sahib
Well-known member
Indeed. Relying on the user. Input and output (TRS) connectors are labelled as inputs and outputs so that user does not try to use it bidirectional. Same applies for relay based mux/demux.
Has anyone ever thought about designing a DIY Digital Patchbay?
- Thread starter canidoit
- Start date
Help Support GroupDIY Audio Forum:
Brian Roth
Well-known member
On the routers (Sigma and Sierra) I am familiar with. the "only one source to one destination" was enforced by the user control panel design. You would first select the destination channel, then the source channel, then press the "Take" button. Any previously selected source would be cleared from that output and the new source connected.
Some panels would have buttons for just a single output.
With video, summing multiple sources was a no-no in the routing system itself. Any required mixing was done in the specialized "production switchers", also called "vision mixers" in Europe. The sales brochure posted below shows an example.
Bri
Some panels would have buttons for just a single output.
With video, summing multiple sources was a no-no in the routing system itself. Any required mixing was done in the specialized "production switchers", also called "vision mixers" in Europe. The sales brochure posted below shows an example.
Bri
Matador
Well-known member
Does anyone know of a chip that supports the signal levels one might commonly find inside a guitar amplifier? Like north of the 20Vpp signal range with reasonable (60-80dB) isolation in the off position?
Brian Roth
Well-known member
I wish I knew as well! That's telcom mechanical relay territory. Or, otherwise some sort of discrete component design for each crosspoint. Nothing available via any IC to my knowledge.
Bri
Any of the chips which can run from bipolar +/- 15V supplies should be able to handle 10V peaks with no problem.
That AD8113 chip discussed earlier in the thread can just about do that running from +/- 12V supplies, but I would assume you probably don't need 16x16 crosspoint inside a guitar amplifier.
Analog Devices has analog mulitplexers with dual supplies up to +/- 22V, you could check some of those:
https://www.analog.com/en/parametricsearch/12829#/
Without knowing how many ins and outs you are looking for hard to suggest anything more specific.
Brian Roth
Well-known member
As I hinted earlier, it's been eons since I looked into FET switching components.
I looked at the Analog Devices matrix linked above and found nothing like a crosspoint chip. Although now I know where all the "DG" chips ended up..lol. The actual AD crosspoint chips are apparently part of a different product line.
https://www.analog.com/en/product-category/buffered-analog-crosspoint-switches.html#category-detail
I actually have a reason to investigate crosspoint choices now. I need something like (minimum) 6x6 and preferably 8x8....but multiplied by four ....as in two stereo sets.
Bri
Matador
Well-known member
In theory, a generic CMOS pass transistor built from common medium voltage MOSFETS would form the basis of a good analog switch.
The rub with getting them to work properly in a discrete design is that level shifting is necessary so that you can switch off the upper P-channel devices when the switch is 'off'.
The common cross-coupled level shifter above doesn't work because PMOS devices with gate-source breakdown voltages above 20V are rare and expensive, and that limits the effective voltages you can pass through the switch.
I suspect that most analog switches must be designed with tiny DC/DC converters to generate the upper voltage that is 10V or so lower than upper rail so that the PMOS can be switched fully on, but not exceed the gate-source breakdown voltage.
Seems to me the selling point of the Flock is the software. Instant and easy routing and recall. Otherwise why not just use an analog patchbay? I would think any DIY or repurposed commercial solution would need similar software to be worthwhile and useful.
Brian Roth
Well-known member
Since Flock has been mentioned several times in this thread I looked around on their website for more details. Alas, the site is heavy on fancy web design and lacking in technical details (as in any spec sheets...unless I somehow missed them).
The only "tech tidbit" I found was a mention of something like "+4 dB" level. Seems strange they also included switchable 48V on each input.
I noticed a US patent number in one of their user docs. It's somewhat interesting to read, and mentions the possibility of using the AD8113 chip.
The Flock products look very nicely designed and probably have very a good user interface (never messed with it myself). However, I'm puzzled about their patent being granted on what would seem to be "prior art".
Patent copy (direct from the USPTO) attached below. The final page has the four claims for the invention.
Bri
The only "tech tidbit" I found was a mention of something like "+4 dB" level. Seems strange they also included switchable 48V on each input.
I noticed a US patent number in one of their user docs. It's somewhat interesting to read, and mentions the possibility of using the AD8113 chip.
The Flock products look very nicely designed and probably have very a good user interface (never messed with it myself). However, I'm puzzled about their patent being granted on what would seem to be "prior art".
Patent copy (direct from the USPTO) attached below. The final page has the four claims for the invention.
Bri
This may have been mentioned, but I believe the Flock unbalances, routes, and then re-balances the signal. I used one recently and it sounds fine. The software was a little finicky (sometimes patches would drop out until I de-selected and re-selected the patch. And the unit has a weird “sleep” mode that was confusing at first. It is very convenient and relatively affordable.
Brian Roth
Well-known member
Indeed, the patent describes the unbal/route/bal sequence. I would believe that any analog audio router (I'm thinking of those used in broadcast applications for decades) also does that process. It makes sense...<g>.
My reason for even bringing up the patent is to avoid accidentally infringing.
Bri
My guess is the patent wasn’t actually reviewed for “originality.” It’s issued and then if someone wants to challenge it, the USPO will review it (for a hefty fee). Been there… no fun… patent trolling abounds
soapfoot
Well-known member
We have a couple of the Flock crosspoint switch arrays in our NYC mix room.
I've never been able to distort them (or 'hear' them at all), FWIW.
Obviously we're not torture-testing them... we're just mixing.
I'm sure they do have a threshold at which they do clip, but for all practical purposes (in mixing use cases) I'm just not sure it's a real-world issue.
pucho812
Well-known member
At the tech brekkie this am there was discussion of fix audio pro patch. A 64 x 64 patchbay that is controlled via software. It contains 1500 relays in it.
That’s a whole lot of relays.
That’s why they are so expensive regardless of who is making them.
That’s a whole lot of relays.
That’s why they are so expensive regardless of who is making them.
Last edited:
Yup lots of dodgy patents out there. The (cheap?) patent lawyer Peavey used for my inventions once suggested not even doing a patent search, asking who wants to find prior art..? he was just trying to rack up wallpaper patent wins (in hindsight maybe thats why Uli was able to get around my most valuable patent , FLS so easily).
IIRC patent law suggests that it is the responsibility of the inventor to inform the patent office if he learns about prior art. Maybe show him examples and put the ball in his court.
[edit- what is the patent issue date if its old enough it might be public domain soon or already /edit]
JR
PS; I learned after the fact that one of my (automatic mixer improvement) patents was covered by prior art, but I discovered this after I was no longer employed by Peavey and nobody was disputing the patent so I just let it be. Why open up that can of worms?
Last edited:
Similar threads
