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.
Post Reply
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 »

I would go the power amplifier route.

I think I have an HP200 with a high level option that might do it.

Do you need +30 dBu or +30 dBm? You can get +27 dBu with a 1646. You might be able to squeeze +30 out of a good one using supply voltages above the absolute max and a heatsink. :geek: I actually was able to make a "fifteen ten" run on 90V. :ugeek:
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 »

Do this with a pair of 5534s for balanced output.

Note that Q1 is drawn with incorrect polarity. It should be an NPN.
The base-collector capacitors, C114 and C115, are also incorrect and are 680 pF in the MCI.

Image
Image courtesy of MCI and http://www.proharmonic.com

The 2003 is a 5534. Use BD139/140s for the Qs.

The MCI "Swinging Op Amp"

Image
MCI Swinging Op Amp Ad dB Magazine January 1979.
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 »

Oh, so you want 100V? I did that when I was 17:

https://www.ka-electronics.com/DC_Pream ... 261974.jpg

This one is 70 V p-p. I made it ~100 Vp-p but it was a follower since more than about 70V exceeds the range of the op amp.
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 »

I'd go with the "MCI Swinging Op Amp" using a solo 5534.

You might also try an NJM2114 there. They've got almost twice the output current of a 5532. Parallel the inputs of both sections and ballast the outputs through maybe 10-47R and you should be able to get 100-120 mA. Have the summed outputs then drive the bootstrap Q's and load.

Pik Power stock the NJM2114D: http://www.pikpower.com/New%20Site/njr_ ... ntent.html

We really should ask Mouser to carry that one.
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Signal generator output ranges

Post by JR. »

+30dBu is right at the limit of what you can get from standard IC technology PS voltages, using differential drive.

The simplest answer is to use an audio power amp.

If you really need 100V swing that is a common nominal voltage for constant voltage background music systems in Europe. US standard is 70V. You can actually buy a dedicated several watt 100V amp but the frequency response and distortion won't be bench grade.

Peavey used to sell an automatch step up transformer to convert normal power amps to constant voltage so one of those would give you plenty of swing, but IIRC they don't come smaller that maybe 25W or 50W. And again going through a transformer will not deliver bench grade LF response.

If you only need to be slightly better than +30 dBm a pair of DOAs run from hotter rails in differential mode would get you there. Depending on that you are testing a true +30 dBm single ended output may be more useful. A decent opamp combined with a discrete output stage buffer with say 10 dB of voltage gain in the buffer could get you there. The buffer could be corrected by the negative feedback loop of the opamp, so you should be able to get bench performance without killing too many brain cells. Note: the "swinging opamp" if it's what I think it is (common emitter transistors driven by opamp power leads) is a relatively simple way to get there. I made a modest 35W audio power amp for my sister back in the '70s using a similar approach.

Lots of ways to skin that cat depending on your needs and personal preference.

JR

PS: I see an opposite trend in things (digital) audio with A/D convertors running from 3.3V power supplies. I am working on my second generation drum tuner and have found a very cool class D audio amp that makes a bunch of output with good battery life.. I'm trading the 9v battery for a few AA cells. Good technology spinouts from cellphone/music playa stuff.
Cancel the "cancel culture", do not support mob hatred.
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 »

Note: the "swinging opamp" if it's what I think it is (common emitter transistors driven by opamp power leads) is a relatively simple way to get there.
Yep - that's what it is. Pretty simple circuit with widely available parts. Wished I had an old RE/P or dB magazine with the original MCI ad. It was pretty cheesy.

MCI had to replace a HV op amp, the HA2-2645, which they called a 2002: http://www.ka-electronics.com/images/pdf/ha2-2645.pdf It's still made and very expensive.

There's also the Jim Williams Booster series: http://www.ka-electronics.com/images/pd ... rcuits.pdf

Figure 4 has the voltage gain stage: http://www.ka-electronics.com/images/pd ... pdf#page=3 Or, you could use tubes (figure 5). See also: viewtopic.php?f=12&t=13
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Signal generator output ranges

Post by JR. »

That jim Williams article read my mind...

I would consider something like his +/- 100v swing booster but I would change it from pure class B to class AB for less crossover distortion at HF. You will need to beef the output stage for the extra current and be careful about class A output stage dissipation, but this will get you beucoup swing and should be decent.

One trick I used to use in these non-inverting voltage boosters was to cap couple between the opamp output to the interstage drivers. This limits your max slew rate to the opamps but helps scrub off excess voltage gain at HF and helps stabilize the external feedback loop.

JR

PS: I was also going to mention the LM391 but that's probably obsolete and doesn't swing very close to the rails.

PPS: I used to write a column in RE/P so have some old issues laying around.
Cancel the "cancel culture", do not support mob hatred.
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 »

I used to write a column in RE/P so have some old issues laying around.
John - Thanks for posting here in this neck of the woods. I think that ad ran in '77 or '78. MCI treated it like a major breakthrough. I think it was a cartoon but everytime I saw it I thought of Little Feat's "Dancing Shoes" album cover.

I've wanted to experiment with adding voltage gain to the 1646 output. There's something to be said for the stiff bias (not just the huge class-A portion) of the 1646 headphone amp. I was looking at Williams' circuit and was thinking how to break it down to take advantage of the 1646's differntial output. That common-base stage at the op amp output is interesting. I'm not real good at this BJT stuff 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 »

Just had an idea. Take this current booster circuit with a 1646 (shown below) and make it a swinging op amp by voltage bootstrapping the rails as in the MCI drawing:

Image

This gives you lots of output current, the bootstrap transistors give you added voltage swing. You can't bootstrap a 1646 with a differential output but you can with the "common mode drive" connection. A 5534 might only be good for 35 mA typ drive, a 2114 about twice that, and a 1646 about 3 times the 5534's capability.
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Signal generator output ranges

Post by JR. »

mediatechnology wrote:
I used to write a column in RE/P so have some old issues laying around.
John - Thanks for posting here in this neck of the woods. I think that ad ran in '77 or '78. MCI treated it like a major breakthrough. I think it was a cartoon but everytime I saw it I thought of Little Feat's "Dancing Shoes" album cover.
I don't know if I saved anything that far back.. I think I started writing for them in 1980.
That common-base stage at the op amp output is interesting. I'm not real good at this BJT stuff though.
Discrete design is one of my tools, and the article design is probably more parts than are needed. The simple approach is replicate the very last stages of an opamp in discrete parts. So we need an emitter follower Class A or AB if worried about efficinecy. Then we need the level shifter, generally a common emitter stage on one rail and a current source on the other. The opamp can drive this level shifter stage directly, or just finish with an differential input stage and make a DOA, but it will be easier to get kick ass performance with an off the shelf opamp front end.

I could scratch something up this weekend if there is interest, but probably not worth it for a one off.

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