Adding Composite Input (Capacitor Question)

[ARC Tech-109]

Friday the 5th and the tuner is once again alive with a new RF amp and osc/mixer. My theory is they were somehow damaged internally when the IF connector brushed against the 120VAC terminal of the fuse holder and the replacement confirms this. Everything is now on the mark up to the video amp and now I have to sort out the chroma problems.

To back up some 40 years the set took on some lightning damage when a tree was hit and this killed both the phase detect and killer diodes, I replaced them with modern silicon small signal diodes but it was never quite the same and this was the real reason the set was replaced with a Zenith S-III and I ended up with the albatross in my bedroom. Now I want to get the set running like it used to and this has been quite the uphill battle. I found a set of what I thought were good NOS 1N270's and 1N34A's but they're all leaky passing tens of mv in reverse with no two quite the same. Plan for the time being is to "soft wire" in a 6JU8 and see how that goes, not the most elegant but they worked before in this service.

I will post my findings here so as it's been said in the past, Stay Tuned.

[kf4rca]

Try 1N914 or 1N4148 diodes. They're good to about 100 MHz. and are cheap. Radio Shack used to sell them for $1.99 for a hundred diodes.

[ARC Tech-109]

Had the 914's in there and the drop is too high for the magnavox circuit so I wired up a 6JU8 in their places and ran thru the entire IF and chroma alignment and everything came together. Yes we have success. A few loose ends to tie up, one seems to be an issue with horiz blanking that's causing the right side of the screen to essentially go into cutoff. Bias and guns are set correctly and the cutoff remains when I run the horiz gun adjustments. Other issue is a touchy AGC and this seems to be a bad AGC pot on the rear apron, have not put an analog meter and run the control thru its range as of this evening.

Ended up replacing all of the tubes on the IF strip with new-old-stock. I sourced the tubes from AxMan Surplus in St. Paul MN which is this bazaar of electronics and other things that's been in business for almost 60 years and they have everything... and I mean everything including thousands of new-old-stock tubes in their original cartons so for $2.50/ea so I invested a good hour scanning thru them and got the everything I needed.

IF strip took a good three hours to get the response back into shape, I'm working with a Heathkit IG-57 sweep/marker generator that has seen better days. The IF sweep works correctly but for some reason the markers themselves are not so I had to basically eyeball the response curve and work from there, I used my Motorola R2670 service monitor to sweep the response of the chroma section as this has calibrated sweep/track and spectrum analyzer functions.

Once I clean things up a little I will snap a few realtime pics of the set in operation, source will be from a Sony VO-2860A editor and whatever tape I happen to grab.

[kf4rca]

Great. I'm glad things are coming together for you. I would think a silicon diode would have better performance specs than a germanium diode or... a tube.
AxMan sounds a little like WierdStuff out in San Jose, Ca.
I've got a Singer FM10CS service monitor but I use it mostly for radio work. Goes to 512MHz.
Looking forward to the pics of this Franken-set!

[ARC Tech-109]

I'm sure the 1N914 is better overall but the problem seems to be riding the crest of the waveform with the small signals and its higher forward voltage. Once I went to the 6JU8 the chroma response was far better. This set never had what I'd call "stellar" color, it was good but maybe I'm spoiled by the sheer number of broadcast monitors. Dunno, it uses the 6HS8 as an X-Y demod and from what I remember this design was an "acceptable" lower cost design albeit not as accurate as the R-Y B-Y or I/Q with what appears to be a slower phase response. Good enough for the 1970's I guess when everything was burnt orange with splashes of primary colors here and there.

So everything was working well last night when I shut things down around 0400 central, just before this I dribbled a little hot wax down the IF coils to lock the slugs in place then checked the response curve and picture once again and everything was on the mark.
This afternoon I resumed my work as the outside temps went to the low 90s with mid-60 dewpoints, fired up the set and noticed a really bad picture, checked the IF response and found it had a very sharp peak going just above the picture carrier that wasn't there last night after I dropped the wax.
Ran thru the IF again and locked the slugs in place and everything is holding at the moment, I'm wondering if this might be a moisture issue as the humidity went from roughly 55% to almost 80% over the past day. Not much for air conditioning unless it's oppressive like this afternoon.

Did verify the AGC pot with an analog Triplett 630-NA meter and found it has several "scratchy" points along the travel so it's another trip on the motorcycle to AxMan Surplus in St. Paul as they have bins of these. Minneapolis/St. Paul area used to be a big tech hub during the early computer days with the likes of Sperry-UNIVAC, Honeywell, Control Data, ADC Telecom and IBM in Rochester and had two dozen surplus stores full of this stuff. By the mid-80's only a few remained and today AxMan is the last one standing after almost 60 years in business, they have enough inventory in that building to go another 60 years and it's always busy.

I got the Motorola R2670 monitor during my days at the regional service depot, employees could buy their test equipment at cost and pay for it out of salary. I optioned this one with the everything I could and it's served me well. Other bench test equipment includes a Tek 475 scope, Motorola R1011B 40 amp power supply, along with the various Bird RF meters. TV service is done with a late 60's Tek 453 scope originally from IBM, it has nuvisters on the front end and can survive a kilovolt hit to the input jacks. I also have the Heathkit IG-57A IF sweep/marker and B&K 1246 dot/bar generator plus 466 and 467 pix tube analyzers. Yes I'm on the trailing edge of 1970s technology.

[ARC Tech-109]

Okay so an update on the Bench Queen; it's running. I still have to iron out a few ripples in the IF section but it's a happy set after 20 years of idle time. Going to run thru the Heathkit manual for the IG-57A sweep/marker generator again and see if I can't find why the markers themselves are dead.

The big issue came down to a bum resistor on the horiz keying pulse, schematic called for a 270K and by color codes that's what I replaced the original with, turned out to be 27K on the meter. Red-Violet-Orange and 5% gold stripe so I went thru and verified ALL the replacements. Guess that's why Ohmite calls them "Little Devils". So once the correct value was in place everything started to act normally, AGC is good but still getting the horizontal bends and trails on high contrast scene details, vertical size is a little challenged despite a new tube and known correct resistor values. I suspect the Boost B+ is low.
Convergence was a challenge that had me stumped for several hours, replaced all the selenium diodes on the board but wasn't getting the correct horizontal actions until I bumped the input connector and discovered a number of traces had broken right where the Molex pins were soldered to the board and lifted the pads. Servicing the set I'd run it with the chassis out of the cabinet sitting on some tape cases then when it was time for the bench would just disconnect everything at the Molex connectors, the sheer number of cycles broke the traces. Once that was rectified the convergence fell into place for the most part, still doesn't respond to the R/G diff tilt so that's on the to-do list.

So overall it's getting there, not quite up to where it was back in the day but now an incentive to keep moving forward. One thing it still has is RF sensitivity. Growing up I remember the DX football games with this massive antenna and Alliance U-100 rotator, there were tape markings for the station positions all around the face and every fall we'd have to check all the flat cables. In the house I can run the modulator on a 1/4 wave antenna with good results.

Got loads of detail pictures of the chassis and operation, give me some time to smash them down to fit the server limitations and I will post them

[kf4rca]

I thought there might some wierdnesses going on, You should not have to do a complete IF realignment. I suspected broken grounds. Many times you can just resolder the board and that fixes it. Another thing I thought of was perhaps whiskers growing in your IF cans. I know there is at least one IF can from the schematic you sent me. And then there is the problem of carbon resistors going off value due to its age. I'd replace them with metal film.

[ARC Tech-109]

The weirdness was the wrong value resister in disguise causing the Chroma, blanking and AGC issues and this is where I mistook the problem as being a misaligned IF strip that came from a T933. The AGC overloading was causing the strange response so I attempted to counter these and made things worse. I then went to replace all the resistors on the IF board along with by pass caps then during the alignment phase grazed the IF cable from the tuner against the little fuse block next to the input jack on the chassis and smoked the tuner. Two months later I have the front end out of a T933 and while double checking things came across the rogue resistor, once that was replaced along with the damaged AGC pot everything started to work correctly leaving me with the task of cleaning up the mess I made.

This all started years ago when I attempted to resurrect the old family set using NOS parts I had cleaned out of a storage room working at a Motorola service center. We had hundreds of those resistor kits in the little file drawers along with more in little jars and on tape & reel, I was the service manager at the time and tasked with the decision making... the decision was made to bypass the dumpster and load most everything in my truck. We had just a few weeks to clear out some 40+ years of history as the depot was closing and lease was done, once the test equipment and other capital assets were removed I got the word out to the ham radio community and they picked the place clean over a weekend.

So during the attempt at replacing all of the resistors in the set I managed to break the original IF strip thru sheer clumsiness and made the decision to go with composite video, with the push to digital and impending demise of analog broadcast this seemed like a good alternative and I worked out a few circuits but a few bad real estate moves followed by a divorce that traded me in for a younger model forced the set into a large refrigerator box and into the barn for long term storage. That was 20 years ago.

Now the goal is to un-modify the set and bring everything back to the way it was in 1970... well the TV chassis anyway. I replaced the original amp many years ago with a Heathkit AA-100 I picked up at a hamfest around 1985 and modded the receiver section for stereo TV audio adding one of those stand alone Recoton "FRED" MTS decoders picking the 4.5 SIF up using a loop.

As it stands at the moment I have a good response curve on the IF strip, not stellar and the task for the afternoon is to get this squared away and move onto the lack of vertical height. The 1K-3W resistor for the plate side of the 6GF7 is on the mark and DC values are within tolerance, the tube itself is new from the carton and I've sub'd in a few others with the same results. I have not checked the boost B+ as of yet but it might be a bad boost rectifier as it still has the original selenium on the board.
Once this gets straightened out it's onto a convergence board issue that has the R/G tilt not responding, I've only replaced the diodes on the board so far as I don't need perfect convergence at the moment. We'll get there.

[kf4rca]

What you could do is add a switch (or relay controlled by a switch) and you could use both video and RF inputs on that set.

[ARC Tech-109]

I could and have given thought to that but honestly I'm quite pleased with the performance using the RF input, I did leave a provision to add a composite input by soldering a two post tag strip to the top of the IF section cover shield and ran a wire to this for testing right off of TP2 and this goes directly to the grid of the 1st video amp.

The video amp is responding well and I get roughly 300 or so lines of horizontal resolution that's actually limited by the dot trio & aperture of the CRT itself, chroma response is good without the shading or fading that I was getting earlier. Convergence is what it used to be "back in the day" which is to say acceptable. It always had a slight bit of blue error on the right side that dipped down slightly and the only time it was annoying was when I was doing the ATARI 800 BASIC thing back in the early 80's.

The one annoyance is with the slight horizontal sync bending and the dark shading following the highlights and sometimes the vertical wants to follow the black bar used to adapt the 16:9 aspect ratio to the 4:3 frame in the modern DTV receiver boxes, wondering if video is sneaking into the sync somewhere but the scope didn't show anything during the last probe. I've gone thru the AGC section several times and everything is as it should be in terms of values, the 6GH8 is new from the carton and my signal strength is a measures -86db/m at the 75/300 ohm matching transformer coming from the test port of a Blonder-Tongue CAMS-60 MTS modulator with the RF drive all the way down. Thought it was an overload so I tried a different consumer grade modulator and got the same results so I'm out of ideas as I'm totally missing something here.

Vertical is still a little challenged and if I pull the height any further it loses lock and goes into this cycle of collapse then startup again only to collapse a second or so later never locking onto the sync. Supply line shows a constant +394V after the 1K/3W resistor (R112) and C5 is known good. Boost B+ is running 1196V give or take a few and I'm not seeing any horiz ripple on the supply side so my conclusion is one of the "new" parts isn't doing so well here. Plenty to keep me busy for a while.


I do have to give a big thanks to all the members who have put up with my long-o tell-o posts and given me the pointers and directions, also want to thank damen and Danny S. for their help with the replacement pieces. Without their help I'd still be at square one.

Danny if you're reading this I have not forgotten about you, just a little short on the spare change at the moment. Thanks again for all you have done.

Going to watch the set for a few more days while taking notes before I pull the back off once again, I'd leave it open but I do have three curious kitties in the house and they don't take no for an answer.

I will post a short video of the set warm-up and picture itself showing the bending and shading, this will be done in HDCAM so everything can be genlock'd together to eliminate the flicker effects.

[Electronic M]

If you don't like Letterbox from your DTV converter get s Zenith DTT-900 (or 901, or the LG or insignia rebadged versions of it)... Those boxes have a zoom button on the remote (also accessible through the menus) that lets you pick whether to down convert 16x9 to 4x3 cropped (pan and scan but with pan fixed in the center), 4x3 with letterbox, or 16x9 horizontally squeezed to fit.

[ARC Tech-109]

I have one of those Apex receivers that also has the zoom/crop/squeeze/fluff buttons on the remote, annoying in that the vertical wants to lock onto this black and not the actual blanking bar forcing me to use the ZOOM setting and cropping the sides or FILL which squeezes everything into the 4:3 box... fluff is reserved for my cats.

Just bought up a complete Magnavox T940 SAMS Set-1144 Folder-1 from stealbay for $14.00 that's going to get scanned in complete as a pdf if anyone wants a copy. My original copy is too far gone to work with anymore and for the money I couldn't go down to the main library and copy it for this price.

Give it a few days to arrive and I will get things together. It also has the (incorrect) schematic for the R228 receiver and the 8-function remote. I say incorrect as the Astro Sonic Imperial uses the R220 receiver so I will include both models plus the original Magnavox service pages making one large folder.

[kf4rca]

That's great. Show us some pictures of what you're doing. One thing to remember about using the RF input versus the a video input is that the video response is limited thru the IF strip. There's a sound trap at 4.5 MHz. that you would not have in a video input. Also there is some group delay thru the IF transformers. The video amplifier response is probably flat out to maybe 6 MHz. The sets I have converted look really sharp using the video input. Good luck!