The 6BK4 regulator does not work by regulating the voltage. It works by regulating the total current load on the high voltage. It measures this current at the bottom of the flyback. That is, when the CRT draws less current (dark picture), the 6BK4 draws more current from the high voltage anode lead, keeping the load constant. Therefore, the load (and the voltage) should be constant, until the picture is so bright that the 6BK4 has reduced to zero current. If the CRT draw is raised higher, then the draw on the high voltage output does increase and the voltage drops.

So: if operation is proper, the high voltage should remain constant as brightness is increased up to a certain point, and then the voltage should start to droop as you try to get more beam current. If the voltage is drooping gradually as you increase brightness from black, the regulator is not drawing compensating current like it is supposed to.

hmmm while getting set the pic starts slowly bloom check the current 230ma and climbing, yikes, tried another NOS tube (elmenco branded) and back to 180ma, going to have to watch this.

This appears to set the relative gain of X and Z. The effect should be most visible on saturated colors, like color bars. This will affect all the color bars in some way. For example, one direction should make magenta more reddish and the other direction should make magenta more bluish.

As Wayne pointed out, I think you're misunderstanding how the 6BK4 does what it does. It doesn't tell the flyback to 'make' more power, which is what I understand when I read your post, it's only there to provide a constant load on the flyback. The flyback always operates wide open, only thing that can drag it down is the CRT or the 6BK4. When beam current at the CRT is low (dark images), the 6BK4 picks up the slack so to speak. When beam current at the CRT is high (bright images), the 6BK4 will be cut off. In this way the load is always constant on the anode, but if you ask me it's a pretty crude and wasteful way to do things.

Regulation isn't the issue here, since the 6BK4 only absorbs 'excess' HV when the CRT doesn't need it. The issue is lack of sufficient headroom in the HV supply to begin with, which explains why the chassis works the same wether the regulator tube is installed or not. It made little to no difference to the images on screen for me, there's just not enough HV to make a difference. Other sets will overshoot by a lot without the regulator, my CTC-9 for example would hit nearly 32kv unregulated. In that situation there would be a difference in images on screen, but not the anemic CTC-5.


Do the variac test, see what happens.

I disconnected my shunt and the HV shot up to 25kv with a MUCH brighter screen, still holding at 180ma... So I will mess with the HV setup and regulation tomorrow.

I just realized there is no HV pot, so I will start with some tube swapping on the 6BK4

really? You must have the one chassis the was built right then, none I have ever seen made more than about 22kv. You can copy the circuit from any of the later sets to add the HV adjust pot, then you should see some improvement in brightness. Set it so the chassis puts out around 23kv with a blank raster, see if that does the trick.

Tube swapping the 6BK4 won't help, you need to change the tube bias with the HV pot.

yea I was kinda shocked by all I have seen with other reports, that's is why I had not even bothered to try disconnecting the shunt tube cap before as I just assumed it would not make any difference.

I hate the idea of drilling a hole in the chassis to add a pot, so if a tube swap does not work I will try upsetting the voltage divider of the two 1.8 meg resistors that are used to establish the bias with a fixed resistor bridged on the lower half of the divider. Will start with a 10 meg just to see. A better way would be to just replace one of the resistors with a 1.5 meg fixed and then a 500k pot is series as done on the 7's.

The schematic shows a 100meg cap in the cathode to grid, 100meg? yikes, anyway if that has gone low it would foul things up to. I did replace the .0033 cap that is there as I have read that leaky caps there have been responsible for regulation problems due to bias issues on that tube.

After talking it over, I think I will just check to make sure that all the existing values are correct and not try to up the HV more than the design calls for.

My concern is HV breaking down the insulation and arcing over in the HV tire.

I would rather have a dimmer CRT than no HV at all with a fly that is so hard to come by. I did try a new shunt, same HV 20kv. the old one was completely dark, which I presume is from Xray bombardment of the glass.

I've been following this thread, I'm about to tackle my Wingate chassis in a few weeks (finally! :sigh:). After studying a lot of early color set designs, I found it really weird that the schematic has 19.5Kv going to the anode while all others are 21-25Kv. I guess RCA was experimenting with different methods and I'm glad this one was only done on the CTC5. I agree with you, Dave. I think when I get into my chassis, I'll just make sure everything is as it should be and live with it. The only thing I'm thinking about doing is that contrast mod just to see what happens. It seems that the mod works better on some sets than it does on others.

Well the good thing about my set anyway is the HV is rock solid (with 5kv of head room it should be). I did not try it without the mod, but with it there is no sign of blooming at all.

Make sure you check the plate load resistor on the horz osc borad, its the 39k.

Won't happen, starting with the 5 there was a little bit of potting and a rubber insulator around the HV winding. Previous designs (CTC-4/21-CT-55) only had some wax dribbled around the bare windings, and they operated at higher voltage than the 5 does. CT-100 was different as well, it has some sort of plastic around the winding instead of wax.

It already makes 20kv, another 3-4 isn't going to hurt it. You don't have to drill a hole in the chassis either, you can use the pot temporarily to figure out what value you need. When you know, replace the pot with a fixed resistor and be done with it. The benefits are worth the potential risk, IMO. If I could get 23kv out of my set, I definitely would do it. Also, chassis for the 5 seem to be popping up lately. So if a flyback did blow, chances are you could find another one. It wouldn't be as hard as trying to locate one for a CTC-4, those are made of unobtanium.

while futzing with it I noticed the green was a bit weak, after screwing with drive controls, I finally decided to check tubes. Normally if they test a bit low on my mercury tube tester I don't sweat it, and one of them was a bit low. swapped in and color looks a lot better. per my tester SHB 3000, was about 1500, new tube was 2000 for what ever that is worth.

another odd thing was when I tried to use my vector scope the pattern was a mess on color bars. found out that if i disconnected the grid leads (so the vectorscope was not ready the chroma) the color bars on the screen look good, with in connected it was a disaster. I have never had a problem with the vector scope loading down the signal like that before.

oh the tubes where the 12BH7 difference amp tubes.

Perhaps its blasphemy to suggest, but would using a separate HV supply to provide the desired regulated 25KV be a solution? Would this also reduce the load on the flyback and lower its temperature?

Dave

Sorry guys. I think I my post was misinterpreted. I guess I wasn't clear.

Since the 6BK4 is a "shunt" regulator, I realize it shunts current to maintain a constant load current, hence supply voltage. Also to regulate, the 6BK4 does not sense current, it senses voltage. The boost B+ voltage is sensed which should be directly proportional to the HV.

What I was suggesting is that if unregulated HV voltage is at or below the point the 6BK4 begins to conduct, the 6BK4 will not shunt any current, hence will not regulate the voltage.

I measured about 23kV with the 6BK4 cap disconnected. (The cap removed effectively removes all HV regulation). When the cap is connected, the voltage remains about the same. This means the 6BK4 is not conducting, probably because the unregulated HV is too low.

The question I have is if the focus voltage remain constant at the point the HV begins to drop (HV out of regulation), this means the problem is the unregulated voltage at the output of the HV rectifier is too low. This in turn implies that the HV pulse arriving at the HV rectifier is of insufficient amplitude. Further, because the focus voltage remains the same (constant CRT load), the HV transformer core is not being over saturated with the HV supply current demand. In other words, the transformer is okay.

On the other hand, if the focus voltage drops with the HV, the transformer core is perhaps becoming over saturated and I would suspect a fault with the transformer.

It is interesting to note that the schematic suggests the HO Tube cathode voltage is 7.5 VDC. This means the cathode current should be 7.5/24 = 312mA. I do not think the current should be appreciably lower than this because I believe the flyback pulse would be reduced which in turn reduces the HV.

while checking some other stuff, mainly I wanted to see if the two 5600 ohm resistors in the chroma matrix used to drive the green were ok (they were I noticed the 5600 ohm resistor in the grid circuit of the green diff amp was about 50% out so I replace it. while check other stuff I found R702 (6800) was reading over 100k. The odd thing was it was not wired as in the schematic. In the schematic it show it across C and D of the chroma take off transformer T701. In mine its connected from pin 7 of V701a to D with no direct wire from pin 7 to D as shown (which is why it was easy to check in circuit).

I replaced it as well. Turn the set on and there is too much green in the highlights now, so a minor tweek and all good again.

for fun I tried putting a 22meg in P with the ground side of the voltage divider on the shunt, increased about 1kv to 21.5. The overall bias with this setup is about 12v but of course it would change with brightness settings. the schematic shows a 15v (375/390) so its all pretty good.

adding the 22meg in p is within the tolerance level of the resistor so I will just leave it at that.

I don't know what happened but the HV started going up to over 25kv and I could see the shunt was not glowing green suddenly. I tried wiggling the tube and even a new tube, still not regulating. finally I pulled the chassis and check the filaments for voltage, all the resistors, put it all back in the cabinet and now its working again, I have no idea what was going on with that.

I am going to try changing the voltage divider resistor to 120k vs 83k (as done in later production runs) just to see if there is any effect on the cathode current in HO tube (the voltage divider resistor is on the horz osc board).

the 82k measured at 94k I went with a 120k. No drive bars, horz lin looks good,
21kv (with a 22meg on the ground side of the shunt) and the cathode is less that 180, about 178, which is about as low as I have seen on ANY of my color sets. there is a 'tiny' hint of a drive line that I can see on a blank raster, so I may go back to the 100k but will leave it be for now, and will check the fly temps as well.

The color looks good, but I still need to go over the purity and convergence again.

duh after looking a bit closer I think I am at 22.5kv the divisions are not 1kv I should have looked closer.

some progress pics

couple more OOT resistors, a common cathode resistor on the 2nd bandpass amp and blanking amp 270 reading 750, and I doubt it will make any difference but will replace anyway.

normally when I am replacing parts on old wafer style tube sockets, I just clip the lead and pig tail, this is to prevent breaking a tube socket pin connector and ruining my day.

In this case it was just the 270 ohm resistor and I could see it was not really wrapped around the hole, so I got out the solder wick and safely removed the resistor. Replaced the resistor all looked good, the only other connection to the tube socket was the wire that went to the blanker tube socket, but I did not disturb that.

I always confirm my work so if its a pig tail I make sure I have good continuity across the pig tail connection (if the wire is not clean its easy to have a good looking connection but no actual contact INSIDE the pig tail.

in this case I did not have to worry about that as I could see I had good flow and had filled the hole in the tab with solder. I did check to make sure the wire that goes to the blanker tube cathode was still good and surprise, it was open. a closer look and I could see the wire had broken right at the pin that I had just soldered to. it was partially corroded and I guess the minor movement done in the install of the resistor must have broken it.

I had to fish the wire out and strip it to get a good clean piece, lucky there was enough slack (just). solder wick out the connection again and redo thru the hole and wrap, solder back. Needless to say I was very careful as the socket tabs break easy.

when done I confirmed that I had continuity back from the pin thru the wire to the other tube socket, AND the 270 ohm resistor was still 270 and connected.

Watch out for those very fragile hookup wires used, in this case I think it was 26 gauge solid so not much to break it.

after fighting poor greens I decided to try the vectorscope again. this time I clipped on the other side of the grid coupling caps for RGB on the CRT. Much better, did not effect the CRT display at all.

While there were color bars, the blue and reds were fine but the magenta was weak as was the green.

looking at the scope you could see the fingers but they were much more of an elipse that an a circle.

My interpretation of this is the phase of the CW from the osc was not correct to one of the demodulators, and the transformer that changes the phase needed to be adjusted. Since green is derived from the X and Y demodulators, then a defect there would really mess up green.

What I like about the vectorscope is the instant visual response (not just that its changing but that its changing in the right direction.

as it turned out about 1/4 turn of that phase adj transformer was able to open up the elipse and give a nice circle pattern, when I looked up at the CRT it now had a very nice color pattern.

This is the second time the vectorscope came to the rescue for me, I highly recommend one. Besides they also have nice dot and cross hatch settings.

funny thing is how good you can get flesh tones even with green so useless, I guess its just the flesh tone does not need much green unless your looking as Spock.

color looks great, but it was a bit weaker than before (which makes sense as the elipse was larger than the circle) a tweek of the ACC (which controls the chroma band pass amp gain based on burst amplitude) fixed that.

I still plan to make a custom top for the fly cage, it will have a plenum with a built in filter to help keep the fly temps low. Even with the low current thru the fly (about 175ma) it gets hotter than I like after extended viewing (hit about 165f after about 1 hr).

You are correct - don't know what I was looking at before.

I also found the vectorscope useful for touching up the matrix balance pot on the back of the chassis as well. this pot varies the cathode bias resistance of demodulator tubes. the effect on the vector scope is to squash down the circle pattern to get it as round as possible. I found that by adjusting the top core of the CW driver trans and the matix pot at the same time was the best way to setup the matrix.

I assume with the vectorscope you are using a gated rainbow pattern producing a "flower petal" vector display.

If the drives required for the red and blue guns were equal to make white, then the pattern would need to be elliptical, with the B-Y gain = 2.03/1.14 = 1.78 times the R-Y gain. (These are the ratios dividing the B-Y and R-Y axes before the chroma is modulated.) However, the red phosphor in the early tubes is weaker by about the same ratio, so equal (approximately) is correct (that is, the pattern is circular).

By a fluke of changing both the green and red phosphors in later tubes, the red gun requires less drive to make white; so there was a fix to reduce the red luminance drive; and at the same time, the yellower green sulfide phosphor means that the extra R-Y drive is still needed, to approximately compensate by making the difference between reddish and greenish colors greater. So, the matrix adjustment does not have to change, and can stay with a circular pattern when a new CRT (e.g., 21FBP22) is substituted.

Yes I am using the gated rainbow pattern. The petals are roughly circular now, but not very pretty (somewhat misshapen pointy, sickle shaped), so I am sure more work could be done, but I am not sure if its a fault of the demodultor or the chroma signal coming from the video detector or maybe just the design. I plan do do an IF and Chroma sweep alingment later with a BK 415.

It may be fun to play with some of the chroma tubes just to see the effect.

I was a little bit surprised at how small an adjustment was required to from practically no green (green would look either blue or brown depending on tint setting) to perfect green, the adjustment was about a 1/4 turn and it is very narrow. I have had quite a few sets that had brownish greens, so next time before I touch anything I am just going right to the vectorscope and the gated rainbow to see whats happening. But you mention that in some cases a round petal design is not optimum. Most of the CRTs I work on are the 21fjp22 or 23vadp22 types. I have a both grey face and green face types of the roundies.

it's hard to say exactly what shape the petals should have. Remember that the signal that's being gated is changing phase across the width of each bar, so even if phase response of the chroma is perfect, the trace will go outward from the center and back to the center while changing phase at a rate determined by both the varying phase and the final demodulated chroma bandwidth. Sickle shape is an indication of some phase distortion (rising waveform is different from falling), but this can be affected by fine tuning, adjustment of the sound trap(s), etc. and can be very hard to get perfect. On the other hand, if it's very bad, it will show up as edge discoloration on strongly colored objects.

If you really want to look for distortions, you need regular color bars, which have the same phase across the whole bar. Then you can make out things like the green-magenta transition not going directly across the gray center point, or other transitions not going straight between bars.

The circular case has been closer to correct than the case of greater B-Y for all tubes, since there never was a tube combining rare-earth red and NTSC green. The early tubes (weak red with NTSC green), and the later tubes (sulfide red or rare earth red with sulfide green), both require something closer to a circular petal pattern.

You are lucky you aren't looking at a Motorola with the single tube chroma section - that got G-Y off the cathode, where it was weaker than nominal and matrixed from the wrong proportions of R-Y and B-Y, resulting in that "brown or blue" syndrome you observed.

I next tried a reg NTSC color bar pattern, pretty good, but very noticeable ringing was observed, it was tunable but as crystalization was approached it got pretty bad. Turned out to be the 1st IF (stared with the 3rd and worked back subbing as I went). I am glad I have a good supply of tubes to sub.

It just keeps getting better in small increments.

I do get a little bit of color bleed that shows up more on the NTSC as there is no gating but most of my old color tube sets do this to some degree.

There are quite a few check points for setting up the chroma, but I am running out of gas and don't feel too inclined to mess with it much more.

the ring is still there, slight but still there (its a ring not a ghost as you can see multiple diminishing rings. Its tunable so I am pretty sure its in the IF. It looks like this set relies on pressure grounding of the pcb's (not the solder stakes like later RCA color chassis). Any one with some tips on what I may look for as far as IF ringing (of if there is another area that could cause ringing? its in both the luma and chroma. I don't think it would be delay line issue since it is tunable.

For the heck of it, try subbing the 3rd IF again (i know you subbed it before and the main culprit was the 1st IF). I've had a number of cases of ringing being caused by the 3rd IF tube.

I will try a few more OC and report back later.

did not get to more tube swapping but did notice some odd voltage on the video amp (6aw8). the screen was way high, like 210 vs 180 and the plate was low 110 vs 150. I checked the voltage divider on that tube and found the 15k was 11k and the 56k to ground was 65k. Not a lot but in those direction it would add up to too much screen voltage. I went ahead and replace those two ( had some old school 2watt carbon comps on hand) and check it again, much better screen at 178 and plate at 138. Seemed to help the video out, brighter and the contrast control is not as sensitive anymore, but still has a good range.

this all stemmed from what I thought was a bit dim of a pic, I started with checking out the video out, then moved to the video amp. Interesting effect of the change was before the fix I thought I had a slight tracking issue, as increased the contrast, the green highlights seemed to worsen (was checking by watching a B&W program. I just assumed it was a CRT issue. After correcting the video amp tube (and the contrast seems to work "better") I don't notice the green highlights happening. I spent a LOT of time trying to setup the the CRT drive pots to get rid of problem before. Perhaps it just not as much drive at the CRT cathodes, but it def is better.



Its fun to check things out, make a few changes check again, to slowly see improvements. I think it may add to overall trouble shooting, by being able to see direct cause effect. Not perfect, and certainly no substitute for good trouble shooting technique, but fun non the less.

more intresting tidbits, looking over the schematic, I can see the brightness control gets it neg voltage for low brightness from the horz out grid (filtered thru a .22 cap and 470j resistor. I am pretty sure I increased this neg voltage from the -40 to something less (more neg) when I was changing the plate load resistor in the horz osc in a attempt to increase drive to the horz out. Later CTC get this neg voltage from the grid of the blanker tube. I think I will check the later CTC-7 and see if it makes any changes for the great neg voltage at the horz out tube shown on that schematic (and if there is any changes in the circuit that taps this to supply the brightness control with its neg voltage).

tried a few more 3rd if tubes, found one that got rid of the rings. I think any more improvement will have to come from a full alignment.

while checking on brightness level I checked the cathode and grid voltages at the CRT. I don't have my VTVM handy, but the green grid was maxed out at 240v while the schematic has it a 255, the cathode voltages were much closer.
I don't know if the load of the DMM is more than a VTVM, but I may just check that 1 meg resistor that feeds the B+ to bias the grid thru the background pots. I would think that I should be able to hit the 255 at something less than max setting, as opposed to the max of 240v. the other drive controls seem to have no problem getting the the correct bias setting. While I am able to get good gray scale, its at a reduced brightness. Later I plan to install my CRT test socket so I can see where the G2 voltages are (they are not easy to get at like the cathode and G1). I am pretty sure I should get a brighter pic, the CRT test fine, and the HV is solid at 22.5 kv, I know the early CRTs were not as bright as later but I would at least like to see how it looks when the voltages are closer to what is in the schematic.

chasing low brightness down the CRT bias voltages turned up some interesting notes. The blue background when maxed out would only bring up the G1 to 240v the schematic calls for 255v and it makes a big difference here. So I started by checking the resistance of the 1 meg resistors that couple the pot adjusted boost voltage to the G1s. I checked the blue G1 1 meg out the circuit and it tested at 6 megs. Replaced that one. The other two G1s had no problem getting two the specified voltage so I will leave them be for now (can't test in circuit and I don't want to replace something that is working, even if it has drifted as the pots can make up the difference).

While I was at it I check the 10k resistor that provides the boost to the pots, it was 16k. I have noticed that when it comes to voltage dividers you need to be pretty close to the correct values to have the correct ratios to divide the voltage by. Since I was low on G1 voltage and a drifted high resistor would make that worse I went ahead and replaced it as well.

turned it on and wow, solid blue screen, I was able to turn the background control to about the 1/2 way, and now since I was not limited to brighness (and greyscale to the weakest color blue) I was able to increase the overall brightness to what is pretty much normal for all my color sets.

I am getting a bit tired so will leave the checking of the screen voltages to later, but one of those (or both) resistors was really a problem (the 10k that was 16k or the 1 meg that was 6 meg).

Still no blooming at all, I don't know why you guys with 5's are having such low HV voltage and regulation issues, this one seems rock solid. All I did was make sure I had good HV tubes and I did change out the 83k to 120k plate load resistor on the horz board, but that was just to get the fly current as low as possilbe with out a drive bar (there is just a hint of one in there now, 100k would prob be perfect). Iam getting a solid 22kv with plenty of head room.



My recommendation to folks having low brightness, remember to check voltages at the CRT.

I don't know if mine has the mods done to it or not, and I'm 950 miles away from being able to find out any time soon. But as I said, yours must be the exception rather than the rule. Every single one I've run into so far has been really weak, even with pots for HV and drive mine fails miserably to sustain 22kv at anything close to a normal viewing brightness level. Count your lucky stars.

I am currently hundreds of miles away as well from working on my set. But I have exactly the same problems with my CTC5 as Nick. The shunt regulator is hardly passing any current because my set is only generating about 22kV with the shunt removed.

I am curious if in the murky past someone hasn't replaced the Horizontal Output Transformer in your set. Are there any traces of resoldering?