NYD eq (was: gain block for a passive EQ)

Sounds like a nice project to me Jaakko.

I'm sorry I don't have much to add but I really fancied doing this very same thing with the TF1 output stage and matching transformer from Pheonix Audio UK.

I'm pretty sure it would work well in a passive EQ design. I think it uses the same pinout as the Neve BA440. Supposedly a nice clean Class-A replacement part for the A/B Neve.

I didn't get that far with this and don't really have the money for it now but thats my plan. After hearing the smooth sound of the preamp I would love to use this gain block in an EQ.....not sure how transparent it would be but the pre sounds very airy...maybe suitable for a mastering EQ.

I'm curious about Joe Malones 99V and forthcoming 99S as well.....

Wouldn't transformerless be a better option if you want clean?

Good luck with it.
Cheers Tom

A NTP M-100 opamp?

My ears are burning... someone must have mentioned my name :wink:

First of all, I've been planning to post a revised table of values for my EQ circuit, values that give more useful, less-broad curves than what I posted originally. I'll try to do it soon.

I suggested a 10K:600 stepdown transformer for my EQ for three reasons. First, it provides a higher input impedance to the source, which is preferable for most modern equipment. Secondly, it reflects a much lower source impedance to the EQ circuit, which maximizes the amount of "boost" available. My EQ works properly with source impedances of 60 ohms or less; with the stepdown transformer, you can use equipment having source impedances of up to 1K. Thirdly, it drops the input level by 12dB, reducing the risk of saturating the inductors in the EQ circuit. The total makeup gain required is 12dB (stepdown of the input transformer) plus the loss of the EQ circuit at its "flat" setting (10dB in the case of my EQ), plus the stepdown of the output transformer, if one is used.

The same concept should work well with other passive EQs such as the Pultec. The Pultec in particular, in its stock form, has an input impedance which can dip to a very low value at certain settings, and most newer gear will not be happy driving it.

As for gain makeup: if you want a clean gain block that can work well at low impedances--remember, my EQ and most of the passive EQs we talk about are nominal 600-ohm circuits--then it's really hard to beat the JH 990. Don't use the Hampt0ne if you're looking for clean gain; it's a "color" amp.

When you said clean the Forssell 992 comes to my mind. Sooo clean...
Do you have any spare or they are all in the Sontec?

Anybody try the Lange PEQ amp?
Link at the Pultec metea.
Just scored an output tranny off eveil bay.
Thanks to Tommypiper for using his account!
(I maxed out on vintage iron, so i got the boot for non payment issues.(

I used to own a Lang. It was great, but not as useful as I would have liked for what I do. I traded it for a pair of API 550As.

:thumb:

yeah, John's 990 is about $50 if I remember correctly. A very good value for a high-performance amp. If I had the money and the free time, I'd build a whole console with 'em

I have to break out the calculator and verify this, but I think you can change it to 1dB steps simply by dividing the values of the resistors in the "L" section by two (and using twice as many resistors, of course!).

I'll post some suggested values for L-C networks later tonight.

The reason I wrote "three bands maximum" is because it would be difficult, unless using very high-Q tuned circuits (which have their own problems), to use more than three bands without excessive interaction between them.

I'm still planning to post a new chart of suggested L-C values. I just haven't had a chance to do it yet. I took tomorrow off, so maybe I'll have some time later tonight or tomorrow.

By the way, one thing that seemed to confuse a few people when i first posted this circuit is that there is only ONE set of resistors as shown on the diagram--not one per band. All bands share the same "L" attenuator section.

The reason I wrote "three bands maximum" is because it would be difficult, unless using very high-Q tuned circuits (which have their own problems), to use more than three bands without excessive interaction between them.

Click to expand...

How does this interaction look like?

Samuel

How does this interaction look like?

Click to expand...

No help on this one?

Samuel

I'd imagine it to be excessive "widening" of the Q as the bands overlap?

Click to expand...

Ah, not as bad as the resonances I suspected. Any explanation for this? I guess our imperfect inductors are to blame?

Samuel

It's not as complicated as you guys think. It's simply due to the fact that when boost/cut curves overlap, they interfere with one another, cancel out to a certain extent. This isn't any weird parasitic effect of the inductors; this is true even of active EQs that contain no inductors, if the bands are spaced so that they overlap.

The reason why you're bound to run into this if you try to expand a passive EQ beyond three bands is because it's difficult to get the cut/boost curves narrow enough without using high-Q tuned circuits, which have problems of their own in an application like this. You don't want your filters "ringing" except perhaps as a special effect in a synth or whatever...

Sorry I haven't been more forthcoming with information, but I'm busy getting ready for the holiday. I hope to have something useful to post here soon.

It may seem like a monologue sometimes, but I (and many others) are lurking back here learning what things are going on. I am very interested in several EQ types and will eventually buils a couple of different types to play with. The NYD EQ seems to be a really good candidate for a versitile, easy to make EQ.

I tend to get easily swept up in design frenzy and overcomplicate things to the point they end up with so many options they are actually useless. That being said, It is usually better for all concerned that I stay back a bit and ask a question or two to clarify a point rather than spew my usual nonesense and complicate things.

Again, I think this will be a good project for me if I can just keep my hands off the controls. I am wanting to use a simple vacuum tube makeup and transformer coupling. What would be a suggestion for something along these lines?

Just want to say with the inductors I used (UTC) the high end boost is not dramatic. In order to get a decent Q or slope with a high shelf the resistance added tends to cut down the level of boost. I feel its a good amount of boost and if I were to use a 2:1 input instead of a 1:1 and more makeup gain I might be able to get a bit more out of it.

the NYD design is very easy to play around with and I changed around quite a few caps before I settled on frequencys I liked.

just make sure the make up gain is sufficient. I used the SRPP stage and with the 600:600 input tx it is just about unity. I would have liked to have more but I was guessing when I started.

Over all I really like this design....

Sorry I haven't been able to offer much input. I've been busy. But here are two drawings which will help to guide your efforts.

Equivalent circuits and notes

EQ schematic

Whoops... I wrote down the wrong value for the series arm in the attenuator in the "equivalent circuits" diagram. I had 12dB stuck in my mind for some reason, instead of 10dB. Anyway, the diagram is corrected now.

For the mid band, you'll find that you'll get nice narrow (but not too narrow) curves by choosing an inductor and cap that each have a reactance of about 600 ohms at the center frequency.

As it says on the equiv. circuits diagram, L or C with a reactance of about 180-200 ohms seems to work best for shelving.

You ended up with values very close to, or the same as, the "revised" values I came up with (and never got around to posting! :roll: ). This isn't surprising since you kept the reactances right around where I advised.

I'm not planning to do nearly as many frequencies, though, and only three bands:
Low: 40Hz, 60Hz, 100Hz
Mid: 250Hz, 1kHz, 3kHz
High: 5kHz, 10kHz, 15kHz

These frequencies are simply ones I consider "useful" in a basic EQ.

The frequencies in your chart are spaced very close together. Are you planning to use all of them? You're not going to hear much of a difference as you switch from one frequency to the next.