Testing a Flyback Transformer (Leader LBO-302 Oscilloscope)

[floobydust]

I use 2CL70 6kV 5mA Vf=23V, fast recovery 100nsec. HV rectifier diode. You know the forward voltage is really high due to the diode stack, so they fail a multimeter diode-test.
For an autopsy, you can disassemble the 3HT-5 diode stack by pulling the end caps off.

The potting compound is also used to stop corona and ozone buildup, which degrades the local air's dielectric. You can buy small tube of MG Chemicals silicone potting compound.

Stay with carbon comp 22k resistors because of the surge currents and voltages. For the amount of hassle, I would replace all the parts. Even disc capacitors age and go low value.

[bsfeechannel]

I use 2CL70 6kV 5mA Vf=23V, fast recovery 100nsec. HV rectifier diode. You know the forward voltage is really high due to the diode stack, so they fail a multimeter diode-test.
For an autopsy, you can disassemble the 3HT-5 diode stack by pulling the end caps off.

Yeah, very important observation. How did the 3HT-5 fail? If they failed "open", chances are the there was not enough forward voltage from the ohmmeter to turn them on.

[Smoky]

Alright, here's a list of what parts I have or what's on order from Digi-Key:

...and FloobyDust, I couldn't find any diodes matching your spec's or that part number yet, so if you have a supplier near you, please send me their phone number or link, and I'll keep checking though. Thanks!

For the 3Kv 1000pf capacitors, C406, C407, C420, and C408 on the main board, I ordered Kemet's GoldMax 5% 125C X7R's (C330C102JHR5HA). They seem to be built pretty stout (I'll attach a picture of their construction, kind'a neat).

For the 3Kv 470pf capacitors, C410 and C420 on the main board, I ordered TDK's CK45 Series 10% 105C ceramics (CK45-B3FD471KYGNA). They're not potted and easy to get to.

For the 22K 1/2 watt resistors R427 and R428, I have a choice of several different ones on-hand:

Vishay 22K 5% 155C 100ppm Metal film 1/2 watt
Dale 22K 2% 175C 200ppm Metal film 1/2 watt
Stackpole 22K 5% 155C 500ppm Carbon film 1/2 watt
TE Connectivity 22K 1% 155C 100ppm Metal film but in a 1 watt size.

For the 470 Ohm 3 watt resistor R316, I ordered a Panasonic 5% 235C Metal oxide film (ERG-3SJ471).

And FloobyDust, the 47 Ohm 1/2 watt resistors R306 and R312, I have Vishay 47 Ohm 2% 150 C 100ppm Metal film or Stackpole 47 Ohm 5% 400ppm Carbon film. Probably doesn't need to be anything special.

I totally agree, I don't want to use any of these old parts. So the delay will be the diodes.

I measured the volume of the potting tray and it is 1.6 cubic inches (26.22ml), so I'll get searching for that Silicone potting stuff (MG Chemicals). It comes in a clear formula by some, so it would be cool to see the components floating in it

I did use my Fluke 8024B DMM to test those original diodes and the user's manual states it uses 3.5 volts or less for the diode test.

And just one other thing that I noticed, power transistor Q401 on the HV Oscillator board, it's supposed to be an NPN 2SC1059, but what's installed is a 2SC1168 X. Who knows, but I do have a new Motorola gold-pin JAN2N3739. Do these transistors happen to interchange?

Fun stuff!

...now the weeds need pulled, the lawn needs mowed, the garage...

Happy Fourth of July!

[floobydust]

Digikey and Mouser stock nothing for high voltage rectifiers. I haven't seen any from North American distributors so I've been buying them off eBay or Aliexpress, under $0.50 each. The Tesla coil fanboy shops also have them.
It looks like the white cartridge type HV rectifier diodes are selenium.

The hardest part with silicone potting is getting rid of air bubbles. But at 2kV this is not super critical. Just keeping part leads apart and the capacitor plates away.

[bsfeechannel]

I discovered the manufacturer of those diodes. It's Origin Company, Limited. They have a subsidiary in the US called Origin Electric America Co., Ltd.. They still produce high voltage rectifiers.

Studying their datasheets, I concluded that their part numbers follow the pattern Voltage(kV)-Speed-Current. So, if 3HT-5 follows their standard, 3 means 3kV, H means high speed, which for them is t_rr<= 500ns, and 5 may signify 5mA, or perhaps 500mA. I'm not sure about T. However, I guess they can be contacted about it if someone gets that curious.

I initially estimated the oscillator to run at around 70kHz, but now that I know that it is a tuned-collector oscillator, given that C403 is 10n and the primary coil is 1.8mH (as measured by the OP), we have a frequency f=1/(2πSQRT(10nF·1.8mH)) ≃ 37.5 kHz. So T= 1/(37.5 kHz) = 27 µs, way over 500ns.

If we go to Mouser, we'll find 3,865 pieces in stock for immediate shipping of a 3kv, 500ns rectifier (R3000F-T) for mere $0.18 a pop.

Of course we'd have to experiment with those diodes to see if they work fine for this application.

[strawberry]

6kV 5mA Vf=23V  rather selenium rectifier
something 20 Si diodes in series sounds silly

MOUSER : BAS170W

[floobydust]

For low current, fast recovery HV rectifier diodes, Asian manufacturers seem to have taken over. I would try get samples.
Microwave oven parts are higher current and capacitance, majority are slower switching, so I would not use them.

I have some extra 2CL77's 20kV 5mA and Vf measures around 34V. PM me and I can send some.
You want no sharp edges at soldered joints and the entire side of a disc capacitor is at one potential, so it would not be against the opposite potential. You have to plan the sandwich for lowest corona.

edi electronic devices inc. NY.
Dean Technology Dallas, TX.
All Electronics surplus GP260 6kV 200mA (R6000) but maybe not fast recovery. Tesla Coil forums found them around 100-120nsec.

[bsfeechannel]

For low current, fast recovery HV rectifier diodes, Asian manufacturers seem to have taken over.

One possible explanation is that after the fall of RCA, the Asians took over the production of TVs and monitors. Those diodes were primarily used around CRT circuits. In fact I suspect that the T in 3HT-5 means TV, because you find that kind of designation with other manufacturers as well.

[Smoky]

Let me show you this.

Here are two pictures of the CRT multiplier board inside of my Sencore SC61 Waveform Analyzer.

This board takes -1500 volts and ramps it up to -4500 volts using EDI DL800 diodes and 6Kv capacitors.

The Sencore also uses several of these diodes to generate the three -1500 volt legs coming out of the HV supply board too.

No potting, just a small metal cover that I unscrew from the chassis.

Someone on Ebay has these DL800 diodes for sale.

What do you think?

...and thank you so much for the help!

[floobydust]

Take a listen to the Sencore CRT multiplier or look at in the dark and you will see why potting is preferred. You don't have to have it but it's insurance. Dirt dust and humidity can be a problem without a cover.

From Paschen's curves, air breakdown is 3kV/mm at sea level and you design HV air spacings for around 10X greater. Some HV diodes are stubby, their lead spacing is designed for them to be potted. The EDI datasheet is saying if the devices operate over 10kV/inch (400V/mm) it needs to be encapsulated.

The DL800 diodes look good to use, to 4kV in air and 8kV potted. I think the HV capacitors have much less spacing though.

[Smoky]

Here's what I did, I ordered 5 2CL70's (6Kv) and 5 DL800's (8Kv)

I was thinking too how close the components are on this little thing, coils, caps, diodes, etc.

One thing's for sure, these diodes will be smaller, so more room to work with.

BSFeeChannel, I followed up on your remarks about selenium rectifiers and I read how they were prone to failure and that they also have a "service life."

I didn't think much about the lead spacing of the capacitors when I bought them, but I did read a warning that they had on the 1000pf caps about market volatility. Can you believe this, the cost of those capacitors almost doubled since I bought them Tuesday!!!

I also carefully cleaned and stripped the two fine strands of wire coming into the tray from the coil. They look good.

I noticed the way the components were installed. Their leads were turned into loops and dropped over the wire and soldered.

[Smoky]

First design of rectifier circuit using 2CL70 diodes:

Btw, all of the 1000pf capacitors measure 1020-1050pf on the Sencore.

[Smoky]

Second design using DL800 diodes:

[Smoky]

I made custom potting trays out of my business cards

I intend to make epoxy "modules" with just the three outer leads exposed. That way, if a rectifier fails, I can quickly un-solder and change it out.

[Smoky]

I purchased MG Chemicals 832HD Black Epoxy potting compound ($10) since I own a small Loctite 50ml dispensing gun.

It should arrive on Thursday.

I intend to solder-up and pour at least three "modules" to make use of the potting material since it will go bad.

[Smoky]

I replaced capacitors C410 and C418 (470pf 3kv's) and C408 (1000pf 3kv) on the main board.

A matched 47uf 250v capacitor bundle replaces the "three-cap-can" (C301,C302,C303) in the power supply and a new 470 Ohm 3 watt resistor replaced the burnt C316.

...and I cleaned all of the switches in the scope too!



And I failed to mention that the old "voltage-doubler" capacitor (C406) measured 1440pf on the Sencore.

[bsfeechannel]

And I failed to mention that the old "voltage-doubler" capacitor (C406) measured 1440pf on the Sencore.

This may indicate that the capacitor is leaky. The higher value means that the capacitance meter took longer to charge the capacitor, due to the leakage resistance.

[Smoky]

Check this out

Someone on YouTube demonstrated how to test a high voltage diode.

I first set my power supply to 25 volts DC at 200ma:



Next, attach a 270 Ohm resistor to the anode side of the diode under test and connect it to the (+) side of the power supply.

Then, ground the cathode side of the diode.

Use your DMM to read across just the diode.

Here are the results of the original "Origin" diode tested both ways:





Here is the result of a new 2CL70 diode:



And here is a DL800 diode:



I read power supply voltage both ways on the original Origin diodes! They're "open."

Now notice the difference in the voltage drops between the new 2CL70 and the DL800 diodes.

This is awesome!

[Smoky]

Last but not least...

...the little epoxy "mudcake" modules poured fresh tonight!

We hit the switch tomorrow night



I'll be trying the DL800 diode design first.

That diode has up to six times the "average and surge" current capacity than the 2CL70 and "maybe" sacrificing 50 nanoseconds of switch time (150 vs 100).

The DL800 lead diameter matched the resistor's lead diameter making the assembly tougher. The 2CL70's leads are thinner.

The new 22k Stackpole carbon film resistor measured 21,800 (under 1%).

[floobydust]

 

[bsfeechannel]

 

[Smoky]

Just home from work and I'm ready to put this thing back together!

Now I did put a small blob of type II silicone over the solder joints, so start up may be an hour or so from now.

[Smoky]

Unfortunately, resistor R316 still gets super-hot as I ramp up the isolation transformer.

What's nice though, I have my Sencore SC61 Waveform Analyzer probe hooked on test point TP402 and it kept an eye on what should be -1500.

It reached -914vdc or so before I felt high heat on the resistor and rolled her down.

No visible lighting in the tube.

So there is still a malfunction somewhere.

Maybe it's in the CRT itself?

I may try a couple power transistor swaps next since I have them.

Thank goodness I don't do this for a living. I'd be feeling really hungry

[bsfeechannel]

But if you disconnect the CRT and try to bring the scope up, what voltage do you get at TP402 before R316 starts "chernobyling"?

[Smoky]

BSFeeChannel,

I disconnected the CRT and read TP402. I'm getting just 2 volts now.

I reconnected the CRT and it reads only 2 volts.

Whatever the problem is, I think it's taking out the rectifier circuit with ease.