Repair and Calibration 7844 Dual-Beam Oscilloscope

This scope had a fault. The power supply had a short circuit with a blown primary fuse. In many cases a blown fuse can be a terrible fault. Before the scope run for approximately 1 hour with four Plug-in under an ambient temperature of more than 30°C. After changing many times the Plug-in (I always switch of before changing) a loud Bang and Bum and the fuse was blown. Nowadays I put a room fan behind the power supply, always very cool.

gehe zur deutschen Version  

photo shows the linear voltage regulator board, consists of +130V, +50V, -50V, +15V, -15V und +5V.

Removing AC-cord, Plug-ins out and remove the power supply from the mainframe. Take care when working with this switching converter power supply, the main capacitors storing a high charge and the rectified dangerous line voltage, wait at least until the small gas discharge bulb had stop blinking, then the capacitor voltage is under appr. 80 volts. When reaching that small voltage discharge the caps with some hundred ohms totally. Follow the service manual for all safety hints.

Use only an AC-line isolated variac transformer when working with the open power supply. Take care also for the CRT High Voltage, this glass-vacuum isolated capacitor stores the energy even for many days! dont't think: "I switched off the scope yesterday, now I am safe" your are not !!!, the voltage is still high enough for a very long spark. Be extremly careful when you remove the white High Voltage plug, discharge it on the chassis, but be careful again when pull of both plugs - keep a respectful distance to any electronic part when the jack is open! otherwise the spark will destroy parts on the PCB. When measuring on the PCB use a 1:100 probe with a specified voltage much higher than the line voltage, for your own safety.

Unfortunately the cables from the power supply to the mainframe are short, but long enough to search for errors. I guessed the fault is located somewhere in the primary circuit in the Inverter PCB, this means the power supply should be separated from the mainframe and semiconductors must be checked.

Some cables from the power supply are spreaded wide in the instrument, here the Z-axis and CRT circuit. Before removing the connectors, write down which connector fits where, it is helpful for reassembling, take care for connector polarity. A torch is a usefull helper. The black dust on the wires energised by the electrostatic force of the HV in the CRT circuit, nominal -2960 volt on the vertical mounted PCB, be careful in that area. Remove all black dust with e.g. Isopropanol and a cotton cloth.

This is the Power Supply Inverter circuit, AC-line voltage rectification, generates the rectangular supply voltage for the high frequency transformer. This DC/DC Converter acts as LC series resonance tank, L=1mH and C=0.03µF resulting in a 29 kHz natural frequency. Depending on line voltage amplitude and secondary load varies the pulse width and switching frequency. This series resonant circuit has a good efficiency and approach sinusoidal primary current, a good precondition for a well secondary rectifiying with low ripple. Clean output voltages are a must for a precision oscilloscope.

On the photo top right the transformer T1230 for supplying the antiphase switching transistor base current (Q1234 and Q1241), mounted on the bottom aluminium heat sink. Top left side, there is the 50/60Hz line voltage sense transformer T1208, supplying the oscilloscope with a "LINE" trigger signal. Middle, left the small white part a gas discharge overvoltage spark protection preventing from excessive high line voltage by blowing the primary fuse.

Middle, right the TO-case transistor Q1246  called the "Over Voltage Stop Circuit" a componemt of the safety concept. This Q1246 were destroyed togehter with a 22 ohms in the base of a main swichting transistor.

Left side from T1230 the small round T1235, supplying the DC-DC Controller U1275 with a dual low power supply voltage, the T1235 has also a overcurrent protection function.

Left AC line diode bridge, a common mode choke, another gas discharge protection, right side aluminium heat sink for the bipolar switching transistors. Before you reach the Inverter, some wires must be resoldered from the PCB, note everthing detailed before removing anything.

Unfortunately both switching transistors Q1234 and Q1241 were defect with an internal short circuit.

AC-line filter and primary fuseholder. Isolation material for the aluminum hest sink of the switching transistors. Normally you should not remove this small heat sink, because reassembling need some special care. Be carefull that both surfaces between aluminium and isolating-foil are very clean, apply new heat-conductive paste - no small foreign particle between foil and metall, very dangerous, possible isolating damage . After a careful reassembling, torque the screws softly and alternating. The Service Manual write: "dont remove that heatsink" - but here it was a must.

Another broken part, the 1mH Series Resonance Coil L1237 and the Series Resonance 900V foil capacitor C1237 (not faulty). The core was broken, some small peices were spreaded in the power supply, but the choke has still 1mH on the inductance meter.

Leaving that part in the power supply would be dangerous, I don't know the inductance value under load current, never know what happens when the crack increase, a further inductance loss can results in an excessive current destroying the power supply again - too risky - replaced by another 1mH choke taken from a old cannibalized 7904 powe supply, that choke were a little small, but should work. If you don't have a spare part, make your own 1mH coil. I think any LF power choke of a similar size should be ok for this purpose.

I really don't know why the choke is broken, too hot, mechanical stress or both?

Reassembling the choke, use heat-conductive crease and torque the screw again carefully, not too much will result in mechanical stress. Ferrit material is very hard and brittle dangered by too much torque and small parts between the mounting gap. This screws can be secured by a loctite when using a low torque.

A typical repair place with AC-line isolating variac, test oscilloscope, curve tracer, probes 1:10 and 1:100, Service Manual and transistor datasheets. Having a Curve Tracer is a very useful tool when repairing instruments, saves so much time and gives a high level of assurance of part function. With a simple DMM transitor tester function you won't find every defect transistor, some transistors are ok with the DMM tester but they are showing funny curves on the Curve Tracer.

No further faults could be located in the Inverter - Reassembling - Power ON - nothing happened - disappointed. What's the matter? Something wrong with the high frequency transformer, oh no, this would be a terrible fault. Reaching the inner double shielded primary winding what a bad job to do, winding everything new, no thanks. (most terrible winding would be the inner HV winding, catastrophe when leakage)

One hour later I was luck because I found a small defect 2N2222A Transistor near the U1275 Controller.

left a new transistor, right a fault one
I assume the high short circuit currents caused the defect switching transistors. Through the high currents flowing in T1230 may be on teh secondary developed a high voltage destroying Q1252. Q1254 were lucky to be protected by C1259 with 2µ2. Also Q1246 could be destroyed thru the 4 windings of T1230, even R1240 were fault. I don't know the real reson which part destroyed which.

List of fault parts

Q1234 NPN Power Transistor
Q1241 NPN Power Transistor
R1240 Basiswiderstand 22 Ohm
Q1246 NPN Transistor vom Over Voltage Stop
Q1252 NPN 2N2222A Kleinsignal Transistor am U1275 DC-DC Converter Controller
L1237 Serien Resonanz Induktivität 1mH

Calibration Start

A full calibration makes sense after a large repair or after receiving a new instrument, this was my case. Most oscilloscopes receiving new having a calibration within specification, some are teribble, but ones with excellent status are seldom - the design is not the problem. It takes time after many years of use such a full calibration, time is something seldom for many companies, they sell it, hobbiests are lucky about that.

Necessary equipment, if you have two Calibration Fixtures it is a luxus still today , decades ago it was a sign for a good status. Without the calibration fixtures it takes much more time. Another nice item for the horizontal part is a Time Mark Generator.  For amplitude calibration a Sweep Sinus  Generator for adjusting the vertical amplifier step response Rechteckgenerator or e.g. a 284.

Another time consumption, searching the test points in the manual. Calibrating this 7844 took me 10 to 15 hours, I guess when used for it 3 hours, Consider with a Dual Beam many calibration steps must done twice due the two guns.

The following series show some photos of the calibration, I didn't took for every calibration step a photo, this would be too much.

Measure the voltages on the testpoints, connect the probe on the recommended points. Use a DMM for DC and also check the voltages with a AC-coupled  test oscilloscope for ripple.

Power Supply Output Voltages AC-Coupled

Power Supply Output Voltages AC-Coupled

Power Supply Output Voltages AC-Coupled

Calibration CRT photos:

Chapter B. Z-Axis and CRT Display
1. Adjust Z-Axis (R1180), set Beam 1 Intensity for a 5V square wave.

Chapter B. Z-Axis and CRT Display
2. Adjust Z-Axis Beam 1Transient Response (C1168, R1168, C1172) , adjust Beam 1 Intensity for square wave >=64V.

Chapter B. Z-Axis and CRT Display

2. Adjust Z-Axis Beam 1Transient Response (C1168, R1168, C1172) at TP1185 , Beam 1 Intensity amplitude reduced by 25% reduziert wurde and adjust for best step response with high rise time and lowest aberrations. In this case it was a good idea readjust the step response. Fix the probe very short, if you don't do oscillations can occurs. Before turning the potentiometer and trim-capacitors mark their original positions with a small pen, gives you a chance going back to the old settings - you never know if you reach at all better setting than before and when there is a chance going back you it's your luck.

Chapter B. Z-Axis and CRT Display
3. Adjust Intensified Z-Axis Beam 1 (C1168, R1168, C1172) at TP1185.

Chapter B. Z-Axis and CRT Display

5. Adjust Z-Axis Beam 2Transient Response (C2068, R2068, C2072) at TP2085 , target best step response. Oscillation occurs due bad probe connection, difficult on this PCB, deep inside the scope.

High Voltage from the multiplier in the power supply.

Setting of Z-Axis Response Beam 2 after readjusting.

Adjusting the CRT geometry is a demanding procedure, in one step it was necessary adjusting the beam to a point round as possible, lenses helped me for a better adjustment.

Chapter B. Item 17B Adjust Beam 1 and Beam 2 CRT Centering.

On a sticker on the CRT shielding the original CRT geometry settings. In one step it is necessary to check the given values.

Chapter B. Item 17B Adjust Beam 1 and Beam 2 CRT Centering.
When it's correct adjusted the beam should looks like here (-0.2 DIV), it was correct no readjustment necessary.

Chapter C.  Calibrator and Output Signal

There are many possibilities adjusting the amplitude and frequency of the calibrator. I followed for the amplitude the manuals instruction, for the frequency I used a 7D15 Counter.,

Chapter C.  Calibrator and Output Signal

Everything seems to be ok, amplitude and frequency. 

Chapter E. Horizontal System

2. Adjust Beam 1 Horizontal Amplifier Gain (R4820)

For this adjustment is a Calibration Fixture a wonderful help. The calibrated step generator allows a precise adjustment of the vertical and horizontal gain. If you adjust all of your 7xxx scopes with the same gain, every plug-in is changeable without a necessary recalibration of the front "Sweep Cal" or "Vertical Gain".

Chapter G. Vertical System

Adjusting the vertical amplifier response is the most difficult calibration step and one of the most important ones. Determines bandwidth and quality of the step response. the calibration fixtures allows an easy gain adjustment and step response, adjsuting for a flat response you'll time, after 20 minutes of trying you'll understand which trimmer causes which effect. For the vertical amplifier it is very important to mark the old trimmer positions - to go back to the old setting if necessary. Fortunately the settings of the step response should remain stable for many years, may be for instruments life time,  I don't know.

Chapter G. Vertical System

vertical gain well adjusted

Cleaning the oscilloscope

Cleaning make sonly fun after a sucessful repair, when cleaning first followed by a unsucessful repair, what a horror, I don't think about that.

Cleaning with Isopropanol (removing again the PCB - no thank you)

My tools Isopropanol, contact cleaner, metal polish for the non-anodized aluminum parts and silicone oil protecting paint and plastic knops, give everthing back a nice shine.

Cleaning the knops with Isopropanol and a brush. I don't like dirty knobs with other people dust.

Some metal parts can be polish for a very nice shining. Unfortunately oscilloscopes without scratches are very seldom, most users in companies didn't take much care for these expensive instruments. Many people were never interested in keeping their equipment clean, they took care on the weekend for their cars but not for the equipment. Nowadays it getting even more worse.

With two Calibration Fixture it's possible to see the full CRT with lines, extremly helpful.

I'am very satisfied with this geometry of a Dual Beam, wonderful, only small abberations from the ideal. After some weeks use get the scope a more detailed CRT geometry readjustment. In this phot one beam draws the vertical lines the other beam the horizontal lines, setting the beams vice versa changes also vertical wiht horizontal.

Don't think all oscilloscopes having a good adjusted geometry, many people don't know the condition of their geometry. Without owning fixtures use the grounded amplifer for a horizontal line check and adjust a time base in the vertical compartment for checking the vertical lines.

For example this photo shows the PG506 step response of a 7A26 amplifier adjusted on this oscilloscope. Took approxiamtely 20 minutes for a flat adjustment. Some improvements are still possible. This amplifier and mainframe combiantion has a specified rise time of 2.2ns, the result is well within the specification. I adjusted for a flat response with less abberations, minimum rise time wasn't the target, a gain of additional 200ps rise time would be possible, but the price for it an overshooting.


I like the 7844, this socpe has a good change to be my Number One for genral purpose working. Using two fast bright beams it is very comfortable. I don't want to talk to much about what you can do with a Dual Beam in comparison with a Dual Trace, there are not many applications where a Dual Beam is really necessary - but one thing for me getting important - it is very comfortable !

Independent beams,
Two independend trigger systems, e.g. Beam 1 with 20ms/DIV and Beam 2 with 1ns/DIV at the same screen, try it with a Dual Trace.
A Dual Beam must not share the intensity time with another channel, a Dual Beam with both beams on 1ms/DIV still looks pretty, on a Dual Trace the screen already starts flickering.
Trace on Beam 1 and Readout on Beam 2, a Dual Trace must share the readout always.

If you don't like  "ALT" and "CHOP" Buttons, try a Dual Beam.        Tektronix Gallery        Index