Point of oscillators in vintage mixers?

[user 37518]

P.S. I remember some engineers used the console's oscillator along with a kick drum side-chained gate to create the '808 effect'. They claimed the console's oscillator sounded better than the Protools osc. There is a basis to this, usually, the console's oscillator was not very high-end, probably a good ol' Wien bridge osc, which produced considerably more distortion than the software version. I guess this is what they liked about it.

 

[ruffrecords]

Well, I’m the old days of crt monitors, was in school and talking to the studio tech basically said there is something wrong in studio”insert letter”. He asked what I was talking about and I explained how it felt weird in there. He laughed and said turn off the monitor. When it doesn’t have signal but is on, it’s black screen, bring on emits high frequency which is most likely what you noticed and way it was weird. He was not wrong.

When I was a kid I could easily hear the TV line output transformer whistling at 15KHz. 60 years later I have tinnitus at the same frequency even though my ears don't get much past 8KHz these days.

Cheers

Ian

 

[Tubetec]

We had a couple of Ns-10' blow on us , it wasnt the console oscillator ,
it was a certain combination of the inline function flip switches on the desk and tape monitor select and the whole thing became an oscillator ,
The Amcron managed to push the voice coil of the woofer out of the magnetic gap where it got stuck on one occassion , the amp didnt even blink .
We lost a few tweeters as well they just go out instantly in tiny puff of smoke ,
We always had spares so Id arrive down with the meter and tool box and we'd be back up and running in the time it took to boil the kettle and make a brew .

 

[JohnRoberts]

The CRT monitor raster noise was something like 15.75kHz... One of the weirdest noise troubleshooting incidents I encountered was when computer monitors in two different design engineers next door offices were interfering (beating) with each other. They were both free running about 1 Hz apart. This caused a very odd and distinct 1Hz signal in the noise floor of a powered mixer amplifier stage. Turning off either one of the two monitors stopped the 1Hz noise.

JR

 

[JohnRoberts]

P.S. I remember some engineers used the console's oscillator along with a kick drum side-chained gate to create the '808 effect'. They claimed the console's oscillator sounded better than the Protools osc. There is a basis to this, usually, the console's oscillator was not very high-end, probably a good ol' Wien bridge osc, which produced considerably more distortion than the software version. I guess this is what they liked about it.

The typical console line up oscillators were cheap diode break sine wave generators... I forget the part numbers now (8038? ) there was more than one... Not low distortion, but good pitch and amplitude stability.

JR

 

[AnalogPackrat]

XR2206 was another function gen IC of that era. Also LM566 could have triangle output "waveshaped" into sine-like for cheaper.

 

[pucho812]

The CRT monitor raster noise was something like 15.75kHz... One of the weirdest noise troubleshooting incidents I encountered was when computer monitors in two different design engineers next door offices were interfering (beating) with each other. They were both free running about 1 Hz apart. This caused a very odd and distinct 1Hz signal in the noise floor of a powered mixer amplifier stage. Turning off either one of the two monitors stopped the 1Hz noise.

JR

At a post production facility I recently was at they purposely had 2 giant machine rooms because different pieces of gear had fans which Would oscillate at different speeds andp get the beating effect.

 

[pucho812]

For what it’s worth.
The neve v series desks can easily have an oscillator in every channel if you recap it incorrectly. Recapped eq modules, and the caps were physically too big causing the leads to come in contact with other leads creating an oscillator that was tunable

 

[user 37518]

XR2206 was another function gen IC of that era. Also LM566 could have triangle output "waveshaped" into sine-like for cheaper.

The XR2206 is still used and can be found regularly on eBay Chinese kits. My students usually go a long way trying to find them. There is a local supplier which still has a lot of them in stock, I could bet that students/hobbyists are the only ones buying them today.

That brings me to another hypothesis of mine: LM741s are still being produced, mostly because it is the academic op-amp par excellence used to teach EE undergrads about op-amps in introductory courses. I have no way to prove this, though.

 

[AnalogPackrat]

The XR2206 is still used and can be found regularly on eBay Chinese kits. My students usually go a long way trying to find them. There is a local supplier which still has a lot of them in stock, I could bet that students/hobbyists are the only ones buying them today.

Also popular in synth DIY projects, especially the Thomas Henry 1v/Oct VCO. I have a small stash of them.

That brings me to another hypothesis of mine: LM741s are still being produced, mostly because it is the academic op-amp par excellence used to teach EE undergrads about op-amps in introductory courses. I have no way to prove this, though.

I think there are some low-end far east fabs cranking out generic op-amps based on several old designs using reverse engineered mask sets and who knows what process. Maybe the same place that cranks out the fake small die "2N3055" transistors.

 

[Brian Roth]

Back In The Day (1977), my then-employer and I decided to get into the console biz. I used an Intersil 8038 chip for the desk oscillator. It was a very new part, but performed quite well if you followed the app notes with added trimmer parts.

That chip was a function generator with square, triangle, and sine outputs. The latter was indeed created with a "diode bender" circuit. Distortion wasn't ultra low (around 0.5% when carefully trimmed IIRC) but was Plenty Good Enough for a "line up" generator.

The main advantage was super flat frequency response over the adjustment range and ultra-fast settling time when adjusting frequency.

FWIW, Jay McKnight (RIP, owner of the MRL calibration tapes company) built his own custom function generator to create the test tones. Jay wasn't concerned about the "slight" amount of THD but was most concerned about flat signal levels when changing frequencies.

I think the XR2206 was a newer chip.

Bri

 

[user 37518]

Back In The Day (1977), my then-employer and I decided to get into the console biz. I used an Intersil 8038 chip for the desk oscillator. It was a very new part, but performed quite well if you followed the app notes with added trimmer parts.

That chip was a function generator with square, triangle, and sine outputs. The latter was indeed created with a "diode bender" circuit. Distortion wasn't ultra low (around 0.5% when carefully trimmed IIRC) but was Plenty Good Enough for a "line up" generator.

The main advantage was super flat frequency response over the adjustment range and ultra-fast settling time when adjusting frequency.

FWIW, Jay McKnight (RIP, owner of the MRL calibration tapes company) built his own custom function generator to create the test tones. Jay wasn't concerned about the "slight" amount of THD but was most concerned about flat signal levels when changing frequencies.

I think the XR2206 was a newer chip.

Bri

IIRC the ICL8038 uses a triangular to sine wave converter, which inherently produces a fair bit of distortion. I don't know if the XR2206 uses one as well; most likely it does since both chips are very similar.

Regarding ultra-low distortion oscillators the Tektronix SG-505 had < 0.0008% THD which, for many years, made it the undisputed champion. The current AP systems can achieve figures lower than that, and there are also some amateur designs which claim much better specs (< 0.00001%) than the SG-505. The problem, as always, is not just making an ultra low distortion oscillator, but rather a computer controlled ultra low distortion oscillator sweepable beyond the audio range, with linear gain vs freq response.

 

[Brian Roth]

I'm 99% sure the XR2206 uses a "diode bender" to create the sine output from the inherent triangle waveform, same as the ICL8038.

For a "setup oscillator" in a desk, ultra-low distortion is down the list. The important things are very flat amplitude and quick settling time as the frequency is adjusted.

I distinctly recall bench oscillators from reputable manufacturers which would take MANY seconds to settle down to final amplitude, especially when switched into the lower frequency ranges.

For that reason, for decades I've used a portable function generator for aligning analog recorders. I want to be able to quickly jump from one end of the audio spectrum without waiting for the damn thing <g> to settle down at the final amplitude level. Many desk oscillators I've encountered were slow to settle....and some would vary +/- a good fraction of a dB between 20 Hz and 20 kHz.

We're not requiring 0.0000001% THD for frequency response checks.

As I mentioned earlier, Jay McKnight at the MRL test tape company gave me the idea. If a function generator with "diode benders" to make a sine from a triangle was good enough for him....good enough for me! lol

A 0.5% THD level doesn't muck with the ability to look at overall frequency response.

Bri

 

[user 37518]

I'm 99% sure the XR2206 uses a "diode bender" to create the sine output from the inherent triangle waveform, same as the ICL8038.

For a "setup oscillator" in a desk, ultra-low distortion is down the list. The important things are very flat amplitude and quick settling time as the frequency is adjusted.

I distinctly recall bench oscillators from reputable manufacturers which would take MANY seconds to settle down to final amplitude, especially when switched into the lower frequency ranges.

For that reason, for decades I've used a portable function generator for aligning analog recorders. I want to be able to quickly jump from one end of the audio spectrum without waiting for the damn thing <g> to settle down at the final amplitude level. Many desk oscillators I've encountered were slow to settle....and some would vary +/- a good fraction of a dB between 20 Hz and 20 kHz.

We're not requiring 0.0000001% THD for frequency response checks.

As I mentioned earlier, Jay McKnight at the MRL test tape company gave me the idea. If a function generator with "diode benders" to make a sine from a triangle was good enough for him....good enough for me! lol

A 0.5% THD level doesn't muck with the ability to look at overall frequency response.

Bri

I agree with you. I was thinking more in the line of measuring THD VS frequency. Which I realize is probably a good topic to start derailing the thread. My apologies.

P.S. I've also noticed some commercial function generators, which achieve THD figures around 0.1% when turned on and slowly start reducing towards around 0.05% after some minutes.

The slow amplitude settling time might still be common with very cheap 100% analog function generators, but even the modest ones today have very fast settling times.

I have a battery operated BK Precision "pocket" function generator which has surprisingly low THD and switchable frequencies. I don't know about the level vs frequency variations, but I haven't calibrated a tape machine for the last 15 years. And if I need more precision I can always reach for the big guns.

 

[Brian Roth]

user 37518, I was on a Heathkit "building binge" in the 1980's because I was recently divorced and poor!

One kit was as shown here:

https://reverb.com/item/56089728-he...r-lights-up-no-leads-no-way-to-test-cipollina
Mine still works perfectly and is my go-to for analog recorder alignment. Yes, some of us still deal with analog recorders in 2023.....lol!

Only downside to that model generator is lack of a calibrated output level. But, with a BNC "T" connector and my external DVM....no problem besides an extra step. I am so familiar with that Heath that I can "blindly" twist the knobs to set a level and be within a couple of dB above/below +4 dBu!

Mr. Fluke nails the final correct voltage for me.

Ultra-flat frequency response.....maybe +/- 0.1 dB or better from below 20 Hz to well over 20kHz. Instant settling as I switch/sweep across the entire spectrum.


Bri

 

[AnalogPackrat]

Back In The Day (1977), my then-employer and I decided to get into the console biz. I used an Intersil 8038 chip for the desk oscillator. It was a very new part, but performed quite well if you followed the app notes with added trimmer parts.

That chip was a function generator with square, triangle, and sine outputs. The latter was indeed created with a "diode bender" circuit. Distortion wasn't ultra low (around 0.5% when carefully trimmed IIRC) but was Plenty Good Enough for a "line up" generator.

...

I think the XR2206 was a newer chip.

Bri

On another forum a guy reports designing a product using XR2206 on '78, so if they were newer it wasn't by much. Also seeing lots of complaints that the cloned parts are not near spec on several counts. Caveat Emptor.

 

[JohnRoberts]

On another forum a guy reports designing a product using XR2206 on '78, so if they were newer it wasn't by much. Also seeing lots of complaints that the cloned parts are not near spec on several counts. Caveat Emptor.

My (fuzzy) recollection was that the 2206 was a later part. These were widely used for general purpose function generators.

JR