Testing a Flyback Transformer (Leader LBO-302 Oscilloscope)

[bsfeechannel]

I redid my calculations.



I_MEAN = 0.42 V / 10 Ω = 42 mA

V_R316 = 42 mA * 470 Ω = 19.74 V.

P_R316 = (42 mA)² * 470 Ω = 0.829 W.

Pretty close to what you are measuring with the multimeter and below the spec'd 3W.

[Smoky]

This has been an incredible experience for me.

Thank you BsFeeChannel!

You too FloobyDust!

Here's a shot from my Tektronix TDS420:





And one last thing for tonight, I'm measuring 154.1 and 154.4 volts AC going to Diodes D301 and D302 from the power transformer. I believe it should be 128 volts AC as read on the schematic.

Please check this because this could be the reason why R316 has to drop so much voltage.

[floobydust]

I redrew the circuit, stared at it and realized the cathode current path is either through R416 or the neons.

The neons are there to provide Q405 plenty of leeway for the voltage at its collector to be up to 100V more negative than the cathode. The heater is also 100V more negative to repel any emissions in its direction.

So the voltage at the control grid should not be more positive than -1500V with the maximum intensity, and not more negative than around -1600V with zero intensity.

Here it's a high voltage glowing zener diode as somebody at Leader intended? It still doesn't make enough sense beyond supplying (when glowing) more current for the two 5V zeners. I see no link from Q405 collector to the lamps.

Higher sweep speed CRT's and TV's use an additional acceleration (second) anode. This is where neon lamps or spark gaps are for dealing with static charges.

[floobydust]

... I'm measuring 154.1 and 154.4 volts AC going to Diodes D301 and D302 from the power transformer. I believe it should be 128 volts AC as read on the schematic.

Please check this because this could be the reason why R316 has to drop so much voltage.

How are TP301 (+150V), TP302 (+130V), TP303 (+40V), TP304 (-30V) ?
It would tell if the power transformer is hyper, or the loads on 150V, 130V are too light and causing the raw DC voltage to be high.
It's a badly distorted sine-wave at the transformer secondary, so your multimeter may not read true RMS.

My poking around in LT Spice I'm still getting high ~5W power dissipation in R316. I don't have all the flyback transformer's data (winding resistances, secondary inductance etc) but made some estimates.
170V raw I get Peak 225mA, Average 50mA, RMS 100mA current through R316 for 4.7W and a 3W part would be past the limit over 200°C.
Careful a thermocouple does not connect to the resistor if paint or ceramic is chipped, you would get a hazardous live thermocouple which is dangerous. A spot IR temperature reading might be enough.

Before going into any notion of changing the factory design, best we agree if things are normal or not for R316.

[xavier60]

Not that it really matters, I think it's not a flyback converter.
There is no damper diode across Q401 to pass reverse current for energy recovery.
I expect the Collector waveform to be rather sinusoidal and Q401 not saturating.

[Smoky]

When looking at the schematic, the power going to R316 appears not to be regulated.

There is slight discoloration on the board in the area of R316 and it's probably from 40 years of heat transfer.

I see no reason to allow this to continue.

Also, the Ebay seller that sold me the neons said that he found a few neons that fire at 59 volts AC. He put them in the mail. He said they are physically different in size.

[Smoky]

Even though I got home from work late, I'm right back at it!

Tonight, I want to do these following test points:



Here is the first one, TP601. The trigger level knob is set to "auto" and -1500 VDC is at TP402.

TP601:



Here's the problem. This screen-shot was taken only when +27 volts is applied to Q406. When I begin to turn VR401 to raise the high voltage, resistor R316 goes into meltdown. I seen vapors from R316 this time!

When the trigger level knob is in the "auto" position, heat from Hell emits from R316 when turning VR401.

I continued with the other test points only for future reference to compare. They don't look right.

[Smoky]

TP602:



I would say "not right."

[Smoky]

TP603:

[Smoky]

TP604:



It appears off too.

Btw, on page 29 of the LBO-302 Operating Instructions and Service Manual, I stand by my use of "Flyback Transformer" in the title of this EEVBlog.com thread

[bsfeechannel]

I need to find out why resistor R316 gets blazing hot.

We have discovered, by your measurements, that R316 is dissipating no more than 1W.

This is a reasonable value. If, instead of that resistor, you had a transistor, or an integrated voltage regulator, the power dissipated would be the same.

However, one thing is power dissipation, another is surface temperature.

If the resistor you chose has a high thermal resistance, it'll bake.

That's why I asked you to measure its surface temperature. You don't need a thermal camera. A thermocouple will do. Many modern multimeters have a range for temperature using thermocouples.

If the temperature is out of what is expected for a carbon film resistor, perhaps a wire-wound resistor (with a heatsink) of a higher rated power would be in order to replace the old resistor.

[Smoky]

BsFeeChannel, the leads of R316 began to melt the solder on the PC board.

The resistor is a 3 watt flameproof 470 Ohm metal oxide film made by Panasonic (ERG-3SJ471).

The service manual states it should be a metal film so I thought this resistor was a good choice, but now, maybe not.

[bsfeechannel]

This is not the first design where I see components heating up so much that they melt their own solder joints.

It's too early to discard other causes, but poor design choices is something that can always be included in the list.

For the time being, let's perform some measurements more. We could hook up a scope probe on C402 and see how good it's decoupling AC. My hint is that it is leaving quite a bit of AC for R316 to deal with, and that could explain the extra heat that our calculations are not showing up to now.

If your scope is dual-trace, we could also hook up the other probe on the incoming lead of R316, subtract the waveforms (if your scope has this feature) and see exactly what voltages R316 is really having to sustain.

By the way, it's always a good idea show more than one pulse (but no more than two or three) on the scope, so that we can calculate the frequency at which the oscillator is operating, because that helps to determine duty cycles and the impedance of reactive components (capacitors and inductors) in the circuit.

 

If you please.

[Smoky]

BsFeeChannel, it looks like there will be a delay for the waveforms since the I removed the .68uf cap from the machine.

Here's that .68uf cap:

[bsfeechannel]

And as I mentioned before, when the "Auto" trigger mode is energized, R316 goes ballistic, so I image there's a little problem with this "merger" of sections.

When you autotrigger, the beam is present all the time, which draws current through the cathode, which is fed by the HV supply, which is supplied by R316. No surprise.

I'm going to get to work on those requested waveforms tonight.

Thank you.

What made me scratch my head was floobydust's simulation on SPICE. Although he used approximated values, it hit me: we are forgetting to consider the current that goes down the drain through C402 and has to come through R316.

We need to investigate that.

[Smoky]

BsFeeChannel,

I have the .68uf 100v capacitor out of the machine from the trigger level pot. I do not want to reinstall it, it shows signs of stress. I imagine this will delay the waveform from C402 until the new capacitor arrives.

From the calculation that FloobyDust performed, which size of resistor would be preferred, a 5 watt or a 7 watt?

[Smoky]

"When you autotrigger, the beam is present all the time, which draws current through the cathode, which is fed by the HV supply, which is supplied by R316. No surprise."


I see now. That's why the heat ramps up on R316 after I start turning VR401 to reduce the voltage on Q406. The current draw increases.

I even see twinkling in the neons when auto-trigger is activated and when the base of Q406 is at +27 volts. R316 feels slightly warmer during this state also.

I may not be making sense to you, but I can recollect the different levels of warmth from R316 and when they change

[Smoky]

FloobyDust,

You mentioned that you needed to guess on some of the data about the Flyback transformer in your calculation.

Well, when I first discovered that I didn't calibrate the leads properly when measuring inductance of the Flyback transformer's windings on the Sencore LC53 Z Meter.

I retested the coils twice and here are the results:

Primary side:

   Red and Orange winding measured 6.8 uH.

   Yellow and Green winding measured 2.27 mH.

And the Secondary winding measured 720 mH.

[Smoky]

I just read something interesting about "Flameproof" and metal-oxide resistors.

Flameproof resistors may not burn but they can get "red" hot and damage nearby components.

Metal-oxide resistors can handle the power and heat of the environment they are installed in, but the resistor gets hotter because of its smaller case size.

The 3 watt 470 Ohm Panasonic's that I installed for R316 were both "Flameproof" and a "metal-oxide film" type.

Just lovely!

[bsfeechannel]

I have the .68uf 100v capacitor out of the machine from the trigger level pot. I do not want to reinstall it, it shows signs of stress. I imagine this will delay the waveform from C402 until the new capacitor arrives.

No worries. It'll be good to investigate that to be absolutely sure that we do not have any other issues.

From the calculation that FloobyDust performed, which size of resistor would be preferred, a 5 watt or a 7 watt?

Go for broke. The only caveat is to see if it fits on the PCB, or it needs some adaptation like a lug/terminal strip, etc.

I may not be making sense to you, but I can recollect the different levels of warmth from R316 and when they change

Ha! It's the old recollecting-the-different-levels-of-warmth trick.

[xavier60]

Because of the small value of C402, some of the AC power that's produced by the oscillator is being dissipated by R316 as well as what's being caused by the DC.

[Smoky]

Check out this observation.

Here's a picture of the original R316.

Since the resistor and the PC board face downward in the scope, notice how they notched an opening in the PC board to allow the heat to rise!

Leader knew this resistor was going to get hot.

[xavier60]

Look for any change in waveform at the down stream end of R316. Hopefully there might be a clue.

[Smoky]

Thank you Xavier60.

I should have screen shots of those locations Monday night. Once digested, any modification ideas are surely welcome.

BsFeeChannel, the hole spacing for R316 is 1-1/4". I was easily able to insert a .039" lead into both holes. The Dale 5 and 7 watt wirewounds are ~7/8" long and share the same lead diameter (.032"), so we're good. I'd like to see a heat sink proposal on my bench ASAP

I have heat sink compound and that aluminum rail near R316 in the picture above appears beneficial. I have small taps and dies down to #4 too.

The Oscillator board mount on the opposite side uses the aluminum rail w/compound to sink-to also.

[floobydust]

You're about to fall into the power resistor trap. Did you notice the Vishay is 250°C at rated power?
Resistor manufacturers are competing based on resistor size and power rating. They have the materials good for even 350°C operation, but this is ridiculous.

It's simply the laws of physics, where larger surface area = larger heat dissipation.
So a cooler running resistor has to be physically larger. No way around it, in convection cooling unless you want to use a fan. You have to be careful selecting a part.

If the flyback oscillator gets stuck and does not start up, because you are slowly raising the variac up, Q401 will sit there on and badly roast R316 and heat up the flyback primary winding. I would expect the mains fuse to blow. To protect the flyback I would have a fuse there, or be using a fusible resistor.
Unless it is oscillating and now R316 is running much hotter than before?
I would need the DC resistance of the flyback windings and the power transformer 128VAC windings to make the simulation better. I can post it for people to try out. It's a delicate circuit, the HV load has maybe 10% effect on current through R316. It seems tuned 0.01uF/2.27mH for 33.4kHz and you are seeing it operate at I think 24kHz. I'm not sure what these are tuned for.