Hallicrafter 509 image shifted horizontally
 

Just finished recap of a Hallicrafter 509 10". I have good vert and horizontal width and locking. I have good vertical height and locking. However as can be seen in the photo attached, the image is shifted by 50% horizontally. I assume the vertical bar up the center of the screen is the horizontal blanking interval that would normally be at the right side of the crt and out of sight off the face of the crt. Because I have full height and width I believe the horizontal and vertical of the deflection yoke is working so I do not suspect the yoke itself.

In the attached photo (although it is hard to see the detail of the image) the half of the image on the right side of the vertical bar is actually the left half of the full video image. (IE: the image is shifted to the left by 1/2 of the screen)

I am not real good at tv theory so I am hoping someone with a better technical background can point me in the right direction.

Thanks for your help!

I haven't looked at the schematic. I'm guessing the horizontal frequency is off by some multiple of what's correct. Any chance you could have installed a cap value that is off by one decimal place? Otherwise is there a course horizontal frequency adjustment that could be off?

Yes. I did adjust the horizontal frequency to get the H-hold to lock. H-hold locks very well and I do not have multiple images.

There is just one image. The image is shifted to the left on the crt by 1/2 of the screen. That is to say; the vert bar in the middle of the screen separates the image on the crt into a left half and a right half. The image on the left half of the screen is the right half of the full image and the right half of the screen is the left half of the image

If I remove the rf antenna input (or go off channel) I get a nice snowy raster. without the vertical bar up the middle. I can see full scan lines from left to right.

Likewise if I pull the 6al5 video detector between the output of the IF section and the input of the Video amp section, I get a nice clean raster. So it appears that what is being fed into the 1st video amp is a corrupted image (maybe)?

With all of that said I am leaning to an issue in the IF section.

I am going to try and scan the original Halicrafters schematic and post it in my next post.

And I will go back and check my cap values as you have suggested.

I already checked all the electrolytics to make sure I didn't switch the polarities, and that all checked out OK.

This is almost certainly not a problem in the IF, as the H sync and blanking looks normal in the picture.

I think the problem would be somewhere in the H sync waveform or H sweep waveform being fed to the H phase detector.

Can you post a link to a schematic for that section?

cant upload schematic
 

VK wont let me upload the schematic. VK says file too big. I downsized the file but it loses too much detail and can not be read.

What do I do?

Tried to post it on VK but file is too big. I looked for it on the web but my set differs from the Sams which is posted at the ETF site and the OEM schematic is of poor quality at the ETF site.

I have an old email address for you. I will try to send it to that address.

Thanks Wayne :D
Bob Galanter

Assuming the scan frequency is correct, could there be a DC bias current running thru the yoke winding? Just a SWAG. :headscrat

A DC current would shift the raster sideways, not wrap it around.

Bob and I have started an email exchange. I think the problem is somewhere around the sync discriminator or its inputs.

I agree. The H osc appears to be locking to the sync, but locking to it out of phase. The horizontal phase detector/horizontal AFC circuit is probably the issue though if the sync or sweep signals feeding it are somehow phase shifted that could also do it.

Right you are. That was a senior moment/brain fart on my part. Dumb, dumb.
:lmao::deal:

-------------------------------------------------------------------------------
This is what Wayne wrote to me................

Thanks. First let me say I never have trouble-shot one of these, so I can only make suggestions based on general principles.

Let me know: Do you want to proceed by email or on Video Karma? If VK, please copy the following there. It might be the best way to get someone with particular experience to help.

Have you tried subbing v11 and V12? Can you poke around this area with a scope?

The left half of V11 separates the sync. I’d guess it’s OK because you have vertical sync.

The right half of V11 is the phase inverter. One polarity is on the anode and feed the vertical circuit and the upper half of the sync discriminator.

The opposite polarity should be on the cathode of the inverter (R74) and feeds the lower half of the sync discriminator. A scope should show equal and opposite polarity waveforms at left sides of C58 and C59.

There should be a horizontal waveform of some kind coming through C61 and shaped by C60 and R79. I don’t know if it should be a pulse or a sawtooth wave or what on C60, but I’m thinking sawtooth more or less. If I understand the circuit correctly, the pulses from the sync phase inverter should be occurring in the center of the sawtooth, creating a neutral control voltage out of the sync discriminator and making the horizontal sweep line up with sync.

If V11 and V12 are good, I would check components in this area. Anything connected to right half of V11; anything connected anywhere on V12.

(FYI Vert sync is fed to the vertical oscillator through R75 and the following R/C integrator components to filter out the H sync and transmit a vertical pulse.)

Regards,

Wayne

A little bit of background
 

This set is quite problematic for a couple of reasons:

1. The set has been worked on and I am not certain what is original and what is factory. I have found some replacement components that are incorrect values

2. I have both the Sams and the OEM Halicrafters schematics. The Sams reflects later wiring and component values, and the Hali is earlier which more closely resembles my chassis. However neither schematic totally reflects what my chassis is.

3. The set has a separate power supply with HV flyback and B+ supplies. There are 4 multi pin connector cables going back and forth and trying to follow what all the wires do is quite a challenge not to mention trying to relate all these wires and connector pins on the schematic.

With that in mind I spent about 10 hours comparing the main chassis against both schematics. I decided that I would use mostly Halicrafter schematic values. However even then, there are a few changes that match only the Sams and are obviously production changes that are not reflected on the Hali schematic. So it's a mishmash but mostly like the Hali schematic.

I did find a few resistors that were out of range and at least one 1meg that was open. But when I powered the set back up, I now only have a raster without any video information.:tears:

I have not been able to work on the set powered up due to the cabling issues between the 2 chassis. My next step will be to devise some method of working on the chassis live. At that point I will attempt to see if I have lost B+ somewhere that may have caused the video to disappear.

I will post again after working on the chassis live so I can take some measurements. I suspect working on it live will be more fruitful.

Thanks for all your help.

I looked at the schematic in Riders and expect it's the same as your factory. Here's a few pages from Kiver explaining this basic horz AFC circuit along with an easier to follow schematic in figure 16. Requires a pulse from the horizontal output transformer.

https://live.staticflickr.com/65535/...cd81aede_b.jpg

https://live.staticflickr.com/65535/...808eefde_b.jpg

https://live.staticflickr.com/65535/...789a1319_b.jpg

Thanks! Shows sawtooth derived from sweep pulse and +/- pulses from sync as I suspected. Nice to know I wasn't hallucinating about how it works. :D

A better, or more recognizable video display might be helpful as well. What you have there is so washed out, hard to tell what's going on. If it were just a phase issue, then I would expect you would have a decent image on either side. Try injecting video directly.

Well, I was able to devise a method to get the chassis powered up and in a position where I can work on the underside. I traced the video path with my scope. It disappeared at the 1st video amp 6v6. During my work I inadvertently lifted the grounded side of the filament ckt at the 6v6. (this is the kind of dumb stuff you do while working at 4AM when you should be sleeping) Video is now back with the video image shifted by 1/2 screen and the vertical bar in the center is still there.

Next step is to use my scope with the info that Kevin Kuehn posted and see what my scope discovers. Hopefully this will lead to discovering the problem.

Just out of curiosity does the horizontal hold work normally, as in can you move the picture slightly left or right before losing horizontal lock? You should have the horizontal hold control centered when you make the horizontal oscillator coil adjustment. Ordinarily you wouldn't need to move the slug far to get lock unless someone previously messed it up. While you have your scope in there verify that the saw tooth frequency at the grid of the horizontal output tube is correct.

Yes the H hold works normally. I have the Horizontal hold control optimized at the center of it's range, and when you adjust the control the stable image (and the vertical band) will shift slightly left and right depending on which way you turn the control. IE: it seems to be working perfectly normal.

Next step is to start looking at the wave forms in the article you posted.

BTW, what should the frequency be, at the grid of the 6BG6? 15.734 Khz?

I may not get to that today. Got some other things I need to do.

Thanks for your help,:thmbsp:
Bob G.

15,750 Khz.

if the image is a single image (it is) and is not skewed diagonally (it's not) [as shown], the frequency is correct. No need to try measuring it. Only the phase is wrong.

I agree with old TV nut.

Some counters are tricked by the non-sinusoidal horizontal waveforms so I typically adjust horizontal frequency by first measuring scope divisions between leading edges of horizontal sync pluses in a digital composite video source (DVD player, DTV converter, etc) then adjusting the vintage TV H osc for the same number of scope horizontal divisions between peaks (and usually the peak slightly higher than the hump on synchroguide based circuits) works well.... Probably not needed here, but useful if your frequency counter gets confused on complex waveshapes.

Is it not possible to have the horizontal oscillator adjusted wrong and still be locked, but out of phase?

There are two things happening.
1) The oscillator is locked. The phase locked loop is basically functioning. The free running horizontal frequency may be off center, but it is within the pull-in range of the loop.
2) It is locked at the wrong phase and is stable at the wrong phase. This should not happen whether the free-running horizontal frequency is correct or not.

Note that when the loop is locked, adjusting the free-running frequency +/- will rock the phase back and forth a bit to generate the correction voltage that keeps the locked frequency correct. This effect can be seen when the circuit locks at the correct phase as normal or the incorrect phase as shown here.

Something is making the control voltage out of the phase detector correct when the phase is 180 degrees from normal.

.

Hi!

If you have a nice locked picture, then what happened is that you made a mistake
with replacing parts and the sync is getting to the H Osc. out of phase, or delayed
because there are 2 parts in the wrong spot and an RC is charging & discharging
in a way to delay the sync and it's showing up in the middle of the screen.
It's locked to the sync pulse as you can see it's pretty perfectly straight.
I think the best way to find this is to Check the wiring, Look for wrong
parts in wrong spots, etc.

.

Component wise there's not much in the path to mess things up. I suppose if the two 1500pf mica?(domino looking) caps that connect the +/- sync inverter pulses to the sync discriminator have not been changed I'd go for those. If one or both are leaky it could potentially offset the DC control bias going to the horizontal oscillator grid. Also check the two 100k discriminator resistors for balance.

Found the problem
 

Thanks to all who made suggestions:banana::banana::banana:

I took Kevin Kuehn's suggestion and measured the grid of the Horiz output tube 6BG6. It had a nice saw-tooth waveform, but when I calculated the pulse width on my scope it appeared that it was running about 1/2 speed at about 7800HZ.

From there I figured that something had to be amiss in the horizontal oscillator section. So I started testing values. I found a bad mica cap across the H freq coil that was shorted, and I replaced it. That did not fix the problem so I kept testing. Then I found a 270K (R87 on the hallicrafter schematic) that had drifted to 780K. I replaced it and the oscilator speed just about doubled. After tweaking the H-Oscilator freq coil, the picture came into sync and I had a perfect full picture.

Again thanks to all that contributed. So many knowledgeable techs that know a lot more than I do.

Bob Galanter