All things G1176 - the new "repost" thread.

[spacecho]

Hello Gachet,

i had a similar problem of no compression in my 1176. I used sockets for the bf245a's and i found that the connection was intermittent. When i soldered these parts in it works fine.

I suspect that you have either burned them when soldering them (i think these parts are quite sensitive to heat so you really need to be careful when heatsinking them....) or if you have used sockets they are intermittent. (i found my sockets a little tight and it was hard to get the transistor legs into the holes.

you mentioned that your input and output gain is ok. The circuit will behave this way without the bf245's in place so i think thats where the problem lies...

hope this helps

 

[fucanay]

[quote author="RogerFoote"][quote author="fucanay"]I just finished the front panel for my first G1176 and I plugged it in and it sounds great even without having done the calibration. I've never calibrated anything and I'm not entirely sure my DMM is right for it. I'm supposed to be reading mV but I think I can only do down to a tenth of a volt AC. Anything less show 0.0 . So I think I have some problems on that front. Anyone care to point me at a decent DMM for this?

Matt[/quote]

Matt

Any "decent" DMM that has a mV range is ok. I use Flukes mostly, but we have a CEN-TECH P35761 ($10 Harbor Freight special) that will do ok for mV measurements. The CEN-TECH has a low range of 0-200mV.[/quote]

Thanks Roger,

My DMM has no Mv range on it. It only has a 250 and 700 AC setting on it. There is a Harbor Freight over in Newark, so I'll go see what I can find.

So when it says that I should get 50mV on the input, that mean at the TRS jack between the the Ring and Tip or does it mean from the Tip and Sleeve? :?

Matt

 

[enthalpystudios]

Hope this helps out, but you want 50 mv between tip and sleeve matt.

How are you generating your signal? I used wavelab to do mine, don't have a signal/function generator yet..... If you're going that route, you can just take an output from your sound card.

Anyway, I plug a trs into an output of the line amp outs on my motu, but a any soundcard output will get you 50mv. Anyway, wherever you're getting your signal, clip your dmm onto tip and sleeve (pin 2 and pin 1 on an xlr, respectively) to get your 50mv.

the important thing is that you set unity gain on the 1176, just as long as it's as low as possible... I believe the idea is to set unity gain on the line amp while insuring that when you switch to GR mode no compression will take place until you raise the input signal (with your computer or signal generator volume, not the 1176) if your dmm only goes down to 100mv, that should work, pending an expert opinion... i'm pretty sure a 100mv signal wont bust the threshold.

Then go ahead and probe the output, tip and sleeve, and set the input/output until you get that same 50mv. Then follow the rest of the procedure to get -6db gain reduction.

Two meters is a really good idea for this step. Its really hard to set an input signal that gets that -6db of gain reduction when you have to jump your DMM from input to output, back and forth, and set the input signal at the same time.


I have a couple questions as well, however...

Also, is there a ballpark range for the input signal that should achieve the -6db? I seem to endup with a pretty damn high input signal to get the compression going.

And can this be done with a scope, hooking up input and output signals to the scope? I just got a decent dualtrace scope, and if I can do it like that I have a feeling it would be way easier. Wouldn't a normal dmm, even a fluke true rms, not really be as good as a dedicated ac mv meter? I seem to recall people mentioning that dmm's don't track all that well as frequency increases.

Billy

 

[fucanay]

The good news is that you all helped me adjust this thing and I'm done. Suddenly the meter makes more sense now too. I went out to Harbor Freight and got the new DMM with a mV setting on it and everything went perfect from there.

Billy,

I used Logic Pro's Test Oscillator to output the signal out of my M-Audio Delta 1010. It worked pretty good. Resolution from Logics faders was a bit hard to deal with, but I eventually got it using a combo of the channel and master faders.

But I think I'm done with this one, I just need to pretty up the wiring and close the case. Then it's onto the next one which is mostly complete. I just wanted to go through all of the wiring and calibrating on one unit to make sure I kept focused.

I love when something gets done.

Thanks for all the help.

Matt

 

[imo]

I have my 1176 up and running, did the ground cut, reduced the noise floor quite a bit. ALso matched the FET's, whiched really helped with the meter tracking.
I am getting compression pretty early on, with the unit properly biased and set up. I find myself having to keep the input knob around 9 oclock if i just want a few db's. I was wondering if anyone had any ideas on the best way to tame the threshold on the unit. Would it be a resistor to the input pot, or would i be better off ammending it somewhere else?
Thanks
Ian

 

[gyraf]

..the only really sensible way to alter threshold sensitivity is finding fet's with different cutoff voltage. The BF245A will get you in the approximately correct range..

Jakob E.

 

[kato]

I used Synthetic's voltage checkplot to create this one for the reverse side of the board - if it's useful to anyone:

Voltage Checkplot: Reverse - right-click, save as...

It's the same as Synthetic's version except it's for the backside of the board rather than the front. The only advantage is that the voltages are printed left-to-right. (easier to read)

Here's the front side. Though I'm not sure what good a frontside voltage check is - there's no way to reach most of the spots.

edit 24 hours after this original post I realized I posted a "mirror image" of the reverse. Sorry if anyone printed it and wasted their time. It's now corrected.

 

[kato]

I did a quick check of my voltages and nearly all of them are low!

Examples:

.86v should be 1.02v
1.73v should be 1.76v
25.8v - correct
0.54v should be .51v (higher)
11.9v should be 11.5v (higher)
12.3v should be 11.68v (higher)
16.68v should be 17.25v
4.26v should be 4.5v
14.82v should be 15.3v
4.0v should be 4.03v
27.8v - correct
30.3v - correct
14.42v should be 14.54v
-9.76v should be -10v
3.4v should be 3.51v
16.75v should be 16.72v
29.1v - correct
14.6v should be 14.06v (higher)
29.7v - correct
28.4v should be 28.6v
4.26v should be 4.98v
4.46v should be 4.54v
13.95 should be 13.35 (higher)
13.98 should be 13.40 (higher)
and
30.3v should be 35.4v!!!! (this one is way off)

Aside from this last one, do any of these voltages look like cause for alarm?

None of the switches, jacks, knobs, meters, or lamps are connected - just the PCB in a case connected to the transformer. (not even an on-off switch yet.)

Also, I checked the volts with only the power supply built up and got -9.7v instead of -10. I went ahead and stuffed the board anyway. Could this low -10 be due to using an underpowered transformer? I'm using the Amveco TE62045 (10VA, .5A) originally specified in the N.American Parts List thread. Some have suggested it's too small for the job and to instead use part TE62055 (15VA, .6A).

The input and output tx are already installed. (Lundahl in, OEP out.)
The IC1 (TL071) is not installed.
I don't know if they affect the voltages.

Any advice? I would love to hear it. I'm still new to this...

Kato

 

[gyraf]

Voltages can be considered correct if within +/-10% - execpt for the +30V that needs to be +29V or more to ensure correct regulation.

What is interesting for disgnosis is the 0.6V base-emitter drop (difference) on individual transistors.

Jakob E.

 

[enthalpystudios]

Right now I've got a stancor wf30 wired 600:600 from the input to the pads for the oep/lundahl input secondary's. Should this be wired 2:1? or 1:2? I'm really trying to figure out how the lundahl is wired, but something confuses me about the lundahl datasheets, I just can't quite get it.

I could see wiring it 2:1 rather than the 1:1 I have now, as my units seem to have too much gain, especially with the OEP output tx's reversed for 1:2.

Hopefully someone knows offhand?

Thanks,

billy

 

[Gachet]

Hi the professionnals

I'm building A dual 1176 with in and out transfos and having one unit working great and the other not.So I started comparing the units?

Without the fet it's passing audio like the other unit.
With the fet,it needs a lot more input gain even when bypassed.
In ground mode same problem and no compression and the needle doesn't go at 0db.
I have cheked Q12 Q13 Q14 Q15 they are OK

The problem comes from the GR stage because I have a positive voltage moving between 2 and 4,5 instead of -1,4V at point 7.
At C17 R60 instead of 4,55V Ihave something moving between 1,9 and 4,6.
And the same at point 22 instead of 0V I have smething moving between 1,9V et 4,6V.
At R60 & R62 I have something moving between 17V 25v

What are the moving voltages saying?I'm not enough experienced to understand that sort of bug.
ANY IDEA??

 

[gyraf]

Have you calibrated the unit properly?

If you have a lot of GR just by putting the FET in, this points towards the FET taking off a great deal more than the 1dB that is the target for the q-bias adjustment procedure.

If you have the problem after adjusting properly, try different FET's - and check for soldering, shorts, components, orientation and interwireing as always.

Jakob E.

 

[Gachet]

hi gyraf

yes I've tried to calibrate the bias but it's not reacting as it should.
I have check for soldering, shorts, components, orientation and interwireing at least 3 times each.
I have tried different fets.
And nothing Ican't find the problem :shock:

Except those moving voltages I've no clue.

...

 

[gyraf]

Gachet,

(Could you please resize your avatar to the recommended 60x60 pixels - it takes up a lot of screen-space all through the pages you participate in)

There's nothing to do but check and double-check. possibly try new FET's.

Jakob E.

 

[kato]

[quote author="gyraf"]Voltages can be considered correct if within +/-10% - execpt for the +30V that needs to be +29V or more to ensure correct regulation.
. . .
[/quote]

Hey thanks for the response Jakob! I appreciate it.

So it looks like most of my voltages are within 10%.
The really *big*error is actually a mistake in the checkplot voltage guide I used. I will update my version of this checkplot (one page back) this evening. I'm a little surprised that Synthetic's checkplot was posted almost nine months ago and no one has mentioned this.

The point marked 35.4v is at pin 1 of the 7824 on the UREI schematic. The voltage checkplot points this value to pin 2 - where I'm getting 33.7v which is a lot closer to 35.4v than 30.2v. (only 1.7v off as opposed to 5.2v off. Within 10%.)

The thing that clued me in was looking back at Mnat's wiring guide where he says to expect +30v at pin 2...

I called it 6-ish because I don't know what's considered correct - and I'm getting 6.11v there. And 30.2v at the spot Mnats calls +30v.

I think I'm ready to stop obsessing over perfect voltages and get this mofo racked and calibrated!

 

[Gachet]

Hi

At last my dual 1176 is fully working and calibrated. :grin:

And I wondered about stereo interconnection.

Is it possible to connect both fet to channel 1 GR circuit in order to have same ratio attack release and bias on both channels(And disconnect wire 7 of channel 2).
To my mind it's possible in the case of matched fets:same bias voltage.(I have some matched pairs)
Will the different current (caused by the shunt) affect the behaving of the fets?
Second I'd like to feed the channel 1 GR circuit (from R26)with a mix of both channels so the compression will be trigged by both channels.
I'd like to add a variable resistor between the shunt and R26 to calibrate the change of level caused by the stereo shunt when passing in stereo mode.

Will that working?
Will the unit tolerate that?

 

[gyraf]

Gachet,

Look through this thread - and the 1176-meta - and you'll find all info you need about stereo linking.

Jakob E.

 

[dukasound]

Hi
Did I miss, is it better choise to use output transfomer with 1:2 or 2:1 ratio.
We have Jakob project wher is Lundahl wired 2:1 and Mnats who is wired 1:2. Also, we disscused about original 12614 transformer and someone told us transformer have 1:1 ratio.
:?
Duka

 

[gyraf]

There's no simple answer to that. Depends entirely on the transformer in question and what you want to acheive.

With OEP's I wouldn't go over 1:1 for power reasons, but with the Lundahl LL5402 you can wire it 1:2 if you feel you need the additional gain.

On the other hand, for a sound closest to the original - that is, the rev#F we have in our own studio - use the design as-is. This has lower gain than the original though, but I valued original sonics over original-gain in this adaptation. We never use the units at full-gain anyway.

Jakob E.

 

[Gachet]

...