[time-nuts] question about HP 5601a harmonic generator

[Kejia Lee]

Dear Richard,

Very nice to hear from you, and thank you a lot for the help!
Indeed, I have cleaned the erosion still not working.

I am trying to now* remeasure the break voltage of the diode*.
I suspect that the doping get diffused in the isolation layer after age,
which make it a good diode, but with different parameter for the step
recovery.

>From what I see in the circuit, I understand that *I need to tune the *
*resonance at a given frequency to have higher voltage to make sure *
*the diode reversely discharged*. I noted the tuning is very tricky, as
the waveguide assembly is also a part of the resonance circuit.

*Do you have a internal dimension of the assembly*, that will help speed up
with the EM calculation to understand which screw for what.

Best regards,

Kj

2017-12-25 2:42 GMT+08:00 Richard (Rick) Karlquist <richard at karlquist.com>:

> I extensively studied the 5061 harmonic generator when I
> was designing the harmonic generator in the 5071A.
> We are now going on 30 years since that work.  The diode
> had some HP part number.  Even if you knew this part
> number, you would need to have the Part Information Report
> microfiche that gave the vendor part number.  The
> chances of this diode still being available are very slim.
> So you are not going to get an exact replacement, except
> by cannibalizing another 5061 harmonic generator assembly,
> assuming it doesn't also have a bad diode.  As far as
> replacing it with a different diode, that too would be
> questionable.  SRD's have magic unspecified parameters
> when it comes to what they do in a circuit.  The 5061
> generator in particular has extremely high efficiency
> due to its unique design.  It was designed by a Korean
> professor named Soon Choi (IIRC) who spent a summer
> at HP.  He left no known documentation as to how the
> design works, or how to specify the SRD.
>
> After a long period of study, I concluded that I was
> never going to match the efficiency in any simple
> circuit.  The diode mounted directly in the waveguide
> is the magic.  Fortunately, by that time, we could use
> a phase locked DRO and get around this problem.  Len
> Cutler in his usual style criticized me for "taking the
> easy way out", but he eventually came around to my
> way of thinking.
>
> I will give you a little hope.  The harmonic generator
> consists of a bunch of silver plated parts that are
> supposed to fit together and achieve electrical conductivity.
> After many years, the silver and tarnish and upset the
> electrical conductivity.  You might try disassembling
> the diode mount and cleaning up everything.  After
> you reassemble it, you might have good news.
>
> I will say that I have never heard of an SRD failing,
> and as you say, it looks good at DC.  What possible
> failure mechanism could selectively make the diode
> stop generating harmonics but other wise be OK at DC?
> Somehow let the pixie dust escape without letting the
> smoke escape.  Seems unlikely.
>
> Also check the drive circuitry.  SRD's need tricky components
> to make them "step" correctly.  Again, poor grounding due
> to corrosion could spoil the harmonic generation.
>
> Rick Karlquist N6RK
> Member of the 5071A design team
>
>
> On 12/24/2017 3:54 AM, Kejia Lee wrote:
>
>> Dear All,
>>
>> Happy Christmas!
>>
>> May I know if anyone knows the part number or the specification of the
>> step
>> recovery diode in the harmonic
>> generator for the old HP 5601A cesium frequency standard?
>>
>> I recently received this toy. The cesium tube seems to be OK, and I had
>> located the problem being
>> no 9GHz from the waveguide. I checked the diode, the DC properties seems
>> to
>> be OK, but just
>> can not generate frequency comb any more. I really hope I can repair it as
>> a personal new-year gift.
>>
>> I will be really appreciated, if you had any suggestions.
>>
>> Best regards,
>>
>> K.
>> _______________________________________________
>> time-nuts mailing list -- time-nuts at febo.com
>> To unsubscribe, go to https://www.febo.com/cgi-bin/m
>> ailman/listinfo/time-nuts
>> and follow the instructions there.
>>
>>
>>