I was reluctant to consider a torch because I haven't done it and didn't want to cause any more damage. The second (and more serious) reason is that I live in an apartment. My office / computer room / mad scientist's lab is actually a bedroom complete with carpet. Soldering irons I can deal with. Open flames however, are not going to happen. In summer I could go out on the balcony, but right now there's a foot or two of snow out there.
In any case, the deed is done. My plan was to use a Dremel with a metal disc coated with diamond dust to cut into the solder seam. However, since the sides of the oscillator were slightly bowed, I had to decide whether to sacrifice the can or the bottom. I chose the bottom. Pretty much destroyed it, but the can barely has a mark on it. I think I can fabricate a new bottom - I'll probably attach it with screws rather than resoldering it. When I look inside the can I can see the solder. It extends about 4 mm from the edge. A torch was the only other way I could have gotten it apart. The problem with the oscillator turned out to be even easier to fix than I could have hoped for. There's a ferrite transformer on the output - possibly a balun. The wire is about the thickness of a hair. The ferrite isn't tied down - it's just held by the leads. I don't know if it took a physical hit or if the solder just dissolved the wire over the years, but one of the leads had broken. I resoldered it and instead of a wobbly level of -20 to -30 dbm into 50 ohms, I now have a much more satisfying level of ~ +12 dbm. And the levels in the rest of the unit now make sense. Instead of hitting a Minicircuit RPD-1 Phase Detector with a level around -30 dbm, it's now seeing a level of +7 dbm - just what it should be. I'd like to thank you, John, and everyone else for their ideas. They were a great help. Ed > Message: 1 > Date: Tue, 20 Jan 2009 09:28:31 -0800 > From: "John Miles" <[email protected]> > Subject: Re: [time-nuts] Wenzel Oscillator Repair > To: "Discussion of precise time and frequency measurement" > <[email protected]> > Message-ID: <[email protected]> > Content-Type: text/plain; charset="us-ascii" > > You probably have a lot more room to play with, given the larger housing on > your OCXO. The ULN series is rather densely packaged, and that was one > reason why I wasn't crazy about going after it with a propane torch. In > your case I'd be tempted to try the torch method before actually damaging > the housing with a Dremel tool. > > While fixing a couple of Ovenaire OCXOs that use a similar form factor, I've > noticed that their PC board edges don't come anywhere near the endcap. The > Ovenaire parts were sealed with epoxy, so a heat gun was all that was needed > to open them, but I'm sure I could've opened them with a torch without > hurting anything, if they'd been soldered shut. > > -- john, KE5FX > > >> -----Original Message----- >> From: [email protected] [mailto:[email protected]]on >> Behalf Of Ed Palmer >> Sent: Tuesday, January 20, 2009 5:59 AM >> To: [email protected] >> Subject: Re: [time-nuts] Wenzel Oscillator Repair >> >> >> Now I understand! I was planning to reuse the case. It didn't occur to >> me to sacrifice the case and put the oscillator in another box. >> >> Thanks for the idea, John. >> >> Ed >> >> >>> Message: 3 >>> Date: Mon, 19 Jan 2009 21:20:54 -0800 >>> From: "John Miles" <[email protected]> >>> Subject: Re: [time-nuts] Wenzel Oscillator Repair >>> To: "Discussion of precise time and frequency measurement" >>> <[email protected]> >>> Message-ID: <[email protected]> >>> Content-Type: text/plain; charset="us-ascii" >>> >>> I used a fiberglass-reinforced cutoff wheel. This can open practically >>> anything, but it puts a lot of vibration and dust into the innards of >>> whatever you're taking apart. My guess was that this was safer >>> >> than using a >> >>> torch. >>> >>> Since I wasn't going to be able to maintain the original hermeticity, I >>> remounted the oscillator in a Hammond box, like so: >>> >>> http://www.thegleam.com/ke5fx/w1.jpg >>> >>> I brought the trimmer (which was also damaged in this >>> >> particular oscillator) >> >>> and oven-status LED out to the box lid, along with the four original >>> terminals. Makes a nice package that can be easily opened for >>> >> maintenance: >> >>> http://www.thegleam.com/ke5fx/w2.jpg >>> >>> The repackaged OCXO seems to work fine. I haven't made any hardcore >>> measurements with it but I can tell just by watching the >>> >> counter that its >> >>> short-term stability is similar to my other unmolested 5 MHz ULN. >>> >>> -- john, KE5FX >>> >>> >>>> -----Original Message----- >>>> From: [email protected] [mailto:[email protected]]on >>>> Behalf Of Ed Palmer >>>> Sent: Monday, January 19, 2009 4:07 PM >>>> To: [email protected] >>>> Subject: Re: [time-nuts] Wenzel Oscillator Repair >>>> >>>> >>>> I hadn't thought of using a Dremel. Did you use an abrasive wheel or a >>>> steel cutter? The case on mine looks to be about 20 ga. tin-plated >>>> steel (~0.04" thick). The gap is so small it might have been >>>> >> a friction >> >>>> fit to start with. >>>> >>>> Ed >>>> >>>> >>>>> From: "John Miles" <[email protected]> >>>>> Subject: Re: [time-nuts] Wenzel Oscillator Repair >>>>> To: "Discussion of precise time and frequency measurement" >>>>> <[email protected]> >>>>> Message-ID: <[email protected]> >>>>> Content-Type: text/plain; charset="us-ascii" >>>>> >>>>> I had a similar problem with a 5 MHz OCXO from their ULN >>>>> >>>>> >>>> series. There was >>>> >>>> >>>>> a bad solder joint on the output connection, easy enough to fix >>>>> >>>>> >>>> once I got >>>> >>>> >>>>> the unit open. >>>>> >>>>> In my case I used a Dremel tool to cut the seam. Suggest >>>>> >> wearing a dust >> >>>>> mask, obviously, and keep your cuts close to the perimeter of >>>>> >>>>> >>>> the can, in >>>> >>>> >>>>> case the PC board comes right up to the edge like mine did. >>>>> >>>>> -- john, KE5FX >>>>> >>>>> >>>> _______________________________________________ >>>> time-nuts mailing list -- [email protected] >>>> To unsubscribe, go to >>>> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts >>>> and follow the instructions there. >>>> >>>> >> _______________________________________________ >> time-nuts mailing list -- [email protected] >> To unsubscribe, go to >> > https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > and follow the instructions there. > > > > > _______________________________________________ time-nuts mailing list -- [email protected] To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
