Help needed bringing a 555 back to life


selinica.harbinger@...
 

I've been unable to track down this error for quite some time.
Starting from factory cal procedures, the primaries of all transformers check out. The resistance readings for the power supply rails (-150, +100, +225, +350, +500) are all very far off. -150 shows as 16.85K and climbs slowly from there. +100 shows as 51 Ohms. +225 shows as 1.2k. +350 shows as 7.7k and drops slowly from there. +500 hits the worst at 108k.
The thermal time delay works at least.
The -150V adjust does nothing at all, and the -150V rail hangs at -119V constantly. I can't check elevated heaters with the rails being that far out.
The regulated heaters were showing as working using an accurate AC meter. I cannot currently retest those values as I made the mistake of letting someone use the meter when they could not find theirs (turns out meters do not like being set on ohms and plugged into mains power, who knew).
I ran a test for shorts on all components in the power supply. C640 and C730 trigger the continuity meter sporadically. R798 seems to have drifted badly.
I realize I'm missing something, but I don't know what. The fact the adjustment does nothing makes me slightly suspicious of the 5651 regulator tube.
I'm tempted to go the brute force approach and replace every carbon comp resistor in the -150V portion of the supply with metal film and wirewound resistors and replacing every paper and electrolytic capacitor in the power supply portion if the issue remains stubbornly elusive. The indicator unit also has both beams way off the screen, with one beam too high and the other too low. The adjustment knobs do nothing and the HV section measured out of spec as well.
Where should I look? Is it worth attempting a re-cap of the non ceramic capacitors with modern film, poly, and electrolytics before doing any further testing? Are there any specific carbon comp resistors that are known for being a nuisance in the 555?
If I'm going to hit the point of having to pull each carbon comp out of circuit to check them individually then I'm thinking I just may as well replace them and not have to worry about how those are.
I'm completely lost and in need of guidance if anyone can help, and thank you.


Morris Odell
 

The first thing you need to do is get the -150 right because everything else depends on it. I assume you've eliminated weak tubes by substitution and the indicator unit is properly connected. Check the voltage across C640 which is the starting point. It should be about 290 volts. If that's OK check the voltage across the 5651 V609. That should be 85 volts and stable. You can check R646 and R647 by measuring the resistance across then with the unit cold. It should be 555 ohms. Since the -150 is low and not changing with adjusting the pot, that suggests the 6080 V647 is not getting grid drive. Check the plate voltage of the 6AU6 V634. If it's more negative than the specified -40 check R 633 (1.5 M). You can then work back - check the grid of V634 which should be -76V. Check the output divider by looking at the voltage at the moving arm of the pot which should be -86.

Start with all that and let us know what you find.

Morris


selinica.harbinger@...
 

C640 has 272V across it in mine. The 5651 has -172V across it. The R646/647 pair reads as 498.8Ohm measuring from node to node. I didn't desolder them to check as maybe I'm just blind but they were hard to locate. The plate voltage for V634 was 0V and the grid was -153.5
I can get the -150V adjustment wiper to -148V at best.
Tube were given a check over initially and the weak ones swapped. The indicator cable was checked for continuity and to make sure there were no shorts between lines that should not be there.

I did note a small scorch mark under L790 where the resistor is and the diodes D642, 702, 732, and 672 all seem to have been replaced by a prior owner. They do not match the look of the D762 set nor the manual images or the pictures of other power supply units.
That's just a visual inspection though, all parts tested fine with diode tests and the resistor checked out.


Morris Odell
 

The voltage across C640 is close enough to normal as is the parallel combination of R646/647. But something doesn't make sense. The 5651 is a voltage reference tube that should be at a constant 85-86 volts when lit. It gets its feed from the -150 line through R608 so even if it's open circuit or faulty the voltage across it can't be more negative than the -150 volt line which you say is at -148 volts. Likewise the grid of V634 is fed from a divider that has the -150 as it's most negative end and a plate of V624 as it's positive end so it can't be -153.5 volts if the -150 line is at -148. If the plate of V634 (connected to the grids of V647) really is at 0 volts then V634 will be cut off and V647 will be conducting hard, making the -150 more negative which is clearly not happening. Under these circumstances the voltage there would be slightly positive as it's pulled up by R633 but V647 could be pulling grid current and sucking it down to 0 volts. Make sure pins 3 and 6 of V647 really are at chassis potential. You need to check that what's in your 555 corresponds to the schematic.

Morris


selinica.harbinger@...
 

I double checked the connection on V647, they're at chassis potential. They're both tied together and directly to the metal mount of the tube socket, testing their connection to ground on any point in the chassis gives 2milliOhms at the most.
I may just re-cap the power supply portion sans electrolytics and swap out every carbon composite resistor in the -150V section. The only other alternative coming to mind is checking thermals given the spot on the bottom of L790 and perhaps there is just a bad resistor somewhere that measures acceptable enough cold.
Could the 5651 be possibly fried from going outside of the voltage it can regulate or overcurrent if another component failed or is that a very unlikely outcome?


Harvey White
 

From what little I remember about VR tubes, as long as it fires, it *ought* to have the correct voltage.  I think I remember hearing about VR tubes that didn't fire,  but then they didn't light up.

IIRC, the 5651 was intended not as a power regulator tube, like the OA3 series (able to handle a fair amount of current as a parallel regulator to a load),  but was intended to have a fairly narrow range of current, to be a voltage reference to an amplifier, rather than a direct regulator itself.

I'd suggest looking at the supply voltage to the regulator, the voltage of the tube itself, and see if you can remove any tubes looking at that regulation voltage (to eliminate loading and shorts, if practical).

If the tube doesn't fire and show the correct voltage, I'd say it's bad.  Depending on what supplies you have, you could test it separately, but I seem to remember somewhere around 5 ma nominal for the current through the tube.   It wasn't designed for much power dissipation at all.

Harvey

On 9/19/2021 12:05 AM, selinica.harbinger@gmail.com wrote:
I double checked the connection on V647, they're at chassis potential. They're both tied together and directly to the metal mount of the tube socket, testing their connection to ground on any point in the chassis gives 2milliOhms at the most.
I may just re-cap the power supply portion sans electrolytics and swap out every carbon composite resistor in the -150V section. The only other alternative coming to mind is checking thermals given the spot on the bottom of L790 and perhaps there is just a bad resistor somewhere that measures acceptable enough cold.
Could the 5651 be possibly fried from going outside of the voltage it can regulate or overcurrent if another component failed or is that a very unlikely outcome?





Morris Odell
 

The way the -150 regulator circuit works is this:

V609 provides a stable voltage reference of -85 volts by being fed current from the -150 line through R608. Twin triode V624 is a differential amplifier that compares the -85 reference with -85 divided down from the -150 by the voltage divider consisting of R615, 616 (pot) and 617. C618 controls the AC feedback loop response. It it's leaky it will throw the whole regulator out of alignment and would cause the -150 output to go low. The error voltage at pin 6 of V624 is amplified by V634 and used to control the grid bias of both sections of V647 which is the series pass regulator. V647 is bypassed by R646 and 647 to provide a base load current. If the -150 is low then V647 is not conducting enough.

The 5651 is a voltage reference tube and is not supposed to carry more than a milliamp or two. It gets its current through R608. The voltage across it should not be more than 85 volts. If it's not at that voltage regardless of whether it's lit, check or replace R608 (33K) and the 5651 and see what happens. Also check C618 (0.01 uF). If it's a black beauty or bumblebee cap with colored stripes, replace it and any others you can find.They are notorious for going leaky.

You need to get the voltage reference right before anything else.

Morris


selinica.harbinger@...
 

Got it. Resistors and caps are on order, there are some of the black beauty and bumblebee caps in the unit, I ordered replacements for all of them in the power supply. There will be some delays since a couple other caps tested bad and the replacement is on backorder.
I'll update when I have the resistors and caps swapped and new tests done and what the results are.
It's going to be a few weeks, but the unit will get a nice set of new resistors and caps. I've got a copy of the component list and I'll just start marking parts off. I'll definitely swap out all bumblebee and black beauty caps for modern film or other caps. Electrolytics probably not. The carbon comp resistors are going to slowly go onto the replace list, knowing the issues they can have. The ones that have drifted badly are getting replacement asap (when the order arrives, someday) and the remainder get assumed suspicious.
The main factor with the electrolytics is just how hard suitable replacements are to find, much less the mount issue. I may put it on the someday list depending on how they reform (if needed). Maybe.


Mark Vincent
 

Selinica,

I sent you an email about what I did in mine last month. I have one of these I fully restored as a birthday gift to myself a few years ago. I had to restuff the cans because I had ones that were shorted, the resistors would burn open when the relay energized. I like mine with the new 1% resistors and new condensers. The horiz. position control is dissipating more power than it is rated. Mine has increased value resistors on each side and I still have enough range. Mine is an early S/N with P11 crt. The plug-ins were also restored and aligned. I did modification of the timebases to allow some plug-ins to be used that need that mod. to work (triggering). I also have the A versions I restored for the time bases.

Mark