Converting pedal input and output for use with mixer

Started by jimladladlooklike, November 17, 2022, 08:07:42 AM

Previous topic - Next topic

jimladladlooklike

Hi folks,

A friend of mine who plays synths has asked me to make them an A/DA Flanger for use with their mixing desk. The desk is an Allen and Heath Mixwizard with 6 TRS balanced Auxiliary outputs, which they are planning to use as an effects send. These each operate at -2dBu line level (615mV??). I would like to know the best way to convert this signal into ~100mV so I don't end up with unwanted distortion at the input of the flanger circuit. I'd also like to know if it's necessary to convert the output signal to a low Z return input after the effect circuit and, if so, how. My understanding is that the output of this pedal would probably be too low as well as having to high an impedance?

I have seen a bunch of threads here and elsewhere that mention different methods of achieving the line level to inst level conversion, the most simple being a voltage divider. I have also seen folks mention a reamp circuit, which I think would be best as it means that the balanced output could be converted to unbalanced. I think I'm leaning towards this currently.

For the output, could I just use an op amp circuit?

I'd like to get some opinions on the best way to tackle this.

Thanks!

ElectricDruid

There's a good ADA flanger project here:

http://www.moosapotamus.net/ADAflanger-MN3007.html

(I'm guessing you don't have a SAD1024 hanging about, but maybe you do).

The schematic is here: http://www.moosapotamus.net/images/FlangerClone_SCH_rev5_MN3007_jan2010.gif

Since the flanger runs on 15V internally, I'd think that it has sufficient headroom for line level signals without much trouble. What you  might well find is that the early stages of the circuit provide gain to the guitar signal going in (e.g. it gets boosted to something more like line level) in order to get the best S/N ratio.

So using it with line levels directly could be as simple as finding the gain boosts and toning them down.

jimladladlooklike

Hey Tom,

That's the one I'm going with indeed. The layout is from Sabrotone:



It uses an MN3007 in place of the SAD1024, so will cost a bit less to build!

Thank you for the pointers. So it's a case of finding where gain is set at op amp stages and messing with resistor values?

I'm still unsure how to go from a balanced out from the desk into an unbalanced input. This could be achieved with an xfrmr based reamp circuit right?

I will have to ask my mate about whether the inputs to his desk are TRS or TS, but if they were TRS would I need to find a way to convert the unbalanced output back to balanced?

Thanks!



anotherjim

For line inputs to FX it's easier to forget (almost) balanced input and work off the hot only. You lose 6dB by ignoring the cold but you are expecting the line input level to be too hot anyway?
The cold can be terminated to 0v at the FX.

There is a pseudo-balanced method which is pretty simple in concept.
The FX input is made to have a line input impedance of 10k. For guitar FX that may have 1M input you just add a 10k resistor to 0v on the immediate signal input and the balanced hot feeds that. If you find it's too easy to overdrive the FX input, the 10k on the input can be a 10k log pot and used to attenuate the input.
The balanced cold goes to 0v via another 10k. Both hot & cold see a 10k load.
The balanced line-level source now thinks it's feeding a 10k balanced input. You still lose 6dB.

Output to balanced can be just as simple except some FX have a fairly high output impedance of over 1k. They will lose volume and possibly bass also if asked to drive a 10k line input. Here I'd suggest adding a balanced output driver using a dual opamp. That should give a possible 6dB boost on the FX output.

Since the ADA output has a 10k pot and a fixed 100R resistor, the output impedance will be 100R with the pot set maximum (Or removed!) and that output can feed a balanced pot. You can either terminate the cold to 0v or pseudo-balance it where the cold connects to a 1uF cap and that goes to 0v via another 100R. It now looks like a balanced 100R output.

jimladladlooklike

Okay, so for the input I'm looking at something like this?



Then for the output a dual op amp circuit such as this?



With this circuit on the output (if I'm on the right track) you're saying I could eliminate the original 100k volume pot from the circuit altogether, use the output to feed this unbalanced to balanced driver, then use a balanced (dual?) potentiometer plus the 1uF cap and 100R resistor on the cold as a volume control?

Am I understanding this right?

anotherjim

The input. It's a nicety to terminate the balanced cold in a matching fashion. It can work fine to ground the cold at the input and then the screen doesn't actually connect at all there - it just goes as far as the jack which should be insulated from chassis. Note that some balanced mixer outputs are only pseudo-balanced anyway and don't actively drive the cold. The functional diagram in its manual should show that.

The output. If using the opamp balanced drive is used (that's the circuit I was thinking of!), the volume control would go before it where it already is. No need to mess with a stereo-ganged one (they don't have perfectly matched tracking between sides)
The Esp circuit is for dual-balanced power supply, you need to make changes if you only have a single supply voltage. That is a half voltage reference for the opamp instead of ground (which will already exist in the FX) and DC blocking caps in the input and outputs of 1uF minimum.

jimladladlooklike

Cheers Jim, I did wonder whether I'd need plastic 1/4" jacks.  Thanks for clarifying that. So the 10k resistor from cold to ground isn't necessary, but only if I don't ground the screen of the TRS jack? Or should I not ground the screen even if I do use said 10k?

Regarding the output, that makes a lot more sense, thank you. 1uF caps, also makes sense, noted. Had noticed the bipolar PS myself as I made the mistake of not doing so on a previous project, but thanks all the same for pointing it out!

anotherjim

You're making me think now!
A balanced mixer output that is fully balanced and actively drives the cold should withstand the cold shorted to ground. But I remember an incident where a powered wedge had to be fed from a mixer Aux with an ordinary TS mono cable. It worked at first and then went silent. Presumably, the dual opamp feeding the Aux went into thermal shutdown from the short circuit to ground on the cold. Getting a balanced TRS cable solved the problem. However, balanced-unbalanced schemes that simply tie cold and screen together on a TS or phono plug abound both DIY and as commercial adapters.

If you use TRS plugs and sockets, you can choose how to connect it up in the FX. To avoid hum loop, the FX only needs to connect to the mixer ground at one end then assuming the FX PSU doesn't also connect to ground, you don't have any loop. You could always add a ground lift switch if you like.



merlinb

#8
Quote from: jimladladlooklike on November 20, 2022, 01:11:23 PM
With this circuit on the output (if I'm on the right track) you're saying I could eliminate the original 100k volume pot from the circuit altogether, use the output to feed this unbalanced to balanced driver, then use a balanced (dual?) potentiometer plus the 1uF cap and 100R resistor on the cold as a volume control?
You may be overthinking it. You really only need it to be impedance balanced, so you can use one opamp and a pair of 100R series resistors + blocking caps. You can give the opamp as much gain as you need. Sure, a truly balanced output would give you an extra 6dB headroom, but do you really need it? KISS.


Processaurus

#9
Quote from: anotherjim on November 18, 2022, 10:22:35 AM
The balanced cold goes to 0v via another 10k. Both hot & cold see a 10k load.
The balanced line-level source now thinks it's feeding a 10k balanced input. You still lose 6dB.


It shouldn't be necessary to present a load to ("terminate") the driving source of the cold signal. The circuit described doesn't hurt (and it is just one resistor) but one can just leave it floating.

Processaurus

You know, if the OP's friend is using it as a bus effect with the aux sends, it might be nice to have a "wet only" option for the output, and the dry happens from the regular channel on the mixer. Also opens some creative options where you could further effect or eq the ADA's modulated sound before mixing it with the dry to get the flange.

jimladladlooklike

Cool, thanks folks. I have a few options to play around with.

jimladladlooklike

Okay more questions...

I plan on using the dual op amp driver on the output, fancy that extra 6dB, but I want to put it on a DPDT switch so it can be used with guitar if need be. What I understand is that if I send the signal through the driver, I don't want to ground the sleeve of the return TRS jack in order to avoid loop hum.

However, what do I need to consider if I make the driver switchable? I guess if I'm bypassing the driver the sleeve of the TS cable must be grounded? Does this call for a ground lift and if so where exactly should it go?

(I'm presuming I can make the 10k on the hot to ground on the input switchable in order to make sort of an inst/line switch? and that plugging a TS into the jack will work as the sleeve will take the path of least resistance to ground?)

thanks

jimladladlooklike


jimladladlooklike

Hey folks, still in the process of building this one. Have also spoken to my friend who asked about wet/dry options.

I found this thread on modwiggler, OP mentions lifting one end of R41 to eliminate the dry signal, which I'm going to add. Anything I need to consider when doing this or is it that simple?

Also any advice on where to put the ground lift switch in the circuit (to be used when using unbalanced to balanced converter) would be helpful. Cheers guys



jimladladlooklike

Hi again folks.

I have more or less finished this project, really appreciate all of the guidance re the input and output. I still need to test whether these work as desired but I'm fairly confident I've wrapped my head around it and wired everything up correctly. The circuit sounds good, but I have some concerns that arose during the calibration process.

I followed the procedure as per the documentation: http://www.moosapotamus.net/files/ADA-MN3007-rev5-Jan2010-Documentation-rev20120225.pdf

Bearing in mind I built the MN3007 version, the lower and upper clock frequencies should be 69.6 KHz and 2.6 MHz respectively, but mine read at 21.6 KHz and 0.96 MHz, even with both T4 and T5 fully CCW. I feel as though the best sounding range probably isn't available as the circuit is at the moment, so I'm wondering where I should start troubleshooting. I have another MN3007, so will try that and see if it makes a difference for starters

Thanks!

ElectricDruid

Changing the MN3007 won't affect the clock frequency. If the BBD was loading the clock in some way, what you'd hear would be degraded signal through the BBD as the clock pulses get rounded by the loading, and in the worst case it wouldn't pass signal.

If the clock frequency is too low, the first place to start is the VCO capacitor, C29/39p. 39p is actually a really tiny value, so it's just about possible you could have stray capacitance affecting things. I'd certainly make sure to clean up the board well around the clock. If any of the clock's tracks on that Sabrotone layout are longer than they need to be, I'd shorten them too. Parallel tracks on a stripboard have measurable capacitance, and it can be ten's of pF, so it'd hit us for this job.

After that, you'd have to work backwards through the 4007 into the range-setting components (IC3c, IC3d, IC2a) and make sure everything is as it should be there. Could be as simple as a wrong value in that part.
I'm assuming that calibration of the range is done with the Manual control and the LFO not playing a part.

HTH,
Tom

jimladladlooklike

Hi Tom

Yeah I set the range with the manual control.

Appreciate the pointers, will check all of the above as soon as I get a spare evening.

Thanks!

jimladladlooklike

Found the problem!

I put a 390pf cap in place of the 39pF.

Circuit is sounding awesome now!

Thanks again