Ring (and/or Balanced) Modulation for the modulator masses

[Dandolin]

I got nothin', but burmp....

[crochambeau]

UPDATE: found this thing on the bay. Could I just chuck it in there or do i have to design a filter for it?

Unless I am mistaken, there is more to it than that. There are two aspects of this that I would be concerned with.

Switching frequency for the boost should be well above audible frequencies at your target load. I've seen a lot of cheap converters that switch in the 1-10 kHz range (and sometimes the controller changes the frequency for nominal efficiency reasons, so a boost or converter solution can play quiet on one thing and whine on another). Any hash on your power rail will make it to output unless your common mode rejection is 100%. I could not see a spec on that listing, so I would expect it to be audible.

Second aspect I would concern myself with, and I just pulled and had a momentary peek at a posted schematic that I cannot confirm applies to your project, is power rails in relation to common/ground. I see a proper bipolar supply depicted on the schematic where we have a positive 15 volt, a negative 15 volt, and a 0 volt common. It's true you can explore a split rail configuration (where a "virtual ground" is planted between 0 volts and your positive rail), but you may need to cut & re-route traces that reference 0 volt common, which would be applied to negative rail. It gets messy, and a preferable solution is to simply feed it the three wire power it needs, which that linked converter does not supply.

I've seen some solutions floating around, but I have not bothered to bookmark them because I'm pigheadedly rolling my own. There are still some bugs to shake out of mine, or I'd happily offer one up. I'm using an MC34063 family topology which allows switching in the 60-80 kHz (and above for some variants) range, resulting in easy to filter rails. But again, I know I saw something offered to the DIY market that performs these functions (0,9 volt in, -rail,0,+ rail out). If I trip across it I'll post here or in DIY, but my slothlike google-fu this morning returns nothing (apologies, but I would avoid that eBay converter if you're looking for a solution).

[qersty]

Thank you for the heads up. I dont think the +/-15v thing applies to mine, it runs of an external power supply. Apprently the newer thing is different than the really old ones (mine is just old as in probably 90s). It may be so that it does some voltage stuff internally but there is no vref from the power supply

http://experimentalistsanonymous.com/di ... er%204.gif

[crochambeau]

Ah, if yours just wants a hefty positive power rail it's probably already set up for a split rail and that eBay thing could work.

If it whines, I would suggest including a choke in your filtering, but by and large it looks like an inexpensive experiment, I would go for it.

edit - I just digested that configuration you posted, and it does call out a ground at the input, putting your - aspect on a different plane. So unless the output of your power module is decoupled from ground you might experience some negative rail/ground interaction that could cause grief.

Apologies if I'm muddying the waters here.

[qersty]

No muddying done! it seems that the ebay thing grounds the negativ output. im only basing this of the datasheet suggested circuit and i doubt it is very different from that

[BetterOffShred]

Based off the schematic they are using a 18VAC or so wall wart and rectifying to +/-15V like Curtis said... they originally had the,1496 chips, that EEA is a 633.. so without seeing the chipset etc. It's anybody's guess.. can you open it up and see what's happening?

[qersty]

Oh it has been open for a year I can take pics no problem. Mine is the reissue, the wallwart is 40v. I could check the voltages too if you can give me the points to check.



one more chip hidden by the middle pot, it is the 633. It its exactly the same as the eea trace
(dont worry about the bent pot, it is broken)

[crochambeau]

the wallwart is 40v.

Begging your pardon, but how electronically inclined are you? Do you own and are comfortable with a multimeter?

I ask because this head scratcher of a puzzle may have a simple solution wrapped around a serious pitfall.

The DC inlet connector looks like a bog standard barrel type, and not a multi-pin, confirm?

My thinking is that they can get away with a two pin power feed because the 40 volt PSU will be isolated from the other power rails because it is its own supply, and that each leg of the PSU is balanced across system common (0 volt).

The issue is that most standard 9 volt pedals will unite power rail low, signal reference, and shield at common. You might be able to AC couple input and output of that board (employ a non-polarized capacitor in line with both input and output wires feeding the PCB) and *disconnect* the "ground" wire from that board to enclosure and replace the 1N4744s with matched resistors to establish your split rail dummy ground. I would grab something like 10K to start at this point, and depending on the quality of the DC power from your boost you may need to stiffen the filtering caps (though I would let the 470 uf ride to start).

What you will want to do is confirm that there is no short between the old ground (green conductor on the EA layout posted above) and system common before applying power.

If there's noise on the + power rail it will be everywhere, so clean power is encouraged. But it can probably be done (you'll need to measure to confirm)

[Dandolin]

why do I feel like I'm watchin' someone line up for a cliff dive into a golf hole?

[qersty]

the wallwart is 40v.

Begging your pardon, but how electronically inclined are you? Do you own and are comfortable with a multimeter?

I ask because this head scratcher of a puzzle may have a simple solution wrapped around a serious pitfall.

The DC inlet connector looks like a bog standard barrel type, and not a multi-pin, confirm?

My thinking is that they can get away with a two pin power feed because the 40 volt PSU will be isolated from the other power rails because it is its own supply, and that each leg of the PSU is balanced across system common (0 volt).

The issue is that most standard 9 volt pedals will unite power rail low, signal reference, and shield at common. You might be able to AC couple input and output of that board (employ a non-polarized capacitor in line with both input and output wires feeding the PCB) and *disconnect* the "ground" wire from that board to enclosure and replace the 1N4744s with matched resistors to establish your split rail dummy ground. I would grab something like 10K to start at this point, and depending on the quality of the DC power from your boost you may need to stiffen the filtering caps (though I would let the 470 uf ride to start).

What you will want to do is confirm that there is no short between the old ground (green conductor on the EA layout posted above) and system common before applying power.

If there's noise on the + power rail it will be everywhere, so clean power is encouraged. But it can probably be done (you'll need to measure to confirm)

I have basic electronic knowledge. I have no problems usin a multimetr

[crochambeau]

The next biggest hurdle is getting any legible information out of me, apologies, I'm sitting here with morning tea while my neurons are firing.

Signal ground will be an issue with the "solution" I spoke of above.

If you have this connected to a daisy chain 9 volt power supply, where the common is common across all the other pedals and cables and amps, and your inboard "signal common" (formerly known as 0 volts) is 15 volts above common it will have 10K (or whatever that half of the voltage divider height you settle on) of DC path to return to actual common (the signal reference for what is happening at input and output jacks). This could induce noise and/or affect your signal level in a detrimental way.

My workarounds for that (potential) issue are becoming increasingly complex, to the point at which a proper bipolar supply would be simpler. This thing may perform reasonably well at +/- 9 volts (headroom would suffer, if that's a concern), then all you would need to include is a voltage inverter module to make -9, then you can sidestep the shifted ground plane mess and call it a day.

I'll get around to muddying the waters sooner or later..

[qersty]

I checked and it does short to ground. Def will if i replace the diodes. REplacing the supply entirely probably is the ideal fix. Headroom may actually be an issue, I have no real basis for it other than that it pushes alot of low end in the filter mode

[Chankgeez]

[youtube][/youtube]

[Dandolin]

Flaming Space Ring Lips? sounds painful, very very painful....

[Chankgeez]

Don't read the words! Just listen to the sounds & look at the pretty pedal.