DC bucking convertor noise

[HaroldBrooks]

A quick background, I wanted to have a "Battery only" power supply for my pedal board, so last year I bought IKEA Ladda AA batteries and put 10 of them in a battery holder, and then purchased a small DC/DC bucking convertor to lower the voltage from 13.5vdc to 9.8vdc or so (like a very fresh 9V battery). It's been working good and I can easily go two gigs without a recharge...

But yesterday I noticed that when I use a rig without a noise gate, and turn my "Boss ST-2" distortion drive up to full, the background noise is huge, and it's from the power supply. I never realize this before because generally I don't play with that high level of distortion. However when I let the box run on an ordinary 9V battery, the background noise dropped to approx. 1/4, and that is at all drive settings.

So I discovered my DC/DC power supply with the dropped voltage was the culprit. I now run less batteries (7 instead of 10) and no converter to yield a high of about 9.8vdc with freshly charged batteries, and things are quite.

I never realized this type of power supply could generate so much noise in my signal chain, but it does indeed !

So I now have the idea to run 7 AA rechargeables in two battery packs and run the two battery packs in parallel to gain some more mAh power, but keep the voltage correct and around 9.8vdc.

Does anyone have experience with these types of small DC/DC convertors ? https://www.amazon.com/MCIGICM-step-...tronics&sr=1-1

Is there a way of getting rid of the buck convertor noise easily ? or should I just use two battery clusters as I just described ?

Thanks for any help !!!

[Old Tele man]

The square-wave rise/fall time-switching generates "infinite" number of odd harmonics. Filtering & shielding are needed.

[uneumann]

The square-wave rise/fall time-switching generates "infinite" number of odd harmonics. Filtering & shielding are needed.

I use them - but ditto++ ^^^^^ on that comment!

[Helmholtz]

So I now have the idea to run 7 AA rechargeables in two battery packs and run the two battery packs in parallel to gain some more mAh power, but keep the voltage correct and around 9.8vdc.

I am no battery expert, but I don't think it is advisable to wire dry batteries in parallel. As the voltages between the 2 packs will hardly be completely identical, the lower voltage pack will permanently draw current from the other one, thus discharging it to the lower voltage.

You might wire decoupling diodes in series with each pack to prevent this. The diodes will lower the voltage(s) by around 0.7V.

So with seven fresh 1.5V batteries you would get around 9.8V.

[nickb]

The square-wave rise/fall time-switching generates "infinite" number of odd harmonics. Filtering & shielding are needed.

I really don't think you can make a blanket statement like that. For example many mixers nowadays have an SMPS, have low level micro phone inputs a kilo wat power amps are are perfectly quiet. Nor can you say square wave, what about resonant technologies?

If we are to use broad strokes I'd say that the parts you can buy on 'the bay' are likely to be built to a low cost rather than low noise so you get what you pay for.

You can change to a simple and very cheap linear regulator and accept the power loss or use a low noise converter

Suggestion, I'm sure there are plenty more.

http://www.electronics-sourcing.com/...sip-8-package/

[Helmholtz]

...and as the switching frequencies of lower power SMPS are typically above 20kHz, the harmonics should be ultrasonic/inaudible.

[nickb]

...and as the switching frequencies of lower power SMPS are typically above 20kHz, the harmonics would be ultrasonic.

This is true but they may switch to cycle skipping under the right circumstances which can produce quite low frequencies so the switching waveform is now chopped.

[Helmholtz]

Anyway, if a SMPS introduces noise, additional filtering would be required.

OTOH, a linear regulator needs a minimum voltage drop and wastes power.

I actually don't see the need for a regulator here.

[HaroldBrooks]

I am no battery expert, but I don't think it is advisable to wire dry batteries in parallel. As the voltages between the 2 packs will hardly be completely identical, the lower voltage pack will permanently draw current from the other one, thus discharging it to the lower voltage.

You might wire decoupling diodes in series with each pack to prevent this. The diodes will lower the voltage(s) by around 0.7V.

So with seven fresh 1.5V batteries you would get around 9.8V.

Thanks Helmholtz, I never knew that but I think you are correct about parallel wiring of batteries, from what I just read.

I took a current measurement from my pedal board with all the pedals on, and running at an idle (no signal running through) and it's right around 250mA, so the draw on the excellent LADDA 2450mAh batteries is fairly low. I wanted to measure this empirically, so I played all day yesterday on and off for 8 hours (the longest gig I would have), and the series run batteries in the single pack still measures around 1.24vdc individually, and according to the graphs I've seen on discharge for those batteries, they should have about 2-3 hours left until the critical voltage and current rolls off.

So I've now concluded I don't need to run two battery packs in parallel, and at least with fresh batteries a full charge will last a very long time. I am a bit paranoid about batteries failing at a gig, as last time one of my 9v batteries did just that, right in the middle of a glorious lead solo ! I felt like quite embarrassed in the process, and vowed to never let that happen again.

I used a power supply years ago, but at least the one I used was again noisy for the signal chain, and there is always the risk of power interruption that I don't like.

[trobbins]

That board uses a buck converter, which provides better efficiency when the input voltage is about twice the output voltage, so that the internal parts operate near a 50% duty cycle. So you may benefit from increasing your battery voltage to something like 18-22V.

That board can also operate in discontinuous and continuous modes, and it may well be that your load is varying or at the point where it is sporadically transitioning between the two modes, and that can cause disturbance signals that are down in the audible range. One test is to add some loads, or reduce the loading, and see if that makes an audible difference to the noise. Switchmodes 'like' to run in to a constant load or a defined load range, as that defines all the internal design and operating conditions.

That board may have a significant high frequency ripple voltage on its output terminals, depending on your load. That ripple voltage could be somehow generating noise in one or more pedals - you may want to check if some/all/one pedals are more susceptible. The internal switching frequency is fixed at 150kHz, so making an output filter for the converter needs some finesse as not all parts (or your wiring and layout) behave themselves as you may expect. To adequately prepare a filter, you really need to know what load current is being drawn (or could be drawn). The alternative is to use a linear regulator as the output filter, and adjust the converter's regulated output voltage to be suitably above (but not too much) the dropout level of the linear reg's input. You can then run a load voltage of say 8.5 to 9V. A low drop-out linear regulator would help, but you still need to know what the load current could get up to, so as to design any heatsinking of the regulator. Linear regulators are generally designed to filter mains frequency, so their filtering performance at 150kHz will be down substantially, but perhaps sufficient for this application - some regulators may be better than others.

[HaroldBrooks]

That board uses a buck converter, which provides better efficiency when the input voltage is about twice the output voltage, so that the internal parts operate near a 50% duty cycle. So you may benefit from increasing your battery voltage to something like 18-22V.

That board can also operate in discontinuous and continuous modes, and it may well be that your load is varying or at the point where it is sporadically transitioning between the two modes, and that can cause disturbance signals that are down in the audible range. One test is to add some loads, or reduce the loading, and see if that makes an audible difference to the noise. Switchmodes 'like' to run in to a constant load or a defined load range, as that defines all the internal design and operating conditions.

That board may have a significant high frequency ripple voltage on its output terminals, depending on your load. That ripple voltage could be somehow generating noise in one or more pedals - you may want to check if some/all/one pedals are more susceptible. The internal switching frequency is fixed at 150kHz, so making an output filter for the converter needs some finesse as not all parts (or your wiring and layout) behave themselves as you may expect. To adequately prepare a filter, you really need to know what load current is being drawn (or could be drawn). The alternative is to use a linear regulator as the output filter, and adjust the converter's regulated output voltage to be suitably above (but not too much) the dropout level of the linear reg's input. You can then run a load voltage of say 8.5 to 9V. A low drop-out linear regulator would help, but you still need to know what the load current could get up to, so as to design any heatsinking of the regulator. Linear regulators are generally designed to filter mains frequency, so their filtering performance at 150kHz will be down substantially, but perhaps sufficient for this application - some regulators may be better than others.

Thanks for your expert help trobbins, sounds like you know about that type of voltage reduction circuit !

Now that you mention it, I did disconnect a few other already bypassed stomp boxes in my "trouble shooting" of the noise, and it did indeed lower the noise a bit at one point when less or more (cant remember which) current was being drawn.

So since I already ordered two 8 x AA battery holders, I will run them both in series, which should give me about 21vdc, and I can drop that to 9.8 vdc and see how that works. I will let you know. I would like to use all the 16 rechargeable batteries I currently have, provided that the loss from the drop in voltage doesn't eat up too much current capacity to heat, and I have a net gain in run time before discharge. If that were the case I would be happy indeed.

I can tell you that the Boss ST-2 power stack box has a good deal on noise inherent with the modeling processing to begin with. I am reasonably sure that someone at Boss too a very good and fairly accurate model of a Marshall Plexi or JCM 800 and just upped the input signal as an adjustable parameter, as part of the input modeling algorithm. I owned and gigged with an early Plexi and JCM 800, and that box is the closest model to that sound I've ever heard, bar none. That being said, the old Marshall amp designs were never built around having that much signal gain, and indeed would make a Tremendous amount of noise and hiss when drive to that level. Many improvement were made to more modern metal amps to avoid those pitfalls I believe. Here's a plexi without any boost, and despite the hum you can hear a lot of other broadband noise along for the ride : https://www.youtube.com/watch?v=zKTi8l8Yuxg

Again, I occasionally crank the Boss pedal to 10 if I want to be extreme, and if my drummer and the rest of the band are loud, you can't hear the noise built into the Boss box. But the power supply noise I was getting put it at a whole new level so if I can "Have My Cake" and eat it too, by reducing the noise back to a straight battery level, I would be overjoyed.

I'll let everyone know how it goes.

[Chuck H]

I did read "infinite number of odd harmonics", but...

Could you run two bucking transformer in series, but out of phase with each other WRT winding direction. I did this with the HF inductor in an attenuator build that would howl whenever single coils were used and you got within nine feet of the thing. It worked a treat for that circumstance.

This would allow the original MO of just "more batteries".

[SoulFetish]

... I totally just forgot what I was going to say

[Chuck H]

... I totally just forgot what I was going to say

Bwaaahahahaha...

Enjoying your shut in time?

Sorry

[loudthud]

My experience tells me that you have to be very careful with grounds on the output of some switching supplies. An additional stage of L-C filtering on the output might be appropriate. Does the little switcher isolate input and output grounds ?