Tektronix PG 506 Calibration Generator

Der PG506 ist ein Rechteck Generator der zur Amplituden Kalibrierung der TEK Scope's Geräte genutzt wird. Die Höhe der Amplitude ist extrem genau, Ausgänge für schnelle Flanken <1ns werden auch geboten um Überschwingen bei den Scopes wegzutrimmen.

The PG506 is a squarewave generator calibrating oscilloscopes. The height of the amplitude is extremely exact, has two flat fast rise outputs.

• 10 Hz to 1 MHz

• 3 squarewave output modes

• readout of scope deflection error

• amplitude calibration mode

• Period: fixed at ca. 1 ms or DC

• Amplitude: 200 µVolt peak-peak to 100 Volt peak-peak

• 1µs to 10 milliseconds in decade steps for period

• calibration current loop of 5mA for current probe calibration

• high amplitude output

• rise time: unterminated: 100 ns or less. Terminated into 50 ohms: 10 ns or less

• Fast-Rise Output

• Rise-Time (terminated into 50 ohms): < 1.0ns

• Polarity, simultaneous negative and positive

•   Direct Readout of Oscilloscope Deflection Error

Unter dem rechtenTransistor mit Sternkühlkörper sitzen die beiden getauschten Dioden (siehe Text weiter unten). Both changed diodes under the most right transistor with the black heatsink (see coming text).

Das Gerät hatte einen Fehler, die selektierte * Diode CR280A war defekt, das hat mich ein paar Stunden gekostet, man sucht meist wo anders. Der Ausschnitt aus Schaltplan zeigt den internen 100 DC Volt Standard, aus dem alle Kalibrationsspannungen abgeleitet werden. Durch die defekte Diode war die Urspannung anstatt 100 Volt ca. 120 Volt. Die beiden Dioden sind werksselektiert, wahrscheinlich auf gleich Junction Spannung bei einem bestimmten Strom um einen Gleichlauf zu garantieren. Ich habe daher beide gegen zwei 1N4811 vom Gurtabschnitt getauscht. Zuvor wurden an einer Stromquelle zwei Stück auf gleich Spannung selektiert. Die Dioden sind auch auf der Leiterplatte direkt nebeneinander montiert um eine ähnliche Temperatur zu erreichen.

Nun läuft das Gerät wieder wunderbar, es ist eine Freude damit ein Oszilloskop einzustellen, gerade wenn man viel die 7000er Serie benutzt muss man doch manchmal beim Tauschen der Einschübe entweder die Verstärkung oder die Zeitablenkung korrigieren, da ist es eine wunderbare Sache, wenn gleich immer ein PG506 daneben steht, der etwas genauer ist als die im Oszilloskop eingebauten Schnellkalibratoren. Kaputt ist jetzt lediglich nur ein einzelnes Segment aus der LED Prozent Anzeige, dieses zu tauschen ist nicht einfach, man muss hierfür die ganze Frontplatte auseinanderbauen, zum Glück habe ich das Ersatzteil nicht.

The instrument had a fault. The * selected diode CR280A was defect. This schematic ist part of the 100 volts standard amplitude, 120 volts applied nominal on the emitter of Q280, with the faulty diode the nominal 100 volts node showed almost 120 volt. After some hours I located the fault. These diodes are factory selected for a similar u vs. i characteristic and on the board they are mounted close together for a similar junction temperature. I replaced both of them with two 1N4148, I took two diodes from a reel and selected two of them using a current calibrator, I choosed two diodes with the same voltage. Now the instrument works wonderful again - but one LED segment from the percent indicator don't lit, changing the seven segment display seems to be a hard work, every knob must removed and the board removed from the front-panel, I'am lucky having not the spare part.

Here are some photos of the fast rise outputs:


Fast Rise Output:


   
Fast Rise Output PG506 with 7S14 Sampler (1 GHz tr<=350ps) in a 7904A mainframe
Right photo: Sampler switched in Low Noise position, left photo: switched in normal position

All instruments are excellent, the PG506 has a fast and very flat step response without overshooting. This is an ideal source for a vertical amplifier frequency compensation adjustment. The sampler works perfect for such repetetive signals, the 7S14 shows no overshooting. A sampler is a powerful instrument, I don't want to miss it.

    
Fast Rise Output PG506 with a 7A29 amplifier (1 GHz) in 7904A and 7104 mainframe

These are the results with a real time amplifer. The 7A29 with it's 1 GHz bandwidth is the fastes amplifier Plug-in of the 7xxx series. The 7A29 has also a very flat response, almost the same flatness as a sampling system - what a wonderful amplifer Plug-in. Also both mainframe showing the same risetime, it seems the 7904A vertical amplifier has more bandwidth then specified, I read somewhere the verticals in both are very similar, when I see these results I can believe. The 7104 Micro Channel Plate CRT has a very sharp focus, compare for example the figure "7".

Direct Comparison of Real Time Amplifier and Sampler

   

Results of two different measurement systems - Real Time vs. Sampler. Both systems showing almost the same flat response and exactly the same risetime.

• Four Division Risetime - measured with Real Time Amplifier - 900 Picoseconds

• Four Division Risetime - measured with Sampling System - 900 Picoseconds

Tektronix Engineers - you did a wonderful work - thank you



Fast Rise Output Falling:

   
again an excellent measurement

Four Division Falltime about 750 Picoseconds

The real rise- and falltimes are a little faster, the sampler and amplifier risetimes are included in these measurements. The only error in this measuremnt is the user !!! He didn't adjusted the vertical gain correctly, that's only a 20 second job, he feel ashamed. That's the reason why he repaired the PG506, he don't want to see gain errors on photos again.

PG506 Rise Time Measurement - S-6 Sampling Head

Measurement on a fast Sampling Head S-6, S-53 Trigger Recognizer, 7S12 TDR Plug-In in a 7834 Storage Oscilloscope. This S-6 sampling head has a risetime of <=30ps equivalent to a 11.5 GHz bandwidth. The 7834 storage oscilloscope operates in a variable persistance mode. A storage oscilloscope combined with an sampling system is a very good combination, allows good viewable low noise and high resolution settings in a sampling system. The curve characteristic (settled shoulder) measured with both sampling systems 7S14 and S-6 looks very similar.

Four Division Risetime - measured with a S-6 Sampling Head - 800 Picoseconds.