Boosting signal generator or op amp outputs to >+30 dBu

Where we discuss new analog design ideas for Pro Audio and modern spins on vintage ones.
juniorhifikit
Posts: 139
Joined: Sun Oct 04, 2009 8:37 am
Location: Oakland/Paris

Re: Signal generator output ranges

Post by juniorhifikit »

Thanks all - I updated my drawing. Still trying to understand all of the "not so obvious" features of this circuit before I breadboard it. Like this from the article:
However, power-supply circuits include transistors, which create junction resistance, affecting the amplifier¡¯s operation.You can use supporting amplifiers to reduce losses and increase the quality of the output voltage of the primary amplifier.
What is the effect of junction resistance on the opamp's performance? Are the protection diodes on the opamp's input on your mic pre circuit necessary?
User avatar
mediatechnology
Posts: 5444
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Signal generator output ranges

Post by mediatechnology »

Not sure I understand the question about junction resistance.

The protection diodes on the input of the input-capacitorless preamp are still necessary.
Though there is no longer a stored charge in the preamp, there can be, and usually is, some type of stored charge in the microphone itself.
A transformer-coupled condenser microphone can still produce a hefty inductive spike when one leg is shorted.
The mic itself can also have output capacitors.
User avatar
mediatechnology
Posts: 5444
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Signal generator output ranges

Post by mediatechnology »

Neither pin 3 or 4 of a THAT 1646 require a DC path to provide bias current.
Going back to your original idea of using the 1646, it may be possible to AC-couple the signal ground reference for Din, pin 4, with a large value electrolytic. (Maybe 47 uF.)
AC-coupling the ground reference for pin 3 may be a little more difficult since it is sensitive to series inductance.
You might be able to overcome that with a small value capacitor in parallel with an electrolytic or a small-value 47R resistor in series with the electrolytic.
My first choice would be the later suggestion.
It's worth trying on a protoboard.

However, even if you get the 1646 with the inputs AC-coupled and bootstrapped rails to work you still need some voltage gain ahead of it.
Unless you need the extra output current the 1646 provides (and you might) I'd go with a single op amp.
You could also parallel two 5532 outputs through 10R ballast resistors for the added load current needed with +/-34 V supplies and have those drive the bootstrap transistors.
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Signal generator output ranges

Post by JR. »

Maybe I'm getting old (and cranky) but it seems a lot simpler, to configure a simple opamps as inverter with gain of 2x or more, then add a discrete voltage gain stage between the opamp output and load. Finally connect the overall negative feedback loop from the output to the opamp - input.

For low distortion I would be inclined to use two common emitter transistors, PNP with emitter toward + supply and NPN with emitter toward - supply. Both devices could be biased on with enough class A current to keep distortion very low. Transistors probably benefit from being heat sinked to drive 600 ohm load and handle static dissipation.
buffer.gif
buffer.gif (11.98 KiB) Viewed 11770 times
I first saw and copied this topology from inside an old BGW power amp, that used a decent opamp front end, this common emitter stage, to develop the voltage gain and signal swing, then the amp used serious power transistors for the power output stage. A simple 600 ohm buffer doesn't need the additional power stage.

The main changes I made for my amp design was that I added proper current limiting to the output stage power devices (the BGW used an overall crowbar protection circuit across the entire PS. My amp is still sitting in my living room, and worked the last time I turned it on... I built it over 40 years ago.

You can figure out the details, like resistor values... You can increase the AC gain by adding RCs or just Cs across the emitter resistors, while just Cs may give so much open loop gain it is hard to stabilize. This circuit comes with no warranty, I have not used it in design for 4 decades.

JR
Cancel the "cancel culture", do not support mob hatred.
juniorhifikit
Posts: 139
Joined: Sun Oct 04, 2009 8:37 am
Location: Oakland/Paris

Re: Signal generator output ranges

Post by juniorhifikit »

mediatechnology wrote:Not sure I understand the question about junction resistance.
Just wondering what effect (they don't elaborate in the article) the junction resistance from the power supply transistors would have on the opamp's performance, that would lead them to their follow-up example with the three opamps. Simply an academic question...
The protection diodes on the input of the input-capacitorless preamp are still necessary.
Though there is no longer a stored charge in the preamp, there can be, and usually is, some type of stored charge in the microphone itself.
A transformer-coupled condenser microphone can still produce a hefty inductive spike when one leg is shorted.
The mic itself can also have output capacitors.
But this isn't for a mic pre - simply a voltage boost for my wimpy function generator.

I get what you mean about the 1646 and voltage gain.

JR - thanks for posting that. I'm going to mull over the two designs before moving forward.

I suppose I could also just find an old hi-fi amp laying in a pile somewhere and use that for my increased voltage output... Not as much sport in it though.
User avatar
mediatechnology
Posts: 5444
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Signal generator output ranges

Post by mediatechnology »

Scott: Your re-posted sch looks fine though you'll probably want to add a Cc and Cfb around the 5534.
One additional thing you can do is bypass the 5534 differentially with a single C rather than C3/C4.
That way the Qs don't have to slew current to charge them.
If you look at the MCI sch you'll see no Cbypass to ground.
My original figure is drawn with that, but the rails are not slewing at audio rates. I later added electrolytic differential bypass shown in the sch as "opt."


The input protection diodes shown on my flying rail voltage booster sch wouldn't be needed in your application.
They are there to provide a discharge path around the op amp input for the servo integration cap.

JR's approach using a voltage gain stage is another way to go about it.

Look also at the Jim Williams Op Amp Booster Stage articles posted here: viewtopic.php?f=12&t=13&p=20

(EDIT: BTW John I think figure 5 in that first citation may be the one that you wrote Jim Williams about.)

In the flying rail application, the "outputs" are the floating Vcc and Vee rails.
That's the reason I used the boostrapped circuit.
The op amp output itself isn't used in the fully-differential design.
It's just there to provide a +/-15V supply bracketed around the DC CM voltage.

MCI built a lot of EQ sections with several of the bootstrapped 5534 stages.

I don't think the effect of the boost transistor's B-E resistance is significant.
In Self's TL07X bootstrap circuits, where he bootstraps them to reduce CM distortion, there's some issues with TL07X load driving capability so he did it with op amps.
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Signal generator output ranges

Post by JR. »

Perhaps... at 5V no problem, but at higher voltages, could draw excess current at idle.

His figure 8 while using a few more parts than I might, can be scaled up to more than 15V rails, and figure 9 scaled down without the final power devices.

For bench test equipment a medium fi audio power amp could work, check it for low enough distortion. The hybrid opamp + gain stage, with feedback completed back to the input opamp should deliver near opamp performance , so useing a modern uber-opamp should deliver near SOTA.

Interesting how dated the Williams article is as modern opamps can now drive plenty of current and closer to rails.

JR
Cancel the "cancel culture", do not support mob hatred.
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Signal generator output ranges

Post by JR. »

http://www.sg-acoustics.ch/analogue_aud ... index.html

http://www.sg-acoustics.ch/analogue_aud ... A-1_r1.pdf

Sam Groner has done a lot of good work, and this +/-40V DOA may fit your bill.

JR
Cancel the "cancel culture", do not support mob hatred.
juniorhifikit
Posts: 139
Joined: Sun Oct 04, 2009 8:37 am
Location: Oakland/Paris

Re: Signal generator output ranges

Post by juniorhifikit »

JR. wrote:http://www.sg-acoustics.ch/analogue_aud ... index.html

http://www.sg-acoustics.ch/analogue_aud ... A-1_r1.pdf

Sam Groner has done a lot of good work, and this +/-40V DOA may fit your bill.

JR
Cool - thanks! Hadn't seen those. Looks a lot like a 990 (which I used to assemble for John Hardy long ago!). I think for my modest test purposes I'll try the lower parts count swinging opamp first. will report back...
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Signal generator output ranges

Post by JR. »

juniorhifikit wrote:
JR. wrote:http://www.sg-acoustics.ch/analogue_aud ... index.html

http://www.sg-acoustics.ch/analogue_aud ... A-1_r1.pdf

Sam Groner has done a lot of good work, and this +/-40V DOA may fit your bill.

JR
Cool - thanks! Hadn't seen those. Looks a lot like a 990 (which I used to assemble for John Hardy long ago!). I think for my modest test purposes I'll try the lower parts count swinging opamp first. will report back...
Yup, AFAIK Deane Jensen (RIP) came up with the concept of sticking an inductor in between the input devices. This is a win-win way to roll off gain at very HF for high slew rate and stability with negative feedback, while delivering the high gain at low (audio) frequency range for low distortion.

Sam is standing on the shoulders of giants, and +/-40v rails is perhaps one application where a DOA is kind of justified. While I expect wrapping a few discrete parts for the voltage swing around a good opamp will work too.

JR
Cancel the "cancel culture", do not support mob hatred.
Post Reply