Getting a scope again

[Th3_uN1Qu3]

As some of you know, due to an unfortunate incident (family problems) i had to sell my scope in November.

I looked around and $300 could buy me a Tek 2235 - however it would've taken about 3 months for me to save that much, and that would be 3 months without buying random parts, and we all know that isn't possible.

So, i found a 2x 10MHz National (Panasonic) scope for $106. That should be plenty enough for SMPS work. I used to have a 100MHz Russian-made single channel scope, but i had no use for the bandwidth as it was too low for FM radio and motherboard clocks anyway, and too high for SMPS and audio. I could have used an extra channel instead of the bandwidth, and that's exactly what i'm doing now.

Before telling me to check ebay... 1) I lost my credit card. 2) With shipping costs, i'd end up around the same price. Scopes tend to be on the big and heavy side so shipping costs are killer. Also this scope is advertised as being serviced and calibrated, and he has more than 1 for sale, so i believe the calibration part.

Picture is attached. The volts/div knob looks like it maxes out at 5v/div so with an x10 probe that would be 50v/div, and with 8 divisions on the vertical this translates to 400 vpp max, but i'll be making my own probes anyway, so no problem making a x20 or or x50 probe if need be. By far more interesting are the numbers on the right side - which although obscured by the angle, logic says they're 50, 20 and 10mV. With a x10 probe that's a minimum of 100mV/div - the scope i used to have had a 5mV/div setting and that was just barely enough for viewing SMPS output ripple with a x10 probe, so this is going to be some trouble. Although, whipping up an amplifier to increase vertical sensitivity shouldn't be too hard.

Anyway, i've closed the auction already and got the seller's details. He lives close to me so that was an extra decision point there. I'll probably go pick it up tomorrow evening. Oh wait, make that today since it's 2:30AM already. Back to SMPS land it is.

[KeriJane]

Congratulations!


I thought that the stated bandwidth of an oscilloscope was kind of a "minimum" bandwidth rather than an absolute one. That is, 100mhz scopes are often used for FM in the US which goes to 108mhz, just so long as the probes are capable and maybe you don't mind estimating fractions of a division on the graticule.
Or so I've been told.

Maybe your 10mhz one can go a bit higher?

[Th3_uN1Qu3]

Stated bandwidth is -3dB point. If you go higher, the signal gets attenuated more. If the signal is a square wave and you don't get at least 5 harmonics it won't look like a square wave anymore, so for a square wave signal of frequency f you need at least 5*f bandwidth on your scope to get a reasonable representation. For a sine wave, you can see up to the rated bandwidth, but it gets attenuated more and more the further you go.

There's an additional issue in that the timebase is usually correlated to the maximum bandwidth, so the signal would be so squished that you couldn't make much of it anyway.

A 100MHz scope isn't enough for checking the output of an FM transmitter, i tried that myself with the one i used to have. Anyway, i ran into some trouble and couldn't pick it up today, i'll tell you what happened some other time, as i had a really hard day.

[ben7]

Hooray! then you can also work on your ESR meter project too!

-Ben

[tom66]

I've got a test equipment store about 10 miles from me. I've bought from them:

- Tagasaki 0-30V 0-3A PSU - £30
- HP 54501A scope (100 MHz) plus one probe - £150
- Gould OS300 scope (20 Mhz) plus one probe - £60

They have towers of scopes where I went, sadly most out of my price range, I saved up my hard earned TV repair money to buy these things.

[PeteS in CA]

Triggered, very cool! Keep in mind that non-sinusoidal signals have significant harmonic frequency components, so above ~1MHz square wave signal edges will look sloppier than they really are. OTOH, you will be able to see signals greater than 10MHz, just not to scale. A signal generator putting out a sine wave signal and a 100MHz (or better) scope can help you plot out what your scope can do above 10MHz.

[Th3_uN1Qu3]

Alright, picked it up today. It wasn't that close to where i live actually, thankfully there's the subway. The seller was a funny old man which owns a TV repair shop - the kind i used to see when i was little, with tons of equipment all over the place. He had everything from CRTs to huge plasma TVs, stereos, receivers, and dozens of speakers. I might drop by sometime later to pick up a remote for the 28" CRT i found on my block a couple months ago, as i couldn't find a fully compatible one yet.

He told me that he'd have work for me if i lived closer... I guess i made a good first impression. He also gave me a few tips that will come in useful.

As about the scope, i got a pleasant surprise: it's pretty small and portable. And no noisy cooling fan either, it doesn't use one. I also got the parts needed to DIY a couple probes, will update later with pics.

[Th3_uN1Qu3]

I played with it some more and here's my conclusions.

The good:

• It works. Duh. I mean, i tested it before i bought it, so this one was obvious.
• The controls feel nice and all of them work fine.
• It has special trigger modes for TV repair. CRT TV that is...
• The tube brightness is usable and focus is good.

The bad:

• The picture has a slight trapezoid distortion, with the slanted edges horizontal. Actually, a bit more than slight. If i put the trace at the very top, the right edge ends up one tick lower than the left. I'll have to look into it. As long as i don't hit the last division it's more or less alright.
  
• The internal calibration source is 50Hz, so it's basically useless. Actually i got the compensation on the 10:1 probe wrong first time and the internal calibrator would have me think everything's fine, when at 1kHz a square wave looked almost like a triangle already. I used my soundcard to generate a suitable calibration signal.

I played the youscope demo on it and it looked great, only for some reason it appeared upside down. Anyway, for now, i'm happy with it. Gotta build a couple more probes. I'll check the caps in it these days, i have the feeling a recap wouldn't hurt.

[PeteS in CA]

If it's more than 10 years old, recapping would be a good idea. Look for date codes on caps and ICs.

[Th3_uN1Qu3]

If it's got ICs, they're not in view. I saw a couple regulators but that's it. There aren't that many lytics to talk about - the two large filter caps check good, and for the small ones i'd need to lift them out and check on the capacimeter as an ESR meter isn't useful here.

It seems like a pretty simple (and cheap) design. It even has a large trimpot in plain sight to adjust the amplitude of the internal calibrator signal. So that calibrator has officially become entirely useless. Wrong frequency and unknown amplitude. There's a coil around the tube that i have no idea what it does, as IIRC oscilloscopes used electrostatic deflection. Maybe degaussing or magnetic field compensation? It has a resistance of about 570 ohms and has 0.783v DC on it when the scope is on. No AC signal could be detected on it.

I still have no idea what's up with the geometry distortion.

[KeriJane]

Maybe it's a delay coil?

Internal calibrators are meant for adjusting the probes. Mine says that it isn't meant as a precision reference. The amplitude and frequency are adjustable on mine... perhaps yours are too?
Of course you have to check it with a calibrated scope while adjusting, so maybe a signal generator could help.

Have fun!

[NormanSaxon86]

http://www.instructables.com/id/Make...t-computer-mo/


Cool link for old CRT monitor

[Th3_uN1Qu3]

I know that. You can do it just for show but not more than that... Also, i don't happen to have any crappy little CRTs lying around, just two 21" Trinitrons and there's no way i'd do such a thing to one of them.

[Th3_uN1Qu3]

And here's a pic of both channels working, checking out the outputs of a SG6105D. That glitch on the other channel when switching shouldn't be there - i'll take a look whether it's still present after the driver transistors. Looks like someone skimped on some filtering on the controller, but it isn't much of a surprise is it?