[BUILD] TB550A

[GregNey]

Nope, quite the contrary, I had everything cranked when I tested it. Will set it up as you suggested tonight.

I can get at it inside an EL500 rack, or rig up a card edge connector with the appropriate connections. I have a signal generator, but not a scope.

 

[GregNey]

They both behave exactly the same with the Freq switches on 0 and the filter down.

 

[horvitz]

That's the same good, I assume?  As in they both pass a signal appropriately?  If so, that means that the problem is in the filter section somewhere.

To narrow down where in the filter section, put a test tone through and then one at a time, turn each band to +2 and then -2.  If no problem, set that band back to 0 and move to the next.  If you find one that doesn't work right, flip through each frequency on the band and see if it impacts all frequencies or just one and also flip through all the gain settings and see if it's some or all of them.  Also, test the HF and LF in both peak and shelf modes.

In short, I'm asking you to test all the settings one at a time and see what works and what doesn't.  Tedious, I know, but it'll help us zoom in on the problem.

 

[GregNey]

No, the same as in my original post. No change. Problem is not with the filter...

 

[horvitz]

Ahh, OK.  Well that's still good information, we're just looking on the other side of the circuit.

First thing I'd like to ask you to do is double check that the small resistors up front are not shorted to the Grayhill switches as described on pg. 20 of the assembly guide.  That can cause your loud problem anyway.  Even if you checked during assembly, it's worth a look again just to cross it off the list.

Beyond that, a good check is to use your DMM and look for DC voltage more than a few mV at the outputs of both opamps and all four DVFs.  That won't show us exactly what's wrong but could potentially narrow down where to look.

Additional testing will require a signal probe or scope.  Since you don't own a scope, a little looking on Google will show you how you can quickly do a DIY probe.  It's going to be hard because that's more subjective but, it will be of assistance in tracing where the signal is getting to.

If you haven't already, also grab the schematic from the tech docs section here.  It's a little confusing but as we're stepping through the circuit, you'll see how it comes together.

We'll get there!

 

[GregNey]

First of all, thank you for your help.. Haven't fixed it yet, but wanted to say thank you nonetheless.

SO, I just probed the LOUD one: I double checked for shorts to the rotary and they're all good. I built a little signal probe, ran a +4, 1k tone in and started tracing the circuit. I started at the end and worked my way backwards. C61 had significantly reduced output. I made it out to C40 on the + input of opamp 2. I had correct signal here and after holding for a second or two, C41 popped... What could that mean?

I did this before looking at the DC readings at the outputs of the opamps/voltage followers.

 

[horvitz]

Yeah, definitely don't thank me yet!  Plenty of work to do still.

Blowing C41 just means it saw a very large voltage.  Better that than to cook your transformer.  I would check the health of your R101 at this point, too.  When C41 goes, that can fry along with it.

When you were looking at C61, which side did you see the reduced output?  You'll want to look at the side connected to the output of A2.

What we are going to find is that either there is a blown trace or a cold solder joint, resulting in a lack of continuity from one place to another or something not being grounded correctly (check R79 and R82 for ground).  That's been nearly 100% of the problems I've seen thus far.  In order to determine where, I'd suggest printing yourself out a copy of the schematic and start testing the different paths and highlight what's good and hopefully determine what isn't.

Where you'll want to start is around A5/S6A/A2 and then to the output.  S6A is the deck in the LF boost/cut nearest the front with the small resistors underneath.

I'm going to send you an email with a copy of that same schematic that I've put through the wringer in photoshop.  Should be a little easier to look at.  Maybe you can send me back some hi-res pictures of both the top and bottom of the main board?  I'm sure you've been staring at it for a while and it never hurts to get some fresh eyes on it.

 

[emrr]

Hey, did you ever think more about a bypass mod using the existing relay?  I found some more variable conditions in my 51X rack that give big level shifts relative to hard bypass.  I'm starting to think there's no solid solution when the unit gets used under varying conditions.

 

[horvitz]

I did think about it, but never got so far as putting it together.  If you want to do it, you're really going to be cutting the thing to ribbons.  Probably around 8-10 traces to cut and then a whole lot of back and forth jumpers.  It would be very messy at best.  If you'd still like to try it, I can get something together for you.

 

[emrr]

Curious to see what you came up with, if not a total PITA to suffer through.  Other option to make bypass closer to universal is probably adding something like THAT 1240 on a daughter board at the input. 

Might be better to just accept condition as is and ignore the bypass when it's not useful. 

 

[horvitz]

OK, you inspired me to at least get it out of my head and onto a drawing.  For everyone reading this, I **DO NOT** recommend doing this mod!  It is untested and probably just a bad idea.  Consider this for educational purposes only.

13 traces to cut. 12 jumpers to install.

Red=traces to cut on top
Blue=traces to cut on bottom
Green=jumpers to install

You will also have to remove the relay where the orange dot is and cut that pin off the relay.  That pad has a thermal to the ground pour and we need it to be open when engaged.  So, cut the pin and probably put some insulation there between the relay and the pad.

You'll need to flip the in/out switch over since it will be backwards.

**EDIT** Sorry but I accidentally deleted this image.

Seems more like something to put on the rev. 2 feature list if that ever happens.

 

[emrr]

Yeah, that looks a little terrifying.  Thanks for the drawing, will mull it over. 

 

[el12123]

Hi,

Thank you for a great kit.  It was fun to assemble, the instructions were great and the layout amazes me in how everything comes together with such tight tolerances.

I am having an issue though and was hoping for some assistance. I am seeing a level drop that seems much higher than what is being discussed in the thread.  I sent some pink noise though the unit and into my DAW was able to audibly and visually verify, via a spectrum analyzer, that all of the functionality of the EQ is working - i.e. I saw and heard changes in the expected places.  However, the drop is much higher than the 4or so db being discussed. I took a couple before and after screenshots of a 1khz Sine wave to show the difference in level:

http://www.flickr.com/photos/el12123/7500110952/in/photostream

I've tried a few different things including different OpAmps and staring at the schematic trying to do some tracing but its time to bite the bullet and ask for help from the experts.

Thanks again,

Elliot

 

[horvitz]

Hi Elliot,

Glad you enjoyed the build!  Yes, it certainly seems that something isn't quite right but we'll get that sorted out.  My guess from you pictures is that it is somewhere around the op amps or DVF boards, but let's not make any assumptions just yet.

First, some questions on your testing.  When you do this test, are all the boost/cut set to 0 and the fltr switch down?  Doing all that eliminates a large section of the circuit, which helps us narrow down where the problem may be.

You mentioned that you swapped op amps, and that's a good thing to try.  The ones you swapped in were known good and working from something else, or were they brand new?

Do you have the ability to do some testing outside of the rack?  If we have to get into more detailed troubleshooting, it can be pretty difficult if you can't poke on it while it's on.

  Brian

 

[el12123]

Hi Brian,

Thanks for the quick response.  The test in the pictures was carried out with boost/cut set to 0 across the board and all switches set to lower position (i.e. filter off).

Op amps were all known good tested in other gear.

I do not have a harness and need to get one - can you recommend one? In the interim I will need to get by with probes, reinsert, test, etc, etc,,,

I also tested each setting of the boost/cut switches together with their Freq switch counterparts, as well as the Shelf/Peaks and Filter - which appears to be some sort of hpf(?) and everything appears to react in the right places audibly and visually.

Not sure that it matters but my signal chain is out of the DAW via a Focusrite Safire to an MP 73 pre DI, out to the TB550a and returned to the DAW.

Thanks for all the help and support,

Elliot

 

[horvitz]

If you want to build an extender, I would suggest either buying the parts from classicapi.com to build one or I just saw a few days ago that JLM is selling a kit to do the same.  Even if you don't end up needing it here, it will be a very handy thing for you to have in the future.

As for the fltr switch, check out my WM thread and there is a description of the functionality and a graphic about halfway down the first page.

So, on to fixing the thing.. first, I would recommend removing the MP73 from your signal chain there.  I assume you are getting a line level out from your Focusrite, and that can go directly into the EQ.  No need to amplify it first.  I don't think this is your problem, but it looks like you are already sending a pretty hot signal out of your DAW and if you make it even louder, you're going to get back unreliable results.  Generate your test tone around -20dBFS and you should be in a good place.

If you put the EQ in the leftmost slot in your rack and remove everything else, is there a chance to at least get some access to it with your DMM?  Simply poking on it on the bench with the power off will really just not be productive since we've got to narrow down where to look before we get to that.  If you can do this, I might start with firing it up with the op amps out and checking for proper voltages at the DOA sockets.  They are labeled +V and -V where you need to look.

Also for easier access, you can actually remove the filter board entirely and the EQ will still work.  Just make sure that all the boost/cut stay at 0.

 

[el12123]

Thanks Brian.

I ordered a kit from JLM - looked the easiest way.

I put the unit in the left most slot, managed to take voltage measurements at the DOA sockets (With filter board in place) and found + and - 15.5v respectively.  (My DMM has been reading a little low of late.) Let me know what you need next and I will attempt ....or we will need to wait on my JLM order....

About the MP73 - It was more about I dont have the right cable on hand (TRS -> XLS Male), so how can I make this work  ....I boosted the signal to 0 to get detail in the pic with EQ in. I will test going forward at -20dBFS.

Thanks you so much for your help,

Elliot

 

[horvitz]

Well, I'd say that a trip to Guitar Center to get the correct cable is probably in order.  I'm not sure what the Focusrite can manage for signal level, but it could just be that you are clipping the EQ into the dirt when it's in.  In fact, that's most likely what's causing the distortion in your second picture.

So I'd say go get the right cable, take out the MP73, run some test tones at -20 (compare in/out again), and let us know what you've got.

 

[el12123]

I was pretty careful when I did this before to verify all my gain stages, but cabling has been re-done as requested - Verified i/o on another module first to verify unity gain.  Then tested with TB550A - The problem is still there:

 

[whomper]

Hi,

Have built two units with was. Great kit and docs!

One of them is not behaving well. When connected and activated, all I hear is white noise.
Have disconnected the filter board, all knobs are on zero and am starting to trace. Have a DMM and Scope.

Where can I find schematics?

Where should I look re.cwhte noise? Signal is not present at all...