Ultrasound DI/Preamp: are these (photo) caps noisy?!

[Jimbo78]

Hi all,

my ultrasound DI/Preamp is a disappointment. Sounds flat, hisses, turn things up (line out/gain/notch filter) and just gets noisier. No "sparkle" on the top end. In fact, tone controls do very little at all. I build a clone of a Barcus Berry buffer/preamp a while ago, and it has a beautiful transparent tone with sparkle, and is very quiet. Just no headroom or DI, hence the DI/preamp 'upgrade'...

So...popped the hood, see attachment.

1.
Caps look cheapish, small drops (tantalum?)...are these my noise problem? Can I swap them out for metalized polyester film "orange drops" or MKT boxy things? (pardon my lack of correct terminology) The value is conveniently written on the board under each component.

2. TL062s...low power but noisier than TL072 s? Is that a straight sub, or will it change the whole thing--i.e. I'll have to change a dozen other parts? It runs off phantom power as well as battery and 9V adaptor, so power consumption not a prob.

3. TL064? Any comments on that, or replacing that?

Thanks for your help!

[pdf64]

TL06X - they just shouldn't be in the audio signal path, horrid hissy things.
I wouldn't use TL07X either for low level signals due to the hiss, but they're everywhere in line level audio stuff, so they must be ok.
There are lots of opamps in there, as a result of all those on board functions, so they need to be lowish current types.
It may be you would be better off with something simpler on board, using discretes or high quality opamps, and have the extra functions in an external unit. Pete.

[Jimbo78]

Thanks Pete.
I've had a look around and found some other options for the opamps: AD8032, LF351, Ne5532, TL082. Using phantom 48V or 9V adaptor, current shouldn't be a problem (or will it?). And there's also the opa5134 , I think that's the right number. Is that a possible sub?
Are there any particular caveats of any of these options? I really just want to swap the tl062's out and reduce the noise and improve the sound quality. BTW, this DI/PRE/EQ is for acoustic guitars, so, high impedence piezo input.
How about the caps? They're everywhere, right through the tone circuits etc. The only other caps on the board are those 50V electrolytics. Will film caps may a difference to the tone?

[Enzo]

Low current IS an issue for something that can run off phantom power. Phantom may be 48 volts, but it is at a low current. If you plan to run off an adaptor, then it is not an issue.

[Jimbo78]

Thanks Enzo. What kind of current can phantom power deliver? For example, if I replaced the 3 tl062 s in the unit with tl072's or 5532's, am I likey to be drawing too much current for a phantom power supply?

[Enzo]

Good question. Have to look up the phantom standard. I can't say off the top how much it CAN deliver, but it is not intended for any sort of power other than the tiny needs of microphones. I may wind up surprised however.

Generally you only find those low power types in things that need to have low power draw.

[Jimbo78]

I see, so if I replace to tl062's I should look for the lowest power option. Actually, the Barcus Berry clone has a single tl062, and that's quiet enough for me. The main difference is the tone, or lack of it. The Berry has mkt and wima caps throughout with the exception of the output cap which is a 50v electrolytic. It is clean, warm and sparkly. The Ultrasound is very middy and flat, no sparkle.

I'm wondering if the a) the caps are just poor quality and not 'musical', b) the 1meg buffered input is not high enough for the piezo (although the pickup maker claims this di box 'matches' their pickup), or c) both a and b.

I'll go through and try to work out which caps are in the signal path for a start. Any suggestions on whether this is a worthwhile mod?

I would like to mod this unit if possible: at the moment I'm running the Berry clone through the DI's preamp and then straight out the FX loop send--just for extra boost, a rather expensive boost!

[Steve Conner]

Well, what is the input impedance of the Barcus Berry clone you built? If it's more than 1M maybe that explains it.

Different brands of capacitor shouldn't make that much of a difference to the sound, unless the designer lost his mind and decided to use high-k ceramic discs for everything. To me, a complicated circuit with lots of functions that has to run off a low current seems like a recipe for lots of hiss. Maybe that is also the reason for the lack of sparkle, the designer had to restrict the treble to hide the hiss.

[Jimbo78]

Yes...that's logical. Can anyone tell me what the caps are in the photo? I've never seen this kind before. So far I'm guessing multilayer ceramic.

[Steve Conner]

Yep, those look like ceramic caps. They really shouldn't be used in the signal path, so hopefully they are just for power supply bypassing.

I hate to get into discussions of "capacitor sound", it's controversial, but there are real scientific reasons why ceramics can sound weird. It may be worth changing any that are in the signal path, if it's not too much effort. I don't think it would cure the hiss, but it might improve the tone or make the tone controls work better.

[km6xz]

They used electrolytics in the signal path.

Calculate the total current available....phantom voltage / 6.8k ohms will give short circuit current. Since many M.I. oriented phantom sources are not 48volts, you need to measure it with a load to see what you have to work with. Only a few units that run on batteries have real DC to DC converters to get 48v, they assume it will be used with amateur mics using electret elements that work fine on much lower voltages. Some phantom sources are 12-18volts. The old 48V standard, with 2ma per mic max current, has been replaced by IEC 61938 which specs 48, 24 and 12 volts so there is little "standard" to a standard phantom source now.
Do a plot of the EQ of the ultrasound, gain and noise to see what you are working with before modifying it. Then do the same thing with the BB unit and design the EQ to match. Also, pay attention to the very high input Z expected by the pickup/guitar tone circuit, you might be loading it too much with the ultrasound, which will also change the pickup/guitar control EQ a lot. Is the hiss still high with a shorted input to the ultrasound?
Before modifying anything, get good measurements of what you are starting with otherwise you are just guessing, hoping that a change is one you like.

[J M Fahey]

Don't know the phantom specs, but if they are limited to anywhere close to those 2mA *total* KM6 mentioned, you are stuck with the TL062.
Each TL072 eats around 4mA, idle, no signal, so 3X that looks like a no-no to me.
NE5532 needs at least twice that.
I think the noise and lack of sparkle comes from the project itself.
For me project is 90% ; parts 10 %
Remember also your Barcus *does* use a TL062 ... with far better results.
I doubt you will "cure" a poor project just by swapping same value "better" parts, they will just do "better" ... the same they were doing before.
Post the schematic of what you have successfully used before (the BB thingie) and what you didn't like from it.

[Jimbo78]

Success

Hi all,

I had a good look at the input, 1Mohm seemed a bit low for a passive piezo (no onboard preamp). The tone issues seemed consistent with a poorly buffered piezo: nasal/harsh, loss of highs, muddy.

So I changed the input resistor to 10M, and now it sounds great! The hiss issue has also gone, for two reasons: 1) I'm no longer cranking up the treble looking for highs, and 2) I worked out my input gain needed to be as high as possible, because line level control was not just attentuation, but amplification, and if that was high, I was amplifying the collective noise of the whole circuit! Now input=5, line=3, =nice and quiet.

Attached is the input buffer section: I was surprise there was no 100pf cap or similar at the input, I thought that was standard practice. I was also surprised the decoupling cap is one of those little mustard things-but having adjusted the impedance with good results, I'm not going to 'fix' what's not 'broke....

Thanks all for you help and comments. I might have managed to fix it myself, but I learnt a whole lot of other stuff...

Perhaps I'll put some jumper pins in with the 1M as well as a 5M and the 10M, so imput impedance can be selected.

[J M Fahey]

Nice to know you could solve it.
As far as RF is concerned, you don't ned an extra 100pF to ground when you already have 10000 to 100000pF already there (the Piezo itself )