[Nomnomnom]

Okay while I've been lurking for years on head-fi now this is my first post so please be gentle lol.
 
I recently assembled a MOSFET MAX kit from glassjaraudio (excellent kit by the way) along with some boutique parts from Beezar and partsconnexion. I followed all of the tips on diyforums.org pertaining to orientation of the parts and initial setup of the trim pots. When I put power to it (supplied Triad walwart w/ 24VAC@1000ma output) I got nothing. No LED, zilch, nadda. Now I shot an email out to TomB on Easter Sunday not expecting a reply but I'll be damned if I didn't have an email from him shortly after. He shot me a couple of ideas to look for and much to my dismay, my fuse was blown. I moved the power supply over to the "no fuse" pads, plugged her in and BAM! LEDs shining bright so I began the biasing procedure.
 
For whatever reason I could not get the power supply voltage over 25 volts so I decided to move on to the MOSFET and tube bias. I got the MOSFETS right about where I wanted them and decided to double check the power supply. Well, I had my DMM set to read mA and I'm not sure if that's what cause the following or not but figured I'd mention it. I put one lead in the GND test point and put the other in the V+ and POOF! A tiny spark and a little bit of smoke came out.
 
Now, Tom brought up a good point before I even started biasing. I emailed him when the MAX powered up but failed to check my email again until I had made a tiny S.O.S. smoke signal. Had I checked it I would have noticed that he emailed me back concerned because something had to have popped the fuse in the first place. So hindsight being 20/20 what I did today was reflowed, cleaned and attempted to power up the board again. Also from his suggestion I went to my local Rat Shack in search of the polyfuse located at PF1 (I got some interesting yet blank looks from the sales clerk there).
 
After doing a little more research however, it looks like a polyfuse doesn't pop? I'm far from an electronic engineer but from what I read it seems as though this is a part that is made to shut down and then reconnect once the current gets back to a specified level. So basically this is what I'm trying to achieve with this post. I hate bugging Tom with email after email. I felt terrible that he even replied to me on Easter Sunday but as far as I'm concerned that speaks volumes about him as a person and is extremely commendable. So what I figured I'd try before ordering every single part in the power supply (that LM317 looks like it would be a PITA to replace) is post my woes here to the community.
 
I have a bunch of fairly decent pictures taken from multiple angles throughout the build so with that being said, I have no clue how to post them and I don't want to be "that guy" who posts these MEGA SIZED PICS that take 6 minutes to view on a broadband connection.
 
Once again I've been lurking on here for quite some time. I saw the rise and fall of Bilavideo (SR60 thread lol), noticed the way the DIY community was and now I'm calling on all of you with infinitely more knowledge than myself on this subject. HELP ME lol.
 
Oh, one more thing, I almost sacrificed my modded T50RPs by hooking them up while messing around with it and got a good 8 seconds of AMAZING MUSIC! And that was through a UDAC2 as a source. I have a grubDAC that's sitting on my desk itching to be built and put in a good home and I believe this is the place for it to go.
 
Thank you all in advance,
 
-Chris
 
Oh and as a side note for anybody who, like I was, is unaware of certain aspects of this DIY community... When I had a question about my shipment from glassjaraudio I called the customer service number and Jeff answered. When I had questions about my order from Beezar I emailed the customer service address and Tom replied. In all honesty this freaked me out a little bit because I thought that these two places were multiple employee businesses but I'll be damned if they're not one man operations so for those of you reading this who are not in the know; these guys are freakin awesome and are bending over backwards for this community. I have had nothing but good luck talking to these men and highly recommend them and their expertise in anyone venturing into this "hobby(obsession?)"...

 

[jdkJake]

There seems to be confusion between the fuse and the poly fuse (pf1). They are different beasts. Regardless, you bypassed the fuse (not the poly fuse) with the input to the terminal block. Not a good idea. As Tom has stated, that was your first indication of problems (most likely a short). You should be able to get 1A slo-blo fuses in that package size from Rat Shack (5x20mm if I recall correctly). By a couple, you will probably need them.

Just post the pictures. The board will resize as appropriate. Copy and paste works fine.

Also advise going out and buying a cheap set of test headphones. No reason to put good gear in harms way until the circuit tests out appropriately.

 

[Nomnomnom]

jdkJake,
 
I'm sorry if I hinted that there was confusion between the fuse and polyfuse. I definitely see the difference. My question concerning the polyfuse was more along the lines of does it need replacement considering my not so good idea of bypassing the actual fuse. I went to Rat Shack today and bought a few slo-blos (@1/2 amp instead of 1 amp) and when I tried to power the beast up again tonight after the reflow I had the same results. Immedaitely blew the fuse. Once again, bypassed the fuse (ok, ok I won't do it again) and the same results. I'm going to attempt to upload my pics and see what happens...

 

[jdkJake]

Nice pics. This is a very clean build.

So, please unhook everything from the board including the power supply. With the board powered OFF, please use your multimeter to measure for resistance (ohms) across V+ and GND. We are first going to check for a short between power and ground.

 

[jdkJake]

So, did you use TO-220 mounting kits for your MOSFETs and power regulator?

It does not appear that you used the plastic shoulder washer correctly. Is that a metal flat washer used on the device side of the heat sink? That would be a no-no unless those devices are plastic cases, which, they are not.

From device side, it should be: bolt, plastic should washer oriented to fit into the device, heat transfer pad, device, heat sink, flat washer, lock washer, nut.

Please remount and perform the measurements again. This need to be done for all of the MOSFETs as well as the voltage regulator. Basically, any TO-220 device.

 

[jdkJake]

For reference:

https://www.aavidthermalloy.com/product-group/accessories/mounting-kits

Scroll down a bit and it will show the proper mounting technique.

 

[Nomnomnom]

Yes, I do believe I used the correct mounting kit as described here http://www.diyforums.org/MOSFET-MAX/MOSFET-MAXheatsink.php. I used this same setup (Bolt, flat washer, shoulder washer, device, flat washer, lock washer, nut) for the four MOSFETS (correct terminology?) and the LM317. The flat washers may be overshadowing the shoulder washers in the pics but here's a couple more that I hope show the correct mounting?
 
 

 

[Nomnomnom]

jdkJake,
 
   For whatever reason my DMM is doing some really crazy crap right now (it's bouncing all over the place when placing the probes into the test points (why do I feel so dirty when I say this lol)) so I'm going to hold off until tomorrow, gonna buy a new one and see what the magic genie says I really do appreciate all your help and I'll continue heeding your advice. Your one of the ones I've been following through my time here on head-fi so I trust you. Thank you for the compliments by the way. After TomB talked to me about my build and the possibility of shorts I reflowed EVERYTHING and cleaned the heck out of it with a toothbrush, rubbing alcohol and multiple rags. I've tried everything and still don't understand why it kills the fuse before anything...
 
Oh hey, my camera is up and running so if you can think of anything that will help, please let me know. I have a little tripod for it too so if nothing else my photography skills can be put to the test also lol

 

[jdkJake]

Okay, those pics of the TO-220 mounting are better. You should be fine unless you torqued the crap out of the screws and broke through the thermal transfer pad. Once you get your DMM online, you can measure that as well (metal tab of TO-200 to the nut, measure resistance and look for an open or infinite resistance vice a short, which is no resistance).

You have a short somewhere on the board. We just need to find it. A functioning DMM will be critical as the pictures don't show anything jumping out.

If you do not have it, be sure to pick up some desolder braid or a solder pump. We might have to remove a part or two to isolate the problem.

 

[tomb]

Here's a thought - did you use the Toshiba JFETs for QB2 and QB3?  If not, and you used 2N5087-88's, then they should NOT be reversed.  I only suggest this because the Toshiba JFETs have disappeared from my supplier.  That means I've run out and haven't sold any in awhile.  I think AMB still has them and I've heard that B&D sells them at quite a price, but I'm pretty sure you didn't try to order them from me.  So, just wondering what you used there and it makes a difference whether they're supposed to be reversed or not (yours are reversed).
 
I agree with Jake - your work looks even better at second look.  It's got to be something simple that we're missing.
 
 
EDIT: Well, they look like the JFETs in some of your pics (they look bigger than the 2N5087-88's).  So, that's not the problem.
 
You'll need to solder a jumper at J1 at some point.  It doesn't affect anything but the signal input and if you're feeding directly to the extra pot pads right now, it won't make a difference.

 

[tomb]

Hmm ... you had one of the rectifiers reversed in the earlier pics, but it looks like you corrected that in the later ones.
 
I'll keep looking ...

 

[Nomnomnom]

I had two of the rectifiers (DR1A, DR1B) backwards... You are corrrect TomB... I'll continue utilizing the advise of you and others gave me tomorrow, I'm crazy tired right now but if anyone can think of anything to post or if any pictures are needed than please throw a message on here. The few seconds of music I got before my poor MAX died were excellent and I'm aiming for a repeat here.... 

 

[beerguy0]

 
Quote:

Well, I had my DMM set to read mA and I'm not sure if that's what cause the following or not but figured I'd mention it. I put one lead in the GND test point and put the other in the V+ and POOF! A tiny spark and a little bit of smoke came out.

 
 
When you have the DMM set to measure current, the meter is basically a short circuit. There is a very small resistance in the DMM, and the DMM uses the voltage drop across the resistor to calculate the current. You may have damaged the LM317 regulator (smoke is never a good sign in electronics

). Also, was the meter dial just set to mA, or did you have the DMM leads in the current measurement position as well?
 
There is also the possibility you may have damaged the meter. Can you still make voltage or resistance measurements? There is usually a fuse in the DMM to protect the meter from over current. If it is blown, I think you should still be able to measure voltage. Try measuring a known voltage on something else, like a battery, and see if you get a reading that looks correct.

 

[Nomnomnom]

Okay gentlemen I raided the local Rat Shack and got some more supplies together.
 
Purchased a brand spanky new DMM (Model # 2200813) to replace my little one (22-820).
 
Oddly enough they had LM317Ts in stock so I grabbed up one of them as well as some rectifier diodes in assorted sizes (now that I'm home I see that none of them match the BOM)
 
A desoldering bulb (I have plenty of braid but figured it may come in handy)
 
Tom - You are correct once again. QB2 is 2SJ74BL and QB3 is 2SK170BL as specified in the BOM
       
        - About those diodes. I powered the board up when DR1A and DR1B were bass-ackwards. Nothing lit up which caused me to double check everything and that's when I noticed them and fixed them. Could this be the cause of the initial fuse being blown and could this have caused more damage down the line? Like I said I grabbed a couple while I was at Rat Shack cause they were cheap as hell and I'm tired of making trips there but one is 200V while the other is 50V. Both are 3amps.
 
        - I'll throw that jumper in @ J1 sometime tonight. The only reason I'm using the pads by the pot is because those punch down blocks seem to let go of the wires everytime I move the board. They will be used with the final product however which will also include a grubDAC that's sitting on my desk waiting to be built (need a new soldering iron tip, stock Hakko 936 tip is freakin huge).
 
Jake - Resistance across V+ and GND = approx. 2.1 kOhms but it took a while to settle there.
 
        - Resistance across all 4 MOSFETS and the LM317 was present but highly variable. I couldn't get a consistent reading on any of them and this was tested with one lead on the large metal tab of the device to the nut on the other side of the heat sink.
 
Beerguy - The miniscule DMM I was using had leads that were permanently attached to the device. It had no dial just a 3 position switch and a "Select" button. Needless to say this little adventure pushed me to purchase a proper DMM with more features than I'll probably ever need.
 
              - I'm really hoping the LM317 isn't fried but like I said I did see one at the Shack so figured I'd grab it up. Any difference in brands as far as that's concerned?
 
To all - Thank you all once again. I look forward to your input.
 
Once again I'm not an electronic engineer so please excuse my noob-ish questions. Hell, I'm googling how to use this DMM right now...
 
-Chris
 
- Oh yes, sacrificial iPod earbuds from here on out. Guess I was jumping the gun a bit from the testing stage to comparison.
 
 

 

[tomb]

Just a reminder relative to what Beerguy posted - never, ever try to measure current with the MAX/MiniMAX/MOSFET-MAX.  Unfortunately, it's such an accepted mode of practice, we don't explain it much around here anymore.  One of the reasons is to actually measure current, you need to splice the meter into the circuit in series.  This is very difficult to do without removing parts and/or chopping up traces, so it's almost never done.
 
Everything is read in voltage, instead.  To bias the output buffer, you want to measure the voltage in mV across the output power resistors.  That's the same thing as measuring at the points identifed on the PCB and the setup page on the website.  The idea is that I = V/R (Ohm's Law).  So, whatever you measure in mV in the buffer, you divide by 2.2 (the ohms of the power resistors).  That gives you the buffer current bias.  For instance, for 100ma, you'd want to set the trimmers so that the DMM measured 220mV.