Electronic Test Equipment Qs

No probes. I am using two-foot lengths of RG-58/U cable with BNC connectors on both ends. The same two cables were used with the two measurements so the only change was the scope itself.

The input on both scopes was set to 50 ohms DC, as the HP 3325A expects to deliver the signal to a 50 ohm load.

 

The Tek 2467 does not show that ripple in the first square wave, only the HP 1725A.

And, it is not just square waves. A similar issue also shows up with sine waves at VHF frequencies.

If the issue is due to not using a probe, why is the problem visible only with the first square wave? I'm familiar with the effect of uncalibrated probes, but in that case, every pulse shows a problem, not just the first.

 

I decided to look at the horizontal and vertical signals produced by the HP 1725A, using the Tek 2467.

The horizontal ramp appears to be OK. The ramp looks more like a triangle at higher sweeps since the fall time remains constant while the rise time approximates the fall time at the fastest sweep times of 10 and 20 nanoseconds.

The vertical signal is where the problem is. The HP has a vertical output BNC connector at the back, which samples the vertical sync signal produced for use by the horizontal sweep. When I connect the Tek to that vertical output, I can see some (but not all) of the vertical square wave pulses have that little ripple.

The basic structure of the vertical amp is as follows: CH A and CH B have individual attenuator boards which also provide some signal amplification. The two boards provide signals to the vertical preamp which has a 40 pin square IC that takes the CH A and B signals and produces a differential output for use by the vertical amplifier and a differential sync output for use by the horizontal sweep.

It appears that the slight signal irregularity is being produced within that 40 pin IC.

Something else which is weird is when I use the Tek, 1 megohm AC input to measure the differential output at the 40 pin IC, the signal amplitude is around 550 mV p-p at 1 MHz and below. However the signal drops down to 100 mV p-p with a 20 MHz signal. Further, the square wave signal looks like a triangle at higher frequencies.

I have one P6137, one P6136, and two P6105 probes. All produce similar results. Since the HP vertical amp has relatively flat response over that frequency range (as it is rated to 275 MHz bandwidth) this must be a measurement artifact...

 

1. I don't earn my professional living with this equipment, it is a hobby. Hence I can't see spending more than three-digit amounts for a particular piece of equipment.

2. I'm trying to see how far I can restore the HP 1725A with a moderate out-of-pocket investment. I had bought that scope in the mid-90s but it has been sitting around unused much of that time. So far the restoration effort has cost around $50 in capacitors plus lots of my time - which is probably healthier than me spending that time drinking in a bar...

3. Regarding analog vs. digital, I am working mostly on analog equipment from audio to UHF frequencies (communications receivers and other ham radio equipment for example) and I like the idea that the analog scope shows the waveform more precisely. The intensity of the trace varies depending upon the repetitiveness of the signal.

I recently bought the Tek 2467 because there's some hope that I can repair it when it fails. I wouldn't be able to service equipment containing surface mount components since I can barely see them.

 

$12K for an oscilloscope is out of my personal price range given the almost infinite claims upon my income placed by my wife and daughter. I spent ~4% of that amount for the Tek 2467 (which had an original $11K purchase price back in the late 80's.)

Agilent Technologies Test Equipment - MSO7034B - Oscilloscopes - Test & Measurement - Allied Electronics

Nice to see though, that you have that model available for your personal use.

 

BTW, a few days ago I bought a Tek calibration station from New York, with the following plug-ins: TG 501 time mark generator, PG 506 calibration generator with 1 nanosecond rise time pulses, and SG 503 sine wave generator. The time mark and calibration generators seem to be the standard for servicing analog scopes of 1980's vintage. I don't really need the SG 503 since I already have two HP signal generators and one HP function generator, but it was part of the bundle.

Assuming that they work (or that I can get them to work) then I will be fully ready to calibrate those scopes. The eBay photo showed all three units' displays nicely lit up, so there's some hope they are actually functional.

I also just bought a Tek 2465A at a highly attractive price today, so analog oscilloscopes will be coming out of my ears. It's not clear whether this unit is working but at minimum the horizontal sweep is functioning and the screen shows the correct indications for voltage scale, sweep speed and trigger voltage, so it passed self-tests and at least the infamous U800 horizontal output IC should be OK.

That unit is similar to the 2467 but does not have the MCP CRT so the HV power supply and the CRT are different. I believe the other components are pretty much identical.

If I live another 20 years and the Agilent digital scopes of current vintage are still working by then, maybe the going market price will fit my budget...

 

Hi Ed,

Thanks for discussing the situation with your compadres.

I've replaced around 30 electrolytics (tantalum or regular, as the case may be) on the vertical preamp board; and the three main sweep integrator, delayed sweep integrator, and holdoff-delayed comparator boards that plug into the main/delayed sweep trigger board. However I have not yet replaced any tantalum capacitors on that main trigger board (which also has the horizontal preamp and Schmitt control.) There aren't that many so I'll work on that soon.

I've recently noticed another problem which may or may not be related: when the AUTO/NORM button is depressed to select NORM, there is not supposed to be a trace until a signal is present. However, on an intermittent basis, sometimes the trace will be there even when NORM is selected.

I just removed the cabinet to measure voltages, and naturally the AUTO/NORM function is working perfectly right now.

 

Time for another update. Richard's msg encouraged me to acquire my latest toy, an HP 8591E portable spectrum analyzer. What an eye opener!

This provides the ability to explore a new dimension and see a chart of amplitude vs. frequency (compared to an oscilloscope which charts amplitude vs. time).

I've attached a photo of the analyzer and another showing the clean sine wave signal produced by my Tektronix UHF leveled sine wave generator. All harmonics are ~ 55 dB down (or more) from fundamental. This shows best case, but average is not too much worse.

The last photo shows the 1 GHz signal produced by my Tektronix TG 501 time mark generator. This signal has a large number of spurious and harmonic signals although they also are ~ 50 dB down or better.

I sold the HP 1725A oscilloscope. The little irregularity that showed up on the beginning of the trace was just due to vertical amp noise, when the 20 MHz bandwidth limiter is turned on then the trace becomes very smooth.