Dynaudio BM6A RF interference [FIXED 100%]

[mhelin]

Isn't it also possible that the various diodes like the input protection or the power amp biasing diodes on the circuit work as detectors?

 

[Kingston]

I have the BM5A mk ii and it doesn't seem to have this issue. Maybe comparing them might reveal something?

You don't know this. You don't live with a 4G cell tower pointing at you. I didn't know this either until we moved.

We have a bunch Genelecs or various ages and models, and all kinds of TV soundbars and such. None have this issue. But for example all modern Genelecs are basically aluminum cannonballs. Nothing can get to them. Whether RFI or meteor attack.

 

[Kingston]

I killed the noise 100%.

All it took was that 0.7mm aluminum faraday cage. I made an aluminum covering for the PCB components section. Looks a bit like some old ATX power supply now. I wanted to use even thicker aluminum, but took a gamble with 0.7mm because it allowed to use Aviation Snips and the sheets can be bent to shape with bare hands. I hate dremeling etc. sheet cutting with power tools because they make such a mess. The snips made the job a breeze with high accuracy.

There is no pin 1 problem. The plastic XLR connectors are not a problem. I had to cut some through-holes in the cage for cabling, no problem with this either.

I'll post some pictures later.

Oh, as a bonus, the very minimum amount white noise that I had always thought were just inherent in the design also went away. It could be only heard up close and was always a non-issue. But now it's also gone.

 

[Cqwet Dbdfte]

Thanks for the update.
RF can enter at any point and get rectified and "detected" as AM radio.
A very good Faraday cage (Lindgren ETS) would have 80dB of attenuation but in this case the level only has to be below the PN junctions threshold.
At the PCB level it is easy to design-in shields, but hard to retrofit.

 

[Kingston]

Minor update. The second monitor cage had some gaps, as retrofitting this thing was not exactly ideal and it was hard to get it right. I nailed the first one, but the second one need some additional hacking with aluminum tape etc. chewing gum level gap balancing.

I wish Dynaudio had done this right to begin with. Really annoying to snuff it out post mortem.

 

[Cqwet Dbdfte]

Back when those boards were designed RFI was not much of an issue.
Expect it to get worse.
New equipment has to comply with a raft of updated standards.
Carbon loaded foam can be used to attenuate RF, but aluminum sheet is less costly. Tin plated steel would be better, it can be soldered together.
McMaster has new tin plated gallon jugs that could be repurposed.
Retrofitting an old design is not fun nor easy.
Finding the actual point of susceptibility with RF injection and adding filtering would be less brutal.

 

[Kingston]

Finding the actual point of susceptibility with RF injection and adding filtering would be less brutal.

Even in this case a Faraday cage would be recommended. I don't understand at all why the original designer neglected this. It was not price, since this was never a cheap monitor - in fact it was from the higher end of things all along. It's not like radio stations weren't a thing when this was on the drawing board, and that a professional monitor wouldn't be sold to environments right under the broadcast mast.

Speaking of design were someone did not want to second guess the hostile world out there. I have Teac V-3010 casette deck from '92. All its internals, support structures, everything is made of copper. It's absolutely beautiful! But also, imagine justifying that cost in any modern design. Back then copper and aluminum were side by side in the do-whatever-you-please-with-this-cheap section.

 

[Kingston]

pics or STFU

The drawing board with initial sheet cut:


Being shoe-horned in place, before through-hole cut in the middle of roof section:


About to be slotted in, right before last aluminum tape here and there. Nothing else was used for fitting besides aluminum tape. Screwed in tight, I ensured everything has good chassis ground connection.

Also visible the heavily modded BM6A mk1 board with LT3081/LT3091 regulation, additional improved smoothing cap fitting for the opamps, modern opamp replacements, and pointless but impressive paper-in-oil caps where ever they did fit. I have a garage filled with Russian/Ukrainian paper-in-oil caps from before the war when they were dirt cheap. And since about a decade I try to think about the most improbable and more and more pointless uses for them.

Additionally visible all the way in the back the kitchen grade aluminum foil that I had already installed, that helped a bit, but am too lazy to remove. No one will ever know.

Further, reflex port is blocked. These are used with a large subwoofer with bass management. They go to 80hz only when managed. Blocking the reflex removes the useless one-frequency fake boost, reduces efficiency a tiny but, but most importantly gives much desirable additional mid bass accuracy. Not a fan of bass reflexes. At all.



Needless to say this isn't the Dynaudio as your mom used to make them. It's a different sound and accuracy level. And back to silent noise floor as well.