM4160 evaluation board which galvanically isolates the usb path between ends...

I will be powering these devices with four LT3045 voltage regulators and linear power supplies. I hope to get it all mounted into a nice chassis of its own..

My experience reflects yours.

There's a Doodlebug Isolator that you can look up which also uses ADUMs. For my DAC, it wouldn't cut it at the time because of the higher speeds needed for DSD128 and 256, so I didn't have access to any chip, just built a new USB connector on pure Electrical principles. Later came a chip for the higher speeds from Silanna if I'm not mistaken and some commercial gear use that.