[Nomnomnom]

Damn... So that being said I never should have had my DMM reading mA to begin with. Could I have killed something somewhere by doing that? I had it set to measure voltage initially but for whatever reason when I started biasing, I switched it over to mA...
 
I'm guessing that the diodes have absolutely nothing to do with this even where I initially powered it up with DR1A and DR1B bass-ackwards?
 
-Chris

 

[tomb]

Quote:

Damn... So that being said I never should have had my DMM reading mA to begin with. Could I have killed something somewhere by doing that? I had it set to measure voltage initially but for whatever reason when I started biasing, I switched it over to mA...
 
I'm guessing that the diodes have absolutely nothing to do with this even where I initially powered it up with DR1A and DR1B bass-ackwards?
 
-Chris

That's correct.  As Beerguy said, a DMM set in current reading mode has almost zero resistance (it would affect the current readings if it did).  So connecting one in parallel - as you would if you simply touched two probes to a part or test points - effectively creates a short circuit.
 
Yes, the rectifiers (the giant diodes) being reversed probably caused the fuse to blow.  I doubt seriously that it would've caused any damage further down the line, though.  As for the smoke that you got most recently, as Beerguy says, that's a bad sign.  I'm hopeful that the smoke came from the DMM or from the test points themselves in contact with the probes.  You didn't say where the smoke came from, so I'm assuming the best.  Most everything is fairly robust in the MOSFET-MAX, save for the smaller transistors.  Even those have to heat up and smell bad before they're ruined.  A momentary spark and puff of smoke is much better than a high-temp meltdown, which can occur if you have the buffer bias set too high and leave it there for awhile. 
 
We just need to find out what you may have damaged in doing this.  I'm not sure where you are with this right now, but I would assume that everything's OK, first, and then start taking measurements - voltage (with power ON) or resistance (with power OFF) only.
 
BTW, measuring resistance on the power supply will involve some capacitors - that's why the meter takes awhile to settle.  The DMM is actually charging the capacitors when it's measuring resistance.  It will be the same with the MOSFETs, since they have a capacitance built-in (charging and dis-charging MOSFETs is a way to test them).
 

Try going back to the test points first to see what they measure in voltage.  Start with the power supply - do you get some voltage and can you adjust it?  Then move on to the tubes, then buffer, etc.
 

 

[Nomnomnom]

So frustrating!
 
I tested the LM317 by briefly powering on and connecting my leads to GND and the metal tab of the device. The readings (in volts) were jumpy as hell but were there. Here's the interesting thing, with RR3 at FULL CLOCKWISE (beyond the clicking noise) I am reading right around 3.6 volts. At full COUNTER-CLOCKWISE I'm at 2.2 volts. Isn't this backwards of how the trim pot is supposed to work?
 
As far as the slight pop, spark and smoke is concerned. I'm not sure exactly where it came from whether it was the V+ pad or the test lead but it definitely popped the fuse in my DMM which was altogether replaced today.
 
Should I just say screw it and start what looks like one hell of a project by replacing the LM317 with the one I purchased from Rat Shack today? Whenever I hear the term "short circuit" I envision two conductive surfaces touching one another that aren't supposed to and I've looked at my work in extreme detail and I can see nothing of this sort...
 
-Chris

 

[jdkJake]

First, take a deep breath and relax.

Stuff like this happens all the time. Your best bet is NOT to start randomly replacing parts that might be perfectly fine.

Slow and steady will win this race. You put a lot of work into this so far, it might be as simple as a nicked trace or solder blob you cannot see.

Let me catch up with the last few posts. I am sure TomB had some good advice (and i want to read it as well). Don't go pulling parts just yet.

 

[tomb]

Quote:

So frustrating!
 
I tested the LM317 by briefly powering on and connecting my leads to GND and the metal tab of the device. The readings (in volts) were jumpy as hell but were there. Here's the interesting thing, with RR3 at FULL CLOCKWISE (beyond the clicking noise) I am reading right around 3.6 volts. At full COUNTER-CLOCKWISE I'm at 2.2 volts. Isn't this backwards of how the trim pot is supposed to work?
 
As far as the slight pop, spark and smoke is concerned. I'm not sure exactly where it came from whether it was the V+ pad or the test lead but it definitely popped the fuse in my DMM which was altogether replaced today.
 
Should I just say screw it and start what looks like one hell of a project by replacing the LM317 with the one I purchased from Rat Shack today? Whenever I hear the term "short circuit" I envision two conductive surfaces touching one another that aren't supposed to and I've looked at my work in extreme detail and I can see nothing of this sort...
 
-Chris

The direction of the trimmer makes no difference.  It all depends on which way you install it or which way it was drawn on the silkscreen.  To tell the truth, I think the power supply trimmer was alway counter-intuitive (to some people), but it all depends on how you define what you want - less voltage or more voltage - and whether either one corresponds to clockwise or counter-clockwise.  I'm not trying to be flippant, it's just that it seriously doesn't matter.  Bottom line, if it goes opposite the way you think it should, then turn it the other way.
 
Now, what means more is the fact that you're talking voltages that only go from 2.2 to 3.6.  If you are truly supplying 24VAC to the PCB, then I would say the LM317 is gone.  Make sure that you haven't blown the walwart, first, though.

 
Wait a minute, though - let me post a pic!
 
 

 

[jdkJake]

BTW, when you do these power up tests, what do you have the board resting on?

Is it on stand-offs that keep the board elevated from the surface?

Is it Just lying on a piece of clean wood? By clean, I mean no chance of metal chips, bits of wire, solder flakes or the like.

Or are you setting it upon the grey foam pad, possibly conductive foam pad I see in some of the pictures?

 

[jdkJake]

I tested the LM317 by briefly powering on and connecting my leads to GND and the metal tab of the device.

Why did you choose the LM317 metal tab vice the V+ test point?

 

[tomb]

Quote:

Why did you choose the LM317 metal tab vice the V+ test point?

Good point!
 
Meanwhile, power it up and measure these two points and let us know the results - in DC Volts on your DMM, please:
 

 
 

 

[Nomnomnom]

Tom - I have measured at the plug itself where the walwart plugs into the board and get a consistent 26.5 volts. It measures the same where the wires are soldered into the board. That trim pot does seem a little weird but it works...
 
Jake - That foam pad is the underside of a mouse pad but I have tested it three ways with the same results powering it up on the backside of the pad, the front side of the pad and on the bare (very clean) piece of oak it's resting on. I have some shorty standoffs kicking around here somehwere. I'll throw 5 of them on and if nothing else it will eliminate a variable.
 
Lol by the way, I just found the smiley faces...

..... Really? Lol
 
-Chris
 
 
 
 

 

[Nomnomnom]

Another Head-fi member PM-ed me and suggested I test the LM317 that way. I'm not sure if I followed his directions precisely though...

 

[jdkJake]

Avoid the mouse pad (both sides) and stick with the wood board.

It is possible that foam is conductive and might have introduced the problem in the first place. Always use wood, or better still, install the stand-offs.

You should see some of the gifs on other forums I visit. Make's these look down right puritanical.

 

[jdkJake]

Another Head-fi member PM-ed me and suggested I test the LM317 that way. I'm not sure if I followed his directions precisely though...

For that part, the tab is Vout for the device, which is why it must be isolated from the heat sink lest it cause a short (the heat sink is grounded). That is the regulated voltage straight out of the device. Not an incorrect place to measure, just unusual for a self-proclaimed noob.

Let's see what you measure at the points Tom suggested. We need to see what is making it into the device first.

 

[tomb]

Forget that - both points are V+ before the LM317.
 
Just measure one of them and put the other probe in GND.
 
Sorry about that!
 
I edited the original post with a new pic.

 

[Nomnomnom]

Lol yes self-proclaimed noob I am but I recognize and research good advice when I see it

 
On to the results: I think life hates me at this point gentlemen, I powered up and attached the test leads at the right hand rectifier location and POP! Little spark, a pop and a puff of smoke. I measured V+ and GND just for hahas and got the same results as before, very low (5V +/-). Just for S and G's I checked the tube bias while I was at it and both were right around 2.7V. The MOSFET bias points had absolutely nothing. I tried the rectifier locations three times with the same results each time AND the tip of the ground lead was the only thing touching the pad location. I wasn't arcing it against the heat sink or anything crazy like that. I now have a nice little scorch mark at that location.
 
Also, it's on standoffs now...
 
-Chris

 

[Nomnomnom]

Ahhhhh got your edited pic. Was I just shorting it by connecting those points in the first pic?
 
Anyways, I'm hoping this is good news but I got 34.5V utilizing both of the rectifier locations and GND.