MixBuzz1 Build/Support thread

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Igor

Well-known member
Joined
Jul 17, 2006
Messages
2,193
Location
Israel
OK...
http://ij-audio.com/downloads/mixbuzz1_docs_v2_4.rar
For the docs.

Very nice BOM - thanx and respect to Evilcat:
http://ij-audio.com/downloads/MixBuzz1BOM.xls


Very important update.

Due to 90% of people who built this compressor did it with 11 steps attack and release switches, the new PCB's are done by this way.

All is same, but easier wiring.

http://ij-audio.com/downloads/MB1_docs_2_5.rar

+12V regulator's heatsink was changed to bigger one.
LM3914 LED current setting resisters were changed to lower power consumption too.

In case of whatever: use 12V relays with 700Ohm...2kOhm coil resistance, to reduce current consumption.

Front for use with Uraltone 23-step switches (the controls are bit moved to get space for 23-pos switches):

http://ij-audio.com/downloads/mb1_2_5ver9cdr.cdr

(saved in Corel version 9)


Please download and take 3 minutes to pass on everything, just to catch the system.
You will figure out the logic very fast.

Than, PRINT and stare into schems/PCB assembly drawings/BOM's.
Yes, you need about 30-90 minute meditation, after this, all supposed to be clear.
Everything is separated in sections, every section's components
have logical order numeration and it's own suffix. Tones of work, BUT this is MixBuzz1 :)

Questions welcome.

I reserve the next 3 posts for kinda FAQ and testproc, they will be filled within a week, hopefully.
 
Reserved for FAQ1
Have to add some things....
For now:


Before starting assembly, take your case and put PCB's inside by the way you like it.
Use 250mm or more case (10' depth).
Don't be greedy on small case, as we already have space-saving brother, MixBuzz500.
Play with pcb's, power trafo, etc, and find a way everything fits fine, mind the place for wires, etc.
Scroll this thread to see pictures to get an idea
http://www.groupdiy.com/index.php?topic=40766
If front panel connected direct to main PCB, first, assemble the case, attach front,
than, connect main, than, mark the holes for main PCB with permanent marker.

Special attention was placed to grounding scheme.
MixBuzz1 have very solid ground scheme, and it is STAR GROUND, not just a ground plane.
(Reason: to minimize all possible leakages of control signal to audio).
Ground plane is not best here, IMHO.
Central hole on main PCB is the GROUND MECCA. It should be connected to case.
XLR's pin1 connected to case; actually, if input/output wire's lenght
(between pcb and XLR connectors) is less than 5-7 cm, no need for shielded cables, just twist 'em.
Check twice that all components are installed right way.
One wrong resistor can cause PITA for some hours; one wrong placed electrolyt' can
explode and make contamination on PCB (need to wash), but,
HOPEFULLY, you will take a 30 min-hour to CHECK EVERRYTHING and
hopefully everything will be RIGHT :)
OK than, enough fear and loathing....let's fire up the baby :)


POWER SUPPLY

will be added soon


After power supply is checked, sonnect it to main PCB.

!!!!!!!!!!!!!!!!!!!
For the next, 1V RMS signal (if not specified) is 1KHZ.

AUDIO
At this stage:
Do not insert any chips into sockets!
Power ON.
Check the voltages with main PCB connected. Let it work 5-10 minutes.
No smoke? Good! Check voltages again.
Power OFF.
Put the VCA's, audio chips (IC1,2,4 L/R).
Pay special attention when inserting the chips into sockets.
VCA's can't go wrong way. See pictures for reference.
Power ON.
Now, check:
Makeup=0, MIX=OFF.
Feed 1V RMS (0DBV) to inputs, press COMP IN (bypass off), check output
is 0DBV (1V RMS). With use of "right" VCA and R6, R3, R100, MF 1%,
the level difference between channels should be less than 0.2db.
BTW, most of people will use soundcards with RMAA to do adjustment;
check your soundcard output 0dbV=1V, some soundcards have 0,2...0.5 db
difference between channels... Some cheap multimters can read less volts on 10kHz AC..
(in other words, me recommend analyzer :)

Same for noise etc measurements, I use analyzer WITHOUT weighting filters.

Adjust VCA's for minimum distortion. SYML, SYMR pots.
Or, leave some distortion at your taste.
@1Vrms 1kHz input, my MixBuzz1 shows 0.0045% distortion on analyzer, when trimmed best possible way;
SINAD -90db, S/N >94 db, 4x2181B VCA's used.
Readings @5V RMS input: distortion 0.0035%, SINAD-95 db, S/N 106 db.
Frequency responce 15Hz-60kHz: -0.1 db max.

Some words about distortion and noise.
More VCA's=less noise. More VCA's does not means less distortion.
Minimal distortion can be adjusted with use of A grade VCA, IMHO, not worth.
To make good sounding compressor with nice amount of coloration,
use ONE VCA and misadjust SYML/SYMR pots for 0.02-0.06 % of distortion.

More "misadjustment" can cause hearable leakage of control voltage to audio,
it sounds bad imo on fast attack times.


After adjustment and checking, turn power off.

MIX (C&B)
IF MIX NOT USED, INSTALL JUMPERS JP2 L/R
Install C&B (MIX) chips, if you use them, NE5532, OPA275, OPA2604 or OPA2134.
Check the MIX potentiometer connected right way (see MIN-W-MAX markings on PCB).
Shielded wire shoud be used to connect MIX pot,
MOGAMI digital (microphone) audio cable IMHO the best.
Check everything again.

Feed 1V RMS (0dbV) to input.
Power ON; check unity gain again;
MIX on; MIX pot = CENTER, adjust R63& (left channel), R64& (right channel) for same +2Db;
check channels are same in mx. DRY and WET positions (about -2Db), adjust for same levels
if needed.
The level difference between channels should be less than 0.5 DB for all range of MIX pot.

After adjustment and checking, turn power off.

CONTROL and FILTERS

IF SC FILTERS NOT USED, INSTALL JUMPERS JP1 L/R

Put the chips in sockets, check everything twice.
When installing ribbon cables, check twice, because it is easy possible
to make a mistake here.

RATIO=10, THRESHOLD=+15DB, ATTACK=FAST, RELEASE=SLOW, MAKEUP=0, SIDECHAIN=OFF,
MIX=OFF, SC MONO=OFF.

Feed 1V RMS to input, check output is 1VRMS.
Turn THRESHOLD CCW, check the signal changes.
Turn THRESHOLD back to +15.
Turn makeup CW, check output is +19...20db.
Turn makeup back to 0.
Feed 1V RMS to input, check output is 1VRMS.

Check the voltages between TP1 and TP2 to ground, they should be 0.00 V,
adjust SC_SYM L/R trimpots.

Ratios

Feed -10db to input; check output is -10db
Adjust THRESHOLD for -1db output (2db compression).
Adjust make-up to read -10db.
Check RATIO=10; feed +10db input; adjust RATIOS trimpot to get output -8dB (+/-0.1 db)

MAKEUP=0, THRESHOLD=+15DB, RATIO=2
Feed -10db to input; check output is -10db
Adjust THRESHOLD for -12db output (2db compression).
Adjust make-up to read -10db.
Check RATIO=2; feed +10db input; check the output is 0dB (+/-0.5 db)

If analog meter used:
Feed 0dbV to input; check output is 0db
Adjust THRESHOLD for -20 or -10 db output (10 or 20db compression, depending on
maximum scale of your analog meter).
Adjust GR_MTR trimpot for corresponding meter's reading (-10 or -20 db gain reduction).

Now, freq resp of sidechain.

RATIO=10, THRESHOLD=+15DB, ATTACK=FAST, RELEASE=SLOW, MAKEUP=0, SIDECHAIN=OFF,
MIX=OFF, SC MONO=OFF.

Feed 0dbV 1kHz to input; check output is 0db ; adhust THRESHOLD for 10 db compression;
change frequency to 50HZ and 10Khz, check for same amount of compression +/-0.5db.
It is OK if compression @10k is 1db less than @1khz, same way "original" comp behaves too.

RATIO=10, THRESHOLD=+15DB, ATTACK=FAST, RELEASE=SLOW, MAKEUP=0, SIDECHAIN=40,
MIX=OFF, SC MONO=OFF.

Feed 0dbV 1kHz to input; check output is 0db ; adjust THRESHOLD for 10 db compression;
compensate with make-up for 0db;
change frequency to 20HZ, check +8 +/-1db output

SIDECHAIN=80
Feed 0dbV 1kHz to input; check output is 0db ; adjust THRESHOLD for 10 db compression;
compensate with make-up for 0db;
change frequency to 40HZ, check +6 +/-1db output

SIDECHAIN=120
Feed 0dbV 1kHz to input; check output is 0db ; adjust THRESHOLD for 10 db compression;
compensate with make-up for 0db;
change frequency to 80HZ, check +7 +/-1db output

Note. All these 40/80/120 markings are very subjective.
I used the capacitor's values which worked best for ear...the actual filter's freq's
can be bit different.

SIDECHAIN=TL
Feed 0dbV 1kHz to input; check output is 0db ; adjust THRESHOLD for 10 db compression;
compensate with make-up for 0db;
change frequency to 40HZ, check +8 +/-1db output
change frequency to 100HZ, check +4 +/-1db output
change frequency to 5kHZ, check -3 +/-1db output
change frequency to 10kHZ, check -7 +/-1db output

SIDECHAIN=TM
Feed 0dbV 1kHz to input; check output is 0db ; adjust THRESHOLD for 10 db compression;
compensate with make-up for 0db;
change frequency to 40HZ, check +9 +/-1db output
change frequency to 100HZ, check +8 +/-1db output
change frequency to 5kHZ, check -6 +/-1db output
change frequency to 10kHZ, check -7 +/-1db output

The output can be different a bit from these values-10% capacitors, etc,
BUT should be very close for both L/R channels.

METER

Meter calibration

1). Check the OFFSETL, OFFSETR trimpots wipers are connected to GND.
If we look to meter driver board and 10-pin connector is at bottom,
central pins of these trimpots should be connected to bottom pins.
I.e. start adjustment without any offset.

2). Set meter switch to OUT; Feed input about 1V RMS, measure output with
AC RMS voltmeter;
Adjust output with MAKEUP to get 0DB, in case of +4dBU it is 1.22V RMS,
THAN adjust SCALEL, SCALEL to get reading of 0db on LED meter scale.

3). Feed about 0.07V RMS to input, adjust output with MAKEUP to get -22dbU (95mV),
and check meter shows -22dB. If no, adjust little bit OFFSET trimpot for corresponding channel.

Recheck, repeat (2) and (3) if need.

4). Without any input signal, set meter switch to GR; adjust GR_OFFSL trimpot to get
reading of 0dB on LED meter;

5). Feed input signal 1V RMS, NO compression NO makeup, i.e.
threshold is CW, makeup is CCW, check the out is 1V as well;
Adjust THRESHOLD for -20 or -10 db output (10 or 20db compression, depending on
maximum scale of meter you want to use).
Adjust GR_MTR trimpot located at the main PCB for corresponding LED meter's reading
(-10 or -20 db gain reduction)

Recheck, repeat (4) and (5) if need.
 
Reserved for FAQ2

Most of VCA-related questions already answered at first part of this post:
http://www.groupdiy.com/index.php?topic=39459.msg487554#msg487554
Please take a look.
 
In the filters bom, there are values missing for CBP12L, CBP12R, CBP14L and CBP14R. Looking at the schematics i guess those can be 100nf Pitch 2,54 as they are filter caps for the TL074 supply rails ?

Any clues on how to implement/calculate CR1, CR2, RR1 and RR2 for the Release ?

Flo
 
In the filters bom, there are values missing for CBP12L, CBP12R, CBP14L and CBP14R. Looking at the schematics i guess those can be 100nf Pitch 2,54 as they are filter caps for the TL074 supply rails ?

Yes, a bit of mess, these caps are in basic kit BOM.

Any clues on how to implement/calculate CR1, CR2, RR1 and RR2 for the Release ?

By ear, take potentiometer and capacitor or 2 pots and 2 caps.

Logically, for 0.9 sec setting release, (between 0.6 and 1.2) use 750 or 820k and 0,47uf.
Rolls music increased a bit maximum release (last setting) and did
"dual release" push-button instead of A setting.
Don't affraid to experiment.
I had some time and printed in Textpad some things about MixBuzz1 calibrating, etc...
2nd post updated.
 
Igor, for a full blown mixbuzz with 4xThat 2181/channel ... 18V/15VA transformer still enough?
 
If 50VA rated toroid fits in the case, why not to use it? A bit overkill, but should work.
THAT2180: never tried to use them in parallel...should work.
 
But you have the provision on the VCA board for 4 That, or did i miss something?

There was one thread about SSL with 4 or 6 Thats, some time ago. Cannot find it right now.

I have enough 2180 on stock to do this, but if its just esotheric i dont have to do it and like to spare them for another project ;)
 
Yes, there is provision for 4x THAT2181 on-board.
I used all grades (ABC) of 2181, 1x, 2x, 4x on board.
I haven't tried 4xTHAT2180 in parallel.
"Original" VCA (there are 10 versions at least) used 8x2151, or older SIP-8, or discrete transistors.
It is more than enough 4xTHAT2181B to get great sound, or, if you like more color, 1x2181C, bit mis-adjusted.
 
Got it. Did read a lot of old posts regarding ultimate SSLs and co yesterday and found out that the dbx modules had 8x2150/81 like you said.

Something to do to use 4x2180? Pin 4 lifting trick?
 
Igor's PSU build,



mine,



I'm not sure for CP1 | CP2 =0.1u but I had two CPs and 47nj :( they are in the bag of PSU kits! I am wrong?
any idea ?
Thanks

EDIT :] I got it --> 2 CAPACITOR CERAMIC OR POLY 47...100NF 100V 5MM PITCH CP1, CP2
 
Mixbuzz1 B_O_M Guide 10 pages PDF
]https://docs.google.com/viewer?a=v&pid=explorer&chrome=true&srcid=0B0nKUsFBLtnPZTI1N2QwYjYtODlmMy00ZGQ0LWJlZTgtMmY5MzEwZmVhYWM2&hl=en&authkey=CJCOiw8]
maybe should be easy!
cheers,


P.S. I got some questions ;

- I will use all options and what do I need to put on  RV1L/R  SCL/CSR parts? or which resistor value I can put?
Edit : /PUT 5K1 RESISTOR INSTEAD AND FORGET ABOUT THIS TRIMPOT

- on C&B RV_1| RV_2 | will have 100r ? if I use all options (SC-MIX etc etc)
- Do I must put Ratio's 500r Trim pot ? if I use all options?
- R6L and R6R TBD ?

Igor, I didn't receive 360k, 750r0, 470k, 200r, and just I got one 2k4! I got a lot resistor M :)
cheers,
 
Igor said:
POWER SUPPLY

will be added soon

Ok, I am trying to do PSU's calibration ! maybe I'm wrong but maybe the PSU  board has a buggy :)
Question(if I can get a response will great! because I tried to contact with Igor but no news );
On the main board PS is,
1 /-15V
2 /+15V
3 /-18V
4 /+18V
5 /AGND
6 /AGND
7 /-12V
8 /+12V
9 / Case

But on PSU board not the same ?
1 /+15V
2 /-15V
3/+18V
4 /-18V
5 /AGND
6 /AGND
7/+12V
8  /-12V
9 / Case

also I have +12V but can't understand why I do not have -12?
cheers,
 

Latest posts

Back
Top