[Doh]

posted 10-27-2001 1:02 PM ET (US) | Msg Admin:
--------------------------------------------------------------------------------
Hey everyone,
Just a quick question about opamp swapping. Is it ok to replace a bipolar opamp with a FET opamp given that the input offset voltage of the FET opamp is equal or lower than the offset voltage of the bipolar opamp?

Here are the specs of the chips I want to swap:
NJM2043(bipolar)
Input offset voltage(typical) 0.3mV
Input offset voltage(max) 3mV
Slew rate: 6 v/uS

AD712 (FET)
Input offset voltage(typical) 0.3mV
Input offset voltage(max) 3/1/1 mV
Slew rate: 20 v/uS

I've gathered that swapping a bipolar for a fet is a sometimes big no-no since the bipolar opamp can create large offset voltages if driving a high impedance load, but don't know if I'll have problems the other way around.

Thanks in advance for all the help!

-Doh

 

[Tomo]

Hello,

Where is this opamp? It is generally ok to do so. However, note FETs are rather weak and easy to break. So those 5532/4 used to biase tube amps are not recommended to be replaced with fet input opamps.

Tomo

 

[Doh]

It's on a solid state amp, not a tube amp. An NAD310, to be exact. Took a look inside and found it used four opamps:

JRC4558 x 1 (FET)
JRC0720 x 1 (FET) - equivalent of TI072
JRC2043 x 2 (bipolar version of JRC4558)

While this amp sounds warm and otherwise "musical" it isn't so detailed and is rather mushy overall and laid back in the high frequencies--good for acoustic stuff, but just plain sounds funny with some pop and electronica etc. I have some opamps on order from Newark and figure I'll install some sockets and try swapping around some combinations to see what happens.

Results so far:
JRC4558 --> OPA2132 Awfully bright...ear-bleeding to be exact.
JRC4558 --> OPA2604 Much more accurate presentation. Good
tonal characteristics.

Gonna play around with some AD746's (like 712, but not unity gain stable) when they get here from Newark since PPL indicated in one of his posts that it was a good replacement for the 4558. I'll need to order some 712's if I'm going to stick them in...maybe some OP249's.

Thanks for the help!

-Doh

 

[Tomo]

Hello,

I can't say too much since I do not understand how they are used. Knowing the names of devices are not enough for me to recommend anything.

AD711 are a little inferior in terms of performance compared to OPA132's. However, I know that OPA604 has better sound characteristics.

Please note that it is NOT recommended to substitute opamps without understanding the local properties of the circuit. For example, it is not recommended to subsitute non-precision high offset opamps for the opamp used in biasing feedback. Also, note the types of opamps to be used. VFB are always good. But then CFB will work better in certain places while it can work poorly in certain other places.

That is why I am saying it is important for us to figure out what circuits do those opamps belong to. Yes opamps are easy to use but that does not mean you can treat them poorly.

Tomo

P.S. Perhaps you can post the picture of the circuit for us.

 

[Doh]

I don't know if you mean by picture of the circuit a photo of the board or a circuit diagram. Below is a picture of the board (though of admittedly poor image quality, as it was done with a scanner)

NADl40 inside

What I know about this amplifier is that it uses both FET and bipolar devices in its output stage. I tried to trace the circuit, but found it to be extremely difficult.


The opamps can be seen to the right of the vertical circuit board. From the top of the screen, we have:

JRC4558
JRC070D
JRC2043
JRC2043

The amp's external inputs sit right above the JRC4558 in the picture and the speaker outputs are to the right of that, with the red, black and white cables leading into them.

Of note, there is a third JRC2043 chip in the device, though it is not shown in the picture, being obscured by the volume pot. That chip is part of the tone control board

 

[Tomo]

Hello,

I can see coils right at the output of the amp. You should see dark copperly colored / enamel colored coils. These often used at the output of speaker amps. It is unusual that input section and output section are relatively close. This means that it is highly likely some of the opamps are involved with DC servo feedback. (If I were to design NFB, I will not make the input so close the output.)

Now pay attention the cables that connected to the tone control board. Most of them should connect back to input section of the amp. One or two batches should be connected to some motor controller. We will ignore this one for now. Notice the extensive network of resistors near the top opamp. These are most likely cable terminators 1K ohm to 100K ohm. Those can-like devices are capacitors that decouple the top two opamps. I am not sure why they needed so many. Perhaps these opamps can misbehave easily or the capacitors are lower rating than required. Look right below the second opamp. You should see resistor-capacitor-network-like section. These indicate this section is a preamplifier/tone conotrol stage. Those top two opamp should be dual opamps.

Look around the large heatsink at the right edge of the picture. Note there are 8 solid state devices. Each channel use 4 channels. (Top 4 are first ch. Bottom 4 are second ch.) Now pay attention the top 4 devices. You should see that this group of 4 devices cosist of 2 different types of parts. two smaller ones and two larger ones. Smaller ones are BJT. Larger ones are MOSFET. These are positioned to push-pull. You should not see large resistors around those devices. The stand-alone heatsinks close to the large heatsink to the right are most like the gain stage for the push-pull output. I know sometimes people make these run at high temperature to ensure Q-point at smacking-middle of operation region.

Please look at the white blocks connected very close the the outputs both right and left ch. Considering the fact there there are output inductors and output cables at very close proximities, it is reasonable to assume that this white blocks has something to do with output stage. Now recall the ways we are accustomed to biase Class-AB amp to Class-A. We make one side to be permanently in the triode region, effectively making that side constant current source. According to Mr Meier and Mr Stokes, we can do this by forcefully offsetting the output by applying voltage directly the output. The white blockes are actually power resistors applying voltage to the output.

This is an important deduction. Since this is a speaker amplifier, Offset voltage is extremely important. Unlike headphone amps, you must absolutely be sure that offset is properly fixed before going the speakers. (Please refer to DC servo discussed by Jorgensen) Those two lower opamps are DC servoes for the amplifier.

So only top two directy effect the sound quality of the amplifier. The bottom two control the reliability of the amplifier but are not in the signal path.

Tomo

 

[Doh]

Wow! I didn't know that much could be deduced simply from looking at a circuit like that!

Let's see here...
Wires from the tone control board do terminate near the two opamps at the top of the unit. In fact, they terminate right below opamp 2.

Below opamp 2 (an below the insertion point of the wires from the tone control) sits a network of small yellow capacitors box-shaped and metal film resistors. These can be seen in the picture. I interpret here that both the opamp on the tone control board and opamp 2 are involved in tone control/preamplifier duties. Hence, both opamps affect the sound quality of the amp? I am uncertain of the role of the opamp on the tone control board, but am a bit weary since it is of the same type that is used as DC servos. I will have to take a closer look at the tone control board and trace the leads from the output to see where they go. Of note, the tone control board appears to have several polypropylene capacitors on it.

Opamps 2, 3, and 4 are run in unity gain configuration. My gf is in the room and I think she would flip if I opened up the amplifier to play with it right now, so I can't say what opamp 1 is. If the gain stage of the amplifier is not found in the opamp section as you indicate, I think it makes sense if it is run in unity gain. Given its position, I believe opamp 1 is the first opamp any signal comes in contact with from the amplifier's input. Will verify this when I open up the amp later.

Tomo, thank-you so much for your help! This amplifier makes much more sense now. Now I'm off to read the Jorgensen article you mentioned. I really want to know how this thing works!

-Doh