Preamp has unwanted delay (and hum)

[EpicFail]

Hi there,

Think this is the right part of forum (sorry if not)...

I was hoping someone might be able to help with this problem:

I have built a preamp for piezo pick-up which is working except that there is a noticeable delay between input and output. From researching this, I think this is because of phase/group delay, but wondered if anyone else has had this problem.

Also getting mains hum - was wondering about connecting ground from sensor to a shielded enclosure, but this preamp is 9V battery operated, so how would grounding/earthing it work? And would you connect the negative nail to the enclosure as well..?

Thanks in advance.

 

[mickeymoose]

Delay?
If you would post your circuit...

 

[Frank Berry]

The low side of the input (or the output) should be connected to the metal enclosure.

 

[EpicFail]

mickey - sorry - forgot to add the circuit. Please see attachment. The only difference between the schematic and the build is that the gain stage after the buffer is now a lot lower, somewhere around x10. I was planning to re-designing this stage so that it is an active volume control and ditching the end passive vol control. It will remain an inverting stage though. May well be reducing gain in the tone control as well in future builds. Turns out piezo signal is stronger than I originally thought!

Frank - thanks for that - so either connect the low sides (I assume that's the ground connections?) of the input OR the output to the metal case - not both? And it sounds like not a good idea to connect the negative rail to the metal case either? Should the low side be only connected to the case and thus not also connected to the negative rail?

Sorry for being pedantic, just trying to get my head around this. I've read quite a lot online about grounding and it still confuses me...

Thanks for your help guys.

Si

 

[SY]

When you say "noticeable delay," what are we talking about? Microseconds? Milliseconds? And measured how?

 

[epicyclic]

Hi Epic

May i suggest changing C3 to 100uF , 100nF is OK decoupling for the input of U1 but will give a high impedance ground for U2 and U5.......this may help with your hum situation .

 

[EpicFail]

SY - I haven't been able to measure it with any testing gear. My term "noticeable delay" is based on attaching the piezo disk to an acoustic guitar and playing a string. There is an audible delay between the string being played (with the accompanying acoustic sound from the guitar itself) and the output from the speakers. It's enough to put you off if you were playing live, for example.

For clarity, the piezo is plugged into the preamp, which is plugged into an M-audio interface which connects to a Macbook.

epicyclic - thanks for the recommendation re. decoupling cap - will try this out.

Cheers

 

[DF96]

Software processing time. Welcome to digital audio!

 

[EpicFail]

DF96 - software processing time was the problem. Just tried it with a normal guitar amp and works with no time delay. Don't know why I didn't do that in the first place - stupid!

Incidentally, the hum is seriously reduced when through the guitar amp.

Is there any way around the whole software processing time delay problem when connecting to an audio interface?

 

[DF96]

Use a much much faster machine or much much faster software.
or
Don't use digital - there will always be a delay, so the best you can do is minimise it.

 

[EpicFail]

OK fair enough. Thanks for everyone's help with this.

 

[Osvaldo de Banfield]

In the resistor divider to bias the opamps input, use two r's shunted both with caps, this way you are creating a compensated voltage divider as used in oscilloscope proves. I used that configuration several times, and the secondary advantage is that the input is bypassed to both rails. Also, you can use one more OP to decouple voltage divider from ICs's input, but not too necessary in this case.

In the schema you posted, the cap is charging to reach exponentially the value of VCC/2, on time the minimum voltage to the opamps can work in class A, in this moment the audio is getting from the output. And, also, when shut down the circuit, this cap continues charged although the opapms are not, and this is very dangerous to the input stage of the opamps.

Best regards.

Good luck!.