Keyboard Shortcuts
ctrl + shift + ? :
Show all keyboard shortcuts
ctrl + g :
Navigate to a group
ctrl + shift + f :
Find
ctrl + / :
Quick actions
esc to dismiss
Likes
Search
BandPass Filter Relay Contacts
|
I wasn't sure where to ask this question, but since I found the problem using my NanoVNA, I thought I'd ask here.
Our club has a multi-band Bandpass filter from ICE, Model 419 that we want to use on Field Day. This unit covers 160, 80, 40, 15 and 10 meters. Complete Unit: Internals: !Ats-2fm84F1OgaUfhmHlYIjlktAx5g?e=JcS9i3 I tested the unit with my NanoVNA and found that the loss varies greatly by simply tapping the unit. Through DC testing, I have determined that the unit has a bad 15m relay where it should be passing the signal directly through (bypass the 15m coils) through the Common to NC contacts. It¡¯s the relay in the upper left corner, next to the small green cap. I¡¯m seeing anywhere from 3 or 4 dB loss at BEST and up to 30 dB loss through the NanoVNA. The loss varies as you tap the unit. When I measure the loss using my transmitter (tune position, 10 watts, into a dummy load with a wattmeter) I¡¯m seeing: 160m -.432 dB loss.. 80m -.414 dB 40m -.914 dB 20m -.801 dB 15m -.964 dB 10m -1.10 dB The faulty contact is only a problem at lower signals, such as the NanoVNA and receive signals. The way the circuit works is that the signal comes in on the input coax connector, and when all relays are disabled (NC position) the signal goes through the NC connections of each relay. When you switch in any one band, that relay is pulled in (NO Contacts) and the signal goes through the Band Pass coil/capacitor network. So, it appears that the 15m relay is intermittent. I tested across each relay in the NC position with an digital multi-meter, looking at the voltage drop across the contacts. Because of the construction practices used, it¡¯s impossible to remove the relays to clean the contacts. My ideas are: 1. Drill a hole in the plastic and spray (lightly) some contact cleaner 2. Remove the coil/capacitors on that relay and try to get the relay out, clean the contacts and replace everything, hoping that the band pass tuning doesn¡¯t change. I¡¯ve been trying to find a way to clean the contacts by passing current through them, but I haven¡¯t come up with much. I tried to contact the manufacturer, but they haven¡¯t built this unit for over 12 years and they have no idea what to do. Anyone have any ideas? -- Mike, W0IH |
|
Bob Albert
That is a common problem with relays.? The only two answers are somehow clean the contacts or replace the relay.? Cleaning is a temporary solution; they will cause trouble again.? And it isn't dirt; it's corrosion and/or carbon buildup.? The limiting cause is the spring tension that holds the NC contacts closed.? And the relay design, which is probably not self wiping.
toggle quoted message
Show quoted text
You are probably going to have to bite the bullet and remove the relay.? Squirting contact cleaning isn't likely to do much good.? Since the problem is more severe with low current, that implies oxidation which breaks down due to the heat of high current. Bob On Wednesday, May 25, 2022, 01:53:14 PM PDT, Michael Foerster <mcfroston@...> wrote:
I wasn't sure where to ask this question, but since I found the problem using my NanoVNA, I thought I'd ask here. Our club has a multi-band Bandpass filter from ICE, Model 419 that we want to use on Field Day.? This unit covers 160, 80, 40, 15 and 10 meters. Complete Unit:? ? Internals:? ? ? ? ? !Ats-2fm84F1OgaUfhmHlYIjlktAx5g?e=JcS9i3 I tested the unit with my NanoVNA and found that the loss varies greatly by simply tapping the unit.? Through DC testing, I have determined that the unit has a bad 15m relay where it should be passing the signal directly through (bypass the 15m coils) through the Common to NC contacts.? It¡¯s the relay in the upper left corner, next to the small green cap.? I¡¯m seeing anywhere from 3 or 4 dB loss at BEST and up to 30 dB loss through the NanoVNA.? The loss varies as you tap the unit. When I measure the loss using my transmitter (tune position, 10 watts, into a dummy load with a wattmeter) I¡¯m seeing: ? 160m ??? -.432 dB loss.. ? 80m??? -.414 dB ? 40m??? -.914 dB ? 20m??? -.801 dB ? 15m??? -.964 dB ? 10m??? -1.10 dB The faulty contact is only a problem at lower signals, such as the NanoVNA and receive signals. The way the circuit works is that the signal comes in on the input coax connector, and when all relays are disabled (NC position) the signal goes through the NC connections of each relay.? When you switch in any one band, that relay is pulled in (NO Contacts) and the signal goes through the Band Pass coil/capacitor network. So, it appears that the 15m relay is intermittent.? I tested across each relay in the NC position with an digital multi-meter, looking at the voltage drop across the contacts.? Because of the construction practices used, it¡¯s impossible to remove the relays to clean the contacts. My ideas are: 1.??? Drill a hole in the plastic and spray (lightly) some contact cleaner 2.??? Remove the coil/capacitors on that relay and try to get the relay out, clean the contacts and replace everything, hoping that the band pass tuning doesn¡¯t change. I¡¯ve been trying to find a way to clean the contacts by passing current through them, but I haven¡¯t come up with much. I tried to contact the manufacturer, but they haven¡¯t built this unit for over 12 years and they have no idea what to do. Anyone have any ideas? -- Mike, W0IH |
|
drill hole and try contact cleaner?? NO NO that will not help
toggle quoted message
Show quoted text
but you can open the relays and pull a piece of paper through the contacts to clean them a bit if dry paper is not enough ... a small drop of cleaner (NOT SPAYPAINT THE RELAY WITH IT) ... then pull some more paper through the relay contacts till they are shiny and dry again contact cleaner alone will not help much but polishing the contacts with paper will do for sure also check how much contact force is on the nc contacts dg9bfc sigi Am 25.05.2022 um 22:53 schrieb Michael Foerster: I wasn't sure where to ask this question, but since I found the problem using my NanoVNA, I thought I'd ask here. |
|
It's pretty normal for relays and switches to get oxidized with time and no use. Just using them with some power through them will probably clean them. The first
review said you can put 300 watts through that filter, I would try that. Good low level equipment that uses relays sometimes uses a "DC wetting" current to keep the relays clean. 73, Gary WB6OGD |
|
Mike, the core problem with relays in such applications is that there are no relays that can handle both large and small signals well!
Relays are available with "power" type contacts, or with "signal" contacts. Contacts intended to carry high current need to have high conductivity, and if they need to switch high voltage they need to resist arcs without burning or melting off too fast. Typically silver alloys are used for such contacts. But these do oxidize. The oxide layer is no problem for power applications, because the high voltages and currents will simply punch through it. But this layer becomes an insulator, or an attenuator, or even a distorting semiconductor junction, to small signals! For that reason relays intended for small signals usually have gold-plated contacts. The gold doesn't tend to oxidize, and gives good, linear contacts even with very small signals. But it cannot handle high currents! So it cannot be used very well in power relays. In ham applications, that involve moderately high current during TX, and very small signals during RX, but usually involves no hot switching and thus no arcs, probably large gold-plated contacts would be best. But such relays would need to be specially made! Normal power relays all have power contacts, not gold-plated ones, and gold-plated relay contacts are usually very small, unable to carry TX currents. It's a real problem. Power type relays with wiping, self-cleaning contacts, as mentioned by Bob, might be a good choice, but even these don't give a guarantee that they won't develop small-signal contact problems. And "wetting" with a DC current, mentioned by Gary, is an old trick that also does work to some extent, but implementing it needs additional chokes and capacitors. My own design approach is avoiding relays for combined large and small signal switching, whenever I can. But in filters like the one you have there, relays are hard to avoid... The alternative is PIN diode switching, but it's far more expensive. Manfred |
|
Manfred, the problem is that I'm not designing this circuit, just trying to get it to work more reliably.
I did get an email from Steve, AK4R that pointed to an article by W8JI on techniques to clean relay contacts. W8JI suggested a "safe electrical cleaning" method of passing about 2 amps of current (using a 12v AC or DC supply) through the contacts. I'm happy to report that this appears to have worked very well. I setup a 12v power supply with some limiting power resistors to about 1.9A and ran it from the input to the output. My initial voltage drop tests (with a DVM) show that this worked quite well. Thanks for the reference Steve!!! -- Mike, W0IH |
to navigate to use esc to dismiss