Tectronix 465 last update

To finish the long story about repairing The 465 is working fine now
I found a solution to prevent damage by possible CRT HV flash over
I just inserted a protective 860k resistor ( could be 1M ) in series with the -2450V  cicuit
Cut the trace direct after the HV diode   see  below the point "test onderbroken"
860k  gave a drop of only 100V    no problem for the HV regulator
the resistor can be low Wattage only 0.0116 W 100V 860k
The cathode current is 116uA neglecting the low parallel load of multiple M OHm
Since then only 1 time there was a breakdown of all voltages "light outage" but just switching off and on and all was working again  no collateral damage...
I use the Tektronix 465 again and live with the occasional break downs until i find a KRT replacement. But  after some warm up time with break downs and return it works stable for hours.
BTW a 100MHz b.t.w.

Hantek DSO5102P USB Digital Storage Oscilloscope 2Channels 100MHz 1GSa/s   can be purchased for 208€  ...

This device has very good reviews  some review mentioned that it was factory unlocked for 200MHz / 2nS even

It continues...
Now the diagnose is the  CRT  suffers internal or external short in the 14kV after 4m  or 24m then some tick is heard the tube display gets dark or shows uncontrollable bright beam.
And the 5V supplied Vertical scale lights  are off.
That flash? damaged the Cascade  and or the power regulation circuit especially U1554 regulating 5V and -8V was damaged several times when testing.
The 465 it worked good for some time and then it broke down.

I let it rest some time last year but now  removed the CRT for inspection
Don't forget to remove the 4 connectors of the X an X deflection electrodes before ...
After 40 years a black dust layer was over the CRT arround the 14kV connection.
But no trace of flashing.

After the last break down the 2450V was also gone, so took out again the vertical amplifier PCB  Now i know how to do it quicker had made cross cuts in the difficult bolts to remove with a screw driver. Still difficult but avoiding un-soldering the input capacitors. This not the the most convenient construction of the 465.
Found Q1414 broken after fruitless replacing CR1421 the 2000V AC rectifier.
I tested the Q1418 collector voltage with my other vintage PM3231 Phillips 2x15MHz
24V DC with a sinus shape 30Vtt  abt. 55kHz  should be the normal working situation.
First i had some HV 1100V  with Cascade disconnected and CRT unplugged.
The  H-voltage was very unstable and way too low. The Q1418 60kHz  oscillation stopped.
Could not measure shorts in the HV circuit. Then i disconnected the HV diode CR1421 so T1420 was without load.. then Q1418 was at full drive and it started to spark around T1420 and Ozon smell..
The AC HV must have been at maximum  way higher then 2000V AC.
I switched off quick hoping that no damage was done.
I measured the HV coil 88 Ohm 432mH that seems still okay.
then I tested transistors in the HV circuit and bingo Q1414 was fault a simple general TUN
Replaced and voila 2450V was back still without the DIY Cascade but i have good hope the cascade  okay its  2n2 10kV and 20kV diodes are safe above normal operational load.

So more or less i am back to where i was before.
Switching the 465 on will result in another flash break down i am afraid.
How can i know if this CRT is okay, it had a good bright beam maybe too bright the brightness control is almost at minimum.
when it was working as long as it did. If the CRT would loose its high vacuum it would not show that bright beam in the first place.
Focus and brightness where normal adjustable the X deflection had the correct width.

I tested the tube at 17kV using and high voltage power supply regulating from 0-20kV  no flash zero current

Now i suspect that the problem is a flash over between -2450 cathode and one of the other electrodes with levels close to 0V

With an internal flashing CRT It is is pulling a dead horse but i learned a lot about the 465 and its electronics.

Next time  (...)  i would  run the 465 with the 24V fuse F1419 pulled and test if the power is okay without HV  and that the triggering on 1kH 300mV Cal works okay
Replace the CRT  and make or repair my DIY cascade is the only solution left.
Replacing this CRT is a thing.. it is not impossible when having another 465 for spare   
Cleaned the tube as good as possible around the 14kV joint.
17kV test with no other electrodes connected had no flash.
Should be okay but i did not found an explanation for the problem yet.
And the experience is that i will happen again if the cause is unclear.

previous :
Tectronix 465 still in trouble
It became an interesting yet frustrating project
Investigating this legendary Tectronix 465 1971  high quality classic electronics.

The diagnoses was:  HV cascade broken that loaded  the  H.V. oscillator so the 2450V was zero.
At that point the amplifiers and triggering  still worked but the CRT was dark.
Now finished a DIY cascade  using 6x 2n2/10kV   6x  20kV 20mA 100nS
Someone shared a picture of his DIY cascade on www.  It worked but when placed the 2450V  dropped to 1200V . CR1421 BY176 was broken just when i tested the cascade, tricky.
Replaced this BY176 with a 20kV 5mA 100nS  eBay type and the  2450V came back meaning the 2000VAC for the cascade was present again.
Tested the HV diode using 33k 2x9V battery and a mA meter. The BY had a very high forward resistance but still showed some weak diode function.
                                                     test by-passing the old BY176 1/2 wave 2000V rectifier
It worked  but not for long... sudden the beam and the V/div signal lights started to blink and the 1kHz 300mV test signal trigger jumped more and more and then stopped.
The 5V was almost to zero, U1554 SN72558 malfunction I used fttb a 5V KIA7805
The SN72558p is now a vintage part  this simple dual opamps are offered on eBay for a crazy 50$ as if they are made of solid gold.
This U1554  dual opamp regulates the 5 and   -8V  the negative -8V was still present with the external 5V it worked okay but again not for long suddenly 5V dropped to zero.
broken U1554  was loading the 15V to much and dropped  to 7.5V.
Replaced U1554 with an 1€ LM358N (Ti)  it worked again ,i could restore the 5V to original using U1554 and Q1556 a medium power emitter follower. The 5V is not loaded much only about 110 Ohms.
But  :wtf: after some time it started to flicker again and now the -8V was gone.
How can that be? The supply for this LM358 pin4= -8  pin8 = +15   The IC is rated +/-16V supply
Then i tried an NE5532 supply specified for max -22  +22V ( the document is unclear if it can handle +/-22V supply simultane but i have to guess it does)
This NE5522 worked but the -8V noise is strong the horizontal trace became very fuzzy.
The 5.060 V  was supplied by the 7805
Forget the NE5532 this opamp is unusable so another LM358N is placed.
What a good thing sockets for transistors and IC in this 465 replacing and testing is very easy.

Now my 465 works but after sometime the triggering jumps few cycles.
That i have seen before as a start of the trouble above.
Possible I overlook something  that can break the 5 and -8V regulator ?
This low power regulation circuit is well separated from the H.V. generation part.
I do not know why the power regulation is done with 2 dual op-amps possible noise consideration.
There must be some high voltage spark involved
One positive thing is that home made 14kV cascade did survive al treatment.
                                             C  2n2 10kV       diode 20kV 100nS  R 47k

Diy cascade build to an example found with Google the 47k resistor  was not explained but it could function as a RC filter where te C is the electrode capacity.When unplugging the 14kV be aware of the HV this CRT electrode holds high voltage long time after switching off !

                           plastic box with HV cascade PUR foam used and the HV cable is now leading over the CRT since the HV-connector can not pass through the small opening under the CRT where it was original  to do that the CRT and u-screen has to be removed aswel !
             Here i used PU construction glue as compound but construction glue is not suitable it will not harden out in sealed boxes. PUR foam will hard out but first it expands and presses through  small openings in the box but then it shrinks again while hardening or curing.
I used PUR foam once to protect antenna cable connection boxes and that worked good to prevent against damp or water damage to cables as long as the PUR is not exposed to UV sunlight.
I have not seen any moisture issues with this PUR foam.
But never seen the effect of shrinking of this foam when hardening.
Think the best way is epoxy based resin.     

The cathode high voltage 2450V is generated by a separate circuit.
Q1418 free oscillating with T1420    powered by unregulated 24V.
So if the 5V or -8V drops still the CRT can produce a beam  but at full strength  since the beam intensity and deflection is no longer controllable.
 I measured about  15uS  60kHz 25Vtt  at the collector of Q1418 Be aware of the -2450V level at the CRT cathode !! and that a normal voltmeter will not measure 60kHz signals.
 The 2 small ionization lamps DS1425 DS1426 i wondered what their function is, they light up short after switching off power. It seems protecting the first electrode voltage. 
I removed the glass fuse F1419  to switch of the 24V for HV generator and cascade to prevent burning the CRT.  while struggling with the low voltage problems.

temporary ??  by passing 5V with an 7805

part of 465 power supply
55V  is the only adjustable regulated Voltage used as reference for the other supply's  +15 +5 -9V
The outputs are current limited. The 110V is composed of 55V + 56V  regulated.
So all supply voltages hang together.  It is a special circuit.
The circuit for  -8 is special
I have not yet a clue why U1554  the SN72558p broke and then it's replacements LM358N and  |NE5532 broke too one afte 1 hour the later after 5 minutes.
By studying  the circuit i hope to get closer to the cause
My guess now is that the SN72558p  can handle the different outputs while the LM358N can't
1 hour  it worked fine and then sudden tick and power down, again U1554 LM358 broken
I did watch 55V on another scope very stable no glitches. and 4.990V  rock stable
This is the frustrating part, making the HV cascade was interesting and the part seems to have nothing to do with the break down of U1554
The SN72558p features a shortcircuit protection the LM358N and NE5532   not
That is a difference at least the chips are not a copy of each other.
The MC1458 also specifies short circuit protection i will try that one too.
Normally i would use a 7805  and a 7908 but how to take the negative voltage
the circuit is reversed ??

few seconds before crash..

This is how it looks now, the horizontal trace okay  it should be a bit smaller some noise in the vertical amplifier.
Until now leaving it running like this always ended in failure in the 5 or -8V  supply i lost trust..