Well, I found my ICD2 clone ZIF socket adapters, so I can easily program lots of PICs. They'll be done one at a time, but at least I won't risk the parts by having to pop them in and out of sockets. So, it'll be easy for me to program the uCs for the group buy. Yay!
Now I just have to wait for the boards to come in... I'm hoping that they'll be here by next weekend.
-Steve
Now I just have to wait for the boards to come in... I'm hoping that they'll be here by next weekend.
-Steve
An externally hosted image should be here but it was not working when we last tested it.
Well, it's no gang programmer, but combined with the latest version of the PICkit 2 software (adds program-on-button-push) it should make flashing all the PICs for the group buy easier on me. There's also far less risk of damage to the PICs or pins themselves.
An externally hosted image should be here but it was not working when we last tested it.
Not long after arriving home this afternoon the doorbell rang, and the UPS driver delivered the anxiously awaited package of nuxx Audio Input Switch PCBs from China.
Well, to keep a short story short, they work as expected. I haven't fully wired mine in yet, but the board accepts and rectifies / regulates power just as I expected, and everything seems to work well. The v1.0-RC of the software works great, so I think it's going to become the full-on v1.0. I guess now I just need to do some more testing (just to be sure) and then start sorting out the pricing for the group buy.
(Yes, I did forget one of the screw terminals... I stupidly ordered five when I should have ordered six. That said, I'll only be using three of the inputs on my board anyway, so it doesn't matter that much.)
If you'd like to see more photos of the board and such, please look here. Of course, there is also the full nuxx Audio Input Switch page available as well.
-Steve
This board looks very nice., great work.
I'm not the best at circuit design(more of a software guy) but I had an idea that I thought I would run by you guys. Was curious if this would work using a sort of "binary tree" of double throw relays. So, one relay could switch between two sources. Put a relay on each of the "branches" of the first relay, then 3 relays could switch up to 4 sources, etc. You could switch 5 sources with 4 relays(the tree would not be perfectly symmetrical in this case). Basically, it saves you one relay as opposed to one relay per input. Don't know if it would really work well with the rest of the circuit design though, or if the advantages outweigh the disadvantages.
Just curious what you guys think.
I'm not the best at circuit design(more of a software guy) but I had an idea that I thought I would run by you guys. Was curious if this would work using a sort of "binary tree" of double throw relays. So, one relay could switch between two sources. Put a relay on each of the "branches" of the first relay, then 3 relays could switch up to 4 sources, etc. You could switch 5 sources with 4 relays(the tree would not be perfectly symmetrical in this case). Basically, it saves you one relay as opposed to one relay per input. Don't know if it would really work well with the rest of the circuit design though, or if the advantages outweigh the disadvantages.
Just curious what you guys think.
I'm familiar with the concept it's used a lot in semiconductor ATE test fixturing, but one source is generally always connected (no big deal but no mute if default source is on) but the big issue for us purists is that it introduces additional contacts in the signal path, and the ones at the top of the tree have lots of extra contacts.
Contacts unfortunately can be very unkind to low level signals, and keeping them to a minimum is considered good audio design practice which is why no one (to my knowledge) uses a binary tree in audio.
Contacts designed for low level switching have special surface treatments and are expensive, even then they can add some measurable distortion particularly as they wear out, and a bunch cascaded add to the effect. The other concern is if dc should appear on the output of a component and appreciable current flows through the relay contacts as they are switched. (Some high performance reed relays will start to degrade when switching as little as 10mA.)
Contacts unfortunately can be very unkind to low level signals, and keeping them to a minimum is considered good audio design practice which is why no one (to my knowledge) uses a binary tree in audio.
Contacts designed for low level switching have special surface treatments and are expensive, even then they can add some measurable distortion particularly as they wear out, and a bunch cascaded add to the effect. The other concern is if dc should appear on the output of a component and appreciable current flows through the relay contacts as they are switched. (Some high performance reed relays will start to degrade when switching as little as 10mA.)
kevinkr said:
Hi Steve,
Care to elaborate? Your fans want to know..
More seriously though, any particular issues?
By the way the site you created for the switch is great.
Kevin
Heh, thank you. The problem I'm having is that when the device is muted all the traces are acting as a giant antenna and I'm hearing a variable buzzing from the output of the amp. I'm also hearing something similar on the inputs.
On top of that, the voltage regulator is getting much hotter than I expected when feeding the rectifier 24VAC, which is the only power available inside the amp I intend to use the board in.
Finally, my original intention was to use the 5V out to power things like a USB audio adapter inside the amp enclosure, but doing that creates a ground loop.
So, the first two things are going to take some thinking. The first... I'm not sure how I'll deal with that yet. Maybe tie the output to ground when the device is muted (this will take another relay) and... Something with the inputs. Whatever needs to be done here will take a bit of learning on my part.
The second may involve a combination switching regulator and then LDO. This would sure cut down on heat and likely eliminate the need for a heatsink.
The third issue is really a non-issue with my design, and is something which should be sorted out in implementation. It won't affect any users of the board unless they are using the 5VDC connector as an output and don't take standard measures to eliminate ground loops.
I'm taking a slight (probably through the weekend) break from working on this all, just because there's a combination of suddenly added responsibilities in my real-life job, and I need to spend some time this weekend getting familiar with OS X 10.5. Then it'll be back on to sorting out the lingering issues with the boards.
So, just hang tight... I've got four unused PCBs here, but I'm hesitant to make them available to anyone unless they really want them for experimentation and whatnot. If you do really want one, it'll be... about $16/each, plus a couple of dollars for PayPal fees and postage. (My cost was $78 for five boards. This would not also include a preprogrammed microcontroller, simply because I only have my three original samples and can't spare any others.)
Personally, I'd recommend waiting for the next version if you want to use one of the boards in an audio device without doing some modifications to it. Still, if you want one as they stand now, let me know.
Bye for now...
-Steve
Just so everyone knows, I ordered the parts to prototype a switching power supply. If this works I should have the overly hot (when powered from 24VAC) regulator issue sorted out.
I'm also looking at having the mute function disconnect the inputs and tie the outputs to common. That should make the amp see no signal and head off any noise while muted.
I'm still not sure what to do about noise when an unconnected input is selected. I welcome any suggestions.
Hopefully the next set of parts will be here by on Monday or Tuesday, and then I can (quickly) figure out the power supply. The new power supply design will increase the cost of completing the board around $8, but it will (hopefully) be a better overall design.
-Steve
I'm also looking at having the mute function disconnect the inputs and tie the outputs to common. That should make the amp see no signal and head off any noise while muted.
I'm still not sure what to do about noise when an unconnected input is selected. I welcome any suggestions.
Hopefully the next set of parts will be here by on Monday or Tuesday, and then I can (quickly) figure out the power supply. The new power supply design will increase the cost of completing the board around $8, but it will (hopefully) be a better overall design.
-Steve
im interested in a group buy for the boards and microcontroler as well. I have a huge need for this product and thought I would have to use a delay circuit plus selector switch on a relay bank. Your project is far more elegant than I can design and if I have the opportunity to benefit from your hard work, I would like to do so.
Something we do in car audio to diagnose noise problems is using shorting plugs on audio inputs. Could that be applied in this case?
Something we do in car audio to diagnose noise problems is using shorting plugs on audio inputs. Could that be applied in this case?
c0nsumer said:Just so everyone knows, I ordered the parts to prototype a switching power supply. If this works I should have the overly hot (when powered from 24VAC) regulator issue sorted out.
I'm also looking at having the mute function disconnect the inputs and tie the outputs to common. That should make the amp see no signal and head off any noise while muted.
I'm still not sure what to do about noise when an unconnected input is selected. I welcome any suggestions.
Hopefully the next set of parts will be here by on Monday or Tuesday, and then I can (quickly) figure out the power supply. The new power supply design will increase the cost of completing the board around $8, but it will (hopefully) be a better overall design.
-Steve
Adding a switcher in audio will give you ( probably ) less heat but more radiated noise. Why not invest some time in rethinking the concept for plain quality ? Just adding a small extra transformer is what you really need. Without wanting to be negative: I think the emphasis is too much on digital/PIC side of the device and not enough on real life performance without negative side-efffects.
Adding a small transformer leaves you the noise question which needs good investigation. Reference to ground of all inputs is also needed. I can assist you via PM if you like.
jean-paul said:
It likely will add noise, although I'm not sure how much yet. The parts to prototype that were only around $10, so I'm giving it a try. I don't think the transformer is the best idea, though, as that will only step down AC. If one wants / needs to power the board from more than 5VDC (which I imagine many people will) then a regulator will be needed, which brings me back to the original problem.
For what it's worth, I'm intending to work on filtering the output of the PS so that the 5V going to the relays has as little noise as I can get.
Do you have any suggestions for noise on non-connected inputs? whiterabbit's idea of shorting plugs (terminators?) works, but it just doesn't seem as elegant as it could be. But, it's simple and it'll work great... I may just go that route.
-Steve
Just adding a small 9 to 12V transformer of let's say 250 mA will do. The bridge and the cap + reg are already on the PCB. Problem solved. The reason why the reg is getting hot is because you feed the board with 24V AC. 24V AC will give around 36V DC so the reg has to deal with 36 - 5 = 31V. No wonder it gets hot !
I have enough ideas but I prefer to do so via PM. Shorting the inputs with terminators is not a good idea if you want the device to be equal or better that normal gear. Maybe better put this way : would you buy a preamp that needs input terminators ?
I have enough ideas but I prefer to do so via PM. Shorting the inputs with terminators is not a good idea if you want the device to be equal or better that normal gear. Maybe better put this way : would you buy a preamp that needs input terminators ?
jean-paul said:
Yes, of course. But by adding a transformer it'll decrease flexibility of power sources and thusly the utility for others. I'm not designing this board simply for my own use.
Besides pulling the switching supply and microcontroller out, do you have any ideas as to how said noise can be better defeated?
-Steve
Adding a transformer increases flexibility when you also add pads on the PCB for connecting plain DC from the device the PCB is going to be built in. If the device has a too high AC voltage one can add a transformer for the NUXX. If the device has let's say 12V DC the NUXX can be powered directly from the device. The K.I.S.S. approach
Pulling the microcontroller out is quite a measure of which I thought you wouldn't want to consider. You would end up with a simply switched input board just like the Elektor input boards. Those are well designed by the way, I use those now. If you don't mind skipping the microcontroller you'll have enough space on the PCB for doing things a very good way.
Pulling the microcontroller out is quite a measure of which I thought you wouldn't want to consider. You would end up with a simply switched input board just like the Elektor input boards. Those are well designed by the way, I use those now. If you don't mind skipping the microcontroller you'll have enough space on the PCB for doing things a very good way.
Hi c0nsumer,
I'm pretty much in agreement with jean-paul about the switcher issue, it really is very difficult to keep radiated noise (not to mention conducted noise) out of surrounding circuitry. It's not impossible but doing it is non trivial. This could be a particular problem with high impedance tube based audio circuitry.
You could add space for an optional series resistor to dissipate some of the power currently dissipated in the existing regulator design when used with supplies >12V.
Using a small transformer to power the board when there is no convenient voltage available sounds like a good idea to me.
Please keep the micro-controller, if I wanted the elektor boards I would have surely bought a set by now..
You're headed in the right direction, none of these issues are or should be considered show stoppers.
I'm pretty much in agreement with jean-paul about the switcher issue, it really is very difficult to keep radiated noise (not to mention conducted noise) out of surrounding circuitry. It's not impossible but doing it is non trivial. This could be a particular problem with high impedance tube based audio circuitry.
You could add space for an optional series resistor to dissipate some of the power currently dissipated in the existing regulator design when used with supplies >12V.
Using a small transformer to power the board when there is no convenient voltage available sounds like a good idea to me.
Please keep the micro-controller, if I wanted the elektor boards I would have surely bought a set by now..
You're headed in the right direction, none of these issues are or should be considered show stoppers.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.