Useful information regarding antennas for SDR products.
Posts: 318
Joined: Mon Jul 24, 2017 8:24 pm


Post by Mike2459 » Wed Jul 17, 2019 3:53 pm

Years ago I purchased a new Akai reel to reel stereo tape deck. It had many features. One of them was auto-reversing. It enabled playback of a full tape, both sides without the need to remove the full reel flip it over and place on the right capstan, and rethread to the empty reel. To accomplish this a relay was used to reverse the audio channels due to the arrangement of the 4 channels on the tape. After several months of ownership the sound began dropping out at low audio levels when playing the reverse side of the tape. A major annoyance. I finally gave up and began playing the reverse side the old fashioned way as outlined above. When I returned to the US several months later I decided to try to fix this problem. It turns out it was due to dirty contacts on a relay that reversed the audio channels. I presume the audio levels switched by this relay was at line levels*, it may not have been. But let's assume it was. It was likely much higher than the RF voltages encountered at a receiver's input. This is the reason for my concerns about using a GP relay in this application.
* A line input level electrical signal typically has a voltage ranging from 0,3 to 2 Volts, while a microphone level signal is more often in the range from 5 to 50 mV (millivolts). Microphone sensitivities range from -60 dBu to -22 dBu referenced to 94 dB Soud Pressure Level (0 dB SPL = 2*10–5N/m2). The consumer line input level electrical signal typically has a voltage of 0,32 V (-7,8 dBu), whilst the professional line input level is typically 1,23 V (+4 dBu).
https://www.kfs.oeaw.ac.at/manual/3.8/h ... de/461.htm
Akai GX285D copy.jpg
Akai GX285D copy.jpg (80.18 KiB) Viewed 1460 times

Reason: No reason

Posts: 128
Joined: Sun Dec 09, 2018 4:38 pm


Post by sdrom33 » Wed Jul 17, 2019 6:53 pm

HI Mike,

Your experience is very informative and clearly shows the risks and pitfalls in circuit design. Reliability analysis, based on MIL specs, was done in the company I was working for several years and the most difficult process was that of finding an optimum balance between component count and reliability. To make it simple: if in order to avert the danger of one hypothetical failure you increase the component count you may end up with a less reliable circuit anyway. So, according to my experience, to avert the danger of failure due to dry current, we are now increasing the component count by 5: two more capacitors, two more resistors and one choke and we are not even sure that the relay will fail, or, better, we do not know how long it will take for the relay to fail. This is a very good example of what I mean.
Furthermore, what problem do we have if the relay fails? Nobody is going to die, no battle in a war is going to be lost, so why increase the complexity and risk a failure in the long term because we increased the number of components? At the worst we shall just replace the relay, when and if it fails.

In its simplicity the case we are analysing proves the necessity of having an excellent circuit designer, because, at the very end of the story, it is his ability that can insure that precious, all important balance.


Reason: No reason

Posts: 662
Joined: Thu May 10, 2018 6:42 pm
Location: Piedmont, Italy


Post by glovisol » Mon Jul 22, 2019 4:57 pm

Now that luckily a lot of confusing information on this thread has been removed, as a service to those who did not follow the cancelled discussion and may be worried by the bits of controversy left over, but still need to build and use the Antenna Swich in question, I can confirm that this switch will work perfectly well and with excellent reliability.

The doubt that was introduced is that since the relay contact, when making and holding in the "receive" position, would not pass any current (this is called a relay contact with "dry current" and it is known by experience that sometimes contacts not passing any current may fail and present high resistance because of surface oxidation with age) then the relay would fail.

While the above is true in the general sense, we must put reasonable boundaries to our problem, these being: (a) projected lifetime of the device and (b) relationship between cost and lifetime.

The relay used in this project is a general purpose encapsulated relay having the following characteristics:

Manufaturer: Songles
Contact material = Silver alloy AgCdO、AgSnO2、AgNi
Contact resistance = 100mΩ(1A 6VDC)
Sealed construction shown by the -SL- designation.

Sealed means that the relay is potted, so the contacs are well protected by external atmosphere. It was stated that these relays have a small hole in the case, but this detail is not true for the relay in question and I doubt this to be true in general, as the hole would defeat the purpose of encapsulation. This relay, complete with driver circuit and mounted on a PC board retails for $1.50 on Ebay.

Even though "potted" is not "hermetically sealed", still this enclosure effectively insulates the contacts for a period of several years. Apart from this, the shelf life of these relays, with the contact alloys listed above, has a rating of 200,000 hours, or in excess of 22 years. While the relay is stored, no current flows through the contacts, but the contacts still exhibit the rated low resistance throughout the rated life. So at the first instance, we can project a lifetime of 22 years.
If we now consider the number of switching operations, this GP relay is rated for a minimum of 100,000 and it can be easily calculated that, even with a large number of operations, the relay will last in perfect operation for a minimum of 4 years . Furthermore, when and if the relay will fail, the only detrimental effect will be de-sensitisation of the companion SDR receiver (but no damage to it) and a replacement relay (costing $1.50) will solve the problem.

Doubt was also passed on the insertion loss of the device. Of course this parameter has been carefully tested in the course of the construction phase and has been measured to be less that 0.2 dB in the range 0.5 to 30 MHz.

Finally it is true that a more expensive relay, with Silver-Palladium contacts and hermetically sealed will last longer than 22/4 years, but certainly at a significantly higher cost, so my conclusion is that the antenna relay under scrutiny is a perfectly good design, not requiring any "improvement" whatsoever.
Last edited by glovisol on Thu Jan 01, 1970 12:00 am, edited 0 times in total.
Reason: No reason

Post Reply