Hi Heater failure is pretty easy to spot. You will be *way* off frequency. No need for a GPSDO in your counter to see if it’s out by 0.0001 Hz. You should be > 20 Hz low in frequency. Since there *are* multiple heaters, there are some odd combinations in a 3801. Your inner heater can be dead, but the outer heater still works. The unit will look like it’s warming up but doing so *really* slowly.
Based on seeing a lot of 10811’s and a lot of modern OCXO’s, a crystal that ages out is pretty rare. The “something else” category is way more likely when diving into one that is off frequency. That wasn’t always true so you can go back far enough in time (like into the 60’s) and find units that have aged out. They also aged out in the first ten years after they left the factory. Things got a lot better over the years in terms of long term precision crystal aging …. Bob > On Mar 3, 2018, at 10:12 PM, Tom Holmes <thol...@woh.rr.com> wrote: > > Bob... > > Interesting point about the heater not working vs the XTAL having drifted too > far. Mine has the same symptoms as the others reported (EFC at the end of its > rope) but have not tackled it yet, figuring I'd have to dismantle the whole > thing. Certainly troubleshooting a non-operating heater would be much more > pleasant. > > Thanks for that insight. > > Tom Holmes, N8ZM > > -----Original Message----- > From: time-nuts <time-nuts-boun...@febo.com> On Behalf Of Bob kb8tq > Sent: Saturday, March 03, 2018 2:07 PM > To: Tom Curlee <tcur...@sbcglobal.net>; Discussion of precise time and > frequency measurement <time-nuts@febo.com> > Subject: Re: [time-nuts] Z3801A OCXO manual trimming > > Hi > > First off some basics about OCXO’s. > > In a single oven design, you have a heater that warms up the entire crystal > and the guts of the oscillator. It is on all the time and > it gets things up to a temperature that makes sense for a given crystal. It > can be adjusted based on manufacturing data or by > trial and error to match the characteristics of that crystal. > > In a double oven design, you have two ovens that are on all the time. One > heats up the other one. They both work together to > achieve the end result. The gain of one adds to the gain of the other to give > an improved result. On some double ovens, the > entire heat range of the inner oven is only 10’s of degrees …. > > In a boosted oven, you have a second heater to get things going when it is > very cold. This is an unusual approach and rarely > seen. Its normally easier to just design a bit more power into the main oven > circuit. In a boosted design, the boost heat goes > away in normal operation at typical temperatures. In normal operation, the > gain of the boost circuit does not count. > > SO …. > > The oscillator in the Z3801 is a boosted 10811. It is boosted to allow them > to hit a spec of -40C on the unit. At the time it was > designed, there was talk about mounting these things in un-heated boxes > outdoors. After they got a bit further into all the > details of the designs … that part went away. The spec still hung around long > enough to apply to very early designs. > > The net result is that you can pretty much destroy the outer heater stuff and > the oscillator will work fine. There is no need > for it in a typical lab. There are some alarm triggers that need to be wired > “ok” when you do so. The details are in the archives. > > But … > > Best guess if your unit is at max EFC = the “real” heater on the 10811 has > quit working. To get at that, you will need to dig > into the guts of the unit. Given the massive EFC on the Z3801 version of the > oscillator, it would take a crazy amount of aging > to hit limit. > > Bob > > >> On Mar 3, 2018, at 1:37 PM, Tom Curlee <tcur...@sbcglobal.net> wrote: >> >> Since the Z3801A is being discussed, I thought I'd ask about an issue I'm >> having with my unit. I use my Z3801 as my working lab standard for the >> usual pieces of RF test equipment. In the past year or so I've had the unit >> drop out of lock and go into standby mode. Resetting/cycling power would >> bring it back into lock for a while, but it generally got worse and now >> stays in hold over mode. LH (thanks Mark Sims!!) reports that everything is >> operating normally except that the it has a PLL unlock. The one highly >> suspicious item is that the DAC is at 99.996902% - full output. >> The unit shows it has over 94.5K hours run time, so I suspect that the OCXO >> has aged to the point that the EFC can't pull it into lock. I disassembled >> the OCXO to see if it had a trimmer capacitor like the standard 10811 units. >> After removing the outer case and foam insulation, I see that the outer >> heater is one of the thin printed circuit serpentine heaters on what I think >> is Kapton. That would need to be peeled off of the case to either get to >> the hole for the trimmer (if there is one) or to remove the cover for the >> inner case. >> What I don't understand is the purpose of what looks like another coil or >> heater wrapped around the Kapton printed circuit heater stuck to the inner >> case. This second coil/heater is 2 layers of 1/8" thick red foam wrapped >> completely around the inner case, with fine copper wires wrapped over each >> layer. At least I think there are wires on each layer. This whole second >> heater is taped down and I don't want to dig any further until I know a bit >> more about what I'm doing or find that there isn't a trimmer on the double >> oven 10811 oscillators. >> Any ideas on the purpose of the outer heater (or whatever it is)? Does the >> double oven 10811 have a manual trimmer that I can adjust to bring the >> oscillator back to the center of the EFC range? Assuming that I can >> manually adjust the OCXO back into adjustment range, will there be any >> issues with the Z3801 performance, things like phase noise, short term >> stability, etc? >> Any suggestions will be most appreciated. >> Thanks, >> Tom >> _______________________________________________ >> time-nuts mailing list -- time-nuts@febo.com >> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts >> and follow the instructions there. > > _______________________________________________ > time-nuts mailing list -- time-nuts@febo.com > To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > and follow the instructions there. > > _______________________________________________ > time-nuts mailing list -- time-nuts@febo.com > To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > and follow the instructions there. _______________________________________________ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
