[time-nuts] HP 105B Problem?

[Ed Palmer]

John,

J. Forster wrote:
>> Thanks for the reply John.
>>
>> J. Forster wrote:
>>     
>>> Likely yes, but check the spec. If the output circuit is a low Zout buffer
>>> amp and two separate roughly 50 Ohm resistors, they should be
>>> independent.
>>> Or, two distinct buffers could be used too.
>>>       
>> Ah yes, the spec that's in the manual that I don't have!  :-)
>>     
> I assume you have checked the Agilent site and with Dave at Artech Media
> for update sheets?
>   
The 105B is one of the products that Agilent sold to Symmetricom which 
means that they don't have any info on it.  Of course, Symmetricom has 
no info on the 105B on their site.  I haven't talked to Artek Media, but 
they don't list it.  I did spend some time with Google to see if I could 
find any other copies or updates, but struck out.

>> I don't know if HP often revises specs or procedures on newer versions of the same model number.
>>
>> I don't think they make major changes w/o changing suffix letter, but am
>> not 100% certain. Certainly they are not likely to degrade performance.
>>
>>     
>> If not, my unit has a problem.
>>     
>>> You can do a simple check with an Ohmmeter, with the unit powered OFF.  Measure the R between the front and rear output center pins. If it's close to zero, they are likely just paralleled, not individually buffered.
>>>       
>> Yes, the front and rear outputs are directly connected - no resistors or buffers.
>>
>> OK.
>>     
>>> As to the amplitude growth, try a different cable length., It might be a resonance w/ the cable length terminated by the 'scope's C.
>>>       
>> I just realized that it's not the device on the end of the cable, it's the cable itself.  Attaching a cable with nothing on the end causes the
>> meter reading to increase.  Not the kind of thing you expect from HP.
>>
>> Oh?
>>
>> Coax (or any transmission lines) at RF MUST be terminated AT BOTH ENDS in
>> it's characteristic impedance to avoid reflections / high SWR. Improperly
>> terminated cables tend to go nuts.
>>     
That's why I repeated the test with the signal generator - to see how it 
responded to an improperly terminated cable.

> It looks at least possible that the 105B outputs have a low impedance, not
> a 50 Ohm source. It's not called out in the catalog page scan:
>
> http://instrumentrentallabs.com/sale/catalogs//HP/1994/499.pdf
>
> I don't have a manual here to look it up.
>   
The old manual doesn't list an impedance either.  However, the responses 
of the 1 MHz and 100 KHz outputs are consistent with a 50 ohm 
impedance.  The 5 MHz looks more like about 7 ohms based on an 
open-circuit reading of 86 and a terminated reading of 76.  The circuit 
check is basically a peak-reading meter.  A 43 ohm resistor in series 
with each output might be very interesting.  But I'd rather find a 
schematic first.

>> I tried the same test on an RF generator at 10 MHz and a similar signal level.  The cable length had no effect on the signal amplitude.
>>
>> The RF generator has a 50 Ohm output impedance, dollars to donuts.
>>     
Yes, it does.

>> Definitely pointing to a circuit trouble.
>>
>> Ed
>>     
>
> I'm not so sure.
>   
I'm not at all sure.  :-)

Ed

>>> [snip]
>>>       
>>>> In summary, should I be able to use both front & rear connectors
>>>>         
> simultaneously without disrupting the other outputs and does it make
> sense that high impedance connections cause the 5 MHz level to rise?
> Sorry for the rather long message.  Thanks for any help anyone can
> provide.
>   
>>>> Ed
>>>>