Limiters & VU meters

[JR.]

Different... .5% distortion, noise floor only -80dB below 1.6V, I am too lazy to look up the NJM part again.

What's wrong with a THAT based solution... At least the audio path performance will be decent.

JR

[carlmart]

The only problem with the THAT solution is that they are not able to answer some of the questions I'm asking. At least until now.

The latest design I made, following their recommendation on which design to start from, still needs adjusting threshold and gain, which even after reading DN00A is not clear to me how to do it and replace those trims with fixed resistors.

Let's take the Threshold trimmer. It goes from 30dBu to 10dBu, controlled by 50K pot. How do I convert that so I get to 20.1 threshold hard knee?

Another thing which is not quite clear to me is the gain setting, as I do not wish to have a trimmer there either.

[JR.]

The only problem with the THAT solution is that they are not able to answer some of the questions I'm asking. At least until now.

The latest design I made, following their recommendation on which design to start from, still needs adjusting threshold and gain, which even after reading DN00A is not clear to me how to do it and replace those trims with fixed resistors.

Let's take the Threshold trimmer. It goes from 30dBu to 10dBu, controlled by 50K pot. How do I convert that so I get to 20.1 threshold hard knee?

Another thing which is not quite clear to me is the gain setting, as I do not wish to have a trimmer there either.

Since I don't think I can teach you electronic design, perhaps put the trim pot in there, adjust it for the result you want, then measure the pot.

You say you have an EE friend, maybe get him to explain it to you.

JR

[carlmart]

Yes, my EE partner is getting into it this week.

I wasn't asking how to replace a trimpot with resistors, as that much I do know.

But what about telling me how do I translate a 20.1 threshold into volts terms that I can perhaps compute with a resistor divider?

Or why do I need that gain trimmer for if I could get my gain in U3A using R17 and R16. Why the need to inject DC voltage into pin 2.

[ricardo]

Since I don't think I can teach you electronic design, perhaps put the trim pot in there, adjust it for the result you want, then measure the pot.

You say you have an EE friend, maybe get him to explain it to you.

I second that.

If your friend can't work out ATTACK, RELEASE & THRESHOLD from the clearly marked THAT schematics, you might be better off with the NJM circuits. There's only 2 things to adjust and they have graphs for threshold which might be easier to understand than THAT eqns.

If you can't work it out, you might as well just use the NJM circuits as they stand. Tell us how they sound.

If the NJM has fast attack & slow release (and I can't think of any reason why they would do differently), it should sound OK with the correct capacitor

[carlmart]

Let me see if I understand it, so there's no confusion.

You do know how to answer my questions but simply choose not to? Why would I then come back here then and tell you how do the NJM sounds?

It's quite likely we will solve these questions with the THAT schematic, and I will certainly compare them with the NJM, but I don't see why I should share this with you.

Because the subjective audio quality is probably the only thing I can really make a difference at, as your technical knowledge (certainly superior to mine) could make a difference. That's my trained ear working, where specs are not the only thing that matters. It looks as if you disregard that by refusing to help me, to solve my issues, as if I was too lazy to do it myself.

It's very sad things look that way to me.

[ricardo]

You do know how to answer my questions but simply choose not to? Why would I then come back here then and tell you how do the NJM sounds?

Carl, some of us have tried very hard to provide helpful answers to your questions.

It is obvious from some of your questions that you need to do a LOT of homework on limiters & compressors (on concepts like ATTACK, RELEASE,TRESHOLD, RATIO etc) before you can make any sense of our poor attempts to educate you.

But I've mentioned GAIN STRUCTURE and again, you seem to have little, if any, idea of what I mean. To take you from this level to one which can sensibly make design decisions on compressors/limiters would be akin to starting with Ohm's Law.

I've also mentioned that for ALL such devices, you have to listen to your latest iteration.

My recommendation of the NJM is cos it is the simplest to build, adjust and listen to. There are NO extra circuit bits to confuse you. If it has fast attack & slow release, it is likely to sound good too.

The missing bit in 'our' recommendations is that neither JR or I have listened to either the THAT or NJM devices. If our experience give us any 'knowledge', it is that you must LISTEN to any such device. The moment YOU build & listen to either the THAT or NJM, YOUR experience immediately becomes more important than our recommendations. But is its possible at that stage, we can advise on how to adjust YOUR circuit further to do what you want.

Presently, what you want is quite ambiguous. In return for accurate technical advice (and the recommendation of the NJM is sound technical advice) we would very much like your trained ear's opinion. I think that's a fair swap.
__________

What don't you understand about the NJM datasheet?

Have you shown any of this to your EE friend?

[carlmart]

I do understand the practical, ON THE FIELD, concepts of attack, release, threshold and ratio.

What I may be lacking is the way to translate them into electronic adjustment on the circuit.

As I see it, and correct me if I'm wrong, I have to organize my gain structure between my input gain, the limiter gain and the output gain. Is that so?

If it is, I do have a specific goal at my outputs: +22 dBu / 9.752V with 100k ohm load for line level; -18 dBu / 98mV minimum with 100k ohm load for mic level.

As the output balanced stage does add very little gain, it's quite likely I will have to add another stage after the limiter to get at my goal. I'm not sure if or how I can do that within the limiter itself, or if or why I should.

The only problems I foresee with the NJM part are the specs, which are not that many on the datasheet, particularly the subjective listening ones, are not that great. It's cheaper and simpler than the THAT part, but 1% THD is not something to be too proud of. The rest of the NJM data I understand quite alright, and of course my EE friend has seen it too.

My objective with this THAT project was to get to a primarily limiter project that can be shown as another Design Note that could be useful to other people. The only difference is that I do not have the electronic knowledge necessary to do it by myself. So from the beginning I saw this project as something that could be shared, at least on what and how better to do that using mostly THAT parts.

I'm sorry if by intending that I'm demanding too much of you or abusing your generosity.

Until now I think I've kept my word, which is that everything I'm advancing forward, with your help or THAT's, I'm sharing it here for others to use.

[ricardo]

I do understand the practical, ON THE FIELD, concepts of attack, release, threshold and ratio.

What I may be lacking is the way to translate them into electronic adjustment on the circuit.

If that is the case, JR's advice is pertinent. Build the THAT circuit with trim pots. Twiddle them until they give the sound you want. Convert the trim pots into resistors. And that's accurate & useful advice.

Without hands on experience with the chips, we can't be more specific than that given your level of electronic experience.

Please consider what you are asking. It's design work of quite an advanced nature for a commercial project from which you hope to make money. Your request for a finished design is appropriate to a chip supplier like THAT .. but I have a very good idea of what they must be making of your request.

But there are ways around this too. The NJM is almost certainly aimed at your application and level. A limiter BY DEFINITION distorts the signal. You won't know whether it is good enough for you until you try it and listen. What I DO know is that you will get it working properly (sound as good as it can) much quicker than the THAT solutions.
_________________

BTW, your 'specific output goals' don't make sense either but they are easy enough to achieve.

... and there's no subjective listening tests for the THAT circuits either

[JR.]

The only problem with the THAT solution is that they are not able to answer some of the questions I'm asking. At least until now.

The latest design I made, following their recommendation on which design to start from, still needs adjusting threshold and gain, which even after reading DN00A is not clear to me how to do it and replace those trims with fixed resistors.

Let's take the Threshold trimmer. It goes from 30dBu to 10dBu, controlled by 50K pot. How do I convert that so I get to 20.1 threshold hard knee?

OK, your question does not seem to line up with the design note, but I will try to answer what I think you are asking.

The 50k threshold pot in DN00A spans from -40 dBV to +20 dBV. If you want +20dBV just leave out the pot and connect the wiper resistor to -15V. If you use odd ball PS you will need to tweak values based on rail voltage you have.

If you read the design note they explain how the threshold works. The RMS detector puts out 0V for -10dBV and rises 6mV/dB from there so your +20dBu is roughly 30x6mV or +180mV. We also need to convert to dBu which is a couple dB less than dBV (-2,2dB) so actual +20dBu voltage is 167 mV. Using 2K resistors like the DN and -15V supply the exact resistor value from - input to -15V is 180.7k . Using the DN value (165k) would give you +20dBV not +20dBu

Another thing which is not quite clear to me is the gain setting, as I do not wish to have a trimmer there either.

If you don't want adjustable gain, just don't add that pot and resistor. All it does is add +/- X dB gain, so without it you will have 0 dB.

I think I said several posts ago, to take something like this DN and just remove the features that you do not want.

I hope this answer makes you happy... Your bill is in the mail.

JR