Millett "Starving Student" hybrid amp
Jan 11, 2015 at 9:54 AM Post #6,886 of 7,277
 
 
This page on the SSMH website has everything you want to know:
http://www.diyforums.org/SSMH/SSMHheatsink.php

If i lost the shoulder washer can i use a standard washer? 


No.  Plus, the shoulder washers supplied with the Starving Student kit are the best you can get - phenolic resin.  I would hesitate to put anything else in there, including a typical nylon one that comes in a standard TO-220 mounting kit.  The MOSFETs on the Starving Student can get very hot.
 
PM me your address or just your real name.  (I can't match up Beezar customers with Head-Fi usernames.)  I'll drop a couple in the mail to you tomorrow - if you promise not to lose them, this time.
wink.gif

 
Jan 11, 2015 at 12:13 PM Post #6,887 of 7,277
One of the parts of this build that has been stressing me out is drilling out the tube sockets. Mainly after looking at the guide, and reading that 2 sockets were broken while taking the photos. I have all the right tools (up to a couple of years ago I had a small benchtop mill), but just holding those sockets screamed fragile at me. I've been putting it off since Thursday, and decided I might as well go for it today.

I did the first socket as outlined in the guide, but after looking at the socket, and the aftermath of the first one (10 minutes of careful, successful, drilling), I decided to take a different approach on the second one. I flipped the socket over and, using a #3 centre drill, drilled off the head of the pin from the bottom of the socket in about 5 seconds. Small tap with a 1/8" drill bit shank, and the pin falls out the front of the socket. Here is a picture of the socket after with the pin, and what remains of the head:

https://lh4.googleusercontent.com/--K6DvEggfLo/VLKrqTBwc4I/AAAAAAAAAeA/9_NxkRORRYE/w1150-h853-no/IMG_20150111_115304.jpg

I highly recommend using a centre drill; the #3 pictured there cost a whopping $0.50 at a local shop that deals in machining equipment. They are extremely rigid and do not deflect like a conventional drill bit, and are used primarily with lathes to setup for a live centre, or to pilot drill for a drilling operation. I happened to have a half dozen kicking around. You need to use a drill press, not a hand drill.

In respect to clamping, I recklessly held the socket against a piece of wood with my fingers while I drilled it out slowly. Slowly being about 5 seconds for the head of the pin to pop off. You could clamp it with some wood and a vice, but honestly there is next to no torque applied to the socket in this way, and I felt more comfortable steering the pin to the drill bit. Disclaimer, using a drill press in this manner could be dangerous; do so at your own risk.
 
Jan 11, 2015 at 12:30 PM Post #6,888 of 7,277
Good thoughts!
 
Some people have been able to simply pry them apart with a screwdriver, but I've found it depends on the socket manufacturing run whether this is possible or not.  Some sockets might as well be glued together rather than come apart.
 
Whatever cleverness or idea works for you - by all means, use it.
 
Just remember when you glue the halves back together, let them cure/dry right side up.  If the glue flows into the top of the socket, the pin clamps will be useless.
 
JMHO, but I wouldn't stress too much over the sockets.  The silver-tinned ones in the kit are quite inexpensive:
http://www.ebay.com/itm/7-PIN-TUBE-SOCKETS-PC-MOUNT-6AU6-6AQ5-0B2-/390343499665?pt=Vintage_Electronics_R2&hash=item5ae2491f91
 
http://www.ebay.com/itm/New-7-pin-Ceramic-Vacuum-Tube-Sockets-PC-Mount-12AX7/150491434014?_trksid=p2047675.c100005.m1851&_trkparms=aid%3D222007%26algo%3DSIC.MBE%26ao%3D1%26asc%3D27673%26meid%3Dcb5d3425e4ac4c40bfe69ecb77e52fa4%26pid%3D100005%26prg%3D11378%26rk%3D2%26rkt%3D6%26sd%3D390343499665&rt=nc (Pay no attention to the 12AX7 reference!)
Those two are US suppliers, so delivery should be fairly quick.
 
I also have a few silver ones left in stock and have placed them up for sale on Beezar again.   (I normally only stock the gold ones.)
 
Jan 11, 2015 at 8:33 PM Post #6,889 of 7,277
Success! I finished the last of the wiring about 20 minutes ago and have been listening to it for the last 5 minutes.

https://lh4.googleusercontent.com/geUdPY64Vk1TYC51LEDEPZObJ9oNlm8VAwe_lcTPJ28=w1151-h853-no

No troubles. Everything went together smoothly. Great kit, well put together with thorough directions. Thanks!
 
Jan 11, 2015 at 8:50 PM Post #6,890 of 7,277
Success! I finished the last of the wiring about 20 minutes ago and have been listening to it for the last 5 minutes.

https://lh4.googleusercontent.com/geUdPY64Vk1TYC51LEDEPZObJ9oNlm8VAwe_lcTPJ28=w1151-h853-no

No troubles. Everything went together smoothly. Great kit, well put together with thorough directions. Thanks!


Congrats!  Looks great!
biggrin.gif

 
Jan 11, 2015 at 9:04 PM Post #6,891 of 7,277
I managed to find the should washer. That fixed the grounding issue. Amps turns on and sounds great. I've done about 10 hours worth of listening and I'm in love.
I'm having some issues figuring out how to get the knob on. I might have misplaced a washer that used to space it but aside from that I'm up and running!
http://postimg.org/image/e4j81k9z1/
http://postimg.org/image/jwbxzee49/
 
Jan 11, 2015 at 9:15 PM Post #6,892 of 7,277
  I managed to find the should washer. That fixed the grounding issue. Amps turns on and sounds great. I've done about 10 hours worth of listening and I'm in love.
I'm having some issues figuring out how to get the knob on. I might have misplaced a washer that used to space it but aside from that I'm up and running!
http://postimg.org/image/e4j81k9z1/
http://postimg.org/image/jwbxzee49/


Looks great!  Just a hint, though - the endplates go on the outside of the bezels.  The holes on the corners of the endplates are counter-sunk expressly for that purpose.  The PCB should be flush with the end of the case.  The headphone jack and volume pot should stick out just enough to be flush with the endplate.  Use the washer and nut that came with the volume pot, then put the knob on.  The set screw goes on the flat of the pot shaft.  You can use a temporary plastic spacer between the knob and the endplate before tightening down on the flat with the volume knob setscrew - or just eyeball it.  That will keep the knob from scratching the endplate.
 
Get those endplates on the outside of the bezels and it'll probably seem more straightforward.
wink.gif
 
 
Jan 11, 2015 at 9:23 PM Post #6,893 of 7,277
P.S. The headphone jack has three black plastic spacers. Then there is one special tapered spacer.  The special tapered spacer goes against the chrome ferrule.  Use one plastic spacer between that and the front of the endplate.  That will look best.  You should not need any spacers between the jack housing on the PCB and the endplate.  (You'll have two black plastic spacers left over.)
 
Pic from Neutrik:
nmj6hcd2.jpg

 
Jan 14, 2015 at 3:00 PM Post #6,894 of 7,277
I'm planning a SSMH build in a wood case, and mounting the heat sinks directly on top of the case.  Can anyone give a recommendation for a way to thermally insulate the case (say, with some sort of washers) to protect it from heat discoloration/damage?  I've looked and looked, and haven't come across anyone that has done something like this.
 
Jan 21, 2015 at 12:15 PM Post #6,895 of 7,277
Hi,
 
Finished my SSMH build some month ago but forgot to post some feedback.
It was my first electronic build, a very instructing experience.
 
Below are some photos :
http://s1363.photobucket.com/user/mbourrin/media/SSMH/SSMH01_zps0e6479e7.jpg.html?o=2
http://s1363.photobucket.com/user/mbourrin/media/SSMH/SSMH02_zpsf39a7926.jpg.html?o=0
http://s1363.photobucket.com/user/mbourrin/media/SSMH/SSMH03_zpsd84e6da2.jpg.html?o=1
 
Since it was before Beezar kit reissue, this build is based on a Fred2004 PCB (many thanks to him & other contributors).

Some infos :
 
- It's a 12au7 version.
 
- My starting BOM was the one in this post : http://www.head-fi.org/t/664191/starving-student-pcbs/165#post_10256930
I made some changes like upgraded output capacitors.
Also, insulation pads for mosefts are not silicon stuff but aluminium oxyde. Anyway it works good, heatsinks are not too hot.
The 4 front LEDs were an attempt to add diode protection for the headphones but are currently disconnected.
 
- Drilling & adjusting the case can take the most time in the build
If you're working with Fred PCB, my advice is to select a larger case than the small Hammond one I selected : tube socket traces are not centered so I had to air wire the sockets to keep the things centered. Also, PCB was a bit larger than the case so I had to file & grind to slide all the stuff inside.
 
- After checking for DC at the jack & testing with cheap buds, I can say that this amp drives my Q701 nicely with plently of power (input resistors are 47k, pot 50k).
I had to ground the pot metal body to get rid of noise when touching the case.
Now amp is silent with full size headphones, but I get some noise floor when using <16ohm buds, don't know if I can solve this.
I'm using the 48V cisco PSU.
 
Miig
 
Jan 31, 2015 at 9:04 PM Post #6,896 of 7,277
Very nice variation on the Starving Student!  I especially like that you used those heat sinks.
wink.gif
  Some people think they're ugly, but we selected them for the Beezar/ECP Audio SSMH kits because they have just about the most power/heat rejection you can get unless you go to a custom machined heat sink.
 
Is that blue paint or is Hammond using blue anodizing with more cases besides the 1455C801? 
 
Feb 2, 2015 at 10:17 AM Post #6,897 of 7,277
hi to everyone.i have two quick question,you will help me?:)if i put pot 100k must put and 100k resistor,i will have problem with 100k pot?and i have two different polyester cap.one noname 5% tolerance and phillips mkt 370series with 10% tolerance.who is better to use.thnks a lot
 
Feb 2, 2015 at 12:37 PM Post #6,898 of 7,277
hi to everyone.i have two quick question,you will help me?:)if i put pot 100k must put and 100k resistor,i will have problem with 100k pot?and i have two different polyester cap.one noname 5% tolerance and phillips mkt 370series with 10% tolerance.who is better to use.thnks a lot


50K is the happy medium for headphone amplifier volume pots, but 100K is ok. As for the additional input resistance, there is no relationship with the volume pot impedance. Rather, the added input resistance is to deal with the high gain of the tubes. Only you can judge how much is needed with the srength of your source and efficiency of your headphones. 50 - 100K is a good starting point guess. If you don't have much volume travel before reaching uncomfortable volumes, then increase the input resistance.

As for the capacitors ... in the signal path, it's more important whether the film cap is polypropylene rather than polyester (not as good). Precision ratings mean little.
 
Feb 5, 2015 at 8:48 AM Post #6,899 of 7,277
I finished my SSMH kit from beezar.com a few days ago and at first things were good but I've now developed a problem. I've got about 3 hours of time on it and just last night something went terribly wrong with the volume. It is inconsistent between left and right channels. I also still have sound when the volume is turned all the way down and it will seems to randomly increase and decrease and can get painfully loud.  Could this be the volume control or do I have a bigger problem somewhere else? If I get a multimeter can I check that the volume pot is operating correcting as I sweep across the range?
 
Thanks,
Luke
 
Feb 5, 2015 at 9:29 AM Post #6,900 of 7,277
I don't think there's any way the volume pot would cause this. I think you have a bigger problem elsewhere.

Carefully disassemble the amp. Check all the wiring to see if there's any chafing with the case. Then carefully inspect the wiring at the terminal blocks. These can come loose over time, especially if the wiring is chafing or binding with case.

Check all of your solder joints - re-flow them where necessary. Look to see if you've got any bridging you didn't notice before. See if the MOSFETs show signs of burning and whether the tabs have contacted the case.

Finally, while it's still disassembled, check the operation again. Let us know if you find anything.
 

Users who are viewing this thread

Back
Top