Reducing HV output voltage from Flyback/LOPT as used in arcade monitors

Ugh! If they are getting other voltages off TAPS on the flyback, then it is going to be real hard to ONLY turn down the UHV without affecting the other voltages. This really requires different winding ratios. Now, if the other voltages are just taken off the point where the transistor drives the flyback, you could fix this by just adding a few turns to the primary winding.

Jon

That would also lower all the other voltages.

Since this is very low current, maybe some proper HV resistors to divide the HVDC down to the proper level.

You can't do anything to the FB as it is most likely potted in epoxy.

Would that it was so easy! This is a sealed flyback in a plastic housing. Late 70s design.

I have gotten a bit further - using a 75R to 100R in series with the Flyback B+ in I can get the HV down to about 20KV, and the +12VDC supply is fine with that. Now the problem is the image is shrunk on the left, but that turns out to be blanking after all. It has to do with transistor X304 and its base feed. Need to muck around with the resistors and caps a bit I expect to try and reduce the horizontal clipping. With X304 disconnected I do get a good, although narrower, image. I will have to expand the width, hope adjusting the caps C513,513 or C515 will give me enough width.

John :-#)#

--
(Please post followups or tech inquiries to the USENET newsgroup) 
                      John's Jukes Ltd. 
MOVED to #7 - 3979 Marine Way, Burnaby, BC, Canada V5J 5E3 
          (604)872-5757 (Pinballs, Jukes, Video Games) 
                      www.flippers.com 
        "Old pinballers never die, they just flip out."

ematic.pdf

I wouldn't worry so much about the HV as I would about the filament. If th e filament voltage stays at or below 6.3, I wouldn't give it another though t. Most of those tubes have no trouble with a few extra KV. If the filame nt is high, drop that or the tube will have a life span emissions-wise in m onths.

Another way of lowering the global output of the fly is add some capacitanc e to the retrace cap. This will of course affect all the secondaries.

I'm not crazy about adding resistance to drop voltage on any load that's dy namic. If you scope the collector of the horiz output with and without a d ropping resistor, you'll see the waveform will be different. You may have short horiz output transistor life. If the monitor uses some sort of serie s/pass regulator (most 70s regulators are) you can drop the voltage by play ing with the regulator and keeping the supply tightly regulated to the fly.

f

o

chematic.pdf

the filament voltage stays at or below 6.3, I wouldn't give it another thou ght. Most of those tubes have no trouble with a few extra KV. If the fila ment is high, drop that or the tube will have a life span emissions-wise in months.

I used to boost tube heaters, and had a Sony with 66% heater voltage boost. It was an experiment on a totally dead tube. Years later it still worked j ust as well - emission ok, but tracking lousy.

nce to the retrace cap. This will of course affect all the secondaries.

dynamic. If you scope the collector of the horiz output with and without a dropping resistor, you'll see the waveform will be different. You may hav e short horiz output transistor life. If the monitor uses some sort of ser ies/pass regulator (most 70s regulators are) you can drop the voltage by pl aying with the regulator and keeping the supply tightly regulated to the fl y.

I don't know whether you can add a new winding that goes in series with the EHT output, bucking its V_out. Wouldn't be trivial given its running volta ge.

Series resistors may not work well, a string of MOVs might be better.

Have you checked the mfr specs of the tube? I doubt it's specced to 30kV bu t it's worth knowing exactly what you need to get it down to.

NT

I would disagree. These tubes were rated with a maximum HV at 22.5KV (safety circuit would cut in), with the wrong flyback they are running at close to 30KV, and I don't think that is a good idea from the X-Ray risk perspective. A family friend of ours was an X-Ray specialist and died of cancer relatively young...

That may work, will try it. Thanks!

Dropping the system B+ reduces the vertical deflection too much to be useful.

John :-#)#

--
(Please post followups or tech inquiries to the USENET newsgroup) 
                      John's Jukes Ltd. 
MOVED to #7 - 3979 Marine Way, Burnaby, BC, Canada V5J 5E3 
          (604)872-5757 (Pinballs, Jukes, Video Games) 
                      www.flippers.com 
        "Old pinballers never die, they just flip out."

of

to

h e e

schematic.pdf

f the filament voltage stays at or below 6.3, I wouldn't give it another th ought. Most of those tubes have no trouble with a few extra KV. If the fi lament is high, drop that or the tube will have a life span emissions-wise in months.

tance to the retrace cap. This will of course affect all the secondaries.

s dynamic. If you scope the collector of the horiz output with and without a dropping resistor, you'll see the waveform will be different. You may h ave short horiz output transistor life. If the monitor uses some sort of s eries/pass regulator (most 70s regulators are) you can drop the voltage by playing with the regulator and keeping the supply tightly regulated to the fly.

Is the vert B+ sourced from the LV regulator?

If you have to use series resistance to lower B+ to the primary of the flyb ack, I might try adding capacitance on the far side of the resistor to stif fen it.

Schematics link is above. The vertical deflection is driven from the B+ (120VDC) directly.

That is probably a good idea, I will check that with my 'scope.

Need to adjust the Horizontal Blanking pulses to transistor X304 as the picture is being clipped way too much on the left side. I'll post some photos later.

Currently working on another project which is related to WPC pinball games and their real-time clocks when using NVRAM. Then back to the GO7!

Thanks,

John :-#)#

--
(Please post followups or tech inquiries to the USENET newsgroup) 
                      John's Jukes Ltd. 
MOVED to #7 - 3979 Marine Way, Burnaby, BC, Canada V5J 5E3 
          (604)872-5757 (Pinballs, Jukes, Video Games) 
                      www.flippers.com 
        "Old pinballers never die, they just flip out."

Hi

According to the numbers you provided, the EHT needs to be reduced by about 7 kV (while keeping other voltages derived from the transformer at their usual levels).

The 7 kV difference is significant, but not insurmountable for a series dropper approach. The current draw of a CRT anode is small, the power dissipation should still be manageable: 7 W at 1 mA (if the CRT ever takes that much current).

There are TVS diodes with voltages up to 600 V (nominal value at 1 mA test current) available (Littelfuse P6KE600A). A series string of 12 such TVS diodes will have around 7 kV of voltage drop across them.

The continuous power handling capability should be enough (each TVS of this size can dissipate over 1 W, with heatsinking even multiple W).

The TVS diodes are reasonably cheap (some 40 cents a piece at Mouser for QTY 10), so the main difficulty would probably be the insulation (potting) of the series string. Probably best to put the whole thing into a long insulating tube and fill the tube with epoxy.

Regards Dimitrij

Interesting idea Dimitri, however I don't see it working out - 12 diodes in series would make the anode wire ungainly at best. Then insulating, etc. where I think the average tech probably won't have HV rated epoxy or silicon glue. Appreciate the idea, and may use it elsewhere!

Another poster had suggested changing the capacitance at the Horizontal Output transistor X01, and indeed that did work! I now have 19KV on the HV output, screen looks normal and all is well.

Obviously I need to brush up on my Horizontal Output theory so I can understand why this worked, but for anyone reading this I added 2 X

2200pf @ 1500V and one 1500pf @ 1500V caps. Took them off a dead GO7-CBO

- 19" chassis.

Thanks!

John :-#)#

--
(Please post followups or tech inquiries to the USENET newsgroup) 
                      John's Jukes Ltd. 
MOVED to #7 - 3979 Marine Way, Burnaby, BC, Canada V5J 5E3 
          (604)872-5757 (Pinballs, Jukes, Video Games) 
                      www.flippers.com 
        "Old pinballers never die, they just flip out."

You (John-Del) suggested changing the capacitance at the Horizontal Output transistor X01, and indeed that did work! I now have 19KV on the HV output, screen looks normal and all is well. Voltages are all nominal, but I will compare with a proper 13" chassis that is still using the original flyback/LOPT.

Obviously I need to brush up on my Horizontal Output theory so I can understand why this worked, but for anyone reading this I added 2 X

2200pf @ 1500V and one 1500pf @ 1500V caps. Took them off a dead GO7-CBO

- 19" chassis.

Thanks!

John :-#)#

--
(Please post followups or tech inquiries to the USENET newsgroup) 
                      John's Jukes Ltd. 
MOVED to #7 - 3979 Marine Way, Burnaby, BC, Canada V5J 5E3 
          (604)872-5757 (Pinballs, Jukes, Video Games) 
                      www.flippers.com 
        "Old pinballers never die, they just flip out."

Nice to find someone willing put put effort into shit like this. I commend it. And to use pretty high tech skills to do it, like about engineers level , modifications and KNOWING what works right and doesn't. Have any idea how hard it is to find people who CAN do that, let alone willing ? Keep it ali ve bro... (het, sounds like a good name for a company)

I'll put my suggestions first and commentary later to be sufferable...

The components around R 516, C 519 etc. have nothing to do with the HV leve l. They are for the ABL, they measure the actual HV current and cut the vid eo gain down if it exceeds a certain level. There is not supposed to be vol tage dropped there like in a CRO.

Notice the difference in the print around C 512 - 514. The 19" version has

2 extra caps there. They downed C 514 down to 2,ooo pF from 2,500. Then the y added C 531 & 532. which slows down the retrace, makes all retrace rectif ied sources lower and trace rectified sources a little bit higher. You can tell which it is by scoping the input the the rectifier, if you see a large negative trace on a positive source it is trace rectified, if the pulse is positive on a positive source it is retrace rectified.

Question : this potted fly, CAN YOU GET TO THE CORE ? You do not have to ge t to the windings, if you can get to the core you can add bucking or boosti ng windings to any winding there. All that area that is encapsulated, can y ou see the ferrite thing around it ? If so, can you see a gap ? All we need is about 0.03 inches for good enough. That is big enough for # 24 AWG to g et through and you can get it with like couple hundred volt insulation.

Increasing the capacitance alone might not do it, but a combination of chan ging the B+ and that might.

Comment :

This thing was built before internal pincusion correction in the CRTs. Ther e is a saturable reactor and another one that is permanent magnet biased. t here are also VDRs. This means that This indicated that voltage and current levels muct be at least close for good geometry. We are not talking moon s hot accuracy here but we must remain within the "linear non-linearitiy" ope rating range of those devices.

It also seems that the B+ is lower in one than the other.; One says (somewh ere) 120 regulated and the other says nothing, but both shutdown use the sa me resistors and Zeners, and that indicates that the B+ level is close if n ot the same.

If, after adding the capacitance I suggested you find that it is not enough then the next thing is to reduce the B+. but don't do it with a passive de vice. The best method would be to add a Zener diode of the voltage desired with the anode to ground and the cathode to to the junction of C 905 & R 9

1. You'll probably want to choose a standard value anyway so choose a lowe r one and put a pot in series. It might decrease the tightness of the regul ation but bucking that we had the gain of the IC itself. Look at most of th em, they try to operate in their linear range. That works to our advantage.

Keep this baby on the road !

True, but MOVs dissipate power.

Have you checked the mfr specs of the tube? I doubt it's speced to 30kV but it's worth knowing exactly what you need to get it down to. "

No 13" tube is speced that high, even in Zimbabwe.

Given the thickness of the leaded glass on those CRTs, I doubt X-rays are a concern. But you could always have the gamers wear tin-foil hats.

It is not just the face of the tube that emits X-Rays, the glass on the bell housing can (I am pretty sure) leak as well. A common use for this

13" monitor was in cocktail tables. So the face of the tube is shown to the players and the bell housing, well let us just say it is pretty much inline with your waist on down...

Not wanting customer gonads to glow in the dark was my main impetus for finding a simple, yet reliable solution that did not defeat the internal over-voltage X-ray shutdown process.

John ;-#)#

--
(Please post followups or tech inquiries to the USENET newsgroup) 
                      John's Jukes Ltd. 
MOVED to #7 - 3979 Marine Way, Burnaby, BC, Canada V5J 5E3 
          (604)872-5757 (Pinballs, Jukes, Video Games) 
                      www.flippers.com 
        "Old pinballers never die, they just flip out."

a concern.

I would tend to agree, but radiation is not the only concern. There is the matter of spacing of the internal components as well as the thickness of th e glass on the bell part. This does determine a limit as well and going muc h beyond it can cause a catastrophic failure likely to wreck the envelope s o it cannot be rebuilt.

It also depends on the customer. If they are corporate and even give a thou ght to ISO compliance they won't like it. Even this modification if it brin gs it completely into specs is something ISO doesn't like. If they are info rmed of the excessive HV and go ahead and allow its use they become culpabl e for whatever happens due to that. If they are not notified then the servi cer might just find out many new things about the legal system. I know it is not the US where we got more lawsuits than people, but they still have t ort laws and there is still such a thing as ISO. Cheating the ISO can even result in criminal charges in some cases, though rare.

Tin is expensive. The shit you buy today is aluminum foil. While aluminum c onducts electricity better than tin (Al 2.65*10^[10-8] v Sn 1.15*10^[10-7]) it has nothing to do with its admittance to "those" rays, or Xrays. Lead ( Pb) would be better. The aluminum can create a reflection zone where the mi nd scrambling rays' effects are amplified. Lead however, fatigues too easil y which makes it harder to fashion acceptable headwear using it. So it is n o mystery why tin is so much more expensive. The damn conspirators are limi ting production and buying up the supplies for their electronic contrivance s to keep us in line, depriving us of the tin we need for those body suits. What else could it be ?

{for those humorically challenged - that last paragraph was a joke}

Simple, lead athletic cups. Like the ones you wear fro sports...

OK, joking aside, to find out if my plan will work you need some actual cap acitors to pad that LOPT collector. Then along with the HV output, watch th e duty cycle of the pulse on the collector, it should be somewhere around 2

0 %.

I just looked up the PAL line frequency and it is not lower than NTSC. We e xpect 12.5 uS retrace pulse so that is probably close to what you need.

HOWEVER, since it is a monitor and has no need for a back porch for a color sub burst, it may be shorter. The design seems like the engineers were con scientious, and shorter retrace is usually more energy efficient. Since the se units were made to run continuously they may have made it their concern.

It is not easy to tell if the retrace is as intended. One way is to use the horizontal phase, as you said it was off. If you return the control to its original position and padding brings it back within spec then you are gett ing there. Another way is to temporarily reduce the B+ and turn up the G2 t o reveal overscan hidden by blanking, and that is a bit subjective.

Modification is fun ain't it ?

And my question stands, the core. Can you see it and if so can you slip a p iece of say, card stock between it and the molded part ?

< catastrophic failure likely to wreck the envelope so it cannot be rebuilt . >

Rebuilt? Seriously? I think the last CRT rebuilder went out of business abo ut a decade ago. Hawkeye. The only rebuilding operation left in the western world is in the Vintage TV museum, and that's only for show.

I think many arcade machines now are repaired by subbing in an LCD panel fr om a computer monitor. I've seen several, and they work great. Sometimes a small circuit board is required to invert polarities, etc., but there are a handful of guys selling those on line.

John, have you considered the LCD sub?

Terry