Scope bandwidth limitation - amp or crt?

The amplifiers are the limiting factor in that case. Timebase is relatively slow, so faster sweep speeds are more easily implimented. But. The trigger circuitry may need a major upgrade to be able to respond to the faster input rise/fall signals you want to see.

As Robert said, the limit is usually in the electronics and in particular in the last stage. Driving the deflection plates at increased speed requires increased slew rate of the output stage. Because of the large voltage swings that leads to costly transistors and a more intricate design.

Regards, Joerg

Thank you. Any idea what sort of voltages are typically involved?

Any rough (order of magnitude) guess on the capacitance of the plates?

The chances of actually changing anything are probably low, but it's interesting to contemplate.

I had an old 5MHz Heathkit scope that I was going to reinvent at one point. I swore the sweep wasn't linear enough (from an RC waveform), and the 5MHzbw could certainly be improved on. IIRC I think it needed something like 170V for a full 10 div deflection! One cycle of 5MHz across the whole screen was then 200nsec. I needed to improve on 170/0.2usec=850V/usec to show a higher frequency! I gave up at that point, and never finished the project! (...and am now a proud Tektronix owner)

Hello Jim,

Mine is an old Hameg 207 and I still have it. Extremely sturdy but drops off at 8MHz. IIRC it needed less but still above 100Vpp for a full swing. Capacitance wasn't the issue.

I thought about the same thing but gave up because opening up the BW of the input attenuator circuitry around that huge rotary switch appeared too daunting. Actually I had the final stage licked, using rather cheap "video transistors". They are (or were) used to drive the cathodes of color TV CRTs.

The reason why mine isn't always linear: It doesn't have a trigger, just a synchronizer. You have to couple it only enough to lock the waveform. If you just leave the synchronizer pot pegged it'll distort badly.

Regards, Joerg

Where's the load impedance come from if not capacitance?

Voltage with next to no load is easy (well, I suppose not trivial when the bandwidth needs to go down to DC)

Basically it seems the vertical amp is an RF linear amplifier... question is what kind of load it needs to be designed to drive.

I've only played with a couple of older low-BW scopes deflection plates directly - and as I recall, they were in the 10-20pF range. Deflection voltages were similar to what's been suggested already.

As has been suggested - vertical bw is primarily limited by the vertical preamp and driver stages. But, once you upgrade those, you'll want to upgrade the horizontal sweep and drive, and the trigger circuits.

Thanks - that gives enough for initial contemplation.

I just realized my question should have been where does the load _admittance_ come from (the capacitance) since the source of the impedance is rather obvious.

IIRC it was a little under 10pF. Not much of a load unless you want to push it into the VHF range. The stray capacitance of your circuit and the parts is often higher than that of the plates.

Regards, Joerg

Joerg wrote in news:1YMNf.17504$ snipped-for-privacy@newssvr27.news.prodigy.net:

FYI,the CRT is a definite limit on bandwidth. The CRT electron gun structure is VERY different between the 100Mhz TEK 465 and the 200-250Mhz TEK 475 models. The 475 uses a distributed deflection plate assembly that is terminated by a separate ceramic termination resistor network. Same goes for the 500Mhz 7904 CRT.

cs snipped-for-privacy@hotmail.com wrote in news: snipped-for-privacy@p10g2000cwp.googlegroups.com:

You neglect the inductance of the CRT leads,and inductance of the CRT deflection plates,along with their capacitance.Higher BW CRTs(TEK) use a distributed deflection plate electron gun structure,actually a transmission line with an end termination.(a differential termination,too.)

Hello Jim,

True, but now you are talking about a whole different market, the high-end. The OP just wanted to get past 5MHz and on those scopes it's usually the amp that limits the BW. They often just couldn't afford another buck or two in production for hotter final amp transistors when the whole scope has to be under $250.

Case in point: I have used the plates of a scope CRT directly for transmitter measurements. I could easily go to 30MHz with the CRT of a scope that was otherwise tapering off at 10MHz (3db). With direct drive there wasn't even one dB of slump in amplitude.

Regards, Joerg

Actually you've both been very helpful.

It seems that in a cheap scope, the amplifier is the first limit, but there's also a limit to what a basic CRT can do - to get really high bandwidth you also need this transmission line plate structure.

Not if you just want to push your 20MHz scope a tad higher but certainly when you want to reach VHF with it. I guess that is one of the reasons why many older Tek scopes aren't repaired once the CRT goes kaputt. A new CRT would often cost more than the residual market value of the whole scope.

If you need a fast analog scope the best bet is to hunt for business closure auctions (that's where I got a 1GHz HP) or other used equipment sources.

One very tough cookie would be the speed upgrade of the input attenuator. This thing is rather complicated with all the little trim caps and the fact that it is all jammed into very tight shielded spaces doesn't make that job easier.

Regards, Joerg

Actually I regularly use high bandwidth analog & digital scopes at work, so what I can do with my home gear aquired during my teenage gears (and just reclaimed from parent's basement) is more curiosity than necessity.

Speaking of deflection, I noticed there was a spot of something on the scope screen, and grabbed a paper tissue to clean it. Talk about electrostatic deflection!

Yeah, I have that to look forward if I go ahead with the sampling scope project I've been thinking about...

How about using the video driver ICs and output transistors from a dead VGA monitor? Use a good op amp to invert the signal to one channel.

Hello Michael,

Sure. I think I mentioned video transistors before, they can be had for even less than those from VGA monitors. However, on a scope that has to retail for $250 even 10c more might not be in the cards because you need two of them, or four if it's a 2-channel.

Personally I am cautious with TV chips such as video drivers. They rpoduction lifetime may not be long and I have seen people get burned big time by sudden parts obsolescence. Basically when they asked for leadtime they were told "infinity".

Regards, Joerg

Joerg wrote in news:wfJOf.44881$ snipped-for-privacy@newssvr13.news.prodigy.com:

No,it's because those CRTs (all TEK-made;single-source) are no longer in production,and no longer available. IOW,there are no "new CRT's".

IMO,if new CRTs were available,people would buy them despite costing more than the present market value of the scope. (assuming the rest of the scope was working or repairable.)

If you are VERY lucky,you can find an unused CRT that was stocked by the US military and sold off as surplus. (and TEK sold off their CRT unit,they no longer make -any- CRTs.)

Always thought glass blowing would be an interesting hobby...

More seriously, either someone in asia is making something comparable, or there are no new high bandwidth analog scopes to be had at all?

My guess though is that affordable real time sampling either has, or soon will surpass the bandwidth of affordable CRT's - at least on the new market, and on the used market eventually.