Help needed with use of NE5534 op amp in -ve tracking supply!

Hi!
This is a follow on from my previous post regarding +- 15 V regulated
supply. As per suggestion from this good group - I built a positive
supply using a LM 723. Works fine.
Now, I have added an inverter follower using an NE5534 op amp. Why
NE5534? That is the only opamp I had in my box and besides I want to use
a low noise op amp.
The op amps power V+ and V- are the 0 and -20.5 output from the filter.
The Vin+ is connected to 0 and the Vin - has an input resistor of 33K
and a feedback resistor of 33K (Unity gain) with a 2K trimpot in between
for adjustment. I have a PNP pass transistor.
RESULT: NO OUTPUT - or - rather -0.07 V output
If I connect a 3.2K to Pin 1 (Balance) and a 3.2K to Pin 8 (Bal/COMP) -
hey presto - everything springs to life!
What is going on? Is the Op Amp powered incorrectly? Should I take Pin 7
(VCC+) to the +20.5? If I have to do that - I have to use zeners or
something so that the IC does not see more than +- 15V.
To +15 Output from the +ve side of the regulator
/|\
| NB: There is a 22pF Comp cap between (5) and (8)
_____|____________________________________0
| | 33K 2K | 33K
| |__/\/\/\__/\/\__|_/\/\/\_
| | | |
| | (7)| |
| | |\| |
| IN- |_ | \ (6) /E
| (2) | \____|/
| | / |\
|_________________ | / \C
IN+ |/|(4) |
(3) | |
---------------------------------------- - 20.5

Please can anyone help me understand this? To me it seems that may be
the input transistors are not in their active region because the bias
current is too low? Maybe I should lower the 33Ks?

Thanks in advance to anyone who even reads this!

Any suggestions will be a bonus!

Regards
Ramendra

 

The only way this could work would be if the input common mode range
included the positive rail. The 5534 doesn't. Power the opamp from
the +15 volt output, instead of 0V.

http://www.onsemi.com/pub/Collateral/NE5534-D.PDF

 

For starters, you've connected the 5534 '+ input' to the positive supply
rail. That's outside the operational parameters ( see input common mode
voltage range on the data sheet ). Therefore it won't work properly.

Why aren't you running the 5534 off the +/-15V regulated supply ?

What are you actually trying to do ? It normally helps to state this at the
beginning.

Graham

 

Texas Instruments makes some extremely low noise op amps that use the
so-called chopper technology, the TLC2654C series.
These achieve input noise of about 2uV p-p right down to DC. They
are available in the old PDIP package which will plug into the Radio
Shack protoboards for breadboarding. Keep in mind the TLC2654 is rated
for +/- 8V supplies only. Use voltage regulators to keep supplies below
+/-8V.
TI has a generous, free sample program. Register and they will
FedEx you up to 5 samples overnight.
The TLC2654 has a bandwidth of about 300KHz so it can be used as a
music amplifier for modest gains (5 or less). In particular the
low-frequency region (< 1KHz is dead quiet).

Charles Gilbert
Consultant
NonDigital.Netfirms.com

 

Thanks for your advice guys. Hmmm... ok, it shows I am a novice. Lesson
learnt - I should have read the data sheet more carefully. It was all there.

Graham, the 5534 is part of the regulated supply. My plan is to use it
as an inverting follower of unity gain to track the +ve regulated rail.
That is why I did not use the +/- 15V regulated supply. Please see the
modified circuit above. Your suggestion should work - I am going to try it.

Once again, thanks a lot for your help.
Ramendra

 

Thanks for your suggestion. I am only a hobbyist. I do build some stuff
for people I know and they reward me handsomely. Are these samples meant
for anybody or for businesses and industry? Dont get me wrong. I like
free stuff too - especially for experimenting. Just want to make sure
this is ok.

Thanks
Ramendra

 

Others have addressed the question of why it isn't working. I want to ask:
do you actually need it in the first place?

Tracking supplies are actually a rather specialized thing. There are very
few cases where they are needed, and in most cases they are actually more of
a problem than a benefit.

In many cases, what you actually want is just a dual regulated supply, where
the -ve and +ve supplies are separately regulated and do not track each
other. This does mean that they won't usually be exactly balanced around
zero (e.g. it might be +15.1v and -14.8v or some such) but for most
circuitry that is not a problem.

The benefit of dual non-tracking supplies is simpler circuitry, better
stability, and lower noise: unless you are careful about filtering, a
tracking regulator takes whatever noise is on the master side and adds it to
the slave side. That noise could be coming from the circuit being powered -
for instance, digital switching hash on the +ve supply can get copied over
onto the negative supply. Not, in general, what you want.

 

Thanks for your response. I hadnt looked at any of the points you have
mentioned. I am making a tracking supply to test it and to study how it
works. That will take me quite some time. After that I wish to use it as
a standard supply for other circuits that I build. I want the +ve and
-ve rails to be balanced so that I can assume that that they are whilst
I work with my other test circuits. The first of these is going to be
sine wave generators of different kinds.
I am starting up a hobby I used to have a long time ago. Lots of things
have changed - I am slowly starting to re familiarise myself.
I have built a dual supply using two separate +ve supplies running off
two secondaries of a transformer. Now, I am working on this tracking supply.
Thanks once again for raising the points you have raised. I will try to
incorporate a RC filter where I am sampling the +ve rail to try and
exclude noise from being replicated by the negative follower as you have
said. I hopr that works.

Ramendra