THAT vs TI

[elskardio]

For what it's worth...
I have compared the INA137 vs That 1246 in the same circuit and I couldn't measure (or ear) any differences.
The TI part is easier to find and usually at a better price too.

 

[emrr]

I don't get it. +6dB voltage gain is +6dB voltage gain, whatever the impedance.
IMO the 1646 has better drive capability.

See my earlier post

 

[Matt Syson]

As with everything you have to look at the surroundings that the parts are used in.
The SSM2142 was/is susceptible to 'static' damage (nylon carpets dry atmosphere etc) and after a while later 'application notes' started saying that a 'bridge' of diodes to clamp the outputs to the supply rails was recommended. I haven't looked at the That corp applications recently but I think they may also suggest this. Then you have to consider 'balance' and what frequency range and of course their ability to handle significant capacitive loads (long cables). The 'quality' of either/any of these devices is likely very similar for the usual studio environment with less than 20 metres of fairly low capacitance cable into 10K 'load as opposed to 150 metres of wet string into a 600 Ohm transformer load (Pultec PQ5 for example). Whether the circuit board tracking is properly 'balanced' at HF, has good RF bypass caps on he power rails and of course decent EMI filtering of the output (so reducing/eliminating RF being put back up the output will all have an effect. There is also an element that 'It's not Burr Brown' so it must be rubbish from some quarters. As they may or may not use capacitors in series with their outputs and to the 'internal feedback points' there are yet more possibilities. I find the 'indeterminate' output polarity of some units disappointing as it forces use of bipolar electrolytics. I suppose one could deliberately introduce an offset which would probably result in one output capacitor being opposite way around to the other which would be a shame. If both output pins could be persuaded to sit a touch negative (of ground) then using a pair of 50 Volt electrolytics would make the part 'phantom power proof (assuming you have the catch diodes as well).
As usual doing something properly is complicated. Having them socketed has an advantage that when you 'blow them up' you can replace them relatively easily and less chance of burning the board.
Sadly I don't have the original papers but at one time the BBC had a 'requirement' that mic inputs on OB gear should be able to withstand 240 Volts AC applied common mode to an input. Needless to say only a decent transformer has any chance of handling that mission. The instruction to be able to have 24dB headroom at all settings was badly written as 24dB above a +24dB line level is getting rather 'tingly' (+48dBu!)

 

[abbey road d enfer]

See my earlier post

My question was about what you say in your earlier post, and doesn't make sense to me. Is +6dB into 10k louder than +6 into 600?

 

[emrr]

My question was about what you say in your earlier post, and doesn't make sense to me. Is +6dB into 10k louder than +6 into 600?

Yes. Apples to apples, they do not exhibit the same gain, at least 10 years ago.

 

[Newmarket]

This part of the market is taken by their line of CMOS switches. See ADG419.

Isn't "Analogue Engine" a registered TM of THAT?

yeah - aware of various analogue switch families. But IIRC the SSM devices had the option of easy ramped switching as well as an 'audio friendly' datasheet since that was the specific market.
"Analogue Engine" (probably "Analog" but you know...Standards ) - thinking about it you right I think.
Maybe SSM had some similar phrase. It's a long time back but I'll see if I can spot the book at home.

I do like the THAT stuff btw and have used some Ingenius and Outsmarts in DIY.
Bought from Profusion in UK.
And really good literature - reminds me of SSM.

 

[Newmarket]

well I can give you all that waffle if you like! I have a earlier BG1 that has the TI equivalent in it that no one uses! LOL, but that might have more to do with the weird choices on the front panel of extremely low HP filter at 33htz and only phase switch on the first channel than it does with the op amps inside but who knows?! The BG8 gets all the love and the BG1 notsomuch

Better, but you'll need to do even better if you want someone to suggest that you go and make yourself at home over in "Gearspace"

Seriously though, what you're getting with those mic amps sounds like a properly executed 'straight wire' approach with high quality devices.
It's kind of refreshing

 

[Gold]

yeah - aware of various analogue switch families. But IIRC the SSM devices had the option of easy ramped switching as well as an 'audio friendly' datasheet since that was the specific market.

The SSM2402 was a good part. I don't think there is an equivalent through hole part available. I haven't looked at SMT. Those are what I'm using in my crossfader. You can choose A or B or Crossfade (which is a VCA). I have a lifetime supply for spares.

 

[JohnRoberts]

THAT is a David compared to Goliath TI. TI acquired the BB brand many years ago. THAT are proudly independant. Les Tyler and Gary Hebbert are great guys (ex-dbx).

For too much information Gary Hebert is also a drummer. I loaned him a RESOTUNE and he gave me this quote

I got the old Ludwigs tuned better than they’ve been in years. Thank You….

Indeed I keep expecting THAT corp to get swallowed up by a major. Their VCA business is a declinining asset, but they have other well respected parts.

JR

 

[dyamakuchi]

Their VCA business is a declinining asset

What's eating the VCAs? Digital?

 

[MisterCMRR]

I can assure you that ICs from THAT Corp are of their own design and fabrication - having nothing whatsoever to do with TI. I was working with them after they licensed my InGenius patent in 1996. They had just bought a wafer fab facility in "silicon valley" and were getting it up and running, which was a factor in the long delay getting my parts on the market. They are a very clever and innovative group of engineers. I should also point out that what sets their ICs apart from the rest is that they use dielectric isolation (DI) instead of junction isolation like other makers of audio ICs. DI allows complementary NPN and PNP transistors in the IC to be as good as their discrete counterparts because they're built essentially on a glass substrate. With conventional junction isolation, transistors are isolated by back-biased diodes in a silicon substrate - and PNP transistors are difficult and cannot reach the performance of a discrete PNP. They originally used DI because the VCA technology requires very well matched NPN and PNP pairs (the original Blackmer VCA is technically called a "current ratioing multiplier"). The DI process makes possible low-noise, high slew-rate designs with fewer transistors in the signal path. I think the process is inherently superior and the main reason the ICs "sound great." I was very pleased that they wanted to license my InGenius circuit design. And, yes, they cost a bit more, but that's because the DI process involves more fabrication steps.

 

[JohnRoberts]

What's eating the VCAs? Digital?

Indeed VCA are obsolete when digital audio paths are being used, a simple digital multiply can vary gain up or down with no distortion or added noise.

We still live in a physical world so mic preamps will remain in use for some time. Of course the first company that combines a microphone with A/D convertor, with no signal integrity compromises will dominate, but that is still a wet dream.

JR

@ CMRR many of us in the industry are well familiar with the history of THAT corp, and before that DBX going back to Dave Blackmer's early VCA development. Dave was working at API at the time but reportedly they weren't interested .

 

[Bo Deadly]

they use dielectric isolation (DI) instead of junction isolation like other makers of audio ICs. DI allows complementary NPN and PNP transistors in the IC to be as good as their discrete counterparts because they're built essentially on a glass substrate. With conventional junction isolation, transistors are isolated by back-biased diodes in a silicon substrate - and PNP transistors are difficult and cannot reach the performance of a discrete PNP. They originally used DI because the VCA technology requires very well matched NPN and PNP pairs (the original Blackmer VCA is technically called a "current ratioing multiplier"). The DI process makes possible low-noise, high slew-rate designs with fewer transistors in the signal path.

That's super interesting!

I wonder if a performance (noise) improvement could be had by making discrete blackmer VCAs with matched parts, extra pads for trimming resistors on a little SMT board.

 

[JohnRoberts]

That's super interesting!

I wonder if a performance (noise) improvement could be had by making discrete blackmer VCAs with matched parts, extra pads for trimming resistors on a little SMT board.

They published an AES paper around the time of the last generation THAT corp VCA

this is a free preprint... THAT VCA preprint I seem to recall more stuff in the full AES paper that Gary sent me a copy of back then.

While you might improve upon the original crude Blackmer discrete VCA don't even imagine you could improve upon the latest generation THAT corp IC VCA technology.

JR

PS: One of my too many patents was for a VCA, a not very quiet class A VCA, with no production trims.

 

[Monte McGuire]

I wonder if a performance (noise) improvement could be had by making discrete blackmer VCAs with matched parts, extra pads for trimming resistors on a little SMT board.

Discrete and "well matched" are at odds with each other. A monolithic approach is pretty much the only way to get closely matched transistors reliably, and a dielectric isolation process is the ideal way to get them onto one die without having them talk to each other through reverse biased diodes whose leakage capacitance varies with signal voltages. Why is it good to get them on one die? So they are much closer to isothermal and thus track over temperature and so they are inherently well matched, despite process variations.

 

[abbey road d enfer]

THAT is actually doing +6.7dB unloaded gain whereas TI is doing +6dB unloaded gain. They are compensating for the drop due to the internal 50 ohm series resistance, so you get +6dB into 600 ohm.

That's not what the datasheet says;
Gain unloaded: 6dB nominal
Gain into 600r: 5.30dB nominal
Practically onfirmed many times.

 

[abbey road d enfer]

Considering that Neumann tried and failed, I don't think anyone else is keen to invest in this.

Have they? I haven't seen any announcement by Neumann that they were abandoning their line of digital mics, which are appreciated by many classical recordists and broadcast organizations.

 

[mjrippe]

Have they? I haven't seen any announcement by Neumann that they were abandoning their line of digital mics, which are appreciated by many classical recordists and broadcast organizations.

Not sure how you replied to a deleted post, must have been very fast! Yes, "failed" was a very subjective choice of words by which I meant they have not become broadly popular. And my experience is not in the broadcast or classical fields

 

[ruffrecords]

So far it looks like the INA137 wins for the input because it is capable of driving a 600 ohm load and I have an EQ hanging on it output that could go that low. For the output my first choice is the DRV134 but the more expensive THAT1646 is pin compatible.

Cheers

Ian

 

[john12ax7]

That's not what the datasheet says;
Gain unloaded: 6dB nominal
Gain into 600r: 5.30dB nominal
Practically onfirmed many times.

I had it backwards, it's actually the DRV134 that is the +6.7dB, at least according to this chart.

http://www.thatcorp.com/Line_Driver_Comparison.shtml