DC Uni-Q. Bugagaga!!!!

[grantlack]

this is going to have an insane amount of relays inside, i thought for sure i'd look at the filter board and have a cmos chip or two staring back at me.  can't fault a man for having his principles!  just out of curiosity, did you test any and find them unsuitable or just go for the relays from the start?

I used 5V DPDT relays guaranteed for 500000 cycles.
They are arround $1.40/100pcs in local electronics supplier store, price falls to $1.10 for 1000 pcs.

fortunate!

 

[Igor]

Me too!

Is this a "noob friendly" project? Probably not but i like to be optimistic

Sounds like an amazing project as usual igor we are not worthy! thank you very much

I tried to keep things simple, BUT the front panel stuff using a lot of SMD.
However, if you have assembled front, the rest is very plain simple.
Stuff motherboard - check twice - fit I/O boards and summing, jumpers for filters - check - than, stuff with the rest of filters - should work.

Say, same job as make 51X rack with 11 slots filled up.

I freaked out myself with taking the freq resp of filters and bought HPIB adapter;
think to put $100 for cheap version of LabView from NI, this combo should work with my over-modified Boonton1120

 

[Igor]

this is going to have an insane amount of relays inside, i thought for sure i'd look at the filter board and have a cmos chip or two staring back at me.  can't fault a man for having his principles!  just out of curiosity, did you test any and find them unsuitable or just go for the relays from the start?

I used 5V DPDT relays guaranteed for 500000 cycles.
They are around $1.40/100pcs in local electronics supplier store, price falls to $1.10 for 1000 pcs.

fortunate!

Well, plain math again.

Pots are the cheapest solution; can't be used for mastering EQ; no recall.

Let's take relatively cheap but still good alternative.

DG506 for GAIN switch, needs another switch BTW to get 31 steps + off:

http://il.farnell.com/analog-devices/adg406bnz/ic-mux-16ch-sp-dip28-406/dp/9605576

The cost $10 in hundreds. +/-22V max supply, measurable distortion (quite low though). Needs to add 17 resistors and external spdt switch.

5 relays and 11 resistors are cool enough for me, instead of DG506. Virtually unlimited headroom, NO DISTORTION at all.

Clickityclack? IMO, funny. Something clinks in my audio unit!!!!!!

However, for GAIN control, there is another good option  - THAT VCA's. Again, VCA + op-0amp (should be good) + cheap DAC08 to implement bipolar Vc,
is more money, less purity.

Freq control. For Sontec, I tested dual optocouplers. Need to be checked at some Iled values for match; adding nice touch to sound.
Matching control circuit to values shown will be the hell due to LDR characteristics.
From 10 of VTL*C*/2 only 4 had matched within 5% half's in 2k...80 k range.

I checked other stuff to control frequency.

Low distortion, good range MDAC is over $15 in qty of 100's, cheap are distorting and low headroom);
VCA's are slow for this purpose.

So, I choose relays as hardcore, pure, affordable thing.

edit1:
The math to DIY whole thing is less than $1300 taking into account good quality components.
Compared to ELMA switches, relays are about 60% less for total cost for same++ thing.
21steps vs 31 steps gain; 21 steps of FREQ vs 24...36 steps;
3RU minimum for 5 band EQ with switches vs 2RU for 7 band EQ
And... all bells and whistles like level (32x0.5db step), 8 steps of 2 db for -8...+4 input pad; MS; etc etc.
BTW, simplified 5 bands version w/o level/ms can be built even for some hundred bucks less.
Just put jumpers instead of unused bands filters...

Edit2:
Ah yep. Forgot. The pipe!!!

 

[andre tchmil]

the pipe is back !!!!

 

[ptron]

Id be really happy to get an assembled front +faceplate and some sort of kit if its possible (maybe we can get some of those relays for cheap) im awful at sourcing parts. No SMD experience or chance to do any of that!

 

[chefducuisine]

Whoa Igor - this looks terriffic !!!

Count me in... this will be the ideal companion for my MixBuzz1.
Already got some ideas for the FP Layout....

HP/LP could be Harrison, huh??

Regards,

Christian

 

[noulou]

Beautiful stuff as usual Igor!!!!     

Any chance of adding a solo button for the different filters? Since the filters are in parrallel and everything is digitally controlled, it might not be so hard to implement. This would make work much faster while tuning the filters.

Any provision for DOAs in the audio path?

What is the widest Q you were thinking for the parametric bands? A Q of as little as 0.5 is many times of great help in a mastering situation.

Really looking forward to this.  8) 8) 8)

 

[Igor]

the pipe is back !!!!

Yep! Why not?

Id be really happy to get an assembled front +faceplate and some sort of kit if its possible (maybe we can get some of those relays for cheap) im awful at sourcing parts. No SMD experience or chance to do any of that!

Yes, this is possible. Prepare to see front panel within a week - two. It's look will be THE killer!!!!!
Morbid and horrible, abandoned and disgusting!!!!!

Whoa Igor - this looks terriffic !!!

Count me in... this will be the ideal companion for my MixBuzz1.
Already got some ideas for the FP Layout....

HP/LP could be Harrison, huh??

Regards,

Christian

TERRIFIC!!!!! No, PURE HORROR, Addams thing!!!!!!

Harrison, mean, these (see attached pic)?
I already draw the filters PCB, it is in-between UREI 546 and SSL's filters.
However, some other filters type which using pots as control can be implemented here.
I thought about ISA110 filters too, but comes up too much relays (have to switch the capacitors),
however, I am open for discussion.

Beautiful stuff as usual Igor!!!!     

Any chance of adding a solo button for the different filters? Since the filters are in parrallel and everything is digitally controlled, it might not be so hard to implement. This would make work much faster while tuning the filters.

Any provision for DOAs in the audio path?

What is the widest Q you were thinking for the parametric bands? A Q of as little as 0.5 is many times of great help in a mastering situation.

Really looking forward to this.  8) 8) 8)

I have changed some things since last week.
Summing can have DOA's or opamps; input/output boards for now are 2 versions in development:
transformers, ll7101 and ll1524+DOA's, and transformerless (THAT/TI bal. line receivers/drivers).
Both types of I/O boards have digitally controlled input pad (-8...+6 dB in 2db steps), this opt can be ommited.

Sontec filter already done with DOA option.
All kinds of VSF better to make with very fast quad op-amps.

All DOA's can be feed from +/-24, and opamps from +/-17V.

Power supply for audio is +/-24V; it is down to +/-17V with regulators on motherboard, the back of case serves as heatsink.
The mother board rails are: +/-24, +/-17, agnd, band in, band out, clip out,
summing to gain +, summing to gain -, summing boost in, summing cut in, than, digital stuff : dgnd, +5V, sdi, sdo, data, clc, etc.
Band in and band out are shorted on parallel EQ filters.

This way, we can build the EQ with series or parallel filters.

Another issue with series/parallel EQ...

For example, SSL channel EQ or ISA110 or Maxi-Q are series equalizers; however,
I preserve the option to bypass each band and go completely pure.

A prob is each of these bands has phase reverse (inverted, unity gain).
In this case, inverting band's "bypass/on" signal goes to simple parity checker on motherboard and
than input's phase reversed in odd/not reversed in even qty of engaged bands.

Series bands simply don't go to the parity checker. As well, summing circuit: in case of Sontec,
there is one op-amp (DOA); in case of Neteq, there are two. Sontec is inverting, Neteq is not.
Summing board has its "parity" too, just have to install 2 jumpers in case of Sontec (inverting).

Widest Q: in case of Sontec, it can be in range approx. 0.3-3 oct. In case of Uni-Q, even bit more
shape'ey (up to 0.25 oct) and less wide, in min pos, than Sontec (about 2 oct).

SOLO. Yes, seems like great option. Has some probs to implement (can be solved, however, comes up as hardcore thing).
Let's take case of series EQ with Sontec topology: I have to feed signal to band (it's input already connected to GAIN),
than disconnect the other bands.
In case of parallel EQ like Neteq, it is easier. Just disconnect summing circuit from all bands and put gain of solo'ed band to MAX.
In case of series EQ, peace of cake - just not feed "direct" signal to band and keep filter in.
All this have to be unified to make solo available for all bands. Again, not easy, but I will sleep on this.
From the other side, each band can be bypassed completely.....is it not enough?

Well, the question: is the SOLO option necessary????

 

[Ben F]

Hi Igor,

Amazing work, this has to be a dream mastering EQ! The option of having output transformers is even better.

I don't think solo for each band would be necessary, however solo for the M/S part would be nice for people without that facility on their monitor router. I suppose you could use the 'bypass' but then would that be pre or post input gain? This would be to hear the M or S signal independently when using the EQ.

 

[ptron]

Hi Igor,

Amazing work, this has to be a dream mastering EQ! The option of having output transformers is even better.

I don't think solo for each band would be necessary, however solo for the M/S part would be nice for people without that facility on their monitor router. I suppose you could use the 'bypass' but then would that be pre or post input gain? This would be to hear the M or S signal independently when using the EQ.

+1

 

[Ben F]

Further thoughts...

I'm wondering if it would be possible to have the frequency, Q, boost and cut on different knobs instead of one? Or even have the frequency / Q knob sharing one control and the boost/cut on another. Is this easily modifiable? In this way it would interact more like an analogue EQ, requiring a more intuitive operation, and users could experiment with different panel configurations.

A/B compare function would be nice, like the summit EQ 200

And the icing on the cake- Stereo Link? One set of controls does both channels? 

 

[zenmastering]

I strongly agree with Ben's comments regarding the use of separate encoders for Frequency, Q and Gain. Ergonomics are very important in a tool that I would expect to use every day. Even well-regarded digital mastering EQs (e.g. Z-Sys, Weiss) have these three parameters on separate encoders. Although I think that these two digital EQs have been equalled, sonically, by one or two plug-ins on the market, the ergonomics of the Weiss and Z-Sys (especially the Weiss) boxes are brilliant and the reason that I stick with them. Please consider this for your DC-UniQ, which looks like an excellent concept.

Other than that, a few more comments and suggestions:

I don't need a solo function

I would like to have the boost/cut steps like this:

+/- 0.25; 0.5; 0.75; 1.0; 1.5; 2.0; 2.5; 3.0; 3.5; 4.0; 4.5; 5.0; 5.5; 6.0; 7.0; 8.0

This is the most (+/- 8dB) I'd ever need in a mastering EQ and the 0.25dB steps actually do make sense when using very wide filters as a 'tilt' style equalizer.

An A/B compare or 'compare current settings with a chosen preset' function would be nice but is not essential.

As Ben also mentioned: In M/S mode, being able to mute either the M or the S signal, between the encode/decode circuits would be very useful. This way we could listen to just the M or S signal but 'in place' in the LR domain.

Simple, switchable stereo linking of both channels would be wonderful.

This looks like a fantastic project. I'm still using and appreciating the passive CRM project, for which I'm very grateful to you, Igor.

My 2 cents worth,

Graemme



 

Further thoughts...

I'm wondering if it would be possible to have the frequency, Q, boost and cut on different knobs instead of one? Or even have the frequency / Q knob sharing one control and the boost/cut on another. Is this easily modifiable? In this way it would interact more like an analogue EQ, requiring a more intuitive operation, and users could experiment with different panel configurations.

A/B compare function would be nice, like the summit EQ 200

And the icing on the cake- Stereo Link? One set of controls does both channels? 

 

[leadbreath]

igor? what do u put in that pipe of yours??? 

 

[goldvasser]

+1 with Graemme !

 

[DOMC Mastering]

+2 Graemme

 

[ptron]

igor? what do u put in that pipe of yours??? 

+1

 

[tekno808]

sounds really interesting!
Count me in!  8)

+1 

 

[JerryAnthony]

Interesting

 

[rafafredd]

Igor, you are a freak! In a good sense, you know.

 

[Ben F]

Igor, hope you are feeling better now. Any progress on Bugagaga?