dbx 160vu clone

[damnyankee]

I'm wrapping up soldering the last few components on the board.

I have a few questions:

On the BOM, R11, R13, R15, R67, R106 are "trim" resistors and C10 is supposed to be a trim/timing cap...but there's no values listed.  Thus, these components have not been installed on the board.  Can anyone help as to what values these components are supposed be?

Thanks in advance!

DY

 

[abechap024]

Hi DY those trim resistors/cap dont need to be used under normal conditions (using 1% resistors).Something maybe that needs to be taken out for the next batch. I let the timing cap trim in there in case anyone wanted to play with the value or install a normal/slow switch.  lol (maybe that needs to be taken out too). but just ignore them for now.

I just finished stuffing the board now just waiting on the THAT chips to show up then test time. I'm super excited to get these going and its looking pretty spiffy with its wima caps and case that Dan made. I'll post some pictures this weekend.

Cheers!
Abe

 

[abechap024]

Reading over the calibration sheet here

http://www.dbxpro.com/product_downloads/Calibration_Procedure/160-161%20Calibration%20Procedure.pdf

turns out r36 (r11 on our boards) is used to trim the symmetry in the signal of the comp pin (labeled offset on out boards) on our boards.

So I'm almost done with everything heading back to my shop for the long weekend. Cant wait to get this puppy fired up and hopefully making some progress with this. Wish me luck!

AC

 

[damnyankee]

Good luck, AC - maybe you'll have the first fully functional unit up and running! 

DY

 

[abechap024]

Okay. Got one up and running last night, except for the output stage and still having issues with meter section... Still need to do all test to make sure everything is right. but unit does compress and seems to sound very smooth.

the meter section is weird! It seems to repsond as spec if I was using a mech zero'd meter, but after the electronic emulation circuit it doesn't work right. I don't know what it is , everything looks like it will work till I put the buffer opamp in and all of the sudden there is +15 v on the output of the opamp! Ive tried cleaning the board lifted the inputs tried different opamps....no luck so thats still a work in progress.

One thing that concerns me, iss the RMS supposed to have huge amounts of internal gain and distort? I was poking around with a pair of headphones and the signal gets much louder in the RMS unit. I know one should expect diode clipping and all that, so maybe thats all it is.

Also need to address a bit of the signal getting in the control voltage path....probably due to the need to trim the RMS unit exactly with r36 and other resistors...

I was having trouble with the that chip at first but got it straightened out....I found that a 1k resistor in r27 on the board helped curb some wierd problems with the VCA distorting out at full makeup gain.

update: 4558 chip is oscillating fixed it with a bypass capacitor. Probably due to to resistors bieng placed too far from the opamp in the layout.
I'll post some sound clips, as Ive never actually used a VU I can only guess what it is supposed to sound like.

I havent looked at the output stage yet but i'm reading +/- 30v which is why I haven't touched it yet...probably fry any opamp I stick in there. so I'll probably poke around that on monday.

 

[haima]

Cool that you made some progress!

what VCAs and RMS units are you using? there's a few options isn't there...?

 

[abechap024]

yes progress! finally!
and after placing a bypass cap on the 4558 voltage control driver opamp it started to sound nice. Still quite a bit of 2nd harmonic distortion, but not in a bad way. sounds ok, but still probably something we want the option of removing. I think it is probably from the signal in the CV.

I'm using a that 2180 and built on PCB RMS unit. I compared it to some files I found of a shootout of hardware and plug in 160 units and it definitely has the same character.

Meter works just from taking a jump off the 220ohm(?)ish resistor before the tl081 mechanical meter emulation.

I'm using a hairball meter and in GR mode it starts at -20 db and goes towards zero the more compression applied. In the other IN/OUT vu settings it gets a little more screwy as the meter does whats only can be described as somersaults....
anyway. anyone else making progress?? btw the calibration for these is a sun of a gun...I still don't understand how to calibrate the RMS entirely but learning..


Oh Ps one of the ABOVE/BELOW LEDs should be flipped around from what was shown in the connections schemo...

 

[damnyankee]

Reading over the calibration sheet here

http://www.dbxpro.com/product_downloads/Calibration_Procedure/160-161%20Calibration%20Procedure.pdf

turns out r36 (r11 on our boards) is used to trim the symmetry in the signal of the comp pin (labeled offset on out boards) on our boards.

So I'm almost done with everything heading back to my shop for the long weekend. Cant wait to get this puppy fired up and hopefully making some progress with this. Wish me luck!

AC

AC, looking at an image of a 160VU pcb (the image is very poor, so these need confirmation):

1.  160VU CR16 (our C10) Trim Cap is not used.
2.  160VU R36 (our R11) is definitely 100k 5%.
3.  160VU R16 may be 1.3k (needs confirmation!)
4.  160VU R52 may be 150k (needs confirmation!)
5.  160VU R64 - no trim resistor (needs confirmation!)
6.  160VU R66 - no trim resistor (needs confirmation!)

DY

 

[abechap024]

Hello.
All the values should be on the BOM and on the board.

But since you are wondering r16 is 1k
and r52 is 100k

what image are you looking at? There is a high res PCB overlay I uploaded. The schematic is a mess but when I get sometime I'll load up a higher res one. Dip trace sucks at exporting images!

Anyway great news. Works! everything works. reading low distortion with a THAT C chip in and even lower distortion with THAT A. sounds great! got the output stage working the LM301 chip seems to be doing just fine even with the +/- 30 plus voltages. So I guess DBX knew what they were doing. Also changed meter emulation circuit around because as is I couldn't FOR THE LIFE OF ME get it to work as it was. So now its at least working. GR it works fine. IN/OUT the VU reads to high in the middle....if that makes sense....the tracking is off. @-10 it reads -5  but near the louds and quiets it reads just fine....going to experiment more with it. Maybe I just need to calibrated it some more.
Anyway thanks to all who participated...should have 2nd rev boards here very shortly...
now I want to try a discrete vca!
haha!

oh ps I'll upload the new meter circuit....its not to hard to change....just need to cut a trace and add a resistor...but i'lll wait till its working 100%

 

[Biasrocks]

Hello.
All the values should be on the BOM and on the board.

But since you are wondering r16 is 1k
and r52 is 100k

what image are you looking at? There is a high res PCB overlay I uploaded. The schematic is a mess but when I get sometime I'll load up a higher res one. Dip trace sucks at exporting images!

Anyway great news. Works! everything works. reading low distortion with a THAT C chip in and even lower distortion with THAT A. sounds great! got the output stage working the LM301 chip seems to be doing just fine even with the +/- 30 plus voltages. So I guess DBX knew what they were doing. Also changed meter emulation circuit around because as is I couldn't FOR THE LIFE OF ME get it to work as it was. So now its at least working. GR it works fine. IN/OUT the VU reads to high in the middle....if that makes sense....the tracking is off. @-10 it reads -5  but near the louds and quiets it reads just fine....going to experiment more with it. Maybe I just need to calibrated it some more.
Anyway thanks to all who participated...should have 2nd rev boards here very shortly...
now I want to try a discrete vca!
haha!

oh ps I'll upload the new meter circuit....its not to hard to change....just need to cut a trace and add a resistor...but i'lll wait till its working 100%

You shouldn't have 30volts on the  unregulated output stage. As I noted in an earlier post a PT with a 20-0-20v secondary gets you the 24v called for in the output section.

Glad you got yours working, I'll have to dive into mine and see if I can make further progress.

When you get a chance could you take some voltage measurements for reference/troubleshooting purposes, it would help.

Mark

 

[abechap024]

Yes I bought my transformer before we noted that a 20v would be sufficient. We'll see how long it works till I'm forced to by another transformer.

I will take some voltage measurements. The first whole night of my build I had to trace the rectified voltages all through the circuit because it wasn't compressing. not fun. But finally found a cold solder joint.

So your using a VCA from an 119 aren't you? Did you calibrate the RMS and all that jazz?? Seems pretty tricky with both discrete RMS and VCA but not at all impossible. Can you trace the voltages from the rms to the vca?
Also what kind of voltages would you want? I know when tracing the contol voltage it maxed out at about 3v in places and got pushed down to mV levels before being brought back up...but If I remeber correctly the CV feeding the THAT chip only measures in mV.

BTW for the TRIM to the RMS I just ended up puting in a 50k (says usually 8k to 70k) resistor because after hours and hours measurements werent really panning out.

 

[damnyankee]

Hello.
All the values should be on the BOM and on the board.

But since you are wondering r16 is 1k
and r52 is 100k

what image are you looking at?

Hi AC,

Matthias posted top and bottom high rez images of the his dbx 160VU pcb on page 5 of this thread.

About the dbx 160VU Trim resistors:

There are five of them listed on the dbx 160 schemo and dbx 160 parts list (as well as our CLX pcb/parts list). However, there are no values listed and I believe the plan was NOT to stuff the CLX clone pcb with them.  

However, after reviewing the dbx 160VU schemo and Matthias' high rez images of his dbx 160VU, the dbx TRIM resistor R36 (the clone's R11 TRIM equivalent) IS installed after all.  The value of R36 is 100k 5%.  There lies the very real possibility this resistor needs to be installed in the clone as well.  And I can't really see Matthias' 160VU PCB very well in some spots (it's dark) so two more of those dbx 160VU TRIM resistors may also be installed.  I'll PM Matthias tonight and ask him if he wouldn't mind checking those 5 resistors out.

As for that TRIM cap next to the tant, you are correct: it is NOT installed on the dbx 160VU pcb so we probably won't need it for the CLX clone.

I hope that clarification helps.

DY

 

[Biasrocks]

Yes, discrete everything here.

Mainly interested with DC voltages around the main components, opamps, vca's, rms detector and transistors.

My output sounds terrible right now, just a pfff mostly, I'll try making the mod to the 4558 chip that may help things.

Where did you end up putting the capacitor to stop the oscillation?

Also, a clearer copy of the schematic would help, I find the one provided is really difficult to read.

Mark

 

[haima]

Also, a clearer copy of the schematic would help, I find the one provided is really difficult to read.

Yeah, i gotta agree - it's pretty tough to follow.

Another thing - it'd be a good idea to give each schematic version a number i.e. "rev 0.3" and a date - in the image and the file name - that way we all know we're on the same page.

Another thing that would be good to do would be versions of the overlay for each option (i.e. what discrete parts to leave off when using a THAT vca for example)...

BTW - what is the reason for C5 being rated at 250v?

 

[abechap024]

Yes r36 (on the actual dbx unit) does need to be installed. In the calibration manual it states how to do it. I just guessed and put about 50k in there and it seems to work OK. But something does need to be installed to correctly bias the RMS unit

Also I'm in the process of renaming allll the resistors and transistors to correspond to the original schematics that will make referencing much easier...

Where did you end up putting the capacitor to stop the oscillation?

And I didn't notice oscillation when the 4558 was driving the 200 ohm series resistor with the 200 ohm resistor to ground when driving the CV for the THAT chip. I did notice that the amount of GR and make up gain was limited. When I installed the 1k series  and omitted the ground resistor (as suggested by arska) there was oscillation so a 500pf (what I had handy) between the inverting input and the output fixed it. and greatly increased the amount of GR available. So I don't know for sure how this applies to a discrete vca. But It shouldn't hurt.

Also what is your voltage reading (dc) around r31? shouldn't be more than a volt or so, and it should be changing in relation to the input of signal.

I would try tracing your audio signal from input to output. I used a set of cheap headphones with one end tied to ground and a probe in series with a large value electrolytic cap. Then you can poke around and listen to everything. If its not getting through the VCA correctly then try lifting R31 (200ohm) that will stop any control voltage from reaching the VCA and then It should at least pass signal. If it doesn't then you will know the problem is somewhere inside the vca, or maybe that external matched transistor.

Also manual says that the +15 rail has to be within +/-.100ma for the RMS to work correctly.
Then R36 (r11 on our boards) needs to be installed.

I hope you can get it worked out. it would be nice to know see you get a working unit.

BTW - what is the reason for C5 being rated at 250v?

There is no reason. That is just what DBX spec'ed but plainly the highest voltage you could possibly get in a correctly functioning unit is 48v difference.

 

[abechap024]

Here is the circuit i came up with...for some reason the circuit as it was wasn't working...so this Is what I have installed in mine at the moment and everything works except the tracking is off a bit, and I don't know if that is a calibration setting or what. But this will at least get the ball rolling...


www.acsoundstudio.com/DIY/160vu/2nd%20Rev/METER_CIRCUIT.jpg

And not explicitly explained in the the picture: there still is the 1.5k resistor to ground between the original meter output and the mechanical zero meter emulation circuit 100k resistor input.

EDIT: Feedback resistor should be 100k instead of 50k to get proper response!

 

[damnyankee]

AC,

Your THAT 2180 VCA unit is working, yes?

And I touched base with Dan and he emailed me his KILLER front panel design!  Do you think Dan would be OK to post the design image?  The silver paint (including simulated top/bottom silver 'bars') is going to be a showstopper in the rack!

DY

 

[abechap024]

Yes my design is working. When I head back to the studio today for the weekend I'm Planing on running a bunch of more in depth test and making some sound clips for my enjoyment and for anyone else interested. I'll post some clips.

Excited to put some bass through it. I ran some sampled bass thru it last weekend and at certain settings it made it sound like I was running it through an SPL Transient Designer! Very cool.

OK and as far as a front panel, here is a file for anyone brave enough to do their own metal work. Really not hard if you have the right tools, but still a pain. I know Dan from Risen drums is planning on making some cases available to the forum for this project.

Here are some templates I made up...cant get autocad to export in a higher resolution....ugh!

www.acsoundstudio.com/DIY/160vu/AC_VU_2ruFACEPLATE.bmp
www.acsoundstudio.com/DIY/160vu/AC_VU_2ruFACEPLATE_dim.bmp

and these are made up with the the Hairbal VU scale meter

edit:
More pics and sound clips.

www.acsoundstudio.com/DIY/160vu/TEST/DSCN0322.JPG
www.acsoundstudio.com/DIY/160vu/TEST/DSCN0339.JPG
www.acsoundstudio.com/DIY/160vu/TEST/DSCN0337.JPG
www.acsoundstudio.com/DIY/160vu/TEST/DSCN0357.JPG
www.acsoundstudio.com/DIY/160vu/TEST/DSCN0364.JPG
www.acsoundstudio.com/DIY/160vu/TEST/DSCN03642.JPG
www.acsoundstudio.com/DIY/160vu/TEST/DSCN0320.JPG
www.acsoundstudio.com/DIY/160vu/TEST/DSCN0349.JPG
www.acsoundstudio.com/DIY/160vu/TEST/DSCN0355.JPG


SOUND CLIPS.  Each example starts out with the uncompressed track followed by a compressed version then a more compressed version.

www.acsoundstudio.com/DIY/160vu/TEST/DRUMS_TEST.wav
www.acsoundstudio.com/DIY/160vu/TEST/808_TEST.wav
www.acsoundstudio.com/DIY/160vu/TEST/BASS_TEST.wav
www.acsoundstudio.com/DIY/160vu/TEST/VOX_TEST.wav

Using RMAA the unit has over 90 db of dynamic range noise is over 90db down and THD is .09%

the noise in the bass recording is from the audio file and not largely from the compressor itself.
Also after recording all test I realized that I hadn't correctly configured the SYM on the RMS unit correctly so I have corrected that and doesn't really seem to effect the sound so much but makes compression more even on the top and bottoms of the waveform.
Enjoy!

 

[damnyankee]

Thanks for taking the time to post this, AC!!!

Sounds great!!!

DY

 

[haima]

thanks AC - sounds like it's compressing  ;D

It'd be cool to do some A/B comparison with a real DBX160 - anyone got one?