Hi everyone,I'm new to this forum.

I'm a sound engineer with very very limited knowledge of electronics.I'm working on a mod for my mixing desk and tried many, mainly audio gear forums, but few people over there seem to have a clue about what I'm trying to achieve.Maybe you can help.
The equalizer section of my desk has three bands: Hi-Frequencies,Mids and Low Frequencies.The Mids are "fully sweepable" semi-parametric, the HF and LF are what we call Shelves or Filters with set frequencies,10KHz for the HF and 100Hz for the LF.I find the lack of frequency choice a bit restraining when comes the time to mix.So I managed to remove the Capacitors responsible for those Frequencies from the PCB(C12 and C13 on the diagram) and replace them by small bread boards with a DPDT on-off-on switch and three Poly. Caps..By putting the caps. in parallel with each other, I manage to get a choice of three different frequencies, for each band.So far so good ,it works.But where I'm stuck is with the LF circuit which is based on a Gyrator circuit.I have read somewhere on the net that you could make that band "fully sweepable" by putting a Potentiometer in series with one of the resistors.I will post that document when I find it.Not to make this post too long, let me show you a diagram of the existing circuit and a few pics. of my work so far.

Any help or direction would be appreciated.Thanks.

 

What is wrong with the microphones or the room to make the sound so bad that you must equalize it?

Maybe your monitor speakers have a bad response and you are equalizing the recording so that they sound better but then everybody elses speakers will sound bad.

 

Audioguru, stop being so doggone grumpy.

Migs,
I'm frankly surprised that an audio engineer would have so little control over the audio spectrum. I would've thought that a pro mixer board would have at least a 10-band EQ in it.

I have an old 4-track TASCAM 424 Portastudio that was good in it's day for hobbyist and demo pro music recording; it only has bass/mid/treble response as well.

I don't know how you could EQ it very well without more than three bands.

You really need 10 or more band control, and you need to at least start by equalizing using a pink noise source in the room where the music was recorded.

This board is mainly oriented towards electronics "n00bs", which category you definitely fit into. However, your project will be quite complex, if you wish to "do it right".

Here is just an example of an amateur 10-band EQ:
http://skema-elektronik.blogspot.com/2009/03/10-band-graphic-equalizer.html

However, if you are attached to a recording studio, you need to use much better components that what are suggested.

Here is another, much simplified Gyrator-based schematic:
http://freecircuitdiagram.com/2008/09/01/10-band-graphic-equalizer-using-gyrator-circuit/

I would not trust either of those schematics any further than I could throw them until I at least simulated them using SPICE to see what they might perform like in a real circuit.

Speaking of SPICE, Linear Technology's LTSpice is very good, and very free.
Google "LTSpice Download" for your reward.

There is good support for it on Yahoo! Groups in the LTSpice forum.
Click: http://tech.dir.groups.yahoo.com/dir/Science/Engineering/Electrical
Join the group. Many, many questions will be answered - and you will have lots more models available for your circuits.

 

Thank you gentlemen for your kind and warm welcome.

This board is mainly oriented towards electronics "n00bs", which category you definitely fit into.

I'm a musician and sound engineer based in London, UK.I have about 20 years experience in that field.Like many "recordists" or "sound guys" we tend to use our gear without getting ourselves, too much into the guts of the machines we use every day.If a problem occurs with one of them, we call a tech guy to sort it out.I want to go further than that and know/learn a bit more.The desk i'm trying to mod is in my home set up,a well respected brand, built according to BBC specks by the old school guys.

Maybe your monitor speakers have a bad response

The speakers I use are Event Electronics made in Santa-Barbara California, again a well established company, those products can be found in many pro studios.I allready have various equalizers,one of them an Orban 672A eight band graphic equalizer, good design. The reason I'm moding my desk is the same as, let's say , you having a 2 bedroom house and wanting to add an extension with a couple of guest rooms.Or a painter wanting more brushes of different sizes and use more extreme shades and colours. Like Sgt wookie said basic recording gear will have Bass-Mid-Treble,as in the Tascam 464 Portastudio I started with.The Eq on my desk is also limited and rather than spend big Money on Outboard gear which can cost alot of Pounds(or Dollars,google Avalon, Manley etc)I'd rater learn something and save cash.My desk's design goes back to the 70's, the use of EQ was limited back then, today's music requires what can be called drastic equalization, think Hip-Hop,Electronic Music, not only instruments with mic's in a room.Graphic EQ's on the other side are seen more in DJ or Live applications, to get rid of feedback etc.Although myself use a graphic EQ on a bass-drum from time to time.Parametric EQ is the favourite tool when mixing in the studio(check George Massenburg).And let's be honnest,the digital plug-ins,softwares made available to us,they just don't have the character of old analog gear.It's a bit like driving a 1977 Mustang VS a 2010 Toyota...
By the way, it's interesting to learn about LTspice ,I'll have a look as soon as I can.
Now,using a formula found on the net and a calculator on the whatcircuits site, I found that. F=1/2Pi Sq Rt of LC. Assuming that F is the frequency in Hertz, L=inductance and C=capacitance.Using the calculator found on this page: http://www.whatcircuits.com/lc-resonance-frequency-calculator/ I found out that, for High Frequencies band, F=10KHz, C=4.7 nanoFarad, therefore L=0.05 Henries(if that is how you spell it)Keep in mind this is a Shelve or filter type, which means that all the frequencies above 10KHz can be "boosted" or "cut".Knowing now the value of L, I was able to find that Caps. with a determined value would give me different frequencies on that band(HF).Think of it as a painter having only blue, in his palette and all of a sudden discovering, red and yellow....lot more possibilities,right?That's what you can see in pic. 2 and 3 of my first post.that's for the HF or Treble.Now,if you can, please have a look at the following two docs.,because my next challenge is the Gyrator(from what I can gather a circuit using a simulated inductor)based LF band used for Bass or Low frequencies.The LF section on my desk is based, apparently, on this Gyrator or Tuned circuit,Resonant circuit, call it whatever you want.The LC filter shown here is not exactly the same as mine,but close enough. http://www.forsselltech.com/downloads/schematics/Gyrator1.pdf
Sorry for the length of this post, I thought I had to explain myself a bit more clearly.

 

Thought it was the classic old Williamson equaliser circuit cropping up here twice in one week (that's the circuit in SgtWookie's gyrator link). Or then again maybe not (it's hard to tell without the rest of the relevant schematic), it looks more like a Baxandall type arrangement with a gyrator(ish) circuit.

Those single op-amp 'gyrator' circuits are popular for guitar amp equalisers where it doesn't matter if the sound is a bit coloured, but they really don't have a place in the studio. But if you want a bit of a go at them, have at it. Some pointers that may be helpful:

• Q is not independent of centre/cutoff frequency, as the two equations in the OP link show.
• The 'gyrator' circuit is not a true gyrator (depending on how strict your definition is), but it's pretty gyrator-like. Electrically it looks a bit like a sideways (the phase is wrong) LC circuit, with series and parallel resistive losses.
• Active filters are more complicated than they look, even with a cookbook circuit and equations. The op-amp bandwidth has to be many orders of magnitude greater than the filter response itself (exactly how many times greater depends on the architecture), there are limitations on the Q and in-band gain (again, depends) and sensitivity to component tolerances is often conveniently brushed under the carpet. As are stability margins, which will turn a filter into an oscillator if deficient.
• Keep the R and C values within sensible limits. I wouldn't use any C less than 100pF, as this is within range of the parasitic capacitance of the op-amp inputs, and the upper limit is bounded by the dielectric. Generally filter-friendly caps above a few hundred nF are both bulky and expensive (don't use electrolytics or tantalums, they're not suitable for an audio signal path). Keep the Rs between about 1k (lower than this can load the op-amp and/or uses more power) and about 100k (higher than this picks up noise and/or reacts unfavourably with the parasitic capacitances of the inputs.)
• As Q increases over 0.5, so the transient response starts ringing like a bell. In an audio application, this quickly fatigues the listener. In the other direction, a Q < 0.5 won't ring, but it will have an increasingly feeble filter slope. A high Q filter will also give a very sharp transition in the phase response, which is undesirable as it muddies the sound. Nice filters (e.g. quality active crossovers) will have additional allpass filters to correct any phase shifts introduced earlier in the signal chain.
• Sallen and Key highpass and lowpass circuits are great for that shelving response, it doesn't need to be a gyrator. Connect to the input only in the type of circuit shown in the OP, just like the gyrator.
• In the OP gyrator circuit, there's a note mentioning a loop from pin1 to pin 2. This is just the configuration for a unity gain buffer - essentially the same voltage at the input and output, but with the added bonus of a high-impedance at the input and a stiffer drive from the output.

If you want a nice studio filter, the "biquad state variable" is the popular choice. It uses 4 op-amps per channel, but why not, they're cheap enough these days, and Q can be varied independently of frequency.

There are lots of filters out there, but don't get too excited. Imagine a chef discovering there are lots of different salts in existence, and then finding out most are poisonous rather than tasty. That said, here's a neat introduction from Maxim.

 

Hello,

ESP has also a nice page on active filters:
http://sound.westhost.com/articles/active-filters.htm

On that site you can find a lot of articles on audio:
http://sound.westhost.com/articles.htm

Greetings,
Bertus

 

What opamps are being used for Z3a, Z3b? I tried simulating that portion of the circuit using NE5534's, but it didn't look too pretty.

Is C12 4.7nF? Is C13 680nF (0.68uF)? I think so, but the image is not quite clear.

Yes, Rod Elliot's site is very helpful. Lots of good stuff over there.

 

What opamps are being used for Z3a, Z3b? I tried simulating that portion of the circuit using NE5534's, but it didn't look too pretty.

The NE5534 is de-compensated for a wide bandwidth at high gains. Therefore it oscillates if its gain is less than 3.

 

The NE5534 is de-compensated for a wide bandwidth at high gains. Therefore it oscillates if its gain is less than 3.

I would probably have known that if I'd read page 1 of the datasheet.

The schematic doesn't say what those two amps are. However, it has NE5534 specified on a part of the circuit that's chopped off.

Tried substituting a TL072; didn't help things.

 

Sgt,

I believe Z3 is the TL072 as per the one picture.

The operators manual for the portastudio 454 doesn't help much, other than it's a four track cassette studio with stereo outputs.

I do have the schematics, somewhere, of an old StudioMaster mixer panel, another four-track, and I'm thinking it didn't have much equalization either. I could be wrong, but I haven't seen that equipment in almost 10 years. If I come across it, I'll let you know, as it's probably typical of the older four track mixers.

 

Thanks gentlemen for all your postssssss,

That is a lot of info for my little artistic brain, it might take some time for me to diggest it all.
But just to clarify the situation a bit, I'm trying to mod the existing circuit of my desk, not build another one, that would be long( 16 cahnnels) costly and I might not have the tech. skills to do it.
To answer Sgt Wookie's question very quickly.I posted the bigger picture of the desk diagram(at least one channel out of the 16,NO master or monitor sections), you might have to zoom in as heck due to format transfer.So, the IC dealing with the HF and LF is a Texas Inst. TLO72 dual Op-amp, from what I understand.One side(pins 1-2-3-4) takes care of the LF and the other side takes care of the HF. It might look simplistic but it is effective and have seen it in other top end desks of that era.Caps. 12 and 13 are metalized poly. caps. C12=4.7n @ 400 (nanofarads I assume) and C13=68u @100 (is that u=microfarads?).Only in the pre-amp section and maybe summing amp(Z1 and Z7) section of this desk, Sanken N5534 are being used, with as you can see on the diagram, a 22Puf cap between pin 5 and 8 to avoid oscillation.Just a reminder,N5534's are single Op-amps not quadruple, TLO72 are doubles, people get confused by the last digit sometimes.But I'm sure you teckies know that allready.On the bottom left corner you can see the Mid-frequency band with a fully sweapable frequency spectrum.The x10 switch (S4) giving you even more choice of frequencies for the mids section only.The rails now, +48V for condenser mics phantom power supply, +15 and -15 are the rails supplying the circuit itself.The other rails on the bottom right corner are to feed the audio signal to the different sub-groups,main mix(Left and Right) and other outputs of the desk.Hopefully the full diagram will give you a better understanding, a fuller picture of the circuit.

 

Is there any chance of you changing the schematic conversion to a PNG format rather than JPG? The jpg puts sloppy blurry compression signals all over the diagram causing quite a reading problem. If your original is a PDF, that would be fine too.

I am having a hard time reading the specs.

 

There you go Sir,
Two documents, two different formats, whatever is easier to read or comes out better.But I'm affraid that it might be the resolution of the original scans giving it that blurry effect and these are the only copies I have.
Hope it helps.

 

Thanks, migs - it's ever so much nicer to have the entire schematic.

No wonder the output from the simulation was so odd; there was quite a bit missing from it.

 

Dear friends,

Thanks for taking the time to post on this thread.With all due respect gentlemen, the mixing desk I have with me is the sort of desk that the BBC (British Broadcasting Corporation) used to have in their OB trucks and some of their hundreds of radio studios they have in their buildings in London(radio 1, radio 2 ,radio 3, Hindi service,Hungarian service, etc etc)The size, although modest in comparaison with most SSL's and the gigantic American desks,API, Harrison etc ,and the three band EQ, has nothing to do with the "professional" status of the product.The EQ on this desk is very very close( I suspect copied) to the ones on the desks at Olympic studios (for those of you knowing a bit about music, that's where the Rolling Stones, Led Zeppelin,later on Bob Marley while in London, used to record and mix) The EMI desks, used at Abbey Road in the late 60's to record early Pink Floyd and the Beatles, had also a mere three band EQ,if not two(can't remember, Geoff Emerick would be able to give more info, he was on those sessions) The bigger desk at the studios where I work, a NEVE ( the Holly Grail of english desks, google that just for the fun of it) has also.....three bands.Closer to you guys, over the Atlantic, Motown studios, have a look what sort of stuff they used... That's how they were building desks at the time.When the Seventies and more sophisticated music came, George Massenburg(google that name please)introduced parametric EQ and more and more bands etc etc.My desk predates that era, and still works fine, just to show how well they built gear in those days.So sorry I'm taking long to explain but I think it's needed. What I'm trying to do is, a bigger choice of frequencies for HF and LF for this little vintage beauty.
Concerning the EQ sends and returns, I allready have that option on the desk, via the insert points on the top left corner of the diagram.They are connected to the 5 times 48 GPO or Type B patch bays, where I can connect a whole lot of sophisticated, parametric and graphic EQ's , I have quite a few of those in racks about 5 foot tall, packed with outboard gear.But all I'm trying to do is....more frequencies on those two HF and LF bands.
Thank you all for your advice, but quite frankly I've been in the recording biz for more than a couple of decades, London is not a small town when it comes to music and recording.I've worked in enough studios to know my options, if there was something else on the menu, I would take it.If I had more space in my house I would have weeled in a bigger console, my wife divorced me and I would die of hunger or brain hemoragea or hemorrhage trying to work out how to use the washing machine on my own...
But I beleive there is another way.And at the same time I can learn something about my gear.What do you think people? Learning something new everyday... wow...In my case, how to add more frequencies to the HF and LF section of my console.

 

Ooops I forgot to include this picture of a mod I did a couple of weeks ago which does, roughly what data 19 was suggesting.The EQ "sends and returns" are allready a standard feature on my desk, where they are called insertion points, which are connected to a heavy brass patch bay, where you can add any outboard equipment you fancy to use.These insert points take the signal just before the fader where you can insert Eq's or other outboard gear into the signal path, some desks it is true, take that signal from a point just after the gain control stage.The switch, an SPDT in this case(data19 was suggesting a DPDT), allows me to switch the outboard Eq or processor in or out of the path.The big yellow wire on the left is the one bringing the signal back from the "Eq returns" or insert returns as it is called here.The green wire at the top is the signal unaltered by the outboard processor. The SPDT switch to the right, choses between processed or unprocessed signal and sends it to the fader, which you can just see at the bottom of the image as a sort of circuit board and a black plastic fader.My desk being an eight buss fully modular thing, I can go slowly, experimenting on a channel at the time.Next channel I might try a push button switch as a selector. I copied that from the ex-Olympic Studios desks.It works fine.

 

Migs,

You already have an out-return with J-2. The out can be plugged into a custom equalizer of however many bands you desire with the return coming back in on J2. The "out" would be the ring of J2 with the sleeve as ground. The "return" is the tip of J2 and again, the sleeve is ground.

 

Thanks for pointing that out Joe Jester,

that is exactly how I'm using it at the momment.And the SPDT switch that you see on the post above your one,gives me access to that path, via J2's like you said which on this desk are called the "insertion points".The ring being connected to the patchbay sends and the tip to the patch bay returns.I allready have graphic and fancy sophisticated multiband parametric EQ's that I can connect to my desk thru those points.That is a standard practice in any studio or even live set up( allthough in live it would be more compressors or gates,graphic EQ being used to "tune the system" or eliminate feedback in the monitors).The switch merely gives me the possibility to listen, with external EQ or any other processor and without it and so on, in order to compare how the said processor is affecting the sound.I allready have those options.I'm happy with the mids section on my desk, they offer me plenty of choice, 100Hz to 1 KHz and 1KHz to 10KHz via switch S4 on the bottom left part of the diagram marked as X10.But what i'm trying to do is, I repeat it and I'm nearly killing myself explaining that on this thread, is learn a bit about electronics and know how to make the existing circuit have a wider choice of frequencies on the HF and LF bands.
Thanks gentlemen for your input.

 

I feel you pain, brother.

I also know your need.

My studio experience is rivaled by yours, but I have had attempted what you are attempting.

I had, given to me for free, 8 channels from a Trident, whos on-board EQing was dead.. (Dont set drinks on the board)
So, I had 8 channels with no EQ, but I did have my outboard gear. Not wanting to use my 31band EQ for the rough-in EQ,
I replaced them. I used the same layout, bought new caps and pots, and jumped the traces out and back in.

I know you have been explaining your butt off, and I have lost a little in the mix ,

By bypassing completely, the HF and LF bands, and reintroducing new filtering caps, you can achieve your goal, which you know.

What I would like to know, is are you still wanting to us the onboard LF and HF POTs?
Or do you want to use external POTs for the adjustment?

 

Hey well done for the Trident channels,
Do you know what series they were?Trident is another British desk that can have a very minimalistic EQ, I'm thinking series 80 and bellow we are talking four band Eq if I remember well, none theless it is a highly regarded design in pro-audio circles.
Back to my one.Despite someone's advice not to mess up with individual "tone controls" on this desk,I sucessfully moded the HF and LF to a basic level.So far I have managed to work out that, keeping the pots for boost and cut (22k Linear pots,R23 and R22) in both HF and LF is fine.All I have done is remove Caps. 12 and 13 which I worked out are the two caps. responsible for the corner frequencies of the HF and LF.I replaced C12 with a small breadboard with a DPDT ON-OFF-ON swt which allows me to put three caps. in parallel with each other to get a capacitance of 3.3nanofarad=12KHz , 8 nanofarad=8KHz and 13.3 nanofarad= more or less 6.3KHz with+/- 15 db of cut or boost.You can see the red and the blue wires going to the points where C12 used to be.Now the red and white wires you see on the pic. go to the place where C13(LF capacitor)used to be. I used the same principle with three bigger caps. put in parallel with each other.That gives me 0.33uF=138/140Hz , 0.66uF=105Hz more or less,1.33uF=more or less 70Hz.The reason why I say more or less is because I don't have an oscilloscope to verify how exact those frequencies are.I just worked it out with a RLC calculator on the web, assuming that in this circuit, an Inductance of 0.05 Henries and a capacitance of 4.7 nanofarad would produce a Frequency of 10KHz and so on.So far it works fine ,I mixed a live recording I did during Easter and can clearly hear how the modified EQ nicely tailored the sound on an electric guitar and a snare drum.Like someone said on a very early post,if the room is right, with good microphones and in this case the recording was done right, almost no need to Equalize that much(well that person on the post was almost saying NO equalization was ever needed,but well, everybody is entitled to their opinion)just a gentle touch here and there, no need for five or ten band graphic EQ, just the right colours added or taken at the right places did the job, while the insert points,"J2" ,were used to process the signal with compressors and some fancy outboard gear.On the pictures bellow you can see PCB, components and switches with caps. on top of it all, of course they don't sit like that during the mixes, that was arranged only to show the whole lot in the pic.