Mini Teardown: Tek 519

[khs]

Hi all,

here a mini-teardown of one of the fastest real-time oscilloscopes until about 1978.

BW ist DC to 1 GHz, risetime is about 300 ps, fastest timebase is 2ns/cm and the sensitivity is about 10 V/cm.
(http://w140.com/tekwiki/wiki/519)

My copy comes from a nuclear research institute near Munich for 10 euros (including cables and adapters). Nuclear reseach requires fast realtime oscilloscopes, because the nuclear events are short and statistical.

It was told me, about 1970 it was very difficult to get a time-slot for a measurement, so for the students the 519 of the institute was available for their work only from midnight until next morning.

Unlike newer instruments this oscilloscope is very very robust. There is no amplifier, so there is more or less nothing to destroy.

A sampling head can be destroyed within ns by a voltage of more than 5V. The internal 50 Ohm termination helps not much, because it is behind the sampling diodes, so before the pulse reaches the resistor the sampling diodes are blown.
 
As a comparison, the 519 accepts pulses up to 100 V, so it is more or less not destroyable.

Switching on requires some patience. During the power on delay all tubes reaches their operating temperature, then a relay switches on the HV power supply with a bang accompanied with a sizzle of the 24 KV acceleration voltage.

So for 10 euros this bang for the bucks is unbeatable.

The display is quite small, just about 2 cm x 6 cm, but the dot size is quite small, so it compensates it a little bit.

A big advantage is the internal pulse generator. It generates a 10V signal with about 100 ps risetime with a reed relay and a delay line. It is adjusted by knobs for drive and frequency, so adjustment is made by listening.

I think even in 2017 10 V in 100 ps is still hard to get for 10 euros.

This instruments helped me to understand low pass filters. I was surprised to see every low pass filter is a band pass filter in reality. The time I buy it (long time ago), it was the cheapest way to get an 1 GHz oscilloscope.

The comparison with the 7104/7A19 shows the relay pulse is not as clean as shown on the 519, but in most cases it's not essential (both pictures 2 ns/div).

Please note in the T519-timebase picture the unsusual time/div switch position to get the switch as near as possible to the 4CX250F tube .

The 519 and the 7104 are no scopes for the daily use. But sometimes it's helpful to have it.

[rsjsouza]

Beautiful pictures! Thank you for sharing them and the story.

I really wish I had more space - so many of these older oscilloscopes have been offered to me but I simply didn't have a decent place to put them... Oh well...