[BUILD] 1176 Rev A - Back to the beginning...

[Echo North]

Great! Just took this measurement and got -9.84VDC. This was consistent with both anodes, but didn't change when ratio was changed.

Bingo!

That's it.  Hold on a sec and let me draw you something.

 

[Echo North]

Ok, here is how I understand it. 

There are two ladders/voltage dividers in the ratio switch.  Referring to the MNATS schematic,  divider A acts as a voltage divider for the AC leaving the preamp and entering the control amp.  As you decrease the ratio 20 to 4 you incorporate more resistors and decrease the AC level.  This effectively sets the ratio.  Without getting too much into it, a smaller AC level in this control amp in conjunction with the threshold, will yield a lower ratio.  This is fine with your circuit I believe.  With an input signal, switching the ratios down, you should see decreased AC at pad 22.

Ok now on to divider B.  This sets the threshold which is basically a negative DC offset voltage.  The threshold is fixed.  It decreases with the 20-4 ratio selection.  See the image below.  The -10VDC rail is connected to this network.  The lower the ratio, the lower the offset (more resistors to go through).  At a ratio of 20, this offset must travel through R64.  In my circuit, I get a 4VDC drop across this resistor.  You are getting none.  This is a huge issue, because having a -10VDC offset will mean it'll be impossible to get over the threshold.  It'll just be set way too high.  The more negative the offset the higher the threshold.

So what's the problem?

Well your negative rail could be shorted to that point somehow.  R64 could be shorted but then switching ratio's would get the other resistors involved.  You might not have a DC path to ground and have no flowing current.  Check your current through R64.  You should have like 3.2mA.

Start fishing around that part of the circuit.

 

[Echo North]

I apologize in advance for anything I'm about to say that may be wrong, I'm venturing deeper into the circuit than I have ever been (I didn't design or layout the board).

I think the path to ground is through the "18" connection (release pot). If I disconnect that I get the -10VDC offset voltage. 

So R58, R61, R62, R63 could be dead. Or R80/R60?

 

[Echo North]

With the unit on, no signal.  You should see your negative voltage decrease as you move down R63, R62, R61, R58.

Maybe your 10M R45 is shorted out?  That would do it too. (edit: Probably not)

Check you Ratio pad 18 -> release pot real carefully.

 

[steve_h]

Thanks for all of that! Here are a bunch of readings and details regarding the components you mention:

-I do see voltage decrease at pad 22 with compression engaged and ratios selected starting at 20:1 down to 4:1 - starts at 0.24VAC and goes down to 0.04VAC.

-When testing DC voltage at R64, I see -9.91 which is exactly what my -10V test point reads in the power supply section. This is the reading at both sides of the resistor, which you know. I reflowed/added a touch of new solder to this one earlier. Current at this resistor is -0.2mA DC. It measures 2.18k ohms.

Some readings (compression engaged, 20:1 ratio, all of the usual settings):

Pad 20: -9.91
Pad 18: -1.11
Pad ground: 0
Pad 22: 0
Pad 21: -9.91
Pad 19: -1.08

Ratio board (20:1 engaged):

Pad 18: -1.11
Pad black: 0
Pad 20: -9.91
Pad green: 0
Pad 21: -9.91
Pad 15: 0

R58: -1.11
R61: -9.91
R62: -9.91
R63: -9.91
R45: -9.91
R21: 0
R20: 0
R19: 0
R22: 0
R78: 0

R60: 0 on one side/-1.12 VDC on the other
R80: -1.12 VDC on one side/-9.91 VDC on the other

Going to look at the release pot now - I did double check these connections, touch them up and check for continuity in the past.

 

[Echo North]

So weird.

Your R60/R80 voltages look fine.

The issue, in my mind, is that you have no current flowing from the negative rail, trough R64 and then the ratio resistor array.  In theory, you should see a drop across R64 leaving you with about -5.5VDC then -3VDC after R63, -2.2VDC after R62, -1.2VDC after R61 (all roughly).  Your switch then taps off this and supplies the negative voltage to Pad 21.  Since you have no current, you have no voltage drop.  This means your offset is always -10VDC which is way too high.

Looking at what I've posted above, you should see pad 21 yielding  -5.5VDC at 20, -3VDC at 12, -2.2VDC at 8, -1.2VDC at 4.

There isn't a lot to this part of the circuit if you look at the schematic.  Check R66 and R76.  They are on the negative rail.  Also what is you resistance between pad 18 and GND with the unit OFF?

 

[Echo North]

You could also disconnect the wire from pad 21 to (in theory) confirm this is the issue.

Disconnecting PAD 21 will remove this -10VDC offset from the diode anodes.  Now your unit should compresses (in theory).  It'll compress WAY too much (in theory).  But you should see some compression (in theory).  If that's the case then we know it's this resistor array/offset issue.

Mike (in theory)

 

[steve_h]

Here's the latest:

Resistance between pad 18 and GND is 1.17k with the unit off.

R66 and R76 both show -9.9VDC. I retouched these as well.  Resistance-wise, readings are R66: 6.4k and R76: 4.2k. Again, a little low, but this is in circuit. I'm pretty certain I pulled them to test out of circuit in the past, as I did with most resistors in the meter driver section - still haven't come across a resistor reading that suggests failure or other issues.

Oddly, disconnecting Pad 21 had no effect on compression (fortunately I have terminal blocks installed that made this easy). That is to say, it didn't register on the meter at least, with the usual test settings. I am double checking that signal is making it into the unit to be compressed at all, each step of the way, by switching to +4 to make sure the meter is reflecting the signal.

 

[kishibashi]

I've built two rev a's in the past without too much hassle but scratching my head on this one. As far as I can tell everything on my build is fine. One problem. Audio doesn't pass.
My Q2 voltage where its suppose to read -2.22V reads 2.22V. All my other voltages are pretty darn close. I'm not good enough at electronics to understand why that voltage is not negative for me? Anyone have a suggestion?

 

[mnats]

I'm not good enough at electronics to understand why that voltage is not negative for me? Anyone have a suggestion?

Any one or combination of the three dot points listed on my FAQ/Troubleshooting page around that point. I was just in your hometown - damn, could have had a look!

 

[supiarmando]

Hi there,

I have a little problem with the 2N3708 for my Rev A. BOM asks min. hfe 250 for some transistors. But I measure none higher than 220 hfe. is this a problem?

measure ment is done with DMM @ 10uA and VCE 2,8V.

Any help would be so great. Thanks

 

[Spadehead]

Hi there,

I have a little problem with the 2N3708 for my Rev A. BOM asks min. hfe 250 for some transistors. But I measure none higher than 220 hfe. is this a problem?

measure ment is done with DMM @ 10uA and VCE 2,8V.

Any help would be so great. Thanks

I had exactly the same problem on a Rev D build. None of the 2N3708s I got were up to spec. I re-ordered from Mouser (40 units) and none of those were good either. Luckily I placed an order for 40 2N5088s at the same time as they can also be used and this did the trick...the first 2 I tested rated around hfe 450. You just need to use the alternate base on the pcb.

 

[supiarmando]

Thanks for the tip. I have ordered 65x 2N3708 and 20x 2N5088. Lets see what will come. Datasheets tells hfe for 2n3708 from 45 to 660/800 and for 2n5088 min. 350 both @ 1mA.

After some google search, I think maybe there is no 2n3708 with hfe 250. has anybody a hfe 250?

or we measure not the way mnats does.

What do you think and @mnats any comment is appreciated! Thanks

 

[Ilya]

I've personally have never had any problems with 3708. All were above 250 (ordered 2-3 times from Mouser)

Thanks for the tip. I have ordered 65x 2N3708 and 20x 2N5088. Lets see what will come. Datasheets tells hfe for 2n3708 from 45 to 660/800 and for 2n5088 min. 350 both @ 1mA.

After some google search, I think maybe there is no 2n3708 with hfe 250. has anybody a hfe 250?

or we measure not the way mnats does.

What do you think and @mnats any comment is appreciated! Thanks

 

[supiarmando]

Hi Ilya,

what DMM do you use? Thanks for your answer!

 

[Ilya]

Hi Ilya,

what DMM do you use? Thanks for your answer!

It was a cheap DMM (10$ approximately). The brand is "Unit", I believe.

 

[germoju]

Hi,

I got some questions about the PCB...

- on the PCB R11=10k but on the BOM R11=9,1k
What is the good one ?

- R43 = 38,3k. i tested all my resistor and the 38,3k from Mouser are 38,8k. Is this bad ? Or that's OK ?

- Do you know where is R77 ?

Thanks

 

[steve_h]

Julien - I believe those questions are answered elsewhere in this thread.

For those of you following along at home, the issue with my build was narrowed down to a broken trace/short on the ratio PCB between R1 and R2. Mike gets the credit for figuring that out - thanks Mike! Got a new ratio board and it's all good. Oddly, the bad board had no issues that could be detected visibly.

Just posting this here for the reference of others having issues with their build where the unit is fully functional, but will not compress. I read through this thread and the Rev D thread, and where others were having this same issue, the fix was something different, so here's another possibility - although the takeaway here should be to check the ratio board connections thoroughly, not necessarily just between R1 and R2.

 

[Echo North]

Steve_H's issue involved some troubleshooting that may help some down the road in the "my compressor passes a nice signal, but doesn't compress" category.  Here's an overview of the troubleshooting, it goes through most of the GR control amp.

He was getting an AC signal from the output pot to pad 22 and the ratio switch was trimming it correctly depending on the selected ratio.  His four gain reduction control amp transistors were lining well with the DC voltage chart and his AC signal was strong at the anode of the control amp output rectifier diodes (CR2 and CR3) and the rectified (DC) voltage was present at the anodes.

All Normal.

This leaves the DC offset voltage which basically sets the threshold of compression.  It's present at pad 21 on the cathode side of the rectifiers.  The DC offset was showing -10V DC which is the rail voltage.  This is not correct.  It should be (very roughly) -2 to say -6 VDC depending on the ratio selected, 4 would be -2ish and 20 would be -6ish (again very roughly).  Basically this offset drags the AC signal down and only the peaks get above rectifiers diodes forward bias voltage and get converted to DC.  This DC control voltage gets fed back to the GR FET.  Now if you have -10VDC present at the offset, you'll drag the signal down so much that the diodes will never be able to conduct and you'll never get a control voltage and/or compression.

From the rail Steve saw -10VDC at R64 (normal) with no drop across it (not normal).  Also no drop across R63, R62 or R61 (again not normal).  He did see -0.75VDC at the other side of R58, but that was not a drop.  It was simply the QBias voltage off the release pot.

So why no voltage drop?  Only on answer...no current.  To see a voltage at those resistors but no current drop there had to be a break between R61 and R58.

Hope that helps some people with compression issues.

Mike

 

[germoju]

Julien - I believe those questions are answered elsewhere in this thread.

thanks Steve H.... i'm at the 16th pages and I will read every pages !