Here are some notes and data on all five of my units - all have been running great for several days now after the fixes/tweaks noted below. I have also power cycled and cooled them and they all lock within seconds of each other.
Borrowed one of TiN's pics to mark some *possible* adjustment points. The board layout on the unit pictured may not match yours perfectly however the pots are all the same even if the 4 caps that need replacement are not.
-The point marked GG appears to set the Rb signal level - if you think you need more Rb level you can tweak it slightly while watching the serial data.
-ADJ. B - is the pot I adjusted very slightly to get one unit that was showing a very low Rb level to suddenly jump back to life and lock. No idea what it really does. If turned too much you will lose lock
-ADJ. C - no idea what it does. I did adjust it and didn't see any change in the serial output levels but changing it too much will cause lock to be lost.
So I recommend the following:
-Power up each unit and verify they output something close to 10 MHz
-Replace the four 47uF 35v volt caps - this is mandatory! I used radial caps instead of axial as high quality ones are much easier/cheaper to source. It was easy to fit them in.
-Fire up the units and see what they do. If they lock at 10.000,000,xxx,x they are doing well - let them warm up a little and if the GG level is really low(like 20 or less) and you're feeling adventurous give it a small tweak to bring the level up to 4x or so. Levels below 20 may show 10 MHz but the lock indicator may toggle or not be stable.
If they don't lock then you may want to give ADJ. B and/or possibly ADJ. C a very small tweak - make note of the original positions!
Some other notes:
-If you find the unit tough to get apart try prying gently at the edges of the serial port.
-I had to drill a few screws on the big outside heatsink that were seized. There were really cheap screws on some and stainless steel on others.
-The units have a SMT 2 amp fuse internally, no comment on how I know this, lol. If you're unit appears completely dead check this!
-The outside pot can adjust the unit very precisely, as can the eeprom fine and course adjustments. You'll need a really good reference to fine tune them.
-All of my units now start at 9.999,500 MHz +-200Hz when powered on, the frequency tends to go down every so slightly as it warms up. When the Rb lamp fires I tend to see a very slow increase in frequency that gets faster over time. Eventually it gets really fast and should lock as soon as it hits 10 MHz - working units will not ramp past 10 MHz and then lower the frequency again, they will lock first shot.
-If you turn the GG level up too high you may see the unit lock/unlock/lock/unlock cycle a few times before staying locked. The GG level tends to start quite high and drops when the units reach full temp. If the level is too high I think it confuses the unit before the level drops.
-When a unit is working properly it should lock and stay locked, no flickering or toggling of the lock indicator at all.
-All five of my units have the course electronic adjustment set to zero, all of them have some small tweaks to the fine electronic adjustment.
-You can safely connect the serial connections direct to a standard RS232 uart/serial port.
-I added a third picture showing where accurate sources of TTL level 5 MHz, 10 MHz and 20 MHz square waves can be found if needed.
-Added 3 pics showing the output on my spectrum analyzer with a 30dB attenuator inline - all 5 units look identical. Spec is 0.5 Vrms output which is 6.99 dBm so the output level is bang on.
Results for each of my units(as pictures below):
unit #5:
This unit has had the 4 caps replaced and the magic pot tweaked a little.
Putting out a stable looking 10 MHz - 10.000,000,001,x - with 1 second samples from my 53132A with GPSDO reference the frequency is only changing +- 1mHz.
Serial output:
79 62 C3 96 85 67 AE DB
79 61 C2 96 86 68 AD DC
79 63 C2 95 86 68 AF DD
79 62 C2 96 86 69 AE DD
79 61 C2 95 86 67 AF DC
79 48 C1 95 86 68 AE DC
79 64 C2 96 86 68 AE DC
79 64 C2 95 86 69 AE DD
79 64 C2 95 86 68 AF DC
79 62 C2 96 85 68 AE DB
79 64 C2 95 86 68 AE DC
unit #4:
This unit has had the 4 caps replaced, no other changes made.
Putting out a stable looking 10 MHz - 10.000,000,001,x - with 1 second samples from my 53132A with GPSDO reference the frequency is only changing +- 1mHz.
Serial output:
8A 75 C3 A8 6E 74 AF DB
89 73 C2 A8 6E 73 AF DB
8A 71 C3 A8 6E 74 AF DC
89 71 C3 A8 6E 73 B0 DB
8A 72 C2 A8 6E 73 AF DC
8A 73 C2 A8 6E 74 B0 DB
8A 71 C2 A8 6E 72 AF DB
8A 73 C2 A8 6E 72 AF DC
8A 74 C3 A8 6E 72 AF DC
89 71 C3 A8 6E 73 B0 DC
89 6F C3 A8 6E 74 B0 DC
89 71 C3 A8 6E 74 AF DD
unit #3:
This unit has had the 4 caps replaced and the magic pot tweaked a little.
Putting out a stable looking 10 MHz - 10.000,000,001,x - with 1 second samples from my 53132A with GPSDO reference the frequency is only changing +- 1mHz.
Serial output:
78 53 BF C1 62 7E B3 DD
77 4F BF C1 62 7E B3 DD
77 3A BF C1 62 7E B2 DF
77 53 BF C1 62 7F B3 DC
78 54 BF C1 62 7E B2 DD
78 53 BE C1 62 7E B2 DE
78 51 BF C1 62 7F B2 DF
78 55 BF C1 63 7F B3 DE
78 54 BF C1 63 80 B3 DF
78 54 BF C1 62 7F B3 DD
78 53 BF C1 62 7F B2 DE
77 54 BF C1 63 7F B3 DE
78 55 BF C1 63 7E B2 DF
78 3A BF C1 62 7F B2 DC
Unit 2:
Caps changed but still no lock at all. It does show a hint of a Rb signal as it passes 10 MHz during a sweep but is very weak.
UPDATE - made a small tweak to what I have labeled ADJ. B - and it has locked just fine ever since. It is almost like the pot was just dirty and needed a hair movement to be make contact again the change was so small. This unit appears to be the weakest in the bunch as the AA level is high.
Serial data:
7A 4C C5 B1 87 68 AE E1
7A 4B C5 B1 87 69 AE E2
7A 4F C5 B1 87 68 AD E2
7A 4F C5 B1 87 68 AE E1
7A 51 C5 B1 87 67 AD E2
7A 4C C5 B1 87 68 AE E1
7A 51 C5 B1 88 69 AE E2
7A 4C C5 B1 87 68 AE E2
7A 4F C6 B1 87 67 AD E2
7A 52 C5 B1 87 68 AD E2
7A 4A C5 B1 87 68 AE E1
7A 51 C5 B1 87 68 AE E1
7A 50 C5 B1 87 68 AE E1
7A 51 C5 B1 87 68 AE E2
Unit 1:
Caps changed and locks with a solid looking 10 MHz - it is wide open on the bench still as my test unit though so no stability can yet be judged. I am sure it will be fine though.
UPDATE - all reassembled and running great so far.
Serial data:
8A 3E C3 B4 6D 72 A8 DB
8A 43 C3 B5 6D 71 A8 DC
8A 44 C3 B5 6D 71 A8 DC
8A 44 C3 B5 6D 71 A8 DC
8A 45 C4 B5 6D 72 A9 DB
8A 45 C3 B5 6D 72 A8 DC
89 43 C3 B5 6D 71 A8 DD
89 2E C3 B4 6D 70 A8 DD
89 43 C3 B5 6D 70 A8 DC
8A 43 C3 B5 6D 72 A9 DB
89 44 C3 B5 6D 71 A8 DC
8A 3E C3 B5 6D 72 A8 DB
89 43 C3 B4 6D 72 A8 DD