does 2520 behave ok in single rail apps?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

strangeandbouncy

Well-known member
Joined
Aug 8, 2004
Messages
2,112
Location
West Sussex. UK
Hi,


   is there any reason why I can't use a 2520 as non-inv amp with single rail, with + input held at 1/2V+ between two 100k rsistors, instead of a NE5524 etc? I have 24V to play with.



    Mnay thanks,



    ANdyP
 
What does the "C" pin really connect to?

The available schematics are flawed, so I just won't guess.
 
PRR said:
What does the "C" pin really connect to?

The available schematics are flawed, so I just won't guess.

The C pin connects to a square of single sided pcb and is presumed to provide a screen.
 
Won't you need to AC couple the output for example, to drive a transformer? If you want to direct couple a transformer, you will need to return the transformer to your "virtual" ground which means it will need to have sufficient bypassing or an active driver to be able to supply the AC currents that will try to flow thru it. These can be as high as a couple of hundred milliamps if you are driving a heavy load like a typical step-up output connection. Of course, if you plan to go to the next stage of a multi-stage device, you just need a well bypassed capacitor.

You should also be sure to provide a DC return path to the virtual ground for both the + and - inputs just for stability.

The ground terminal of a 2520 doesn't connect to much inside except two .1uf bypasses to the + and - power pins and perhaps, in some units to some kind of a shield as Peter Purpose mentioned.
 
As Peter said, and in all the ones I dissected, the common pin only connects to a shield over the bottom (top?) of the pcb (i.e. solder side).
Only in the copies are there any .1uF bypass caps.
 
strangeandbouncy said:
Hi,


   is there any reason why I can't use a 2520 as non-inv amp with single rail, with + input held at 1/2V+ between two 100k rsistors, instead of a NE5524 etc? I have 24V to play with.



    Mnay thanks,

     ANdyP

API2520 can run with single supply voltage but this voltage must be in range +24V to +40V.
Your power supply voltage is +24V so it is able for 2520, but you should consider that this op amp's quiescent current is strongly dependant by supply voltage, so for +24V (or dual + -12V)this current is at minimum. It would be that 2520 work with more THD ( I've studied the 2520 schematic and I know this one) than with the typical + - 15Vcc or single +30Vcc , that the Vcc often used for 2520.

You can use the virtual ground at 1/2 Vcc , that in this case is +12V, but I don't council you to use 100k resistors to make this voltage reference but , in this case, 6.8k. Moreover, the 2520 positive input should be connected to this point (virtual ground point) across the input transformer's secondary. Moreover, the resistors connected to gnd, of the virtual point circuit, should be bypassed with a 220uF cap with a 100nF cap. The output transformer should be connecetd to the "real" gnd through a 220uF cap (with the positive connected to 2520's output and the negative to the trasformer) and not to the "virtual" gnd to avoid that the 2520 offset voltage change the virtual ground's voltage value, that must be 1/2V.

If you want a reference schematic look the JLM baby animal.

JLM uses 10k resistors for the virtual ground, but since you have +24V supply than 6.8k 1% is better value in this case.

Pier Paolo
 




 
Many thanks for all your replies, guys!


    6k8 is probably a bit too low an input impedance. Looks like I'm gonna have to go ic, not discrete. oh well.


      Kindest regards,



          ANdyP
 
strangeandbouncy said:
Many thanks for all your replies, guys!


    6k8 is probably a bit too low an input impedance. Looks like I'm gonna have to go ic, not discrete. oh well.


     Kindest regards,



         ANdyP

in the schematic I suggested to you (JLM schematic with 6.8k instead of 10k used by JLM) has not an input impedance of 6.8k because the input impedance does not depend by the 6.8k resistors. In this schematic the input impedance is the input inpedance of the op amp. Because the input transformer secondary is connected between the 6.8k resistors and the op amp's positive in  :)

If you would want other help tell me, please.

Pier Paolo



 
Pier,


    many thanks for your reply. I am a little confused, since any dividing resistor(6k8 for instance) is clearly in parallel with the input of the opamp! Unless I am just not understanding you correctly. I also understand that the feedback resistor also affects the input impedance. - So much to learn . . .

  If the input impedance is greater than 6k8, I will go ahead and experiment anyway.



    Kindest regards,


      ANdyP
 
No, because the DOA's positive input is connected to the input transformer secondary and not to these 6.81k resistors, is the transformer secondary connected to these ones. Moreover, since the resistor of the divider connected to gnd is bypassed with a large value cap, the positive input DOA sees only the transformer like is in the 312 preamp.

look my schematic , please.

If you look the JLM Baby Animal's schematic you can see that between the 10k resistors that do the voltage divider (I have suggested to you the 6.81k value for these resistors being a lower supply voltage than in the baby animal) and the DOA's positive input there is the input transformer's secondary

http://www.jlmaudio.com/Baby%20Animal%20Mic%20Pre%20PCB%20Full%20Circuit.pdf


 

Attachments

  • schematic suggested.GIF
    schematic suggested.GIF
    11.3 KB · Views: 63
Hi,

Thank you so very much for the explanation.

I was trying to describe something like the circuit attached. There is no input transformer. I need a make-up amp for my Pye mixer channels. I intend eventualy to build a ba284 type discrete amp for the 16 channels. I only need to drive 10k inputs to protools. Pye has passive eq that seriously attenuates the signal. I do not want to add higher output current amps TO EACH CHANNEL, and load my powersupply. I DO want to quickly convert a couple of channels to having direct outs, and have a couple of 2520's hanging around.


    KIndest regards
 

Attachments

  • Glensound .jpg
    Glensound .jpg
    47 KB · Views: 42
this is the first schematic, the other schematic with the 2520 I would want to send to you by email, please.

Pier Paolo
 

Attachments

  • a buffer 2.GIF
    a buffer 2.GIF
    41.5 KB · Views: 44
> something like the circuit attached.

Where does R3 go?? Can't give good advice if key chunks are missing.

OK, I know R3 goes to your + rail(?).

This plan gives little filtering from rail to amp input. If source and C2 are ZERO impedance, rail-crap is rejected; otherwise it sneaks in. Sometimes your rail IS that clean. Do you know?

Pier's plan is very popular. Use a couple 5K-10K resistors to split the rail, a big cap to clean it, then your input bias resistor. You want the divider current to be much more than input current, yet (if possible) not so high it sucks power. Battery jobs may like 100K to conserve battery. I have seen values as low as 470 for a divider feeding lots of amps, but the divider filter cap has to be sized large. Probably any orange-stripe resistor will be fine.
 
Hi Guys,


  Pier, thank you so very much for the post. That looks just the ticket to me! If you have another schematic with 2520 implementation, my email is [email protected]
I am extremely grateful for you time, help and generosity.


  PRR, you are quite correct to admonish me. I was foolishly "grabbing" images from a larger schematic, whilst simultaneously recording a vocal! And they say men can't multitask . . . You are quite correct, R3 goes to +ve rail(21V in the glensound mixer schematic I was lifting it from). It is interesting that you mention battery equipment, since the Glensound in question is a BBC OB mixer that was designed to run on mains, or a battery on location. Perhaps this is why they chose such a high value? it is the post masterfader output amp.
As ever, I am in your debt. I didn't appreciate that the current in the divider was significant, but the mention of it stimulates the old grey matter no end!


    cheers, and kindest regards,



      ANdyP
 
PRR said:
Pier's plan is very popular. Use a couple 5K-10K resistors to split the rail, a big cap to clean it, then your input bias resistor. You want the divider current to be much more than input current, yet (if possible) not so high it sucks power. Battery jobs may like 100K to conserve battery. I have seen values as low as 470 for a divider feeding lots of amps, but the divider filter cap has to be sized large. Probably any orange-stripe resistor will be fine.

PRR, if Mr. Schwarzenegger will ask me and JLM by phone (or by email) to elevate the divider resistors' value saying me (and JLM) that the environment is in danger for the power dissipated by these two res, I promise you to do this one........don't worry........that it will be done..... ;)






 
PRR said:
Pier's plan is very popular. Use a couple 5K-10K resistors to split the rail, a big cap to clean it, then your input bias resistor. You want the divider current to be much more than input current, yet (if possible) not so high it sucks power. Battery jobs may like 100K to conserve battery. I have seen values as low as 470 for a divider feeding lots of amps, but the divider filter cap has to be sized large. Probably any orange-stripe resistor will be fine.

Joking aside, I've considered the possibility to use the same voltage divider by others opamps, since that ANdyP has the intention to upgrade his console with several outs. One of the problem in a console is save the space, moreover the better choise is use the same voltage divider to avoid to repeat unnecessarily the same circuit and moltiply the components required.

However, considering that the input sources' impedance will be much probably <1kohm, this simple (and popular) circuit and the 100k resistors (I consider the several opamps) give a very good rejection vs. supply noise (more complex circuit would be useless). For completeness will post the schematic with higher resistors.






 
Hi Pier,

  Actually, there is quite a lot of spare room in each channel, and no spare slot on the edge connector! there are three spare xlr cutouts on each channel, which is mighty handy. I was thinking of using a separate voltage divider for each channel. I am only interested in building a direct out post passive eq/prefader at the moment, but will maybe add a pre-amp out(pre-passive eq), and a postfader output later, space and finance permitting! I have no idea what the input source impedance is. Unfortunately, the (staggeringly rare, mint) manual/schematics are not here at the moment. They are 150 miles away at my friends house. I am trying to get him to scan an input channel schematic for me. Fingers crossed!


      KIndest regards,



      ANdyP



       
 
Back
Top