RFI on audio when using Left Line In RCA 7000 DLE MKII

[N8RWS]

Hi,

My rig uses the EP4CGX150F896 Cyclone IV series FPGA

I am running rack audio gear through a JDI Passive Direct box then to the left line in RCA jack on back of rig. With this configuration I get a very slight RF echo on my transmitted audio.

If I connect the exact same audio chain to the mic jack on front of rig there is no RF on the audio.

The RF on the audio occurs when the 7000 internal PA is set above 10 Watts or so and when I turn on my amplifier the RF level does not increase.

I have performed extensive RF shielding in the radio (cleared paint from contact points front & rear covers, Bound center plate to solid ground on case, installed ferrite chokes on all ribbon cables and even input a bypass cap on the left audio in line circuit just before it hits the codec pin.

Additionally I have tested with different mic pre amp and several different mics. I was running protocol 2 version 2.2.2 (for my version FPGA) and switched to protocol 1 and then back to P2 2.1.18 along with setting a brand new database.

I believe the problem is being generated within the radio itself and I am not sure if it is RFI or some kind of database / firmware timing issue.

This is one of those things that really should not be a focus for me considering that I can input the audio to the mic jack on front and it works great .... but ... things like this pose a challenge to me and I want to solve it.

I attached a recording a friend made of me "hissing" into the mic.. the slight echo after each "hiss" is what I am working to eliminate.

Jeff RF in Larry's recording 12 4 2025(1).wav

(746.7 KiB) Downloaded 25 times

Wondering if anyone else has run across this kind of issue and or has any ideas for a solution.

Thanks

Jeff

[w-u-2-o]

Echo or reverb on the transmitted signal is 99.999% of the time related to RFI being fed back into the mic input. Depending on what processing is set up in Thetis, this is because the latency from mic input to antenna output can be tens to a hundred or so milliseconds.

Unless something has come loose inside the 7000, it is extremely unlikely that the RFI is occurring inside the unit. The only design to date that has a known internal problem is the first generation G2's, which did not have proper voltage regulation on the balanced mic input amplifier.

It is not a software or firmware timing issue. If it was, then there'd be a great many reports of such behavior, and there are none that I've ever seen.

"Hissing" sounds are not terribly easy to discriminate. Why not use a pulsed tone generator? There are plenty of excellent smartphone app's for generating those.

You say you've done "extensive RF shielding". Have you tried the simple expedient of adding a ferrite on the cable where it enters the line input on the 7000?

To avoid these issues entirely, and to eliminate the low-end mic CODEC IC from the audio signal chain, have you considered using a high quality USB audio interface that supports the ASIO driver? This will deliver the audio directly into Thetis rather source it from the CODEC in the 7000 over the network.

[N8RWS]

Thank you for the reply !!

I understood all points and will use pulsed tone in future !!! And yes I used several ferrite's on that line input cable !

I took your advice on using a USB audio interface. I am now using the balanced line input on my Focusrite 6i6 usb interface & the ASIO driver to introduce the rack audio to the computer. I then configured VAC in Thetis and this it solved the problem completely. No more conventional cables required for receive or transmit audio with my 7000 !!!

Thanks again for the support, I really appreciate it.

Jeff

Echo or reverb on the transmitted signal is 99.999% of the time related to RFI being fed back into the mic input. Depending on what processing is set up in Thetis, this is because the latency from mic input to antenna output can be tens to a hundred or so milliseconds.

Unless something has come loose inside the 7000, it is extremely unlikely that the RFI is occurring inside the unit. The only design to date that has a known internal problem is the first generation G2's, which did not have proper voltage regulation on the balanced mic input amplifier.

It is not a software or firmware timing issue. If it was, then there'd be a great many reports of such behavior, and there are none that I've ever seen.

"Hissing" sounds are not terribly easy to discriminate. Why not use a pulsed tone generator? There are plenty of excellent smartphone app's for generating those.

You say you've done "extensive RF shielding". Have you tried the simple expedient of adding a ferrite on the cable where it enters the line input on the 7000?

To avoid these issues entirely, and to eliminate the low-end mic CODEC IC from the audio signal chain, have you considered using a high quality USB audio interface that supports the ASIO driver? This will deliver the audio directly into Thetis rather source it from the CODEC in the 7000 over the network.

[w-u-2-o]

Even better than VAC is cmASIO. Take a look in the directory where Thetis.exe lives and you'll find a document on how to use the cmASIO interface.