2520 - FABIO's - DC Offset !!!

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

khstudio

Well-known member
Joined
Dec 27, 2005
Messages
2,116
Location
New Jersey, USA
::::::EDIT:::::::::::::::
I ONLY tested the offset on these in a PULTEC, SS AMP CIRCUIT!
The conclusion of this thread is that the "Circuit design" could be the cause & NOT the Amps themselves.
:::::::::::::::::::::::::::::


DC offset is about -144mv :shock:
- FABIO's 2520's - from the last batch.

Doesn't this sound a little high???

Has anyone else measured theirs?

I have 12 of these - 8 are in my 8 channel API 312 (which is 99% complete)
...I planed on building it stock - NO Output CAP
But now I'm concerned.

I should mention that I'm using Fabio's & Cinemag's STEEL CORE output TX's.

I know what some of you are going to say... just use a CAP... I know this but REALLY would rather not. I prefer the more open sound I'm getting without the caps BUT I don't want to damage the AMP or TX.

I spent hours listening to MANY different Caps last night (In my Dual Pultec that uses the 2520 for the output amps) they ALL change the sound but the Panasonic FC seem to be the closest to no cap, with the Nichicon PW not too far behind. (I also tested MUSE & other high end caps.)
 
[quote author="Greg"]Are you putting any DC in?[/quote]

If I'm understanding your question... then NO.

I have coupling caps between the filter & amp sections of my Pultec.

Now, the SS Pultec isn't set up exactly like the 312 but I wouldn't think it would cause the offset to change. :idea: :?:

But I'm NO expert & could be wrong!!!!!!!

Here is the Circuit:
EQP-1A3Schematic.jpg
 
[quote author="Samuel Groner"]What's the voltage at the opamp noninverting input?

Samuel[/quote]

It reads about 12 mv DC on the (-) input. (Pultec circuit)

I haven't tested my "Actual" 312's for the Offset so... Maybe this circuit does affect things? :oops: is that possible?
 
I was told Fabio makes the best 2520 clone. Using it in a Pultec is another story. I didn't know Pultec's had 2520's? I thought they were DOA's used to make API style stuff or upgrade mixers?
 
It reads about 12 mv DC on the (-) input.
I was asking for the noninverting but anyway--the gain of this stage is above 10 so there is little surprise that you end up with more than 100 mV at the output. Put a ~220 uF cap in series with the 1k47 resistor, it might affect the sound of the unit less than at the output (but you'll still see some mVs at the output).

Other options might include a larger cap at the output--about 1000 uF seems right, 150 uF is too small for my taste. Perhaps you could try a back-to-back pair as well.

For your info: a good part of the offset will come from the different DC resistance seen by the two input terminals which causes non-canceling offsets from the input bias currents.

Samuel
 
[quote author="Samuel Groner"]
It reads about 12 mv DC on the (-) input.
I was asking for the noninverting but anyway--the gain of this stage is above 10 so there is little surprise that you end up with more than 100 mV at the output. Put a ~220 uF cap in series with the 1k47 resistor, it might affect the sound of the unit less than at the output (but you'll still see some mVs at the output).

Other options might include a larger cap at the output--about 1000 uF seems right, 150 uF is too small for my taste. Perhaps you could try a back-to-back pair as well.

For your info: a good part of the offset will come from the different DC resistance seen by the two input terminals which causes non-canceling offsets from the input bias currents.

Samuel[/quote]

Sorry:
Channel 1
(+) input is -10.3 mv
(-) input is -10.5mv

Channel 2
(+) input is -10.7 mv
(-) input is -11.6mv

The good thing is THIS circuit is supposed to have an output cap.
I was really more concerned about having this offset with the 312's I've built. Maybe this circuit creates more offset that the 312???

Hey... since the offset is Negative (- DC)
shouldn't the cap's (-) side face the circuit???

I DID try this & to my ears sounded better/nicer.
For now, I'm installing Panasonic FC 470/10v.
They sounded the best in there.
 
(+) input is -10.3 mv
(-) input is -10.5mv


Actual input offset is 0.2mV, which is frikkin good.

You have 10mV of common mode DC, very likely base bias current in a large base resistor. 100K is a large resistor for low-noise audio BJT inputs. You show a DC voltage gain of 15, which times the 10mV gives 144mV near-enuff.

If you really have a 100K input bias resistor, it is NOT the op-amp's fault.
 
[quote author="PRR"](+) input is -10.3 mv
(-) input is -10.5mv


Actual input offset is 0.2mV, which is frikkin good.

You have 10mV of common mode DC, very likely base bias current in a large base resistor. 100K is a large resistor for low-noise audio BJT inputs. You show a DC voltage gain of 15, which times the 10mV gives 144mV near-enuff.

If you really have a 100K input bias resistor, it is NOT the op-amp's fault.[/quote]

Thanks for your input. :thumb:

I just got done a session & saw this post.
Cool... so you're saying that THIS Circuit's Design is causing the increased Offset at the output :?:

If so, good... the circuit requires a cap anyway.
Like I said, I wasn't sure & was mainly concerned about using these in my 312's without output coupling CAPs.

When I get a chance I'll test the 2520's in my API 312.

Thanks again.
 
Hi guy´s,

I will take the opportunity this time when the 2520 discuss come up again.

This 2520 schematic that always pop up on the web, where come this re-drawing from in the beginning ?
I have a feeling this is not the true 2520 "A5A" version that you can find in the mostly of API products trough the years. (1980 and forward)

Have somebody done a compare between a DIY and a original 2520.

Interesting parameters are:
Current consumption
Open loop gain
Noise
Slew rate
Behavior and swing before clipp in 600 ohms load.
And of course a listen test.

It would be very interesting if some of you have opinions about this.

--Bo
 
Last edited:
Yea... since I have 12 of these I'd like to know how "Close" they are to the REAL 2520's :?: :?
I was under the impression they are supposed to be clones of the Huntington's.

After taking to a few people (offline) it seems the Transistor count is off but who knows, there were several version of the 2520.

I can say one thing... I asked Fabio for a schematic of the version he sent me & he refused. :roll:
I don't quit get this... for one, they're un-potted & one of the reasons people liked the idea of being un-potted was they could repair them if needed... but how the hell can you do that without a schematic... i don't get it.

Has anyone REALLY done a DETAILED comparison :?:
 
I have work with API gear since middle of 1970´s, and I like the 2520 version from this time, because this was the only versions that work with out problems.(10 of 312 cards still work perfect after 30 years)
I can not tell if they also are the "A5A" version because I haven't removed the thin plastic top yet, where you can find the circuit version on the solder side of the PCB.

The good 2520 version i have from the middle of 1970´s was manufactured in Melville N.Y.

When the 2520 problem start for me, it was when DataTronix in Reston V.A. takeover the API company, and from this time there seem to be a lot of durability problems from time to time.

Any way, if I remember right,
the "A5A" shall have total 10 transistors, and shall begin with a NPN differential pair together with a "CCS" constant current source, and then following with a PNP diff. pair. (with emitters up towards +v)

--Bo
 
Last edited:
I have work with API gear since middle of 1970´s, and I like the 2520 version from this time

What do you mean... 2520 from this time?

You mean Fabio's latest batch???

The ones I have DO have 10 transistors on them... is this a good sign?

Thanks,
Kevin
 
Kevin,

Bo probably meant, from the mid 70s. Which is the "same" as Fabios.

take care
Gil
 
Kevin,

I mean those orginal 2520 I buy direct from API factory, Long Island out side N.Y. in 1975.

I think 2520 was mark "Melville, N.Y." up to 1975 and "Huntington, N.Y." from 1976 ????

--Bo
 

Latest posts

Back
Top