[time-nuts] RF mixers for oscillator characterization, some questions

Magnus Danielson magnus at rubidium.dyndns.org
Sun Apr 19 23:54:18 UTC 2009

Hej Bruce,

Bruce Griffiths skrev:
> Hej Magnus
> Magnus Danielson wrote:
>> Bruce Griffiths skrev:
>>> Arnold
>>> Arnold Tibus wrote:
>>>> Bruce,
>>>> a very good summary!
>>>> I do not know where I can get the proposed MinCircuits DBM here, I
>>>> looked to all the big and international well known supplier of
>>>> electronics parts with negativ result.
>>>> If I have to import them from USA they could end up even above 100 $
>>>> ea. we have in Europe to take the high shipment costs and the 19%
>>>> VAT into account. 
>>>> Is it likely to get used mixers with damaged diodes?
>>>> Did somebody alredy look inside the metal case of the 10514A?
>>>> Will it perhaps possible to reconnect the common ground connections
>>>> such a way that the mentioned loops can be opened?
>>> Not without changing the IF port connector or at least insulating its
>>> shell from ground.
>>> It may be easier just to construct you own mixer if you cant easily
>>> obtain the PCB mount Mincircuits mixers that have separate IF and RF
>>> grounds.
>> Bruce, many SMD mixers also have separate LO, IF and RF grounds. It is
>> just not documented in the datasheets. The recommended footprint
>> pattern (on say RMS-5+) actually ties them together. The MiniCircuit
>> SMD mixers I have been able to investigated (some of them have open
>> bottom, so looking in just takes a sharp eye) is not hooked up with a
>> internally forced common ground. A simple DC resistance measurement
>> would indicate if a particular model has their grounds connected
>> together, but really, they should fix their datasheets. I guess it
>> just haven't been important for them before.
> Yes its likely that at least some SMT mixers have isolated grounds,
> however not all will.
> In particular all Minicircuits SMT phase detectors appear to have a
> common ground:
> The bottom view of the SYPD series phase detectors (attached) seems to
> indicate that all grounds are connected together internally (or at least
> externally on the package).

Appears is the key word. I have a Minicircuits SMT mixer which does have 
separate ground in real life but this is not visible in either the 
datasheets internal schematic, pin list or layout. The SYPD could infact 
also have them. The SYPD-1 does not seem too unreasnoble to actually 
have separate grounds. If it has separate grounds, the pairs should be 
1-6, 2-5 and 3-4.

The lack of these details in the datasheets rules out being able to make 
conclusive decisions on this. Sample them and measure rules.

The way these are manufactured, having to bond the grounds together 
within the encapsulation would be a nuance.

Now, to make it clear, I do _NOT_ say that all of them has isolated 
grounds, but I say that some of them do have them, even if not 
explicitly documented in the datasheets. This also makes it hard to 
select amongst them to find those which do have isolated and those that 
don't. The plain and ugly truth is that we do not know, except that we 
may not be having the correct information. We just have to be careful 
not to draw conclusions from information which was not designed to 
convey this detail.


More information about the time-nuts mailing list