Tek TDS 420A Score/Questions

[Forced Perfect]

So I picked up a dead TDS 420A (4 ch, 200 MHz, digital scope, has the FFT math option, VGA output) on eBay for $60 because it would power on for a moment and then die. I figured it's probably just a supply rail or something and I -might- be able to fix it and it might make an interesting YouTube video of me failing at it (keep in mind I'm an electronics newbie ).

The damn thing works! Man, just when you thought Dave was the only one with the repair curse. The CRT brightness was really low and they saw the flash of green on power on, but didn't look closely and wait for it to start up. I just opened it up and adjusted the brightness pot. Passes self tests and everything. Hell, there wasn't even much dust in it.

This being my first digital scope (I have a 20 MHz, 2 ch CRO), I have a couple minor questions. I'm not going to be doing anything really fancy on this thing beyond messing around with Arduinos and whatnot. So mostly low frequency PWM troubleshooting, etc. I have two onehunglow 1x/10x 60 MHz eBay probes from my CRO.

Should I invest in getting a real Tek probe or two used on eBay (vintage of course )?

Is there anything else I should know about this model? I can't seem to find too much information on it beyond specs and service manual. Common failures, neat tricks, etc. I found you can actually save a color image to disk by picking the EPS color format instead of TIFF. It gives the channels different ones while keeping the UI the same black and white.

It came with the rack adapter, but I took it all off and tossed it. Too big and I can't be bothered to resell it. One cal sticker was broken, but the other one was intact. Back has printer, serial, GPIB, and VGA interfaces.


Frontend board along with all the ADCs, RAM, etc. The cards include the main CPU card along with the option boards (VGA out, etc.)



Little bit of adjustment and I get a nice, bright image.



Saves screenshots to floppy with a button press.

[tautech]

Well you can be right pleased with yourself. 
Now you can afford whatever probes you desire. 
The cheap P6100 probes on the 'bay will be just fine for a bit better than those you have, although if you don't need to accurately measure rise times and the like, stick with what you have.
Many EE's have a "working" set of probes for everyday use and the P6100's are perfect for this.
If I were you I would also get some cheap 250-300 MHz probes, although not renown for their performance at their max bandwidth, will serve you well at 200 MHz.

[T3sl4co1l]

I have a '460 -- the, I guess, earlier model without the floppy and extra options, and 350MHz / 150MSa.  Main complaint I have is how clunky the menus are -- sadly, something Tek still hasn't fixed to this day.  (Anything TDS and up is post-glory days -- the 2000 series units are probably the best, with the 2465 being top.  Tek's new management in the 90s really screwed themselves over; what's more amazing is how their brand somehow continues to be lauded, despite their poor instruments.)  That, and the "signal path compensation" is a bit drifty.

Maintenance: recap it.  You can inspect the boards and caps, you can wait until it fails, whatever.  But sooner or later, it's probably going to need it.  I had to replace every SMT electrolytic in mine, because they all leaked electrolyte, causing corrosion.  Which fortunately doesn't seem to have caused any faults (..yet?).  PITA, but better now than later when the joints are crusty and corroded (if they aren't already).  Only the Tek boards seem to be a problem; the PSU and CRT modules are 3rd party and don't fail the same way.

Tim

[Forced Perfect]

Maintenance: recap it.

I was looking at the calibration settings and all but the signal path test have additional requirements (ranging from "software from Tektronix" to NIST sine wave gens, etc.) The service manual says the cal is lost when disconnecting the batteries. (seriously, the scope has NVRAM for waveforms, but they put the cal in damn volatile memory?) So I'm hesitant to remove the cards to really inspect things. I assume the batteries can simply come out with the board and keep everything going? From what I can see, the caps are in good shape and the power supply filter cap has a sticker with "IN-CIRCUIT TEST JAN 16 2002" on it, so it's possible mine has a new power supply or is somehow that new? As I said, the thing had hardly any dust on it.

My "lab" is pretty sparse right now. I don't have a function gen, etc. to even do a sloppy recal if it is lost. Although I really would like to do a proper teardown.

[nctnico]

Your oscilloscope says 'copyright 1996'. That means it is a later model which doesn't have the leaky capacitors. I'd leave the oscilloscope as it is.

[Forced Perfect]

Your oscilloscope says 'copyright 1996'. That means it is a later model which doesn't have the leaky capacitors. I'd leave the oscilloscope as it is.

Ahh, excellent. I suspected it was released pretty late in its life from reading a web page on the floppy interface. It said serials from B050100 and up include the floppy drive (option 1F) by default and mine is in the 80s.

[tggzzz]

If you are buying probes and want to look at high speed digital signals, consider the accessories that come with them.

In particular, consider the self resonant frequency of the tip capacitance and the inductance of the ground lead (~1nH/mm), e.g. 20pF+100mm -> ~100MHz. For a given probe the only way to raise the SRF is to reduce the inductance - i.e. keep the lead short.

Other than direct connection (see Bob Pease's & Jim Williams' application notes on probing), the best ground leads I've seen looks like an articulated 20mm long needle on the side of the probe. IMHSHO the difference between those and the croc-clip-on-wire is significant.

[Forced Perfect]

If you are buying probes and want to look at high speed digital signals, consider the accessories that come with them.

In particular, consider the self resonant frequency of the tip capacitance and the inductance of the ground lead (~1nH/mm), e.g. 20pF+100mm -> ~100MHz. For a given probe the only way to raise the SRF is to reduce the inductance - i.e. keep the lead short.

Other than direct connection (see Bob Pease's & Jim Williams' application notes on probing), the best ground leads I've seen looks like an articulated 20mm long needle on the side of the probe. IMHSHO the difference between those and the croc-clip-on-wire is significant.

Although I probably won't be doing stuff quite this fast (at least while I'm learning), this video is quite interesting (as are all of Mike's videos ).

[tggzzz]

If you are buying probes and want to look at high speed digital signals, consider the accessories that come with them.

In particular, consider the self resonant frequency of the tip capacitance and the inductance of the ground lead (~1nH/mm), e.g. 20pF+100mm -> ~100MHz. For a given probe the only way to raise the SRF is to reduce the inductance - i.e. keep the lead short.

Other than direct connection (see Bob Pease's & Jim Williams' application notes on probing), the best ground leads I've seen looks like an articulated 20mm long needle on the side of the probe. IMHSHO the difference between those and the croc-clip-on-wire is significant.

Although I probably won't be doing stuff quite this fast (at least while I'm learning), this video is quite interesting (as are all of Mike's videos ).

Although you might not be, the circuit logic might. 25 years ago sub-nanosecond risetimes were common, and some modern logic (e.g. FPGAs) will be much worse, of course. Don't forget for a "mere" 100MHz square wave, you need >500MHz to see the edges (a rule of thumb, of course)

An example of the probe ground spear I like can be seen in figure 1 of http://cp.literature.agilent.com/litweb/pdf/54006-90002.pdf . I find that easier to use than the common spring mechanism shown in the video, but the objectives and overall result can be the same.

[maxxim]

Hi, what is the value of the tantalum capacitor on the underside power supply pcb (hybrid pin 13-19).
Thanks