Capehart 661-P

[Phil Nelson]

I replaced the three 680K resistors, whose values all had drifted too high. I ordered replacements for the 20-meg resistors, but I'm still waiting for them to arrive. Their values range anywhere from 21 meg to about 26 meg.

Phil Nelson

[old_coot88]

Just curious how strong the horiz lock is on a weak/snowy signal.

I see where i misspoke slightly in a prior post where this was said:

The pic is expanding with increasing brightness and shrinking when brightness is lowered, the opposite of what you'd see with a weak rectifier.

Coffee hadn't kicked in yet. Doh. The pic is doing what you'd expect with a slightly 'soft' HV supply.

[Phil Nelson]

I just checked the horizontal with a strong signal from a DVD player and the much weaker signal from my in-house transmitter. It locks fine either way, with plenty of stable area between the ends of the range where you lose sync. Possibly the locking range is a little wider with the strong signal.

Turning the brightness up and down definitely makes the image grow and shrink. Since I've ordered the 20-meg resistors, I'll go ahead and replace those when the time comes. I could try another 1B3GT tube, as well. The TV is watchable in the meantime.

Phil Nelson

[old_coot88]

So it uses 1B3s instead of the 8016s shown in the print?
Doggone li'l set gives one the warm fuzzy-wuzzies. Cabinet-wise it's just painfully cute. And circuit-wise it's got such novel circuits and they seem to work really well.

[Phil Nelson]

It came with 8016s, one of which burned out. My book says a 1B3GT will sub for 8016, and it works fine.

Phil Nelson

[old_coot88]

By golly you're right, they will sub. I never ran into an 8016 before.
Still trying to visualize how the 6L6 circuit generates a sawtooth rather than a sinewave. Maybe some of the mavens of the deeper mysteries could offer some insights.
oc

[Phil Nelson]

They label the 6L6 "horiz osc," but perhaps it only does output. In an old Milton Kiver book, I found a circuit (first scan) very similar to the one in the 661-P's horizontal/AFC subchassis (second scan). They both follow the general scheme shown in the 661-P functional diagram (third scan).

If I understand correctly, the sync signal arrives at the 661-P's horizontal/AFC subchassis through pin 3 of the adapter plug. One half of the 6SN7 amplifies the signal, which is compared by the 6AL5 to the output of the 6K6 local oscillator. The 6AC7 control tube corrects the frequency. The signal passes to the second half of the 6SN7, and then to the 6L6 output tube through pin 7 of the adapter.

Or something like that. In any case, my scope shows a sawtooth at pin 2 (plate) of the 6SN7, which goes to the grid of the output tube. Not the hump-and-peak waveform you get in a synchroguide circuit, but it makes a very stable picture, all the same.

Kind of a bare-bones system, otherwise: no damper, no horizontal linearity control, no drive control. The linearity isn't great, but I'm not going to worry about that until I've worked on the HV level and width some more.

Phil Nelson

[old_coot88]

Hokay. Now it makes a lot more sense. What threw me was the 6L6 being designated "oscillator". So i hadn't bothered to eyeball further back into that H.sync subchassis. Turns out it actually uses a pretty decent phase comparator circuit, rather than "direct injection" of sync into the 6L6 as i first thought. But still wondering how the flyback is kept from ringing without a damper.

From the FWIW department, here's a few thoughts pertaining to the HV..

The three 680K 1W resistors in series sure look to be a post hoc engineering fudgery to limit the HV. As mentioned before, they form the lower leg of a voltage divider. The capacitive reactance of the 500pf 10KV cap forms the uppermost "resistor" of the string. If t'were me, i would open the lower string and insert a fourth resistor and experiment with its value, starting with 680K 1W. Finally, if it looks to be warranted, maybe open the string entirely and see what the HV does.

If the HV still seems a little "soft" (bloomy), i would first try a 680K 1W across the aforementioned cap and then experiment with values, to see if it'll stiffen the supply up (ie., give it better regulation, making the pic less bloomy).

And those four 20 meg resistors in series, totaling 80 meg, serve no obvious function other than to bleed the charge off the CRT when the set is turned off. They could be dispensed with altogether (IMHO), as most sets never used a bleeder at all on the CRT anode.
oc

[Phil Nelson]

Originally Posted by old_coot88

open the lower string and insert a fourth resistor and experiment with its value, starting with 680K 1W.

Do you mean substitute a single resistor for that string?

Originally Posted by old_coot88

And those four 20 meg resistors in series, totaling 80 meg, serve no obvious function other than to bleed the charge off the CRT when the set is turned off.

Yes, the raster disappears instantly when you power off. No glowing dot, for sure!

Phil Nelson

[old_coot88]

Originally Posted by Phil Nelson

Do you mean substitute a single resistor for that string?

No, just open the string and insert a fourth resistor in series with the other three. Probably the bottom end would be the best place to open it because that's where it's at the lowest potential.

Yes, the raster disappears instantly when you power off. No glowing dot, for sure!

I dunno, seems like it oughta take several seconds for the charge to bleed off through 80 megohms. There might be some additional factor(s) killing the dot.

[Penthode]

Phil,

I was looking at the schematic and I see the horizontal hold is called "framing". The control called "horizontal sync" seems to me would help frame the image. Very quaint nomenclature on this set overall. Curious.

On the AFC subchassis, the 6K6 is indeed the horizontal oscillator, the 6AC7 reactance control and the 6AL5 is the horizontal phase detector. The set up is almost identical to the classic RCA Synchrolock horizontal AFC. Perhaps the subchassis was a later production add on when they found the 6L6 as an oscillator/output combination wasn't up to snuff? Perhaps Farnsworth bit the bullet and later obtained a licence from RCA to use this circuit? Who knows?

What I find particularly interesting is its almost prewar design using so many octal tubes.

Great work. I am especially impressed by your CRT fix!

Terry

[Phil Nelson]

Yah, I was curious about the framing control, too. In fact, the horizontal sync control does what you'd normally expect: it controls the horizontal until you lose sync at one end of its range or the other.

The framing control has the same effect as (drumroll . . .) the horizontal centering control. Turning it simply shifts the entire image a bit left or right. Don't ask me why they wanted this control in front -- or wanted it at all, since it acts like a second centering control, unless it's a leftover from some earlier design that wasn't as stable.

My favorite quaint term is "beam relaxor transformer" for flyback. I wonder where they came up with that!

Phil Nelson

[Phil Nelson]

Originally Posted by Penthode

The set up is almost identical to the classic RCA Synchrolock horizontal AFC.

I'll be darned. I have never paid much attention to those SynchroXYZ names, but I looked up my RCA 630TS and that circuit from the book is a carbon copy of the 630TS horizontal AFC. The RCA manual has a nice explanation of how it works, too. Should have done my homework when I worked on that set a few years ago Oh well, you can't learn everything at once.

Phil Nelson

[Phil Nelson]

Well, the cabinet's refinished and the audio & picture are pretty peachy. I'd call this project done except for the imperfect horizontal linearity seen in the most recent photo.

This set has no horizontal linearity control, damper tube, or drive control, so there ain't much to play with apart from the width & centering. This degree of non-linearity isn't a big deal when you're watching a movie, but this is not the only TV I've had with exactly the same issue -- left side too wide despite all adjustments -- so I wonder what the deal is.

Phil Nelson

[Phil Nelson]

T'aint perfect, but every project must end, and 3.5 months is long enough for this one. Here's an article, with a few items not covered in this thread:

http://antiqueradio.org/CapehartFarn...Television.htm

Thanks, as always, for the advice & encouragement.

Phil Nelson
Phil's Old Radios
http://antiqueradio.org/index.html