The AM Forum

Why does my very expensive Tektronix digital scope suck? It great for looking at waveforms, digital signals and pulses, has all kind of fun displays on screen like frequency and time, DC offsets and all other things but just try to feed it a RF signal. I like using a scope to look at the carrier of an AM transmitter to set modulation. My 60 MHz dual trace Tektronix TDS-210 displays something that does not begin to look anything like what my old Tektronix seven thousand series analog scope or any of the other scopes I have in the shop show. I am just looking at a sample of my RCA at 1.885 and would think that’s well within the capability of a 60 MHz scope but looks to me like the scopes input is not capable of dealing with it. I have used all different types of attenuators on the input and probes but no matter what the digital scope still display something that looks like serious over modulation and distortion but with the older analog scopes everything looks clear.  The Tektronix TDS-210 not a cheap scope, it cost well over two thousand dollars and am wondering what will happen if I get a more expensive digital scope, let's say a 200 MHz version.

Probably the digital processing within the 'scope to get the incoming info to the display is where it falls apart.
It cannot deal with a real-time input
Fred

Usually the sample rate is tied to the horizontal sweep rate setting on the front panel. When you set the sweep rate to best view the audio/modulation envelope, the sampling rate is way to low to properly sample the carrier.

It sucks because you can not control the refresh rate of the display.

It sucks because if you could control the refresh rate, either the scope "video" memory or the LCD memory, row and column addressing, and panel could not handle the necessary refresh rate per individual pixel.

It sucks because it does not refresh the entire screen in an instant but wants to draw that line. while pretending to itself that the display is the type lines are supposed to be drawn on, instead of the kind that should be refreshed in an instant. Software maybe?

It sucks because as you seem to get close to looking at what you want, how you want it, artifacts and discombobulation fill the screen.

It sucks because because it has rigidly fixed pixels with no individual contrast based on electron beam dwell time or Z axis modulation.

It sucks because because it can eat every last data-point in a real analog signal made of two vastly different frequencies but not display them promptly at all, compared to their real timesacles.

It sucks most of all because it might not have the most important thing, the one single tube (and the necessary concessions made in the digital hardware) that keeps even a digital scope from not totally sucking, the vector-scan CRT.

I'm sorry if this seems mean, it is not at you. This kind of thing gets in the way a lot where it's necessary to observe random waveforms but the industry seems to have decided how things are going to be.

There are some really new DPOs that might be better, TEK makes them.

Don't feel bad about the TDS-210 BTW. The 5GS/s unit in the lab at work has similar performance when viewing an amplitude modulated AM signal. The stupid thing knows what picture it took, it just can't get it onto the screen before the next one is ready, and I suppose it feels obliged to throw that one on there in the middle of the other one. Bah! If it's not Scottish, it's ....

I use an old HP 500Khz scope.  I feed the envelope waveform directly to the vertical plates.  Works perfect.  The plates alone are good to about 80Mhz.

I had to make some minor mods to the scope's CRT connections to make it work.

Fred

That's I have always done. I disconnected the vertical amplifier in my HO-10. Since I don't use it, it saves drain on the power transformer, which has a reputation for crapping out.

From the comments above, I take it no-one make an analogue scope with CRT display any more?

Tektronix dropped their analog series several years ago, I think the only analog devices they sell now are aimed at the video monitoring market.  For new analog you are probably forced into offbrand choices like this one:  http://www.mpja.com/prodinfo.asp?number=17522+TE

I am staying with my Tektronix 7000 series.  I think in the last 10 years I have replaced some tantalum caps in several of the plug-ins but my 7623A, 7854 and 7904A mainframes have been trouble free.  Even my very vintage 551 and 556 scopes have not given me any trouble but they don't get used very much and are overdue for a full calibration procedure whenever I have the time.

Yea, have to agree with everything above and did not realize about the time base setting the sample time. Think that may be the biggest problem, somehow assumed the scope always sampled at a high rate but by linking time base to sample time that will always be an issue. Wonder if on the higher end scopes like the new ones allegedly good out to 1 GHz if you can change this? I have an old 7603 mainframe but it's huge and heavy and I dumped my older small Tex scopes back after I got the new digital one so looks like I may be looking for an old 465 or something like that or maybe make a hole for the 7603 and its "scopemobile" in the shop. One good thing about the 7603 is you can buy them all day long at hamfest for $20 each so may pick one up for spare parts too.

Rodger,
I really enjoy my Tek, 7000 mainframe.  I have the vertical plug-in's to use when I need a scope and also the 7L13 spectrum analyzer plug in when I need a spectrum analyzer.  It has always worked very well.  Sometimes on my bench I also use an old Tek 475 which I like as well.  It has the Tek DVM which also measures temperatures with the temp probe.  That is the scope I started with back in 1973.  My Tek 2465 which is the newest generally sits on the shelf.   

Joe, W3GMS

The first one I owned was a RM 45. I got a CA plug in and entered into the dual trace world. Here is what they look like.

http://www.flickr.com/photos/56635145@N00/3440051130/in/photostream/

klc

I picked up a 7904 at Deerfield that will see a 1GHz sine wave

Maybe this explains why the ham radio appliance manufacturers discontinued monitor scopes almost a generation ago. I had just figured demand had fallen off as more and more Joe Bloe Hammy Hambones couldn't figure out how to use one.

Assuming it is a dual trace digital scope ...

Use the audio input to the transmitter as one of the two channels' input signal. Trigger off of that channel / signal. Set the triggering to "normal" (not "auto"). Don't use any waveform averaging. Set the horizontal sweep to around 400us per division (not critical). Adjust the triggering level so that you trigger off something close to the positive (or negative) peaks of the audio signal (also not critical)

Use the other channel to view the rf waveform.

What you should see is the rf waveform's envelope faithfully following the audio waveform.

Also, since you are triggering off of audio peaks, and using "normal" triggering, both waveforms will remain on the display until the next triggering event.

I usually scale and offset (vertically) the audio waveform so that it is superimposed right on the top of the rf waveform's envelope.

Stu

I use a 7603 in the shack with a freq counter plug in.  One channel goes to the IF output of the receiver, the other to a loop around the coass going to the tooner.  The last Howard Cty hamfest I scored another 7603 with a bunch of plugins including a 1GHz plug in with integral probe, a 7CT1 curve tracer, a digital pattern recognizer, and some assorted 2 channel vertical amps and timebases for $75.  Works perfectly.  I've said this before (perhaps to the point of annoyance) but I can't figure why anyone would mess with a hammy hambone monitor scope when superb quality scopes are available for lunch money...

And yep, digital scopes aren't the best for displaying a mixed analog waveform like an AM signal.

HP made some good scopes in that class that did X-Y stuff very well. I'm an RM503 fan myself.

Did you do it by isolating the deflection amps from the plates with something like 1M resistors, and then capacitively coupling the plates to the transmitter?



I'm a 7000 man too, but I do have one tektronix portable CRT-based 150MHz digitizing scope and a 453? 100MHz unit. haha it's a bit rough to carry a 7904 around.

Here are some digital photos of the screen of my dual trace digital oscilloscope showing: the audio input and the rf output of my single-FET Class E transmitter (with a Class H modulator).

The yellow trace is the audio input waveform to the transmitter.

The blue waveform is the rf output of the transmitter... picked up by a current transformer type coupler... with a 30:1 voltage step down between the transmitted rf signal and the output of the coupler. [The peak rf voltage at carrier is 1.13V/division (channel 2 scale) x 1.5 divisions x 30 volts/volt = 50.85V. Also, 50.85V peak rf into the 50 ohm antenna/load = 26 Watts of rf power at carrier]

The first photo has the horizontal sweep set at 400us/division
The second photo has the horizontal sweep set at 2ms/division

Note that the triggering level in the first photo is set to near the -80% modulation point. The triggering level in the second photo is set at around +80% moduation. In both cases, the trigering is off of the yellow (audio) waveform.

The vertical scale of the blue waveform has been adjusted to 1.13 volts/division to match the vertical amplitude of the yellow waveform. The yellow waveform has been offset, vertically, by 1.5 divisions to cause it to sit on top of the upper envelope of the blue waveform.

Note how faithfully the rf output envelope (blue) tracks the audio input waveform (yellow)

Since this is a digital scope, it continues to display the waveforms when I stop talking, because there are no new triggering events.

Stu

Stu, that's a digital phosphor scope, not a digital storage scope as was being discussed by the OP.  2 completely different architectures.

JN

Hi!

The Tektronix literature indicates that the TDS-210 is a "real time digital oscilloscope"...

http://www.metrictest.com/catalog/brands/tektronix/pdfs/40W_10992_4.pdf

If it is, then it should do fine for displaying the stuff I showed in my earlier post... if it is triggered as I suggest.

I do agree that there are digital storage oscilloscopes around (I have one of those) that work like classic sampling oscilloscopes (i.e. only useful for displaying periodic waveforms). For applications below 200 MHz, I think that most of the newer products are real time types.

As an aside, real time digital oscilloscopes (some that work in conjunction with a personal computer) are becoming increasingly capable and affordable.

Best regards
Stu

Fred,

Do you just connect your RF probe directly to the vertical plates? I have an old Dumont 500kc scope that has direct connections to the plates in back and this would work perfectly.

My xmtr is completely HB and it has a circuit that is for the scope.  Its a tuned circuit that works off a 4 turn coil that is in line with the coax.  The 4 turn coil is coupled to a calibrated pick-up coil that feeds the antenna current amp meter.  In that section there is another tuned LC circuit that picks up signal off the 4 turn coil.  The coil of the tuned LC circuit has two link coils.  One link coil feeds the built-in receiver that can be used to listen to my own signal.  The other link coil feeds the scope plates.  The scope signal is feed on coax (RG-58) to the scope.

Because you need a lot of voltage to feed the vertical plates, I have RF peaking coils that are in line with the RG-58 that feeds the plates.  The LC tuned circuit also has another capacitor that is used to adjust the voltage feeding the plates.  This is used to adjust the height of the waveform on the scope.  So, you can see that I can adjust the scope pattern right from the xmtr.

You can see how complicated a simple task can become.

To further answer your question,  you may not get enough voltage off a simple RF probe to feed the vertical plates, you need somewhere between 40-80 volts.  You have to experiment with your set-up.  You may have to use a small LC tuned circuit near your tank coil to pick-up enough voltage.  Or maybe just a small untuned link coil near your PA tank coil.  You may have to use balanced twin-lead (TV 300 ohm) to the V plate terminals on the Dumont.  Keep in mind, you will need some way to adjust the height of the waveform on the scope.  There may be some further help in one of the Handbooks.

Fred

I also have a RM-503, I use it over on the audio bench.

As I remember, I added RFCs in line with the vertical plate wires near the back of the CRT.  From one of the plates I used a .01 disc cap to a BNC connector.  The other vertical plate I added a .01 disc cap to ground.  This gave me an unbalanced vertical plate connection that I feed with RG-58.  The RFCs are needed to maintain  the correct bias voltage on the vertical plates.  I added a toggle switch to bypass the mod to return the scope to original.

Fred                   

In all honesty I have to admit that I don't really know how to use one either.  I have gotten by with the following method:   Hook up scope, turn it on, and flip switches, twist knobs and basically jack with it until I start to see something.  Continue until I get a display signal I can interpret.

I have heathkit monitor scopes I recap in the power supply and use without really knowing if I'm putting them to best use or not--I get a trapezoid display I have figured out how to get information from (neg and pos modulation peaks and linearity) and recently acquired a tek 7000 series from where I work via eBay that is bewilderingly complex.  I should read one of the many how to use an oscilloscope books but there never seems to be time for it.

Using the external trigger or by feeding the audio to channel two solved the problem, at least for when I am feeding the transmitter from the oscillator. Also now the frequency display reads the frequency of the modulation and not the frequency of the carrier.
Ray F.

That is a good idea, you and AB2EZ did it so I'll try it on the one at work and see how she's gonna act.

That sure wraps it for me!

Just the reasons I kept two Tek 2467B analog 'scopes with the enhancement plate displays....random  events that might be at the sample rate or masked by other 'modern features'. Also took my own 7000 series mainframe to work with plugins. Young guys bring customers through my lab to show them Ed's 'old school' stuff and its good for some giggles....but guess where they come for nasty problems. I do like the new real time spectrum analyzers though. The old Hp 180 series scopes are cheap, easy to find, and make good monitors


EdZ KG6UTS
now in Baja trying to annoy fish

If I knew my boss wouldn't make a fuss, I'd take one of the 7904's to work. They have maybe 15-20 scopes in the local shared lab at work, all digital. I have two digital scopes in my little lab, which is in a corner of their lab.

One is 1Gs/100MHz, a lower/older model than the OP's scope, and one is a 5Gs/1GHz unit. Before I got the 5GS unit, I traveled for training to my department's 'big' lab up north, and spotted a 100MHz TEK analog scope in the corner. I basically begged my boss to let me have that one but I was only allowed to choose from a list of digital scopes excessed from another company facility. It seems they keep that analog beauty around for special occasions! Very wise.

Does anyone have a digital scope with a raster scan color CRT? The 5Gs TEK unit is one of those, but I found out it really has a monochrome CRT and some kind of switchable LCD color filters in front of it. That is very interesting as I thought LCDs were either 'clear' or 'black' due to their polarizing function, not 'clear' or 'tinted'. It's cycling through the colors at some rate. That explains the weird headaches and dizziness from staring at it and lack of image brilliance too.