Pedal version of DIYRE Colour Palette

[bowanderror]

I really dig the DIYRE Colour Modules, but the 500-series form factor is expensive & I'd rather have them in pedal form. I've made a version of this schematic, but without the balanced I/O and with the addition of a high-impedance input stage:


I have it working on a breadboard, and once I got the balance right between the dual-ganged gain & attenuation stages it sounded really good with the TM79 Multistage Saturation Module in it. It's the first time I've built something with a dedicated bipolar supply, so what's the best way to deal with the ground from the input & output jacks vs. the PSU ground?

Take a look and let me know if I should make any changes before I make a perf layout.

Thanks,
Mike

[duck_arse]

what's the difference between ground?

[bowanderror]

I guess that's what I'm asking: can I just treat PSU ground & signal ground as the same thing in this circuit?

[duck_arse]

in that circuit, with no pwer amps or digital guff or tikking oscillators - pretty much yes. ground is ground is a fairly common saying.

[antonis]

Just place C5, C6, C7 & C8 as close as physically possible to ICs respective pins..

Or delete them and add a 100nF cap between +15V & -VC..
(it hepls to prevent "garbage" entering GND from individual +/- voltage sources..)

[bowanderror]

Thanks for the replies guys. I'll keep in mind the cap placement for the perf layout.

I realized that the output voltage divider feeding the opamp gain stage could be simplified to a voltage-divider feeding a buffer, so I tweaked the values & it sounds the same w/fewer components:


I made a +-15V PSU for powering a few of these + Colour Modules and I noticed that some of the modules have resistors (for current limiting?) between each power rail & opamp V+/V- power pins. Should I be adding something like that to this?

[bowanderror]

So a while back I ordered some PCBs for a Colour Palette Pedal build based on this updated schematic:



After getting everything assembled, I was getting a lot of unpleasant clipping as I increased the dual-gang Drive control (AMP on the schematic). A distorted sound that wasn't the saturation characteristic of the various modules I tried with it.

That was a while ago & having learned a bit from other builds, I think I see some issues with the current schematic:

• The standard Colour Module output impedance is around 1k, so I'm wondering if the 5k Attenuation trimmer is loading down the output of the module.
• As the Colour Palette is a pro audio device expecting a +4dBu input, I had to really boost the signal to drive the Modules, so I'm also wondering if my single opamp gain stage is not happy with the amount of gain I'm asking of it.
• I'm also using an LT1054 bipolar charge pump board, so there is also a chance I'm pulling more current than it can provide.

Based on that schematic, do any of those stand out as the most likely culprit? Let me know what else I may have messed up, I'm aware of the limits of my own knowledge here!

[PRR]

> I'm wondering if the 5k Attenuation trimmer is loading down the output of the module.

The 5k is beyond the 10k fixed, right? So the worst that can happen is the 1k module sees 10k. Which meets the common 10:1 criteria. Anyway isn't "loading down" the whole point of an attenuation knob?

> wondering if my single opamp gain stage is not happy with the amount of gain I'm asking of it.

Do you know how to compute gain vs frequency? (IMHO this is not a problem.)

> a chance I'm pulling more current than it can provide.

Do you have a voltmeter? That's *really* where I would look first.

[bowanderror]

> The 5k is beyond the 10k fixed, right? So the worst that can happen is the 1k module sees 10k. Which meets the common 10:1 criteria. Anyway isn't "loading down" the whole point of an attenuation knob?

Totally forgot to take that series resistor into account... And I guess you're right on the function of an attenuation knob, still feels naughty for some reason? Either way, I probably just should have buffered the attenuator for peace of mind.

> Do you know how to compute gain vs frequency? (IMHO this is not a problem.)

I don't think I do for stages without compensation caps, but I'd like to learn. Is it calculated based on the GBW of the opamp?

> Do you have a voltmeter? That's *really* where I would look first.

I do have a voltmeter, but use the current measurement mode with embarrassing infrequence. In this case, would measuring current draw be as simple as inserting a DMM in series with my +/- rail during operation? It would be interesting to see how different modules stack up in terms of their current draw anyway. The voltages on my host board & the module boards look fine, fortunately.

[PRR]

What does it say, + and - 15V or something?

Do you get that Voltage when it is running?

(Measuring current is a pain.)

[bowanderror]

I took voltage at both the charge pump output, and on the Palette host PCB (after the 100R current-limiting resistors). The first measurement is without a module installed (jumper bypassed), and the other two are with a module installed, with the dual-gang AMP control at 0% and at 100%.



The module gets its power through the host PCB, daisy-chained via a 10-pin ribbon header. My first thought is that those 100R current-limiting resistors are possibly way too big, as they're dropping a lot of voltage with the module installed. I attempted a current calculation based on the voltage drop across each of those 100R CLRs, but I don't know if that is a valid approach here.

I definitely don't understand why the current would go down as the AMP control is turned up. That was not the case when I moved that control with no module installed. Could it be an interaction between the pre-gain opamp and the input of the module?

Here is the module circuit I used to take the measurements with (based on this DIYRE schematic):

[niektb]

Could be that the module sees a lower output current and the output buffer delivers a bit more current. But the difference isn't big enough to worry imho
Anyhow, 30mA is nowhere near the LT1054 rated output current (which is 100mA) so this shouldnt be a problem

[bowanderror]

So I replaced the 100R current-limiting resistors on the V+ & V- rails with 10R, with an improvement in voltage drop (current remained about the same):



I'm still getting unpleasant (and uncharacteristic of the modules) distortion as I turn the AMP knob up. The difference in voltage between AMP at 0% and 100% is still there, but is less drastic, I guess due to the lower overall voltage drop across the resistors.

I still need to get a handle on why I'm not overdriving the modules, but I'm going to check some AC signal levels and report back.

Could be that the module sees a lower output current and the output buffer delivers a bit more current. But the difference isn't big enough to worry imho
Anyhow, 30mA is nowhere near the LT1054 rated output current (which is 100mA) so this shouldnt be a problem

Thanks for the input, niektb! I'm glad the LT1054 is likely not the issue. Can you go into a bit more detail on the module seeing lower output current and output buffer delivering a little more current? I have some trouble imagining currents in these situations.

[ljglmail]

The "ground" of a circuit is the common point for 0-volts and 0-amps. The most direct way of making this ground point is to tie it to a pure conductor (like a plate of a huge power supply) that can handle any spikes in the circuit.

However, if you're building your own effects, you might not have access to a 50A power supply. And if you're using digital circuits, you might not want to tie them directly to the "pure" ground. For example, if I'm running at low voltage (9V or less), I don't have much choice of what kind of op-amp to use. If I want high gain, my only option is a non-inverting amplifier (which needs a negative power supply).

How do we get around this? Use an *isolated* power supply! This can be as simple as a battery or wall wart that's wired up with the device's "ground" terminal and the rest of the circuit's ground terminals all tied together, with one exception - they have their own separate wires going back to the battery/wall wart.

[iainpunk]

the ground is the 0v point indeed, but quite significant currents can flow through ground wires and nodes. the conductor for ground in pedals need not be to extreme or big, as the currents in a pedal aren't much in most cases. generally less than 100ma per analog pedal. digital is different, in that they sometimes require a bit more power, but that rarely exceeds half an amp.

For example, if I'm running at low voltage (9V or less), I don't have much choice of what kind of op-amp to use. If I want high gain, my only option is a non-inverting amplifier (which needs a negative power supply).

you can use a whole range of opamp configurations and types, not just non-inverting, with only 9v.
you really don't need bipolar or negative power supply to run a simple gain stage, all you need it bias, a 'fake' ground that sits at Vcc/2 so the signal can swing 4.5v up or down, from that 4.5v midpoint. another common method is a charge pump, like the Klon Centaur uses.

isolated power supplies aren the solution to not having bipolar power. adding bias to the signal is tho.

cheers

[PRR]

> common point for 0-volts and 0-amps.

A robot-assisted mass-poster.

https://www.google.com/search?client=firefox-b-1-d&q=ljglmail

https://www.debwan.com/ljglmail

http://www.insidehoops.com/forum/showthread.php?472967-Best-Girl-Scout-cookie&s=be80056640a4882239897df2ff3b0add&p=14497479&viewfull=1#post14497479

It's official – Thin Mints are America's favorite Girl Scout Cookie.

The sales results for the 2013 cookie season are in, and Thin Mints topped the list of most popular girl scout cookies. It was followed by Samoas/Caramel deLites, Peanut Butter Patties, Do-si-dos/Peanut Butter Sandwich, and Tagalongs/Peanuts Baked with Fudge.

Trefoils are the least popular cookie among Girl Scouts and their customers, according to the national organization.

Girl Scouts sold $781 million worth of cookies during the 2012-2013 sales season, a 2 percent increase over last year. The group says that's equivalent to more than 8 million boxes of cookies sold to customers outside of troop meetings.

Girl Scouts set a goal to raise $760 million this past year, but surpassed it by $21 million. All profits stay within the local council to fund community programs that benefit girls.

https://www.imybkexperience .com/ https://www.mykfc-experience .com/ https://www.mybpcreditcard .one/

Note payload^^

[amptramp]

I would add small-value capacitors on the order of 20 to 100 pF from the op amp outputs to the inverting inputs.  In particular, the 5532 amplifiers are wide-bandwidth and could oscillate at high frequency (into the tens of MHz).  This could cause clipping where the more extreme parts of the oscillation (closest to the power rail voltages) are clipped and this is reflected in the average signal output.  You could look at this with an audio scope and never see it but it would cause clipping distortion to start at much lower levels than it would otherwise.

I also support the idea of putting the bypass capacitors right across the V+ and V- pins.