HP 8567A YTO Unlock message!

I believe the 8567A was simply a reduced-cost version of the 8568A. It is probably OK to use the 8568A's manuals to troubleshoot it.

I am about to start tracking down the same symptom on an 8566B, so we'll have to wish each other luck! (That's a pretty-different machine, though.)

-- john KE5FX

------------------------------------------------------

------------------------------------------------------

Hi Thanks for the reply. Well I have put it on test mode yesterday (pressing "SHIFT" + "w") and the analyzer worked with a normal refresh rate!!! Made some test on the measurements it makes and ALL of them are inside specs!!! Correct frequency accuracy of the applied signal from the R&S generator, correct amplitude level, all buttons reacting instantly on pressing (before they reacting every YTO message flashing time)!!! Problem is that the error messages are still there. And I have more now! Except "YTO unlock" I have also "Oven Cold", "EXT REF" (thats the most strange as I have put the INTERNAL REF switch right!!!), and 2 more related to the YTO unlock message! Trying to figure out if thats a hardware failure of the YIG Osc, or a failure of the ,master time base TCXO, as it looks like only the refresh rate is "blocked" or delayed for some reason. But also it might be a mis-programming fault, and not a hardware failure... Also I'm thinking that the error it may be located on the 85662 display unit and not on the RF/IF base unit, as it looks that the display unit provides the RF unit some synchronizing signals thru the Bus cable and also some IF signals thru the Multi-Coaxial Interconnection. Not sure where the main TCXO is located, but it looks its on the 85662 unit and not on the RF/IF base... So if the refresh problem is related to the locking of the YTO that will lead to a TCXO REF fault. The EXT REF input is located on the 85662 unit and that guide me to think that the INT REF TCXO is also based on this unit... as it would be strange to pass a

Also for your equipment may be it should be usefull to download this:

Here's a link to the manual:

Don Cleveland

I don't remember well and I don't have a set of schematics in front of me, but a couple of suggestions:

It's unlikely that you are running without *any* 10 MHz reference. That reference goes into locking everything in the box and even if the instrument were in ext ref mode with that loop unlocked, the error is probably small enough that the other loops wouldn't have a problem. Normally the 10M tuning range is only a few ppm.

The YTO loop, (the "outer loop") probably needs several other properly locked references to operate correctly. In fact, it's "YTO unlock" error may only be a logical OR of inner loops. The idea being that if any of those other loops are not locked properly, the YTO signal cannot be trusted either. Oven Cold and Ext Ref are certainly affiliated with the precsion 10MHz reference and probably a detector circuit that looks for

Are there rear panel BNCs on that instrument with adjacent "10M out" and "10M in" labels that should have a short jumper cable that is missing? That might almost explain all of your problems.

The slow refresh rate is probably related to error reporting of the unlocked condition. It will probably go to normal rate once all the loops are properly closed. I wouldn't be surprised if test mode ignores the warnings so that diagnosis can proceed more or less normally.

Glenn

It sounds like a problem with the diagnostic circuits. Or: "Oven Cold" could imply a bad oven-control circuit in a 10811 OCXO module, if it has one. That's a pretty common fault, usually caused by a thermal fuse that's opened up due to old age.

I'd be somewhat surprised if the 8567A had a 10811 OCXO, though. If it has a TCXO with no oven, then the presence of the "Oven Cold" message makes me wonder if there's an interconnect problem between the 8567A and

It's almost certainly not the YIG oscillator, if the analyzer works OK in all spans in Shift-w mode.

It'll be in the RF section.

Hmm, are you sure? That should be on the RF unit, not the display unit, if I remember correctly.

Will do.

-- jm

------------------------------------------------------

------------------------------------------------------

Hi Thank you for your answer. To correct my self :)) rear input of external 10MHz ref is on the RF unit and not on display... I've done some more tests... Switched on external reference and checked with a +3dbm 10MHz sinewave . When i applied the +3dbm signal it works like with the switch on internal position. When i removed the signal from input, all system freezed... So the internal ref must at least works. Dont know about the freq accuracy of it yet (I'll measure it), but its working. So thats good.

Do you have the service manual of the 8567? If you have it, can you make some copies of specific parts of it (I'll pay all the cost for them of course)?

Yes thats should be true. However I still cannot locate a 10MHz TCXO inside that instrument. Have download some service (verification & adjustment) informations for the bigger brother of it (the HP 8568) and at the chapter that reffers to the 10MHz time base has a figure showing the location of the A27A1 (10MHz TCXO). The strange is that at my HP 8567A there is nothing there!!! Only empty space with some nuts to install something and a PCB text near it saying A28 BATTERY... I'm starting to believe that thats one of the differences between the two models... Not sure however, but there isn't a 10MHz time base at the suggested location. It might be installed on a different location?

Oven Cold and all other messages except the "YTO unlock" are only come on during the error correction routine. At normal operation mode, only "YTO unlock" & VERY slow refresh rate of sweep are the symptoms. Even at that refresh rate can make measurements of an input signal... I understand the YTO loop as you explain it. The strange thing is that during the test mode almost all measurements of inputs signals are correct both for frequency and amplitude. I have also checked an AM & an FM modulated carrier up to 1250MHz. And I say almost because I've noticed that when I set a Span 1.001MHz signal appears again...

There is a BNC for 1st Local out and another one for frequency refference IN/OUT depending on an INT/EXT switch aside of it. I've set it on Internal position under normal operation.

Yes thats true as I've read that during error corection routine some signal flags are ignored! Well I must found a manual now! Any other suggestion will be excellent... My regards Sotiris pdmtr

Correct, sounds like it is using either the internal or external 10 MHz reference.

Sorry, I don't. I'm doing all this from very old memory!

Given the symptoms, I wouldn't worry about that at the moment. Since it works in "Int 10 MHz" position, it's finding something.

I think that may be a clue. Narrow spans probably use somewhat different phaselock circuits, in order to improve phasenoise characteristics close to the carrier. The 8568 was very different from the 8566, as someone (John Miles? Hi John de n6gn) has already noted.

The 8568 used a "pilot channel" which was sort of a twin of the main signal path but there only to allow improving the phasenoise of the LO. I wouldn't be at all surprised if things are switched differently for narrow spans where narrow RBW filters mean that the improved phase noise was important. However, the fact that it sweeps slowly and issues an unlocked error on wider sweeps leads me to suspicion that there is still something more general wrong with the YO loop.

Judging from what you've written, I'm shifting focus away from suspecting the 10M reference.

I agree with you that you really need at least an accurate functional block diagram so you can unravel things.

Sorry I don't have any documentation and that I can't remember more detail.

Glenn

Hi John Thank you for the follow up message. Please read also my other answer to Glenn.

About the Oven Cold message I reffer at it on reply to Glenn. It only displayed running the "Error correction routine" mode.

Well I've found that at narrow span modes ( > But also it might

Yes it should be there!!! Only thing is I cannot find it! :))

Nope, you're right, I've corrected that already... :))

My major problem is that here in Greece we don't have a lot of HP support and to find something (used or not) we must take our chances with the "SLR" (shipping lose risk) parameter as I call it :( Have allready lost 3 manuals for test equipments ordered from US and another one delivered to me after 5 months from shipment!!! And mostly here, there is a total lack of (second hand) spare parts sources for things like test instruments... I have seen in the past, your tracking generator for the 8566. Nice mod you've done there on the 86222A plug-in. Will something similar work with 8567A/8568? My best regards to all that helping at anyway... Sotiris Pdmtr

PS. any follow up message is welcome ;)

Sounds like there's a TCXO module buried somewhere, instead of the standard 10811-60111 oven. I wouldn't worry about it any further.

In the case of my broken 8566B, the problem turned out to be A20C7, a 10 uF electrolytic that was open. The YTO main-coil driver circuit was oscillating at about 1 MHz, which kept the outer YTO control loop from locking up.

I started troubleshooting by verifying that the M/N and 20/30 MHz synthesizers were both working properly, using a frequency counter. That meant the problem had to be the YIG or its driver circuitry, including the pretuning DAC. I went through the YTO loop adjustments section in the tests/adjustments manual (which should be freely downloadable from Agilent), and found that while the pretune DAC adjustments were OK, I couldn't bring the actual YTO output frequencies into range. There was a 1 MHz oscillation in the voltage-to-current driver circuit that clearly didn't belong there, so I poked around with an ESR meter and found an open electrolytic (A20C7). Replacing it and recalibrating A20 took care of the issue.

Thanks! Spare parts are a problem everywhere, unfortunately. I was glad that I didn't actually have a bad YTO, or I would have been in trouble. The 1.5-GHz analyzers are pretty different with regard to their tuning/conversion schemes, so no, the 86222A mod won't work with the 8567/8 directly. Check out the article linked from the DKD Instruments page at

-- jm

------------------------------------------------------

------------------------------------------------------