To Make A Square Wave

This has to do with that clunker HP 1725 a we got kicking around. If you kn ow the history, we got it with a makeshift vertical output amp for the uno btainium vertical output IC. Well we got the unobtainuim obtained out of an other unit. In fact I changed the whole board, figuring it would be tighter in alignment. Unfortunately someone had already screwed with the front end alignment. I should shoot that person's fingers off with a .45 but have no idea how to find him.

At any rate, got time on my hands and got back into aligning it, and seem t o be coming along really. Then I run into a problem. This is about a 1 mHz square wave on it :

formatting link

Now it's pretty hard for a fault in the scope to do that unless it has some thing to do with the delay line I would think. Or maybe lytics...

Here's the setup :

formatting link

Better shots of the waveform :

formatting link

^ That is with the amplitude rduced.

formatting link

^ That is at X10 mag on the time base.

Now the ringing is worse at a lower setting from the generator. It is a Wav etek 103. That is the 50 ohm output which is wiper fed to the output level pot. Seems a bit unlikely it would affect the ringing like that but...

Here's the but ; I looked at the BNC to BNC cable and noticed it uses RG59U wire. That is 75 ohms not 50. Is that causing all this ringing ? Other thi ng funny is a longer one of the same type of cable does the same thing, I w ould think the ringing should be different no ?

Anyway, I need to make sure the ringing is not in the scope, so I am rthink ing of getting a couple of ultra high speed diodes and soldering them right to the BNC jack on the scope to assyuure I am getting a square wave. One l ittle problem with that is this is best done at the max setting for vertica l gain, which is I think 0.5 mV/div. or something.

I think, ideally, this should be done at maximum vertical gain to remove th e attenuator from the picture, so to speak. Therefore, should I maybe go wi th a couple of Schottys, or what ? What I want to do is to make absolutely sure the scope is getting a square wave. Once I am sure the ringing is not in the scope I can move on.

The 50 ohm output on the Wavetek goes up to at least five volts. If running into the 50 ohms on the HP, say I use a 100 ohm resistor and two back to b ack diodes, what diodes would clamp this just right - be high speed and jus t clamp, leaving a perfect square wave right at the input jack of the HP ?

Or should I approch this a different way ? Maybe use an even higher voltaag e, clip it and then break it down resistively ? Whatever happens I really h ave to get that cable out of the picture.

All I can figure is with the BNCs on RG59U, they were surplus from a cable TV company or something. Maybe.

Any ideas ?

Reply to
jurb6006
Loading thread data ...

You _did_ twiddle the little calibration adjustment on the front panel, yes? Assuming it's there it'll be a little hole with a slotted shaft buried inside, that you can adjust with a trimming tool.

If it's the cable, it'll have a ring period of twice the delay in the cable, and with just a 50:75 mismatch it'll die out fairly quickly -- in less than 10 cycles, certainly. It'll also be square-edged, if you can look at it fast.

If it's impedance mismatch it's more likely the generator -- the whole pot thing isn't going to give you a very well-controlled impedance. Using some fixed (or switchable) 50-ohm attenuators should, however.

There are 75-ohm BNC connectors out there. IIRC, the pin smaller on the

75 ohm connector (BNC connectors try to maintain controlled impedance through the whole connector, so the dimensions are all controlled by the inner diameter of the outer shell). I know that one impedance plugs into the other, but not the other way around -- I just can't remember which is which.

Rather than doing all the messing around you suggest, I'd start by generating a clean square wave terminated with 50 ohms -- I'd suggest a

555, buffered by a 74HCxxx device, with an L pad that pencils out to 50 ohms. If you're really feeling paranoid about ringing, get a BNC connector that'll go straight onto your board, and just hang the whole square wave generator off of the O-scope input.
--
Tim Wescott 
Wescott Design Services 
 Click to see the full signature
Reply to
Tim Wescott

[snip]

I sure wouldn't trust a function generator to have a clean square wave! A cheap way to make a known clean edge with the function generator is to drive a 50 ohm resistor via a Schottky diode. Set the generator levels so that the diode is "off" on one phase of the waveform. The edge that you trust is the edge where the diode is off (or at least in the act of turning off). The smaller the wiring and components - and the faster the risetime of the generator and smaller diode capacitance - the better the corner. Build this into a connector that attaches directly to the scope input.

This is how the old style Tek "Type 106" and PG-506 made their 'fast edge' outputs. If you need the other edge, build the circuit with the diode pointing in the other direction. Never trust the side where the diode is off.

There are more sophisticated, faster-responding ways to do it but for this scope you don't need those; though you will probably need a faster current generator if you want to tweak the

(old Tek guy)

Reply to
Frank Miles

Just as a check, does the amplitude double if you go to the high impedance input? vs the 50 ohm setting. Mikek

--
This email is free from viruses and malware because avast! Antivirus protection is active. 
http://www.avast.com
Reply to
amdx

This one doesn't have that.

As you cna see it does. What puzzles me is the cable length doesn't seem to matter. those WFs in the pics are on like a 2 foot cable. WWe have a four or whatever foot cable, and he ringing looks the same. this leads me to bel ieve that the ringing is at the front, right there in the Wavetek because I T does not like the impedance.

I am not rally paranoid about ringing per se, I just would like to see this scope be in the best alignment possible without the factory treatment. thi s is a one meahertz square wave, if the bandwidth is to the twentieth harmo nic, which would only be 20 mHz, it shouold give me a good square wave.

And watch my next post.

Reply to
jurb6006

Wow. I thought my little Waveytek was the cat's ass, albeit a bit 1960ish.

Really though, I have the print, how much harder can it be ? This is only o ne meg.

I actually did something almost like that, but dammit didn't have the camer a handy. I uised two 1N914s back to back off about a 1K I think. It was wor se. The ringing was gone but I had rounded edges and all that. There really may be something rong with that scope, outside of what has been in conside ration. (naw, never happened befroeeee lolol)

I want my Tek 7603 back. Now THAT'S a scope. Big as a damn horse, but one h ell of a scope. (what really hurts is that is one time I had THE genuine Te k scopemobile for it)

Excuse me while I go check out some self immolation websites...

Actually NO scope I have ever owned as a fuckng cool as the TekTRONIX (NOT TEK, this is with the TRONIX on it) 310. Thing opened up like a book.

Reply to
jurb6006

Yes.

Reply to
jurb6006

Well, i seriously doubt the cable impedance mismatch is having any effect at the slow speeds you are using. To be polite, adding diodes is not exactly the way to "solve" the perceived "problem". If you still think the input to the scope is a problem, then make a

6dB pad and use it AT the scope input. Betcha dimes to doughnuts that will make ZERO difference.
Reply to
Robert Baer

So you're saying you think the problem is in th scope, right ?

Sounds plausible, but if so it has to be in the output section, or whatever affects both channels.

Reply to
jurb6006

The people I work for, at one time made scope probe coax wire. I believe the conductor was a very small copper clad type of wire.

This solved a couple of issues, one of them was to introduce a small DC resistance in the wire, there by lowering the Q a bit.

High Q causes ringing and even with the cables they made, there still was a small amount of it, hard to get rid of without sacrificing other properties.

Have you tried using a 10:1 probe or even real scope coax wire ?

Also, you may want to entertain the idea of putting a 50 ohm non inductive R at the scope end. I have a couple hundred 50 and 75 ohm carbon composite for this. Your scope may also have the 50Z switch on it, I know my Tek 350Mhz does.

If after the use of a 10:1 probe, that has been calibrated via your scope reference, you still see the problem? Then you have an issue with the generator.

Jamie

Reply to
Maynard A. Philbrook Jr.

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.