Topics

Repairing my Tektronix 454 oscilloscope need some advice

Victor
 

Hi members,
I have since many years a tek 454. Few years ago the trace stop to show. Since I have more spare time to dedicate to my small lab I would like to repair it.
The first step was to repair the power supply witch is working now fine. Then I discover that the CRT circuit is not working (very low voltage on the TP 1469 (-1960V)) and no voltage on the CRT heater.
I test all the parts on the Z Axis circuit board feeding the HV voltage and I couldn't find any trouble. I open the plastic box containing the HV transformer for testing and all the parts seems fine except a 1 Mohm resistor which I replace it. Is not easy to work on this circuit with few information but I think the HV transformer is dead.
A very well know Canadian supplier for Tek parts doesn't have any spare available for the HV transformer. So I'm looking for some advice or any members experience on this scope or additional test to confirm my suspicion. Also any specs about the HV transformer will be welcome.
Thanks in advance.
Victor

Albert Otten
 

Victor,

Just to be sure, did you test the fuse F1437 or the voltage across C1437. Also did you test as far as possible the driver transistor Q1430? Q1430 is not in the HV box. I think B, E and C can be reached between the back panel and the HV box.

Recently there were very long discussions here concerning 454 HV problems.

Albert

Victor
 

Albert,

Thank you for your comments.
I check again the CRT circuit and fuse F1437 is ok, the capacitor C1437 is new (I have change it last week) and Q1430 is ok also. Nevertheless I believe something is wrong on the CRT circuit. I measure voltage on Q1414, Q1413 and Q1423 and some voltages are not as the indicate on the manual... transistors has been tested and are on the specs,so may be some diode or resistance is bad....
Unfortunately, during my investigation I face another issue. The Line EMI filter just dead. So I'm looking for a part that can be installed in the exiting space,and then I can restart my circuit test.

Regards,
Victor

Dave Wise
 

If I understand correctly, the 454 HV is similar to the 453. I successfully rewound a 453 transformer using a low-tech process which I documented in topic and photo album titled "453 HV Transformer Rewind".

HTH,
Dave Wise
________________________________________
From: TekScopes@groups.io <TekScopes@groups.io> on behalf of Victor via Groups.Io <vmcfer=aol.com@groups.io>
Sent: Saturday, February 15, 2020 10:40 AM
To: TekScopes@groups.io
Subject: [TekScopes] Repairing my Tektronix 454 oscilloscope need some advice

Hi members,
I have since many years a tek 454. Few years ago the trace stop to show. Since I have more spare time to dedicate to my small lab I would like to repair it.
The first step was to repair the power supply witch is working now fine. Then I discover that the CRT circuit is not working (very low voltage on the TP 1469 (-1960V)) and no voltage on the CRT heater.
I test all the parts on the Z Axis circuit board feeding the HV voltage and I couldn't find any trouble. I open the plastic box containing the HV transformer for testing and all the parts seems fine except a 1 Mohm resistor which I replace it. Is not easy to work on this circuit with few information but I think the HV transformer is dead.
A very well know Canadian supplier for Tek parts doesn't have any spare available for the HV transformer. So I'm looking for some advice or any members experience on this scope or additional test to confirm my suspicion. Also any specs about the HV transformer will be welcome.
Thanks in advance.
Victor

Albert Otten
 

Victor,

Before accusing the HV transformer ... some remarks.
How "very low" was HV at the test point? Is the oscillator amplitude at Q1430 also (in ratio) very low?
What voltage did you have at Q1423-emitter (or pin "P")? The shown voltage -4.35 V is not more than an indication; also there is ripple of 1 V or so. The oscillator starts with base current to Q1430 provide by R1425. Amplitude increases rapidly when Q1423 conducts at the negative going peaks of "P" voltage (at about -1 V). With too high load that situation might not occur.
Did you check HV pot R1401?
Did you measure the current via F1437? Normally about 0.27 A.

It might be possible to test the transformer (and load) to some extend. Remove F1437. Apply + 10 V (w.r.t. GND) or so at C1437. Apply about -2 V at "P". Apply a signal generator output (via DC blocking cap) to Q1430 collector, keeping the amplitude below about 4 V. This way no semiconductors will start to conduct. There should be resonance at about the normal oscillator frequency, about 25 kHz. Turns ratios can be checked when when you can access the secondary winding ends or rectified voltages. (Just an idea, I have never tried this in my 454.)

Albert

Albert Otten
 

(continued)
I did an easier and less risky test. F1437 tied to GND, wire to "P" disconnected. Sine wave generator FG504, same GND , 50 ohm output applied to Q1430 collector (no other termination). Viewed waveform at Q1430 collector at a scope. Kept amplitude pp up to only 1 V so Q1430 would never conduct.
Measurements:
Resonance at about 23 kHz, not very sharp but clear enough.
Amplitude 1 V pp at collector resulted in 77 V DC at the HV test point as measured with a 10M DMM.
Impedance was about 10 ohm.

Albert

Victor
 

Albert,
Thank you for all that great tips that you sent yesterday and now.I was busy today with the  interference filter dead. I couldn't find any spare on the market with the same size as the original. So I decide in the mean time to install one found on my spares parts box. It will be out of the scope but temporary will be a solution to continue the job. Tomorrow morning I will be able to energize my 454 and make test. I will let you know about them.
Thank you,Victor

-----Original Message-----
From: Albert Otten <aodiversen@...>
To: TekScopes <TekScopes@groups.io>
Sent: Tue, Feb 18, 2020 5:44 pm
Subject: Re: [TekScopes] Repairing my Tektronix 454 oscilloscope need some advice

(continued)
I did an easier and less risky test. F1437 tied to GND, wire to "P" disconnected. Sine wave generator FG504, same GND , 50 ohm output applied to Q1430 collector (no other termination). Viewed waveform at Q1430 collector at a scope. Kept amplitude pp up to only  1 V so  Q1430 would never conduct.
Measurements:
Resonance at about 23 kHz, not very sharp but clear enough.
Amplitude 1 V pp at collector resulted in 77 V DC at the HV test point as measured with a 10M DMM.
Impedance was about 10 ohm.

Albert

Victor
 

Dave,
Thank you for the tip. I read and see the photo album and this is a great information and a great job ! I believe also that HV transformer is the same on 453 and 454.  Actually I am trying some additional test with the help of Albert Otten. If the test fail then I will be sure that my HV transformer is dead. Then I would like to ask you some additional information or details about your rewound. 
Thank you,Victor

-----Original Message-----
From: Dave Wise <david_wise@...>
To: TekScopes@groups.io <TekScopes@groups.io>
Sent: Mon, Feb 17, 2020 9:05 pm
Subject: Re: [TekScopes] Repairing my Tektronix 454 oscilloscope need some advice

If I understand correctly, the 454 HV is similar to the 453.  I successfully rewound a 453 transformer using a low-tech process which I documented in topic and photo album titled "453 HV Transformer Rewind".

HTH,
Dave Wise
________________________________________
From: TekScopes@groups.io <TekScopes@groups.io> on behalf of Victor via Groups.Io <vmcfer=aol.com@groups.io>
Sent: Saturday, February 15, 2020 10:40 AM
To: TekScopes@groups.io
Subject: [TekScopes] Repairing my Tektronix 454 oscilloscope need some advice

Hi members,
I have since many years a tek 454. Few years ago the trace stop to show. Since I have more spare time to dedicate to my small lab I would like to repair it.
The first step was to repair the power supply witch is working now fine. Then I discover that the CRT circuit is not working (very low voltage on the TP 1469 (-1960V)) and no voltage on the CRT heater.
I test all the parts on the Z Axis circuit board feeding the HV voltage and I couldn't find any trouble. I open the plastic box containing the HV transformer for testing and all the parts seems fine except a 1 Mohm resistor which I replace it. Is not easy to work on this circuit with few information but I think the HV transformer is dead.
A very well know Canadian supplier for Tek parts doesn't have any spare available for the HV transformer. So I'm looking for some advice or any members experience on this scope or additional test to confirm my suspicion. Also any specs about the HV transformer will be welcome.
Thanks in advance.
Victor

Victor
 

Albert,
With the 454 energized today I could make some of your recommended test. I start with the "less risky test" and the results was far from my expectations. After follow your process I applied to Q1430 collector 1V pp and the voltage at the HV test point was only - 3 to - 4 V. I change the frequency from 23 kHz until 27 kHz and no results and no voltage increase to 77 V DC.
Otherwise: The current at F1437 measured is 0.22A. 
Also adjustment on pot R1401 give any variation on the HV test point.
The voltage on HV test point actually (in normal operation) is  - 64V. 
I could'nt make other test mentioned on your message on 02/17.
My test equipment is : scope TEK 7603, wave generator Sound Technology 3100A, meter Fluke 87III.
Any thought from your side ? 
Thank you 
Victor

-----Original Message-----
From: Albert Otten <aodiversen@...>
To: TekScopes <TekScopes@groups.io>
Sent: Tue, Feb 18, 2020 5:44 pm
Subject: Re: [TekScopes] Repairing my Tektronix 454 oscilloscope need some advice

(continued)
I did an easier and less risky test. F1437 tied to GND, wire to "P" disconnected. Sine wave generator FG504, same GND , 50 ohm output applied to Q1430 collector (no other termination). Viewed waveform at Q1430 collector at a scope. Kept amplitude pp up to only  1 V so  Q1430 would never conduct.
Measurements:
Resonance at about 23 kHz, not very sharp but clear enough.
Amplitude 1 V pp at collector resulted in 77 V DC at the HV test point as measured with a 10M DMM.
Impedance was about 10 ohm.

Albert

Albert Otten
 

Victor,

Obviously there is not enough secondary HV, or a diode is bad. If you remove the plastic cover of the HV box you can access at least one diode which is connected to a hot winding end (I don't remember which one, I think it has been mentioned in another recent 454 thread). With 1 V pp over the primary you could view the wave form at that secondary winding, using a 10 M (and 10X) probe. Then a bad (open) diode plays no role.
I estimated the primary impedance (about 10 R) by comparison of amplitude (1 V pp) with the open-circuit output amplitude of the 50 R generator. Did you also do such? Maybe tomorrow I will (just for fun with my 7854) view both voltage across and current into the primary. With the 7854 keyboard calculator it's easy to determine reactive and real input power to primary. I guess it will mostly be real power because of the (cold!) CRT filament load. (Of course the phase shift between voltage and current can be viewed on an arbitrary scope, without storage or calculator).
You might also view the collector waveform during normal operation. The 0.22 A current seems to be small and probably indicates that Q1430 can not deliver enough power to increase the oscillation to normal level; either because a fault in the feedback circuit (resistor string to Q1414 etc.) or a too heavy secondary load (shorted HV cap?) or ...(?)

Albert

On Wed, Feb 19, 2020 at 07:55 PM, Victor wrote:


Albert,
With the 454 energized today I could make some of your recommended test. I
start with the "less risky test" and the results was far from my expectations.
After follow your process I applied to Q1430 collector 1V pp and the voltage
at the HV test point was only - 3 to - 4 V. I change the frequency from 23 kHz
until 27 kHz and no results and no voltage increase to 77 V DC.
Otherwise: The current at F1437 measured is 0.22A. 
Also adjustment on pot R1401 give any variation on the HV test point.
The voltage on HV test point actually (in normal operation) is  - 64V. 
I could'nt make other test mentioned on your message on 02/17.
My test equipment is : scope TEK 7603, wave generator Sound Technology 3100A,
meter Fluke 87III.
Any thought from your side ? 
Thank you 
Victor     

-----Original Message-----
From: Albert Otten <aodiversen@...>
To: TekScopes <TekScopes@groups.io>
Sent: Tue, Feb 18, 2020 5:44 pm
Subject: Re: [TekScopes] Repairing my Tektronix 454 oscilloscope need some
advice

(continued)
I did an easier and less risky test. F1437 tied to GND, wire to "P"
disconnected. Sine wave generator FG504, same GND , 50 ohm output applied to
Q1430 collector (no other termination). Viewed waveform at Q1430 collector at
a scope. Kept amplitude pp up to only  1 V so  Q1430 would never conduct.
Measurements:
Resonance at about 23 kHz, not very sharp but clear enough.
Amplitude 1 V pp at collector resulted in 77 V DC at the HV test point as
measured with a 10M DMM.
Impedance was about 10 ohm.

Albert


Albert Otten
 

Victor,

Today I did more waveform measurements. I always viewed the primary voltage and primary current ( A6302 current probe) waveform and each time I tuned for resonance. Last time I made a frequency reading mistake (when you turn the FG504 knob cw the the frequency deceases...). Without other probes attached it was 27 MHz. Probes at the secondary side lower the resonance frequency somewhat.
From the visible transformer right side 3 very short blank wires go to ceramic supports. From front to rear these are connected to winding terminals 7 (tripler circuit), 9 (cathode, D1452) and 10 (grid, D1440 visible). Resistances to GND: 547R, 243R and about 20k respectively.
Findings:
Always nice resonance with simultaneously primary voltage maximal, current minimal and no phase shift between these two. In hindsight this could be expected when the primary side acts as a nice parallel RCL circuit with not too low Q factor. R was about 10 Ohm. Also each time the secondary voltage was in phase with the primary.
Amplifications, crude values:
from primary to 7: 375, to 9: 185, to 10: 195.
I did a more precise simultaneous measurement of negative peak at 9 and the DC voltage at HV TP (with DMM). The "loss" was about 10 V (of value about 80 V), looks realistically.
Anyway, it could be very informative to view those secondary waveforms.

Albert

On Wed, Feb 19, 2020 at 09:36 PM, Albert Otten wrote:


Victor,

Obviously there is not enough secondary HV, or a diode is bad. If you remove
the plastic cover of the HV box you can access at least one diode which is
connected to a hot winding end (I don't remember which one, I think it has
been mentioned in another recent 454 thread). With 1 V pp over the primary you
could view the wave form at that secondary winding, using a 10 M (and 10X)
probe. Then a bad (open) diode plays no role.
I estimated the primary impedance (about 10 R) by comparison of amplitude (1 V
pp) with the open-circuit output amplitude of the 50 R generator. Did you
also do such? Maybe tomorrow I will (just for fun with my 7854) view both
voltage across and current into the primary. With the 7854 keyboard calculator
it's easy to determine reactive and real input power to primary. I guess it
will mostly be real power because of the (cold!) CRT filament load. (Of course
the phase shift between voltage and current can be viewed on an arbitrary
scope, without storage or calculator).
You might also view the collector waveform during normal operation. The 0.22 A
current seems to be small and probably indicates that Q1430 can not deliver
enough power to increase the oscillation to normal level; either because a
fault in the feedback circuit (resistor string to Q1414 etc.) or a too heavy
secondary load (shorted HV cap?) or ...(?)

Albert

Albert Otten
 

Victor,

Previously I wrote "I did an easier and less risky test. F1437 tied to GND,...". Actually, with the signal between GND and collector, this doesn't help at all because of the high inductance of L1437. It works since C1437 and R1149 provide a low impedance AC path to GND. I noted this later on. The right way is to tie C1437 "+" to GND. Luckily in practice I hardly noted any differences in amplitudes between tied to GND or not. C1437 is located in a clamp at the rear panel, next to Q1430.

Albert

Victor
 

Albert,
This afternoon, I was reviewing the measurement done yesterday and found a problem (intermittent connection) with the cable connecting the wave generator to the Q1430 collector. After repair it I test again my 454 and now with a 1Vpp on the collector of the Q1430, I get 52V at the HV test point.So, with this result I believe my HV transformer is Ok. I have a doubt about the HV diodes... what do you think ? By the way how do you test HV diode,I understand you cannot do it with a regular tester.Tomorrow before made the test that you recommend today. I would like to review the HV CRT circuit to see if any resistor or capacitor are out of specs (transistors have been test and look good).
Thank you,Victor

-----Original Message-----
From: Albert Otten <aodiversen@...>
To: TekScopes <TekScopes@groups.io>
Sent: Thu, Feb 20, 2020 3:19 pm
Subject: Re: [TekScopes] Repairing my Tektronix 454 oscilloscope need some advice

Victor,

Today I did more waveform measurements.  I always viewed the primary voltage and primary current ( A6302 current probe) waveform and each time I tuned for resonance. Last time I made a frequency reading mistake (when you turn the FG504 knob cw the the frequency deceases...). Without other probes attached it was 27 MHz.  Probes at the secondary side lower the resonance frequency somewhat.
From the visible transformer right side 3 very short blank wires go to ceramic supports. From front to rear these are connected to winding terminals 7 (tripler circuit), 9 (cathode, D1452) and 10 (grid, D1440 visible). Resistances to GND: 547R, 243R and about 20k respectively.
Findings:
Always nice resonance with simultaneously primary voltage maximal, current minimal and no phase shift between these two.  In hindsight this could be expected when the primary side acts as a nice parallel RCL circuit with not too low Q factor. R was about 10 Ohm. Also each time the secondary voltage was in phase with the primary.
Amplifications, crude values:
from primary to 7: 375, to 9: 185, to 10: 195.
I did a more precise simultaneous measurement of negative peak at 9 and the DC voltage at HV TP (with DMM). The "loss" was about 10 V (of value about 80 V), looks realistically.
Anyway, it could be very informative to view those secondary waveforms.

Albert

On Wed, Feb 19, 2020 at 09:36 PM, Albert Otten wrote:


Victor,

Obviously there is not enough secondary HV, or a diode is bad. If you remove
the plastic cover of the HV box you can access at least one diode which is
connected to a hot winding end (I don't remember which one, I think it has
been mentioned in another recent 454 thread). With 1 V pp over the primary you
could view the wave form at that secondary winding, using a 10 M (and 10X)
probe. Then a bad (open) diode plays no role.
I estimated the primary impedance (about 10 R) by comparison of amplitude (1 V
pp) with the  open-circuit output amplitude of the 50 R generator. Did you
also do such? Maybe tomorrow I will (just for fun with my 7854) view both
voltage across and current into the primary. With the 7854 keyboard calculator
it's easy to determine reactive and real input power to primary. I guess it
will mostly be real power because of the (cold!) CRT filament load. (Of course
the phase shift between voltage and current can be viewed on an arbitrary
scope, without storage or calculator).
You might also view the collector waveform during normal operation. The 0.22 A
current seems to be small and probably indicates that Q1430 can not deliver
enough power to increase the oscillation to normal level; either because a
fault in the feedback circuit (resistor string to Q1414 etc.) or a too heavy
secondary load (shorted HV cap?) or ...(?)

Albert

Chuck Harris
 

Use a 9V battery, or two, in series with a 10K resistor, and
a low current meter.. then test in the usual way.

-Chuck Harris

Victor via Groups.Io wrote:

Albert,
This afternoon, I was reviewing the measurement done yesterday and found a problem (intermittent connection) with the cable connecting the wave generator to the Q1430 collector. After repair it I test again my 454 and now with a 1Vpp on the collector of the Q1430, I get 52V at the HV test point.So, with this result I believe my HV transformer is Ok. I have a doubt about the HV diodes... what do you think ? By the way how do you test HV diode,I understand you cannot do it with a regular tester.Tomorrow before made the test that you recommend today. I would like to review the HV CRT circuit to see if any resistor or capacitor are out of specs (transistors have been test and look good).
Thank you,Victor


-----Original Message-----
From: Albert Otten <aodiversen@...>
To: TekScopes <TekScopes@groups.io>
Sent: Thu, Feb 20, 2020 3:19 pm
Subject: Re: [TekScopes] Repairing my Tektronix 454 oscilloscope need some advice

Victor,

Today I did more waveform measurements.  I always viewed the primary voltage and primary current ( A6302 current probe) waveform and each time I tuned for resonance. Last time I made a frequency reading mistake (when you turn the FG504 knob cw the the frequency deceases...). Without other probes attached it was 27 MHz.  Probes at the secondary side lower the resonance frequency somewhat.
From the visible transformer right side 3 very short blank wires go to ceramic supports. From front to rear these are connected to winding terminals 7 (tripler circuit), 9 (cathode, D1452) and 10 (grid, D1440 visible). Resistances to GND: 547R, 243R and about 20k respectively.
Findings:
Always nice resonance with simultaneously primary voltage maximal, current minimal and no phase shift between these two.  In hindsight this could be expected when the primary side acts as a nice parallel RCL circuit with not too low Q factor. R was about 10 Ohm. Also each time the secondary voltage was in phase with the primary.
Amplifications, crude values:
from primary to 7: 375, to 9: 185, to 10: 195.
I did a more precise simultaneous measurement of negative peak at 9 and the DC voltage at HV TP (with DMM). The "loss" was about 10 V (of value about 80 V), looks realistically.
Anyway, it could be very informative to view those secondary waveforms.

Albert

On Wed, Feb 19, 2020 at 09:36 PM, Albert Otten wrote:


Victor,

Obviously there is not enough secondary HV, or a diode is bad. If you remove
the plastic cover of the HV box you can access at least one diode which is
connected to a hot winding end (I don't remember which one, I think it has
been mentioned in another recent 454 thread). With 1 V pp over the primary you
could view the wave form at that secondary winding, using a 10 M (and 10X)
probe. Then a bad (open) diode plays no role.
I estimated the primary impedance (about 10 R) by comparison of amplitude (1 V
pp) with the  open-circuit output amplitude of the 50 R generator. Did you
also do such? Maybe tomorrow I will (just for fun with my 7854) view both
voltage across and current into the primary. With the 7854 keyboard calculator
it's easy to determine reactive and real input power to primary. I guess it
will mostly be real power because of the (cold!) CRT filament load. (Of course
the phase shift between voltage and current can be viewed on an arbitrary
scope, without storage or calculator).
You might also view the collector waveform during normal operation. The 0.22 A
current seems to be small and probably indicates that Q1430 can not deliver
enough power to increase the oscillation to normal level; either because a
fault in the feedback circuit (resistor string to Q1414 etc.) or a too heavy
secondary load (shorted HV cap?) or ...(?)

Albert




Albert Otten
 

Victor,
I would first of all test the waveforms at terminals 7, 9 and 10 and check that the amplitudes have more or less the same ratios to each other (and perhaps to the 1 V pp) as in my case. Your 52 V is really to low, so check it against the (negative) peak voltage at terminal 9. The secondary waveforms may show a slightly flat negative peak because of diode conduction, but only slightly. At terminal 9 the negative "flat" peak was about 5 V smaller than the "round" positive peak.
Can you measure the primary current? Perhaps your generator can show the current drawn?
For test HV diodes see the message by Chuck Harris or see message #163125).

Albert

On Thu, Feb 20, 2020 at 09:42 PM, Victor wrote:

Albert,
This afternoon, I was reviewing the measurement done yesterday and found a
problem (intermittent connection) with the cable connecting the wave generator
to the Q1430 collector. After repair it I test again my 454 and now with a
1Vpp on the collector of the Q1430, I get 52V at the HV test point.So, with
this result I believe my HV transformer is Ok. I have a doubt about the HV
diodes... what do you think ? By the way how do you test HV diode,I understand
you cannot do it with a regular tester.Tomorrow before made the test that you
recommend today. I would like to review the HV CRT circuit to see if any
resistor or capacitor are out of specs (transistors have been test and look
good).
Thank you,Victor

Brad Thompson
 

Victor via Groups.Io wrote on 2/20/2020 3:42 PM:

<snip> I have a doubt about the HV diodes... what do you think ? By the way how do you test HV diode,I understand you cannot do it with a regular tester.

Hello, Victor--

The high-voltage diodes likely consist of several diode junctions connected in series
to obtain a high reverse-breakdown voltage. You'd need a specialized test instrument to
measure reverse-bias leakage current and breakdown voltage. For example, stacking eight
junctions reverse-rated for 1000 volts each would produce an 8000-volt diode.

In this example, you can measure the high-voltage diode's forward-voltage drop by
connecting a current-limiting resistor (e.g., 1K Ohms) in series with the diode and a power supply
( 10 volts) whose voltage exceeds the estimated forward voltage drop of eight junctions
(e.g., 0.7 V *  8 = 5.6 volts.). Measure the voltage across the diode you're testing.

If the diode  doesn't exhibit forward conduction, it's  defective. Unfortunately,
this test won't diagnose abnormally low or intermittent reverse breakdown.

73--

Brad  AA1IP

 

Hi Chuck,

Some HV diodes may need more than 9V to turn on. The ones I used in my HV supply need about 17V for a 2mA forwards current. they are some Chinese 20KV diodes used in microwaves (an overkill but cheap, at $7.49 for 8 of them at Amazon). He can use the +100V of the scope itself through a 100K resistor for a short measurement of 1mA forwards, then reverse the diode and measure zero current.

Ernesto

Chuck Harris
 

Indeed, which is why I said a 9V battery, or two...

The diodes in these supplies are a little special in that
they need to be fast recovery diodes. They are working at
about 50KHz.

A common ordinary microwave oven makes its HV at 60Hz, and
can use very slow recovery diodes effectively. A slow diode
will drag a 50KHz supply down, appearing as a short circuit for
several tens of microseconds after the transformer's polarity
reverses.

-Chuck Harris

Ernesto wrote:

Hi Chuck,

Some HV diodes may need more than 9V to turn on. The ones I used in my HV supply need about 17V for a 2mA forwards current. they are some Chinese 20KV diodes used in microwaves (an overkill but cheap, at $7.49 for 8 of them at Amazon). He can use the +100V of the scope itself through a 100K resistor for a short measurement of 1mA forwards, then reverse the diode and measure zero current.

Ernesto

 

Hi Chuck,

You are very right about the importance of the diode recovery time. The specification on Amazon for this cheapy diode is Trr = 100 nS maximum. Other "fast" HV diodes specify 60 or 70 nS, not a big difference.
At 50 kHz the period of the rectified sine wave is 20 mS, 200 times longer than the Trr. During these 100nS the peak of the sine wave will barely comes down, so the operation should be safe. Indeed when I installed these diodes the efficiency of the supply increased, and it behaved correctly.

- Ernesto

Jim Ford
 

Hi, Ernesto.

Actually, 20 ms is 200,000 times longer than 100 ns. Just sayin'.

Jim Ford

------ Original Message ------
From: "Ernesto" <ebordon@...>
To: TekScopes@groups.io
Sent: 2/20/2020 6:41:48 PM
Subject: Re: [TekScopes] Repairing my Tektronix 454 oscilloscope need some advice

Hi Chuck,

You are very right about the importance of the diode recovery time. The specification on Amazon for this cheapy diode is Trr = 100 nS maximum. Other "fast" HV diodes specify 60 or 70 nS, not a big difference.
At 50 kHz the period of the rectified sine wave is 20 mS, 200 times longer than the Trr. During these 100nS the peak of the sine wave will barely comes down, so the operation should be safe. Indeed when I installed these diodes the efficiency of the supply increased, and it behaved correctly.

- Ernesto