Raindirk Preamps, too dark and undetailed. Change TBA231 IC to NE5532?

[Ted Krotkiewski]

Hi friends.
I have a lovely old Raindirk Mini Mixer from 1973 (See attached Pic), wich sounds fantastic when using the line in and it´s inductor EQ´s.
The Mic preamp, sounds a bit on the darker side and lack a bit of detail for some sources. The dark and thick tone of the desk is the reason I fell in love with it in the first place. However, sometimes, tracking a whole band, I would like to have the clarity and detail on some sources.
For example, I also have 2 Raindirk Channelstrips from a Series 3 desk (from Olympic studios), wich sounds a bit more open and detailed while still feeling dark and woody in a very nice way if that makes sense?
Both the Series 3 and my Mini Mixer was manufactured during the same time period.

* The Mini Mixer uses an old IC called TBA231 (UA739). that I suspect is the culprit here. Line input also uses the same IC, but when feeding it with already high-end material I love the sound it imparts.
* The Series 3 uses the TDA 1034 (Equivalent to NE5532 from what I know)
(see attached schematics below)

So I suppose my 1st question is, do you think I will notice any major difference in swapping the old TBA231 Op-amp to a new NE5532???

If Yes (or maybe) on the above Q:
The next question is about the common oscillation issue that might occur when updating to new Opamps.
There is a super interesting article from Eddie Ciletti who once cerviced an almost identical Mini mixer, where he talks about oscillation when replacing the old IC. Its really an enjoyable read in general! Read it HERE

In the article he says the following:
One obstacle to upgrading op amps is that their extended high-frequency response at high gain settings can emphasize other shortcomings in the design, like the circuit board layout or power distribution. Troubleshooting Mystery One’s asymmetrical headroom issues revealed oscillation when cranking the gain and/or boosting the HF EQ. The typical approach to solving oscillation is adding power supply bypass capacitors (0.1 uF to 10 uF, and sometimes both) at the IC sockets, which reduced the problem but didn’t eliminate it.

So in this case of me adding a NE5532 at the preamp section, where should I add the Cap? at the + rail? on both? or where?

Again, I really REALLY enjoy the sound of this little console, So I just want wee more "Resolution" in the mic preamp on 4 channels or so.

I have also thought about the alternative of building 5 channels of Helios 2128 mic preamp for 5 channels, so that I can toggle between different flavors... But thats for another day,... and thread.
The Series 3 mixer replaced the Helios mixers at Olympic back in the day.

If you have any ideas, pleas just throw them at me.

Thank you for your time and help!

 

[Voyager10]

Describing something as 'dark' or lacking detail suggests that the high-frequency response is limited. Personally I'd try to measure it before heating up the soldering iron.

Are you able to do this? The simplest thing would be to use a phone with a signal generator app, feeding in to a DI box, through the desk and then into an audio interface line input. You set the gain for some fixed level (e.g. -10dBFS) using a 1KHz tone, then look at the level for say 10KHz and 15KHz. If it starts dropping below the -10dBFS level, there could be an issue.

(Obviously there could be problems with the phone, DI box or audio interface, so you might need to repeat the process with another desk, or with the DI going into the interface directly).

 

[k brown]

Hi friends.
I have a lovely old Raindirk Mini Mixer from 1973 (See attached Pic), wich sounds fantastic when using the line in and it´s inductor EQ´s.
The Mic preamp, sounds a bit on the darker side and lack a bit of detail for some sources. The dark and thick tone of the desk is the reason I fell in love with it in the first place. However, sometimes, tracking a whole band, I would like to have the clarity and detail on some sources.
For example, I also have 2 Raindirk Channelstrips from a Series 3 desk (from Olympic studios), wich sounds a bit more open and detailed while still feeling dark and woody in a very nice way if that makes sense?
Both the Series 3 and my Mini Mixer was manufactured during the same time period.

* The Mini Mixer uses an old IC called TBA231 (UA739). that I suspect is the culprit here. Line input also uses the same IC, but when feeding it with already high-end material I love the sound it imparts.
* The Series 3 uses the TDA 1034 (Equivalent to NE5532 from what I know)
(see attached schematics below)

So I suppose my 1st question is, do you think I will notice any major difference in swapping the old TBA231 Op-amp to a new NE5532???

If Yes (or maybe) on the above Q:
The next question is about the common oscillation issue that might occur when updating to new Opamps.
There is a super interesting article from Eddie Ciletti who once cerviced an almost identical Mini mixer, where he talks about oscillation when replacing the old IC. Its really an enjoyable read in general! Read it HERE

In the article he says the following:
One obstacle to upgrading op amps is that their extended high-frequency response at high gain settings can emphasize other shortcomings in the design, like the circuit board layout or power distribution. Troubleshooting Mystery One’s asymmetrical headroom issues revealed oscillation when cranking the gain and/or boosting the HF EQ. The typical approach to solving oscillation is adding power supply bypass capacitors (0.1 uF to 10 uF, and sometimes both) at the IC sockets, which reduced the problem but didn’t eliminate it.

So in this case of me adding a NE5532 at the preamp section, where should I add the Cap? at the + rail? on both? or where?

Again, I really REALLY enjoy the sound of this little console, So I just want wee more "Resolution" in the mic preamp on 4 channels or so.

I have also thought about the alternative of building 5 channels of Helios 2128 mic preamp for 5 channels, so that I can toggle between different flavors... But thats for another day,... and thread.
The Series 3 mixer replaced the Helios mixers at Olympic back in the day.

If you have any ideas, pleas just throw them at me.

Thank you for your time and help!

If your description of the sound is fairly accurate, I'd expect the particular op amp used to be least reason for it.

 

[ccaudle]

do you think I will notice any major difference in swapping the old TBA231 Op-amp to a new NE5532

The packages are different, you can't just swap one for the other. uA749 is in a 14 pin package with external compensation (the resistor and capacitor connected between pins 3 and 4), while a NE5532 is an 8-pin package, internally compensated so it doesn't need the additional pins to connect external compensation components.

 

[Voyager10]

If your description of the sound is fairly accurate, I'd expect the particular op amp used to be least reason for it.

Yes, I'd tend to agree. Only if there is a HF rolloff which varies with the gain setting would you suspect the GBW figure might be insufficient.

I'd take a look at that 12K resistor across the input transformer secondary. If the tx is giving 10dB voltage gain its impedance ratio will be 1:10, so the load presented to the mic will be 1K or so, which is low-ish compared to many preamps. I can imagine ribbon or tube mics sounding different here.

 

[Dylan W]

Generally agree about diminishing returns in op amp rolling, but uA749 is from the same era as the 741 and has a slew rate of 1V/uS, so would not be super surprised if changing to a 5532 opens up the sound.

You might also look into adding/redoing the decoupling scheme. When I’ve done that on legacy products, I would describe the improvement as “more detailed” or less veiled. Doesn’t change the character of the circuit, just makes it so that you can actually hear the detail in reverb tails, existing distortion, etc.

There’s a great thread on it here:

https://groupdiy.com/threads/opamps-and-local-decoupling-of-rails-some-questions.37307/

 

[Migs 31]

* The Mini Mixer uses an old IC called TBA231 (UA739). that I suspect is the culprit here. Line input also uses the same IC, but when feeding it with already high-end material I love the sound it imparts.

Hi Ted,
Why don't you ask:

@Cyril Jones

He is a member here and would definitely know what to do.

M

 

[Voyager10]

Relevant to slew rate is this plot from the datasheet:



The ".0082" value for the compensation capacitor puts us somewhere between the first two curves, so slew-rate limiting is certainly a possibility in this circuit.

It doesn't really match "dark":though. I'd still like to measure things rather than just guessing.

 

[Ted Krotkiewski]

Thank you so much friends, you are awesome.
I will start by doing some Sine sweeps through the channels and report back with my findings.

 

[Michael Tibes]

I did swap 741s for 5534s in a different mixer from the same era and it was a night and day difference. Just as encouragement to try that route

 

[Somkereki]

I built a 2 x 600 W power amplifier back in 1983. I chose the TBA231 IC for its input because it achieved the best (126 dB) signal-to-noise ratio. (Back then, they liked to use it in phono preampliphier stages because of its low noise.) They are still used today because everyone likes its sound. The NE5532 is much easier to use because it doesn't require so many external components. I couldn't analyze the distortion spectrum image back then, but now it's made me curious, I'll take a look I also had similar problems with high-frequency oscillation. I used ATES - SGS (Italy) ones because they were better than the UA739. (Back then, the parameters were very different depending on the manufacturer.) I wish you much success and a happy new year!

 

[ruffrecords]

The problem with the TBA231 is its 1V/uS slew rate. A 20KHz rail to rail signal needs to swing from zero to the positive rail in 12.5uS. At 1V/uS it will only swing 12V. This probably explains the "dark" sound.

Cheers

Ian

 

[Somkereki]

I agree! And the input impedance is only max 150 KOhm. Although the 0.5 usec 741 IC was also used for audio purposes. I remember we scolded the manufacturers a lot.

 

[JohnRoberts]

Slew rate is not as much of a limitation as many expect. More is always better unless it is way more than you need.

A good, while still conservative way to look at this is power bandwidth, or how much rail to rail high frequency signal can you support? Another consideration is how much rail to rail 20 kHz audio will you ever encounter? Perhaps if close miking a crash cymbal but no real music has that much HF, that loud.

Start by measuring frequency response of all the inputs. Look first for weak LF response, evidence of dried out electrolytic capacitors. Experiment with upgrading a few op amps and compare. You can perform null tests between two different strips by inverting the polarity of one than summing them together into a bus. Of course try this with unmodified strips to learn a baseline of what to expect.

JR

 

[Somkereki]

It's good that you mentioned electrolytic capacitors! I had a lot of problems with them. I don't like using them for audio signal coupling. I prefer foil capacitors. If it's unavoidable due to the high capacitance value, I always put a good quality ~1uF foil capacitor in parallel. They can cause a large phase delay around 20 KHz.

 

[Newmarket]

It's good that you mentioned electrolytic capacitors! I had a lot of problems with them. I don't like using them for audio signal coupling. I prefer foil capacitors. If it's unavoidable due to the high capacitance value, I always put a good quality ~1uF foil capacitor in parallel. They can cause a large phase delay around 20 KHz.

What sort of value electrolytic are you talking about wrt bypassing with a 1uF folio cap ?

 

[JohnRoberts]

It's good that you mentioned electrolytic capacitors! I had a lot of problems with them. I don't like using them for audio signal coupling. I prefer foil capacitors. If it's unavoidable due to the high capacitance value, I always put a good quality ~1uF foil capacitor in parallel. They can cause a large phase delay around 20 KHz.

I've shared this before... I did some bench work back in the 70s and found the paralleled film cap needs to be at least 10% of the total value to improve ESL (cause of HF phase shift). IMO This is not much of an issue with modern electrolytic caps lightly loaded.

JR

 

[Newmarket]

I've shared this before... I did some bench work back in the 70s and found the paralleled film cap needs to be at least 10% of the total value to improve ESL (cause of HF phase shift). IMO This is not much of an issue with modern electrolytic caps lightly loaded.

JR

Yes. I've seen you report this previously. And other sources staying similar. And was thinking about that wrt 1uF bypass. Since I'm expecting electrolytics to be , say, 47uF up into the 100s.

 

[JohnRoberts]

Yes. I've seen you report this previously. And other sources staying similar. And was thinking about that wrt 1uF bypass. Since I'm expecting electrolytics to be , say, 47uF up into the 100s.

10-20uF is large enough as long as not loaded down too heavily. Another reason to tune the DC blocking poles way below audio.

The circuit that lead me to investigate this was an RIAA gain stage with something like a 22 uF feeding a 360 ohm resistor in the gain leg... I measured (didn't hear) tens of degrees of phase shift at 20kHz caused by the electrolytic capacitors ESL. I repeat modern capacitors are much better, and I have avoided loading electrolytic capacitors with impedances like that ever since.

JR

 

[Newmarket]

Def agree that modern electrolytics are much better. And yes tune the DC blocking low. My stuff on this is for my DIY stuff so I tend to "go large" with the value using pretty much whatever I have to hand. Not that I've done much recently tbh due to time and facilities.