[INRAD] Switching diodes in roofing filter for the IC-775DSP

[Thomas]

Bill,

Disclaimer - by no means am I an expert on the subject of PIN diodes, 
but I understand this to be the case:

The PIN diode differs from standard PN junction diodes due to the 
intentionally wide intrinsic layer between the P and N regions of the 
diode. Yes, the reduced capacitance aids in the OFF state isolation, but 
it is the minority carrier recombination time that is of real value in 
RF switching applications.

Appropriately applied PIN diodes have very long recombination times 
compared to standard PN junction switching diodes. Most any switching 
diode can be used in an RF switching application, but as you mention 
they must be heavily forward biased to place their operation well into 
the exponential forward biased region. Doing so allows a linear signal 
to pass with minimal distortion. The PIN differs in that, because of the 
intrinsic region with its long recombination time, appears more as a low 
value resistance (<1 Ohm) than as a PN junction. The PIN diode typically 
required much less forward current than a PN diode used in the same RF 
switching application.

The mistake that many hams make is assuming that any PIN diode will work 
in most any application. The MPN3404 is rated for VHF switching 
applications, because its recombination time is short; a few 
microseconds. Yet, I often see the MPN 3404 used in HF RF and IF 
applicaitions. Not to say that it won't act as an RF switch, but it will 
not provide the isolation or low distortion of a more suitable HF PIN 
device. By a more suitable HF PIN device, I suggest a device like the 
HP5082-3081, with its 2500nS recombination time (1 tau).

The reason people often use the IN4007 in PIN applications is it has a 
doping profile quite similar to that of HF PIN diodes. Elecraft 
successfully applies the IN4007 as an RF switch in their K2 transceiver 
design. They are able to achieve very good IMD performance with this low 
cost device. That said, it is hard to beat a good RF relay for isolation 
and IMD performance.

'73, Thomas - AC7A


Charlie Mazoch Jr. wrote:

> Charlie Mazoch Jr. wrote:
>
>> Hi Bill:
>>   The 1N4007 is a 1 amp 1000 PEV PN junction silicon diode mostly 
>> used in HV power supplies. I agree about IMD when PN diodes are just 
>> beginning to forward conduct. That's what diode mixer circuits are 
>> all about. But if a diode is hard forward conducting as in switch 
>> applications how does one get mixing?  It seems to me that one would 
>> have to overcome this forward bias before mixing can occur. In my 
>> Kenwood TS-2000 the unused band pass switching diodes have a reverse 
>> bias applied. They use a resistor in an SMT transistor to accomplish 
>> this novel bias scheme.
>> I do agree about PIN diodes offering lower capacitance than a PN 
>> junction diode in their reverse bias configuration. That's what the 
>> intrinsic layer is all about.    73's
>> Charlie  W5VIN
>>
>> William E. Sabin wrote:
>>
>>>In my homebrew rcvr (see QRZ.COM, W0IYH) I use 1N4007 PIN diodes in the 
>>>front end AGC circuits.  A problem with significant IMD occurs when these 
>>>diodes (or any other PIN diodes or ordinary diodes) are just beginning to 
>>>forward-conduct.  I use 3 diodes in series at each AGC location and my IMD 
>>>measurements show that this eliminates the diodes as a source of *increased* 
>>>IMD, regardless of their state of conduction.
>>>
>>>I use MPN3404 PIN diodes to switch the 12 crystals in the first local 
>>>oscillator because of their forward-biased storage time and low back-biased 
>>>capacitance. I don't want the turned-off xtals to influence frequency via 
>>>the turned-off diodes due to turned-off diode varactor effects. The 
>>>interactions between the 12 xtals from 10.4 MHz to 38.4 MHz are negligible.
>>>
>>>For the nine hamband RF filters I use miniature Radio Shack (RS 
>>>275-241)relays which are inexpensive and very reliable (over many years zero 
>>>failures in many low power applications). The simplicity of this method and 
>>>the reduction of DC power consumption are desirable (the off-relays consume 
>>>zero power).
>>>
>>>Bill W0IYH
>>>
>>>----- Original Message ----- 
>>>From: "Charlie Mazoch Jr." <w5vin at earthlink.net>
>>>To: "Anders Janis SM4RNA" <sm4rna at telia.com>; <list at inrad.net>
>>>Sent: Thursday, November 24, 2005 9:29 AM
>>>Subject: Re: [INRAD] Switching diodes in roofing filter for the IC-775DSP
>>>
>>>
>>>  
>>>
>>>>Hi:
>>>> If you read W8JI's notes on switching diodes the 1N4148 is just fine
>>>>as a switch. 
>>>>    
>>>>
>>>
>>>
>>>
>>>_______________________________________________
>>>INRAD mailing list provided as a service by International Radio
>>>INRAD Web Site: http://www.qth.com/inrad
>>>
>>>To Submit Message to the List: list at inrad.net
>>>List Help:
>>>http://mail.inrad.net/mailman/listinfo/list_inrad.net
>>>
>>>  
>>>
>>
>
>------------------------------------------------------------------------
>
>_______________________________________________
>INRAD mailing list provided as a service by International Radio
>INRAD Web Site: http://www.qth.com/inrad
>
>To Submit Message to the List: list at inrad.net
>List Help:
>http://mail.inrad.net/mailman/listinfo/list_inrad.net
>  
>