Millett "Starving Student" hybrid amp
Jun 15, 2010 at 7:36 AM Post #5,296 of 7,277

 
Quote:
SNIP  
This next pic is a shot at my power area. it has the upper tube (not working) as well as the power jack and the wire running to the switch. the green terminal is used to ground wires. all wires in that connection are solidly screwed down.

SNIP

 
Thanks for all your patience and help regarding my project. I believed i could do this before I started, and its very frustrating to be stuck and feel so close. Any help or ideas to help clean up my wire mess would be appreciated. Where can I find those connection terminals/braces that lots of people have used in their build? its the one where people usually bolt it down to their tube socket. If i knew what that was called i would definately use those for my next build.


The connection points in a terminal block DO NOT connect to one another. Unless you somehow bridged the different connections together, this defeats your grounding.
 
Indeed it seems you're very very close, so I'd recommend you do not give up!  Just give it a rest, for a few days, even a week so you come back to the project with a clean state of mind. Rushing it is the best way to mess up an electronics project, particularly when it's your first :)
 
cheers!
 
Jun 16, 2010 at 10:33 PM Post #5,298 of 7,277
Well.
 
I fixed it. But now somethings broken. Just my luck.
 
Anyway, if you read my former posts, i found out there was a short, put my amp in a new abs box blah blah.
 
I lost an RCA jack along the way (silly me)
 
So i didn't use it when powering up the amp for the first time. So, it ran PERFECT. Played the first track of Fear of a Black Planet through it. But then, i left it on for quite a few hours. I mean, we're talking 4-5 hours, i had to go to my fathers birthday dinner.
 
So i went back to play another track before bed to celebrate my own personal victory, and well....I put my earbud that worked in (test set of cheap buds, didn't wanna risk my cans!) Well, all i heard was a pulsing BUZZ, over and over. Only on one of my tubes though, it would pulse with my tube heater (you would see the orange flash in unison with the buzz
 
Other tube wasn't hot, the MOSFETs weren't giving off heat, nothing. Just the one, firing over and over, making a buzz. Turned it off, waited like 10 minutes for it to cool, turned it back on, same thing
 
What the heck could be wrong? It worked for the first 10 minutes, then left it for a few hours, now it's broken, what happened?? I know this shouldn't happen in that period of time.
 
Any help you could provide, that would be appreciated, AGAIN. I hope this is the last of the issues.
 
Jun 20, 2010 at 11:46 AM Post #5,299 of 7,277
I just added the optional 1K resistors R14 and R15 to my build. My SSMH is now less noisy than before. Touching the metal casing of my iPod Nano with the music paused doesn't introduce noise anymore. Both resistors are in fact grid stopper resistors and help to filter out high frequency (radio frequency) noise. I soldered them as close as possible to the tube socket.
 
IMO these "input resistors" as the_equalizer called them aren't optional at all! 
 
Jun 20, 2010 at 5:34 PM Post #5,300 of 7,277
I finally got a new PSU yesterday and the amp  is back to working as it should.  It still takes about 15 seconds for the tubes to light up but I think that is because of the 680uf caps for C1 and C6.  I'm going to keep paying attention to the start-up time and if it gets to be to long I'll install a switch to start one side then the other.  When I took one tube out just to see what would happen it came on in about 3 seconds. 
 
Thanks for all the help on this.
 
Bryant
 
Jun 20, 2010 at 10:59 PM Post #5,302 of 7,277
If I got four relatively cheap (~$5 each) 12AU7 tubes (or specifically, 5814s, which are the milspec versions, with no manufacturer listed), what are the chances they're going to be mismatched? I'm wondering if I should pay the premium for those 19J6s...
 
Jun 21, 2010 at 2:02 AM Post #5,303 of 7,277
Quote:
If I got four relatively cheap (~$5 each) 12AU7 tubes (or specifically, 5814s, which are the milspec versions, with no manufacturer listed), what are the chances they're going to be mismatched? I'm wondering if I should pay the premium for those 19J6s...

If I were you I'd skip the on 19J6's and go for the 12AU7 instead. The 19J6's are increasingly hard to get and will command an ever increasing price. Basic supply and demand. The 12AU7 however is still made today, very common and thus easy to get. Price for a current production 12AU7 is ~$10 each. AFAIK the 19J6 tube is nothing special at all, Pete Millett used them because they were cheap and plentiful. Now that they're very hard to get and expensive IMO it's time to look at other tubes. The 12AU7 or it's family members are a safe bet. If you wanna go cheap, look into the 12SR7 mod. I can get NOS 12SR7 tubes for €2 each at a Dutch webshop.
 
Jun 21, 2010 at 9:47 AM Post #5,304 of 7,277
Why do you need matched tubes for a single ended amp?
you dont.
 
Quote:
AFAIK the 19J6 tube is nothing special at all...


The 19j6 is actually a fairly special tube. What is neat about it is that it has a single cathode shared between the 2 plates & grids. It would work outstandingly well in a long tail pair.
 
Jun 21, 2010 at 11:45 AM Post #5,306 of 7,277


Quote:
Why do you need matched tubes for a single ended amp?
you dont.
 
 

With respect nikongod, that could be viewed as an opinion.  Typically, the standard mantra is that power tubes need matching, but not pre-amp tubes.  However, with our sensitive headphones (as compared to guitars or speakers) and the typically high gains in most tubes - a mis-match can be detected.  It manifests itself as an imbalance between channel dynamics.  One headphone driver may sound a little dead compared to the other.  This can happen even with manual bias adjustments that will center the sound between channels.
 
Granted, it could be that it's peculiar to tube hybrids (the limit of my experience).  It may be that the differences are magnified simply because there's only a singe tube providing all of the gain within a channel.  That's only conjecture, of course, but in my experience with Millletts/Maxes/MiniMAXes and the Starving Student, tube matching can make a difference.  The Starving Student is a bit better in this respect because of the single cathode and paralleled grids/plates tending to level the differences, but some people are still irritated by a tube mis-match.
 
I agree that the chances are still fairly slim that two random, un-matched tubes will cause a noticeable effect.  It all depends on whether someone wants that assurance, relative to the cost of matching.

 
 
Jun 21, 2010 at 1:21 PM Post #5,307 of 7,277
 
[size=medium]
Has anyone here done a P2P build with dsavitsk's additions? If so, was it worth it?
 
Also, I want to put a gain switch on this build that selects between two sets of input resistors, 50k and 100k. And as I see it, there are two ways of doing this, with either a power switch and a DPDT switch, a power switch and a rotary switch, or just a (very special) rotary switch (off, low, high). Does anyone have any input on the best way to do this, or good parts (preferably from mouser) for doing so? I would love to be able to do it with a single rotary switch but I'm not sure how that would be possible since the power would be SPST and the gain DPDT unless I'm mistaken.
[/size]

 
Jun 21, 2010 at 2:16 PM Post #5,308 of 7,277


Quote:
 
[size=medium]
Has anyone here done a P2P build with dsavitsk's additions? If so, was it worth it?
 
Also, I want to put a gain switch on this build that selects between two sets of input resistors, 50k and 100k. And as I see it, there are two ways of doing this, with either a power switch and a DPDT switch, a power switch and a rotary switch, or just a (very special) rotary switch (off, low, high). Does anyone have any input on the best way to do this, or good parts (preferably from mouser) for doing so? I would love to be able to do it with a single rotary switch but I'm not sure how that would be possible since the power would be SPST and the gain DPDT unless I'm mistaken.
[/size]


I've done some thinking re: implementing a gain switch and I came up with a few ideas. The easy way to do it would be to simply wire two wires, connected to resistors, straight from the input jack, both connected to a pole of a DPDT switch, which would pick between the resistors, i.e. a gain selector switch.
 
I've toyed with the fancier idea of using a linear sweep pot as the gain selector, but I don't know how efficient this would be seeing as how it'll be difficult to replicate a certain gain setting unless you're using max or min gain every time. Still, might be worth looking into...
 
if you want to get even fancier, you can use an Alps RK097 as that gain pot, because it has a built-in SPST power switch. This would save you the cost of a power toggle, but it might make a mess of your wiring depending on how you have your case laid out.
 
In any case I can't think of a decent way to wire up more than two gain selections, as the more gain possibilities you have, you need to add at least two more wires to the switch and inputs, which will get messy quickly. Hopefully someone else has a better idea regarding gain switches...
 
Jun 22, 2010 at 11:57 AM Post #5,309 of 7,277
I found an NKK DPDT rotary at mouser.
 
http://www.mouser.com/search/ProductDetail.aspx?R=MRT23-A-ROvirtualkey63300000virtualkey633-MRT23-A-RO
 
As far as I can tell though it would only be suitable for selecting gain, not my ultimate rotary switch dream.
 
Can anyone point me to a rotary switch that can do this? (pardon the poor illustration)
 

 
EDIT: Would something like this work? http://www.mouser.com/ProductDetail/NKK-Switches/MRF206-RO/?qs=sGAEpiMZZMvNbjZ2WlReYmt/ckRDMbNxDxWhl99ESfc%3d
 
Jun 22, 2010 at 12:02 PM Post #5,310 of 7,277
Hi,
 
I've got two small questions about the build of my MSSH!
 
First, ArtemF told it was possible to differenciate the pins of th power jack using a multimeter to solder the wires in the right way, how can I do that?
I also wonder how to put the tube sockets. I'm not using the PCB, and I bought a Hammond Case, but I don't know how to put it in the case, so that it won't move, I tried to find some pics of builds, but I couldn't find any help, could someone give me a solution?
 
Thanks!
 
Antoine
 

Users who are viewing this thread

Back
Top