TEK 485

[mansaxel]

I guarantee I can break this if I have access to add DNS records

You (and other interested parties) do not have that access, and it is not only a matter of admin access to DNS; there are other checks and balances in place too. But in principle, I do not doubt there are vulnerable points in this.

Possibly a classic case of trusting conformance rather than validating conformance in there.

It is a validation, of sorts and it requires several things to be not wrong. I'm certain you can poke holes in it. The computer as worker node is expendable because there will be another one.

[MED6753]

More 485 fun.

I was about to put my original parts mule back into storage and I noticed these 2 options listed and was curious as to what they were.



Option 4 is additional EMC shielding and option 78 is P11 (blue) phosphor! Really?

So I pulled the CRT which surprisingly isn't too difficult. Sure enough. TEK 154-0652-09. P11 phosphor. I'm going to transplant this CRT into one of my functional 485's.

[Cubdriver]

Sweet!

[tggzzz]

Yeah, the blue is delicious

[MED6753]

Papa Smurf sez...."Let there be blue light". And there was blue light. P11 CRT installed and working.



With the intensity turned up to normal viewing levels the camera really reacts to the blue phosphor.



Now to put everything back together and I'm going to touch up the calibration.

[Zenith]

Sweet!

That cat has an air of brooding malevolence and barely contained rage.

[tggzzz]

That cat has an air of brooding malevolence and barely contained rage.

That's standard: they all do.

At least it appreciates a good scope. I hope hairs don't get in any of the mechanical mechanisms

[MED6753]

Started tweaking the calibration on the P11 485.

First on the agenda was to clean the attenuator switch decks, especially the 1M path. They were dirty. A dab of deoxit then flush with 100% IPA and blow dry.

The 50 ohm path on both channels looks pretty good. 10MHz square. The trailing edge ringing is a function of the source.....Heath IG-4244 Oscilloscope calibrator. I actually like to see that ringing. It tells me the high frequency response of the scope under test is damn good. On some of the 1M attenuator settings that ringing is muted indicating compensation issues. So a full compensation needs to be done.



For comparison the 2465 50 ohm path.

[tggzzz]

Started tweaking the calibration on the P11 485.

First on the agenda was to clean the attenuator switch decks, especially the 1M path. They were dirty. A dab of deoxit then flush with 100% IPA and blow dry.

The 50 ohm path on both channels looks pretty good. 10MHz square. The trailing edge ringing is a function of the source.....Heath IG-4244 Oscilloscope calibrator. I actually like to see that ringing. It tells me the high frequency response of the scope under test is damn good. On some of the 1M attenuator settings that ringing is muted indicating compensation issues. So a full compensation needs to be done.

Unlike most scopes, the 485's cal out has a good 1ns risetime into 50ohms. It will be difficult to improve on that.

Into 1Mohm the trace is far more entertaining: a stepwise exponential. The reasons for that are in some ways quite deep.

[MED6753]

Started tweaking the calibration on the P11 485.

First on the agenda was to clean the attenuator switch decks, especially the 1M path. They were dirty. A dab of deoxit then flush with 100% IPA and blow dry.

The 50 ohm path on both channels looks pretty good. 10MHz square. The trailing edge ringing is a function of the source.....Heath IG-4244 Oscilloscope calibrator. I actually like to see that ringing. It tells me the high frequency response of the scope under test is damn good. On some of the 1M attenuator settings that ringing is muted indicating compensation issues. So a full compensation needs to be done.

Unlike most scopes, the 485's cal out has a good 1ns risetime into 50ohms. It will be difficult to improve on that.

Into 1Mohm the trace is far more entertaining: a stepwise exponential. The reasons for that are in some ways quite deep.

The risetime of the IG-4244 is also 1ns and that was verified because it was a required spec to calibrate of counter on the 2465 DMS.

[EC8010]

That correspondence between Tek 485 and 2465 is impressive. In theory, we should expect them to be identical, but at 10ns/div., all bets are off. I wonder what is causing that slight "highish" frequency loss after the leading edge of the square wave in the 485?

[MED6753]

That correspondence between Tek 485 and 2465 is impressive. In theory, we should expect them to be identical, but at 10ns/div., all bets are off. I wonder what is causing that slight "highish" frequency loss after the leading edge of the square wave in the 485?

I think there is a 1MHz adjustment on the 50 ohm path which probably needs tweaking. I'll know for sure when I dive into the calibration.

[EC8010]

I'll be interested to see what you achieve; I used to have a 485 at home (bought with my own money) and bought a 2465B for work (with work's money), but never compared the two.

[MED6753]

Completed the calibration of the channel 1 & 2 attenuators, vertical pre-amp, and vertical final amp as best I could with the equipment I have. Some adjustments require a tunnel diode pulser which I don't have.

This is the 50 ohm input at 10MHz square. Slight improvement with a small amount of peaking.



Same 10MHz square on the 1M input with a 50 ohm feed thru terminator. It's actually flatter than the 50 ohm input which is surprising.



So how come? Would have expected better response from the 50 ohm input. There are no adjustments to the 50 ohm attenuator. And beyond the attenuators the 1M and 50 ohm path are identical. I suspect some fixed resistors or capacitors in the 50 ohm path have drifted. And channel 2 exhibits the same aberrations as channel 1. Am I going to rip apart the attenuators and figure out the fault? Nope. In the end I really don't consider the differences in wave forms significant anyway.

[tggzzz]

I'd compare that with the 1ns risetime cal out.

There is a slight capacitive (downwards) blip 2 divisions after the rise. I'd check whether the position of that depended on the cable length.

My preference is to have a pulser attached directly to the input, to minimise degradation due to cable attenuation and possible cable impedance mismatch (many are 52 ohms, for reasons that elude me). That's not possible with the cal out, and presumably not with the IG4244. Trick: putting a small (3dB/6dB) attenuator pad on the input can reduce the consequences of mismatch.

But overall that doesn't look bad. Here's mine with the cal out

[MED6753]

So the P11 485 joins my other 485 getting some power up burn-in time.

[EC8010]

Nice to see caps keeping the inside of BNCs clean.

[MED6753]

Nice to see caps keeping the inside of BNCs clean.

My cat loves to steal them.

[AVGresponding]

Nice to see caps keeping the inside of BNCs clean.

My cat loves to steal them.

I discovered some cheap but decent ones on AliX, they fit over the socket, with channels for the bayonet pins

https://www.aliexpress.com/item/1005005583218593.html

[EC8010]

My current cats are actually very good in the lab and haven't chewed 'scope probe cables (Cathode, a previous cat was particularly attacted to them), but Heidi likes to steal and play with the stylus my wife uses on her mobile 'phone.

Rather more expensive, but less likely to emit corrosive vapours (think PVC cables and expanded polystyrene) are:

https://www.amazon.co.uk/YINETTECH-Diam ... 121&sr=8-4

Variants with chains are also available but the chains always seem to tangle on something. I put caps on almost all my BNCs.

You do not want to know the price of the equivalent in triax. Glad I bought my triax caps when they were a sensible price.

The 50 ohm terminations on my 'scopes are better than my feedthrough terminators, but (oddly) I get better results with a T-piece and term than the feedthrough. My feedthrough terms must not be very good.

[Zenith]

The 50 ohm terminations on my 'scopes are better than my feedthrough terminators, but (oddly) I get better results with a T-piece and term than the feedthrough. My feedthrough terms must not be very good.

Checking quickly with a 1MHz sine 500mV ptp from the internal AWG in my scope, I find the scope's internal 50R termination, a cheap Aliexpress feedthough terminator (reckoned to be good to 1GHz) and a tee-piece and terminator, the scope reports a ptp of about 475mV in each case. At 20MHz it's 465mV ptp with the tee-piece and terminator being more twitchy.

Swapping the cable around, because one BNC connector didn't seem to get on with the AWG output, the results were 490, 490 and 497 for the tee-piece, with little or no jitter in the ptp reading. That's probably within the specs claimed by this set up.

Of course it isn't a very methodical or exhaustive test. For instance things may be very different beyond 100MHz.

I'm inclined to use whichever is the most convenient, bearing in mind that I really don't want to damage the internal termination on the scope.

Maybe I need to invest in some of those little plastic BNC caps, and investigate my BNC cables rather than assume if they look good then they are good.

[EC8010]

I had to think about why I ended up looking at my terminations, then I remembered. A while ago, I bought a Leo Bodnar GPS disciplined oscillator. It produces 10MHz and 96kHz. When I plugged the 10MHz as external reference to my 'scope, it didn't like it. It turns out that the Ref. In on a Tek MSO54 is not terminated, causing horrible ringing on the square wave leaving the oscillator. And that was when I discovered the differences, plugging into the (300MHz bandwidth) input of the 'scope, terminating with either the 'scope's internal termination (best), T-piece and term, through-termination (worst). I don't suppose it makes much real difference, but it was observable.

[tggzzz]

I thin you'll find that if you put an inline terminator on a 1M//15pF scope, the termination will be 50ohm//15pF. That 15pF is a bugger at >50MHz or so, of course.

A better termination would include a 3/6/10dB pad, on the principle that gain is cheap but you can't recover a mis-terminated signal.

[EC8010]

Doh! I should have thought of that. Thank you.

[Zenith]

I had to think about why I ended up looking at my terminations, then I remembered. A while ago, I bought a Leo Bodnar GPS disciplined oscillator. It produces 10MHz and 96kHz. When I plugged the 10MHz as external reference to my 'scope, it didn't like it. It turns out that the Ref. In on a Tek MSO54 is not terminated, causing horrible ringing on the square wave leaving the oscillator. And that was when I discovered the differences, plugging into the (300MHz bandwidth) input of the 'scope, terminating with either the 'scope's internal termination (best), T-piece and term, through-termination (worst). I don't suppose it makes much real difference, but it was observable.

I have a Chinese TM4313 GPSDO. The Leo Bodnar models may have the edge, but the TM4313 has been adequate for my needs.

A teardown.

https://tomverbeure.github.io/2023/07/0 ... rdown.html

A comparison of inexpensive GPSDOs including a couple of samples of TM4313 and a Leo Bodnar GPSDO.

https://reeve.com/Documents/Articles%20 ... DOComp.pdf

I connect the TM4313 to the 1M 25pF low frequency input of Racal frequency counters and don't bother with proper termination. It doesn't seem to cause reflections so bad as to create spurious triggering. I have looked at the waveform on a scope and it looks a bit ragged, but no worse than expected. I rarely connect the GPSDO to other kit as an external reference.

As the frequency increases the slapdash approach can catch you out. e.g. Cheap Chinese BNC cables that have a dip at 80MHz. I'm not sure about my no name feedthrough DC-1GHz 50R terminators. They've seemed OK so far. However, I'm sure that if you buy an HPAK or Tek feedthrough terminator, attenuator, etc, there's a reason it's expensive, apart from paying for the name.