diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Analog Line Level (http://www.diyaudio.com/forums/analog-line-level/)
-   -   Need orientation with tweaking a TASCAM MF-P01 (http://www.diyaudio.com/forums/analog-line-level/217036-need-orientation-tweaking-tascam-mf-p01.html)

Betadine 30th July 2012 02:15 PM

Need orientation with tweaking a TASCAM MF-P01
 
Hi,

Iīm trying to tweak a TASCAM MF-P01, this unit has 4 channel recording system in a single analog tape (using both sides A and B at once)

The matter is that the output is only stereo, and I need 4 channel independent output for digitalize the 4 channels at once with a computer...

The unit (as you can see) has 4 potentiometers, which will control the volume of each track for later mix them all into a stereo output for the listening..

Because I donīt have the schematics, I have put a cable into each potentiometer output with the hope of getting individual outputs with Line Level (10K output impedance, approx 1Vpp max) for multitrack digital recording.

Signal Volume at that point (potentiometer output) is around 100-150 mVpp max...
while line level output is at 1V-2Vpp max...

The idea is to use an LME49710 (or LM4562 Dual) for that, but I need orientation cause Iīm quite lost in this... :warped:
What I know is that I need a preamp with:
- infinite impedance at his entry (for transparence to the rest of circuitry)
- gain of (1V/0.1V = 10 more or less)
- 10K ouput impedance for line level output (a resistor of course)

Can any expert iluminate this project? I mean how can I build the block system? one buffer followed by one amplifier? or maybe it can be done with only one IC?
any recommendation for circuity, IC types, etc?

any comment will be much appreciated!

Regards,

Mooly 31st July 2012 01:16 PM

It sounds like you just need one gain stage per pot.

Recommends... I would go with an OPA2134 (dual FET opamp). This will eliminate any DC offset issues and it's one of the best sonically. It will also allow for direct connection to the pot wiper if needed without causing "noise" as the pot is turned, the FET opamp having essentially zero input bias currents.

So you need to be clear on what you want and what the set up is.

Vital info is...

Is the pot output (wiper) ground referenced with zero volts DC present or is it an AC coupled feed ?
Do you want the preamp to run on a single or dual rail ?

Each gain stage is text book stuff. Two resistors to set the gain keeping them in the 10K max region to minimise noise. A resistor of around 47 ohms in series with the output to isolate the opamp from capacitive loading that can give rise to instability. The input to the opamp at its most simple connects directly to the pot wiper if the feed is suitable. At its most complex it is isolated via a cap and the input pin of the opamp biased to ground (assuming a split supply) via a 470K or thereabouts resistor.

Betadine 31st July 2012 11:11 PM

Yes, the pot is referenced with GND, I have thought to use a unity gain stage as a buffer and next to it a tipical amplifier Vo=(R2/R1)*Vin
The luck is that superquality is no needed, itīs only for digitalize old stuff (analog tapes), so I suppose there are more complicated circuits for better response, but for the moment I will test this easy configuration...

I suppose dual rail means to treat negative values (yes of course)

- I will keep in mind that R1=1K and R2=10K for example... that will be indeed sufficient low impedance for noise...
- 47 ohms resistor in series with the output... I need to put 10K for "Line Impedance" so I dont understand... do you mean to put 47 ohms in series with the 10k ohms? I will need the 10K in the output!

- The cap and 470K seems good idea...

- I will use caps next to the vcc and -vcc pins... and watch the signal In and out with a sine generator...
Iīm trying now with a general pourpose tl082/tl072 it has low gain in high frequencies but for testing pourpose will be nice for later replace with OPA2134... what Iīm wondering if itīs I can use only one stage (direct amplifier with very high impedance in itīs entry) thats why I use first a unity gain buffer first... tomorrow will post results...

thanks for the tips, building prototype right now...

Mooly 2nd August 2012 08:33 AM

Sorry for late reply.

The TL072/82 are absolutely fine for this at the gains you are using. The quality will be excellent. The OPA type opamps allow a lower impedance load to be fully driven such as 600 ohm.

The 47 ohm goes directly to the output pin and just ensures that whatever you hang on the end doesn't cause instability. Although you might be connecting to 10K load if the wiring/cable added a few 10's or hndreds of pf capacitance then that is waht can cause the trouble. The resistor stops it completely.

Quote
Vo=(R2/R1)*Vin

That sounds like the inverting configuration (not that it really matters for this)

The non inverting configuration is Vo=(R2/R1)+R1*Vin so you would need 9K and 1K for a true gain of exactly 10.

If you can say what power supply you have available eg whether its a single 12 volt or a dual -/+ 15 volt etc then we can advise better on the final configuration.

Betadine 2nd August 2012 11:14 AM

1 Attachment(s)
You are true about the output gain, Vo = (R2/R1) + R1 * Vin

I have attached a pdf with the basics of the circuit, including the 47 ohms at the output...
For transmission lines is good to have the same impedance in the output and the input... so if you have 10K load in the end of the wire itīs good to consider to put 10k in series in the beginning of the wire. Do you think that removing the 10K in series (in the output) would not be dangerous? the small capacitance due to the wiring lenght...would need 47 ohms or at least 47ohms? thatīs why I have drawn two resistors in the output...

About filters...should I put 1uF in series in the output for removing DC Offset?

Is it then recommendable to put in the Input of the first AO a decoupling capacitor with 470K for removing glitches? testing now with TL072.. i have a bandwith with constant gain of 90 KHz so good good, the sound seems clear also...

All the Best

Mooly 2nd August 2012 12:52 PM

1 Attachment(s)
I've looked at your circuit and its basically fine theoretically.

In practice...

You don't need use two opamps. Just the second one is all you need.

Transmission lines really apply more to RF engineering than audio to ensure no power loss in transmission. The 10K resistor would actually cause a few problem. The output of a preamp is usually likened to a perfect voltage source, that is to say one that is unaffected by loading. The 10K resistor makes the circuit hugely dependant on what is connected to it. It could also cause premature HF roll off if any capacitance were present too. Also a 10 K resistive load would half the available voltage output.

DC offset isn't an issue with FET opamps but it would be with bjt types like the NE5532.

In the attached circuit the 0.47uF input cap ensures that you can connect the preamp to virtually anything.

It can be argued that the output should be AC coupled to in order to ensure total compatability to whatever you connect it too. If you know it will feed an input that is at zero volts DC nominally then there is no problem.

Betadine 2nd August 2012 01:52 PM

Get it,

so wondering...

that cap with 470K its a high pass filter I suppose then that you remove any dc 0-10Hz from the input... thatīs the only thing I see...

If I choose to put another cap in serie with R-47ohms in the output should I put also the 470K resistor to ground like the input? I suppose then that putting that filter in the ouput is to be protected from any dc level when connecting the recording device... ?

I need R1 for the gain equation to work! Vo=(R2/R1) + R1 * Vo I suppose that by mistake you have removed that resistor from the + input of the OA

Ah! Also I like the 47 ohms solution for protecting shortcircuits in the output connector hehe
I will remove then the 10K resistor... in power circuits for speakers I see it a lot for maximum power transfer but itīs true that here is no needed. Thanks again itīs almost done :)

Mooly 2nd August 2012 08:29 PM

Yes, the 0.47uf cap and 470K do form what is termed a "first order" high pass filter. The values ensure that the circuit passes all audio transparently and in fact the -3db point (voltage) is around 0.7 hz.

The 470 K also biases the opamp by providing a ground reference to the non inverting input pin. It also defines the input impedance. We normally make line inputs high impedance so that they don't load the circuit driving them. The noise contribution isn't an issue as it gets "shorted out" by the low impedance of the source driving the preamp.

If you wanted to put a cap in series with the output then its value would depend on the load impedance the amp has to drive. If we said 600 ohm minimum (which the TL072 won't drive fully) then to keep the "0.7hz -3db" would need around 400uf. That sounds way too large and in practice it is. Keeping to a more realistic 10hz -3db point and say 2K load which the TL072 can drive fully gives a much more reasonable 8uf. So in practice I would go for a 10 or even 22uf electrolytic (electros aren't all bad) if you feel you need for a cap at all.

A 470K (or lower) from the output end of the cap to ground is good practice because it defines a ground reference for the cap and ensures the output is at zero volts DC whether or not the amp is connected to anything.

R1 and the gain :)

I think I posted that before the circuit appeared and I was assuming R1 and R2 were the feedback components. So for the circuit above its

Vout = R3+R2/R3 *Vin which gives a gain of 11 of course.

Betadine 3rd August 2012 09:26 AM

1 Attachment(s)
Ouch! :mallet: "Never leave FET gates in open circuit"
The university told me thousands of times and yet I still forgotting that :p

While reading your reply I realize that the OA (because of his FETs) needs these protection of course! thats one of the 470K mission. Other way would be a cap to ground, but being carefully for high frecuencies attenuation... Good good, this is getting a very much clear project, again thanks a lot!

I think that with R1 I have mixed the previous inverting amplifier with the non-inverting one... :scratch: ups!

EDIT: In the pdf I attached there are wrong values in C1 and C2 in fact C1=470nF and C2 ... C2 ... electrolitic 22uF?? but that is a polarized cap, we have a bipolar wave, mmm should I take 1uF tantalum better or put two caps of 22uF one for positive and the other for negative semicicles? What options we have?

When finished (I pretend to do it today or tomorrow) I will test it with the sine generator and oscilloscope, will post results

Cheers

Betadine 3rd August 2012 11:35 AM

I have 1uF ceramic caps as the maximum non-polarizaed caps...

C1 will be one of those 1uF cermaic cap (instead of 470nF) would be ok
C2 ... :S


All times are GMT. The time now is 10:15 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright Đ1999-2014 diyAudio


Content Relevant URLs by vBSEO 3.3.2