DIY equalizer based on Studer 169 (standard and mastering)

[benlindell]

I found a NE5532 subcircuit and I'm still getting the bass roll off.

Here's my netlist:

* 2010-02-18 17:31:55 -0500
.AC dec 50 10Hz 30kHz
.print ac v(10)
Vin 1 0 AC .774 0
R23 2 1 3k3
C3 2 3 15n
R24 3 10 3k3
C8 4 5 39n
C13 3 5 39n
R25 5 6 560
R20 1 6 2340
R11 6 10 2340
R10 4 7 11.2k
R1 7 3 11.2k
C16 11 7 220u
R21 11 8 220k
R22 11 9 220k
R11 10 0 22
*X1 14 0 12 13 11 National.LM741
X1 0 11 13 14 12 21 22 NE5534
C18 21 22 22pf
C17 11 10 10pf
R26 12 10 10
Vdc2 13 0 DC 15
Vdc3 0 14 DC 15

* NE5534 OPERATIONAL AMPLIFIER "MACROMODEL" SUBCIRCUIT
* CREATED USING PARTS RELEASE 4.01 ON 04/10/89 AT 12:41
* (REV N/A)      SUPPLY VOLTAGE: +/-15V
* CONNECTIONS:  NON-INVERTING INPUT
*                | INVERTING INPUT
*                | | POSITIVE POWER SUPPLY
*                | | | NEGATIVE POWER SUPPLY
*                | | | | OUTPUT
*                | | | | |  COMPENSATION
*                | | | | | / \
.SUBCKT NE5534  1 2 3 4 5 6 7
*
  C1  11 12 7.703E-12
  C2    6  7 3.500E-12
  DC    5 53 DX
  DE  54  5 DX
  DLP  90 91 DX
  DLN  92 90 DX
  DP    4  3 DX
  EGND 99  0 POLY(2) (3,0) (4,0) 0 .5 .5
  FB    7 99 POLY(5) VB VC VE VLP VLN 0 2.893E6 -3E6 3E6 3E6 -3E6
  GA    6  0 11 12 1.382E-3
  GCM  0  6 10 99 13.82E-9
  IEE  10  4 DC 133.0E-6
  HLIM 90  0 VLIM 1K
  Q1  11  2 13 QX
  Q2  12  1 14 QX
  R2    6  9 100.0E3
  RC1  3 11 723.3
  RC2  3 12 723.3
  RE1  13 10 329
  RE2  14 10 329
  REE  10 99 1.504E6
  RO1  8  5 50
  RO2  7 99 25
  RP    3  4 7.757E3
  VB    9  0 DC 0
  VC    3 53 DC 2.700
  VE  54  4 DC 2.700
  VLIM  7  8 DC 0
  VLP  91  0 DC 38
  VLN  0 92 DC 38
.MODEL DX D(IS=800.0E-18)
.MODEL QX NPN(IS=800.0E-18 BF=132)
.ENDS

 

[benlindell]

Hope you guys don't mind me continually posting to this. I'm just trying really hard to figure out what values to use to get the scale I want.

So I've kind of abandoned the simulator for the moment and back to the spreadsheet. Found out Apple's Numbers can find equations to data when you plot a chart. Like audiox said this isn't really log or lin, but a 4th order polynomial fit it pretty darn well. I think I'm gonna use this and get the resistors ordered and test this bad boy out once I get the channels. I can finally sleep and stop thinking about this darned thing.

Here's my Values for the high shelf, working on the low at the moment:

Step dB	Resistor Total	Resistor Value (exact)	Nearest Resistor
-6	1440	630	634
-3	810	400	402
-1.5	410	410	412
-0	0	0	0
1.5	410	410	412
3	810	400	402
4.5	1146	336	332
6	1440	294	294
7.5	1676	236	237
9	1870	194	196
12	2170	300	300

 

[benlindell]

Here's my Values for the Low shelf

Step dB	Resistor Total	Resistor Value (exact)	Nearest Resistor
-6	5370	2640	2610
-3	2730	1366	1370
-1.5	1364	1364	1370
-0	0	0	0
1.5	1364	1364	1370
3	2730	1366	1370
4.5	4100	1370	1370
6	5370	1270	1270
7.5	6560	1190	1180
9	7655	1095	1100
12	9670	2015	2000

 

[benlindell]

two more things to share/verify. Here's the changed schematic:

 

[benlindell]

This is the one that I need to double check. The Low + side needs two since the signal path goes through the switch.

 

[benlindell]

hey guys I had a chance to rethink the trace cutting and I'm sure I've gotten the high shelf figured out but I'm not sure I've gotten the low end correct yet, could anyone give me some guidance, would be much appreciated.

 

[benlindell]

But I think this maybe it, the green is jumper wires, connecting r24 to r22 and r21 to r23 but putting 1.54k and 5.9k between R1 and everything and R10 and everything.

 

[benlindell]

Sorry for the bump guys but i'd like to double check this before I start cutting.

Thanks.

 

[bobine]

I've been looking for caps that fit the PCB. There aren't many choices that cover the range of values needed AND fit in the space allowed.
Lead spacing is 5mm. The max width (or thickness, T) is about 2.8mm. L(ength) is less than 10mm; 8mm is better.

I only found these dimensions in metalized polyester caps from Wima and CDE.

Are these caps ideal to use, electrically? The description of polypropolene caps seems to match the application better but they are too fat to fit.

Here are lists of part numbers for two different brands (both available from Mouser).
Quantities are for freq selection caps for two boards.

Anyone know of other caps that will fit the PCB?

Wima MKS2 metalized polyester.  5mm LS,
W/T= 2.5mm, L=7.2mm 

2--    220nF      .22uF    MKS20-.22/63/5  (W/T=3mm)
8--    100nF      .1uF    MKS2-.1/63/5
2--    68nF        .068uF  MKS20-.068/63/5
6--    47nF        .047uF  MKS20-.047/63/5
6--    22nF        .022uF  MKS20-.022/63/5
4--    15nF        .015uF  MKS2-.015/63/5
4--    10nF        .01uF    MKS20-.01/63/5

Or...

CDE Cornell Dubilier
Metalized Polyester Type 168, 5mm Lead Spacing
W/T=2.5mm,  L=7.2mm

2--    220nF      .22uF    168224*63C-F  (W/T=3.5 L=7.2)
8--    100nF      .1uF      168104*100H-F
2--    68nF        .068uF    168683*100H-F
6--    47nF        .047uF    168473*100C-F
6--    22nF        .022uF    168223*100A-F
4--    15nF        .015uF    168153*100A-F
4--    10nF        .01uF    168103*100A-F


caps are for these freqs:

Lo

#1  C1  28hz    220nF
#2  C2  60hz    100nF
#3  C3  85hz    68nF
#4  C4  120hz    47nF
#5  C5  230hz    22nF

Hi

#1  C6, C11      2.0 K    100nF
#2  C7, C12      4.1K      47nF
#3  C8, C13      8.8K      22nF
#4  C9, C14    12.7K    15nF
#5  C10, C15  18.9K    10nF

Let me know if I have something wrong... thx

 

[Davo]

I used MKS2 wimas mostly, except for the odd vishay or cornell in the non-wima values.... works and sounds great.  Try and stick to the
2.5mm width or your gonna have to hang the odd capacitor off the back, which isn't really a big deal.  Looks like your doing just fine.

cheers,
dave

 

[joshco]

Hello,
If we want to replace the Standart Version with Pots instead of Volume Switches, which Pots should we go for, 10k or more?

I building this Version with stock Parts and i do not have all the resistor values.

Can I hear equalisation if I only connect the passive part of the circuit, or do I have to power up the whole circuit?

Thanks
T.

 

[benlindell]

You should try 5k for the high and approximately 22k for the low, as you can see in the graphs I made the scale is very unlinear, they will work though

 

[joshco]

Allright, thank you, I have tested 10k and 20k but its not working, have to debug it. Audio is passing through without effect on any frequency.

 

[benlindell]

Wheree are you connecting the pots?

 

[joshco]

i connect the pots one one hand the wiper on the middle pad of the rotary switch (R25), on the other hand on the outer pads of the switch(R23/R24).

 

[joshco]

Finally its finished, thanks to Audiox and the Forum for this nice EQ!

Its built with self etched PCB and from parts I had laying around, caps from futurlec value pack and some Panasonic electrolytic caps. It sounds really nice and is a good addition to the GPultec. The Curves between diffrent Frequency Steps are quite narrow, maybe because of the Mylar Capacitors. Case is hand painted and looks rather ghetto style, compared to other units. I will use this for tracking VA Synth and Samples.

Lessons learnt:
- Triple checking Ground wires, this was the reason why it was not working properly
- Never buying cheap cases again, was horrible to work with
- More accurate wiring the next time, i will post no inside shots



 
   

 

[strangeandbouncy]

Hi,


  I have basically built a version of this eq onto the 500 series board from Gustav. I have no balancing in or out, using just one DOA. I am getting a 6dB loss, and cannot for the life of me see why. I am connecting cold to shield in and out. Am I right in thinking that this should be unity with no eq dialed in?


    Sounds amazing, even if I am having to make up 6dB elsewhere.


    Any Help most gratefully recieved!


    ANdyP

 

[strangeandbouncy]

So I strapped in a 2503 at 1:2, and it sorted it out. Still don't know why I am down 6dB. It doesn't seem to be a bal/unbal issue. Oh well . . .


  Just noticed that without tranny it is ruler flat from 20Hz to 20kHz, but with 2503 it has very very gentle slope , starting at @250Hz, and is @1dB down at 20Hz. It doesn't sound bass light, just a tad more forward. Is this normal for a 2520 and 2503? I have never seen it before . . .


    Lovin' it doh . . . .



       ANdyP

 

[[silent:arts]]

So I strapped in a 2503 at 1:2, and it sorted it out. Still don't know why I am down 6dB. It doesn't seem to be a bal/unbal issue. Oh well . . .

it is. the input stage is -6dB
the balancing adds 6dB
that's all.

I'm thinking of using an 1:2 output transformer too.

 

[strangeandbouncy]

Hi Silent-Arts,


  I am not using the input stage, just going in unbalanced to C3 . . .



      Kindest regards,



          ANdyP