Millett "Starving Student" hybrid amp
Oct 25, 2010 at 5:17 PM Post #5,551 of 7,277


Quote:
Hi there...
Oh well, just finished soldering all the parts together, what a mess!!
Thought using a Star Ground like this  http://mark.rehorst.com/LM3886_amp/amps_chips_and_star_ground.jpg
would be a good idea.. It wasn't. Loads of extra wire...
frown.gif

 
Would it be a bad idea if I used a solid copper plate (1,5mm thick, found that @ work) , connected to the Star Ground
and this plate to the Case or would a strip of it fit ?
 
btw, I went a little to generous with the wire lenght which resulted in some nice noodle salad, guess that's ok? 
But i got spare parts...if anything's not working (and i bet there is) I have plenty of lenght to cut out and resolder
smily_headphones1.gif

 
Cheers


the more area for ground, the better. you can attach your star ground to a copper plate, then to the case, but try not to make too many paths - things can go bad if ground takes unpredictable paths. just connect at one point, and the plate to the case at one point.
 
I did the same thing as you and used lots of extra wire; it may cause some extra interference but it shouldn't be a really big problem.
 
Nov 2, 2010 at 5:44 AM Post #5,552 of 7,277
I haven't been on this thread in a long while, but I've finally got time to put a Starving Student together, but I've become stuck with a few things in planning. Firstly, I'm building a point-to-point 19J6-based Starving Student. I brought the tubes when I was suppose to build it last year, but hence, as a Year 12 student, I was stuck down with school. My current BOM is here (Excel format). Here are the things I'm having problems with:
 
1. R16/R17 on the BOM: You can choose between a 50K resistor, and a 100K "low-gain" resistor. If I was to use a 50K pot, which would be the best one to choose when using a pair of Grado SR-60's and Ultimate Ears Triple.Fi 10's (both rated at 32ohms impedance)?
2. Regarding the pot, heatsink and heatsink mounting kit, I have had trouble finding them. Does Farnell/Element14 stock them, and if so, could someone please supply a catalog code?
3. Generally, can you use IEM's with the Starving Student? I remember reading early in the thread that someone was advising against using them.
 
I'm pretty sure the BOM parts are right, but I am still looking for a few specific parts, such as the pot, heatsink and the TO-220 mounting kit which I haven't been able to find. Is there anything else I need to know before I start ordering the parts?
 
Nov 3, 2010 at 1:22 PM Post #5,553 of 7,277
#1 : I thought you have to match your pot resistance with R16/R17 resistance. so if you were to use 50k ohm pot, you have to use 50k ohm resistor for R16/R17. As for your headphone impedance, i am not really sure about this but low impedance phone needs high gain. I would say 10k ohm pot and 10k R16/R17 would be good. 
 
#3 : from what i have gathered from this thread, starving student will have some "popping" sound during power up and shut down ( with headphone plug in). headphone can withstand such "popping" but it is not advisable so i would say IEM is not really suitable because IEM might not be able to take the "popping"
 
i am thinking of build this amp as well and i am sourcing for parts as well. Need some advice on these parts,
 
10k stereo pot will this work? 
 
is there any difference between metal film resistor and carbon film resistor? can i substitute the resistor with a similar ohm but rated at 1W? 
 
i am using a D2000 headphone (25 ohm) so should i substitute the nichicon muse es to a 470uf? it helps with bass response right? or is it the other way round. 
 
Thanks. 
 
Nov 3, 2010 at 6:18 PM Post #5,554 of 7,277
 
Quote:
 
#3 : from what i have gathered from this thread, starving student will have some "popping" sound during power up and shut down ( with headphone plug in). headphone can withstand such "popping" but it is not advisable so i would say IEM is not really suitable because IEM might not be able to take the "popping"
 
i am thinking of build this amp as well and i am sourcing for parts as well. Need some advice on these parts,
 
10k stereo pot will this work? 
 
is there any difference between metal film resistor and carbon film resistor? can i substitute the resistor with a similar ohm but rated at 1W? 
 
i am using a D2000 headphone (25 ohm) so should i substitute the nichicon muse es to a 470uf? it helps with bass response right? or is it the other way round. 
 
Thanks. 



Plug in after powering up, unplug before powering down.
 
That pot will work but you will need two (one for left channel, one for right). You can also buy a stereo pot.
 
Carbon film or metal film will both do. Some people prefer carbon, some metal film. YMMV. Using 1W resistors is not a problem. Output caps of 470uF will do fine, large output caps help with bass response. Build it per spec first, then start tinkering with it or build a second one differently. Enjoy.
 
Nov 4, 2010 at 8:46 AM Post #5,555 of 7,277


Hello,
 
I have a question about the MSSH schematic, especially where the red circle is. In my build, I have all the four wires directly connected to the pin of the switch, one to each mofset, one to C1, and one to R13. Is that right? The  schematic makes a differences between the wires, I mean, there are two distinct points, and I was wondering if the right wiring was switch -> wire divided in two, one to C1, the other one to both R13 and the mofsets.
 
I hope you'll get me, it's hard to find the words!
 
Thanks you very much for your help!!!
 
Antoine
 
Nov 4, 2010 at 9:35 AM Post #5,556 of 7,277
Using separate connecting points in the schematic like that is good drawing
practice and done to prevent confusion but does not mean those points
need to be connected separately in reality. What you did was fine.
 
Quote:


Hello,
 
I have a question about the MSSH schematic, especially where the red circle is. In my build, I have all the four wires directly connected to the pin of the switch, one to each mofset, one to C1, and one to R13. Is that right? The  schematic makes a differences between the wires, I mean, there are two distinct points, and I was wondering if the right wiring was switch -> wire divided in two, one to C1, the other one to both R13 and the mofsets.
 
I hope you'll get me, it's hard to find the words!
 
Thanks you very much for your help!!!
 
Antoine



 
Nov 6, 2010 at 12:43 AM Post #5,559 of 7,277


Quote:
#1 : I thought you have to match your pot resistance with R16/R17 resistance. so if you were to use 50k ohm pot, you have to use 50k ohm resistor for R16/R17. As for your headphone impedance, i am not really sure about this but low impedance phone needs high gain. I would say 10k ohm pot and 10k R16/R17 would be good. 
 
#3 : from what i have gathered from this thread, starving student will have some "popping" sound during power up and shut down ( with headphone plug in). headphone can withstand such "popping" but it is not advisable so i would say IEM is not really suitable because IEM might not be able to take the "popping"
 
i am thinking of build this amp as well and i am sourcing for parts as well. Need some advice on these parts,
 
10k stereo pot will this work? 
 
is there any difference between metal film resistor and carbon film resistor? can i substitute the resistor with a similar ohm but rated at 1W? 
 
i am using a D2000 headphone (25 ohm) so should i substitute the nichicon muse es to a 470uf? it helps with bass response right? or is it the other way round. 
 
Thanks. 


 
I don't see why it would be necessary to match them. Can anyone enlighten me on this?
 
Nov 6, 2010 at 11:52 AM Post #5,560 of 7,277
the BOM which is created from Dsavitsk stated that the ohm for the stereo pot should be adjusted according to R16/R17. 
 
i am not really sure why as well but maybe this gives a smoother volume control ? 
 
Nov 6, 2010 at 5:03 PM Post #5,561 of 7,277
There is no need to match the input resistors to the same value as the pot.  What you're trying to do with the input resistors is "push" the active band of the pot into something with a reasonable travel over the volume range.  The gain of the tubes is much too high for most headphones, so that if you build the amp without those input resistors, your total volume range - between not hearing and having your eardrums blown out - is about 1/4 turn or less.
 
A great many of us used input resistors that are 2X the value of the pot impedance.  It just depends on the sensitivity of your headphones.  There is no hard and fast rule - either 1X the value of the pot impedance, 2X, or there were some builders who went all the way to 3X.  (Note that there were some who used -0- input resistors, because of hard-to-drive vintage orthos.)
 
Pot impedance, as a general rule, has nothing to do with the gain.  Generally speaking, 100K introduces more noise, while 10K may mean the source will have trouble driving the amp.  It also presents a much lower input impedance to the source, which may necessitate changing output capacitors if the source has any (to prevent a loss of bass).  That's generally why we settle on 50K most often around here - but 10K or 100K will still work OK.
 
About some other comments from previous posts:
  1. Low impedance phones do not necessarily need high gain.  Gain is function of voltage and with low impedance phones, they're more interested in current than voltage for a given power output.  So, it's quite possible that high gain with low impedance is the worst combination.  There are always exceptions, but as a general rule, you want gain low if your load is low impedance.
 
  1. Popping is your headphone being exposed to DC voltage spikes.  Some headphones can tolerate it, some can't.  Why take the chance?  In the case of the Starving Student, learn to plug your headphones in and out only when the amp is on and the tubes are warmed up.
 
  1. The headphone's impedance combines with the output capacitors on the Starving Student to form a bass cutoff filter.  If you use Denon's at 25 ohms, you'll probably need 680uf or 1000uf at the output or you'll notice a loss of bass.
 
Nov 6, 2010 at 5:15 PM Post #5,562 of 7,277


Quote:
There is no need to match the input resistors to the same value as the pot.  What you're trying to do with the input resistors is "push" the active band of the pot into something with a reasonable travel over the volume range.  The gain of the tubes is much too high for most headphones, so that if you build the amp without those input resistors, your total volume range - between not hearing and having your eardrums blown out - is about 1/4 turn or less.
 
A great many of us used input resistors that are 2X the value of the pot impedance.  It just depends on the sensitivity of your headphones.  There is no hard and fast rule - either 1X the value of the pot impedance, 2X, or there were some builders who went all the way to 3X.  (Note that there were some who used -0- input resistors, because of hard-to-drive vintage orthos.)
 
Pot impedance, as a general rule, has nothing to do with the gain.  Generally speaking, 100K introduces more noise, while 10K may mean the source will have trouble driving the amp.  It also presents a much lower input impedance to the source, which may necessitate changing output capacitors if the source has any (to prevent a loss of bass).  That's generally why we settle on 50K most often around here - but 10K or 100K will still work OK.
 
About some other comments from previous posts:
  1. Low impedance phones do not necessarily need high gain.  Gain is function of voltage and with low impedance phones, they're more interested in current than voltage for a given power output.  So, it's quite possible that high gain with low impedance is the worst combination.  There are always exceptions, but as a general rule, you want gain low if your load is low impedance.
 
  1. Popping is your headphone being exposed to DC voltage spikes.  Some headphones can tolerate it, some can't.  Why take the chance?  In the case of the Starving Student, learn to plug your headphones in and out only when the amp is on and the tubes are warmed up.
 
  1. The headphone's impedance combines with the output capacitors on the Starving Student to form a bass cutoff filter.  If you use Denon's at 25 ohms, you'll probably need 680uf or 1000uf at the output or you'll notice a loss of bass.


Thanks. I was confused since I used a 50k Blue Velvet and 100k/50k input resistors with a gain switch. It would make it awfully difficult to adjust gain if we had to change the pot as well!
 
Nov 7, 2010 at 3:55 PM Post #5,564 of 7,277

 
Quote:
  1. Popping is your headphone being exposed to DC voltage spikes.  Some headphones can tolerate it, some can't.  Why take the chance?  In the case of the Starving Student, learn to plug your headphones in and out only when the amp is on and the tubes are warmed up.

There is a simple way to get rid of this problem: diode protection. The diodes should be switched in parallel to earphones and each channel
will need either 2 of them (to kill both forward and reverse spikes) or 4 of them (to double the threshold above which the spikes/sounds will be cut).
Assuming that the opening voltage of a regular diode is 0.7-1.5V and that the earphones can tolerate a short-term 3V overload, I would recommend
using the latter variant (I applied it in my amp since I'm 100% sure that there will be one day when I will forget to unplug the earphones
smily_headphones1.gif
).
It softens the spikes dramatically and it doesn't distort the "normal" sound at all.

 
 
Nov 8, 2010 at 3:39 PM Post #5,565 of 7,277
Oh oh....
thanks to a pic Aflac has on his picasa acc. I had it working yesterday , had something wired completely wrong before ...after rewiring , both tubes lit up, mosfets got hot, tested as the_equalizer posted @ page 259, seemed fine...
Today I was double checking the voltages before casing it up and managed to short something ... There's 2 possible things (or both together) : for cleaning up the wire mess I cut one of 'em , leading to a mosfet pin 2.
I just drilled it by hand yesterday ..Was just about to start measuring @ the tubes , the probe slipped and the tube I was about to check lit up like a christmas tree ...
Power off ... took a look at the mess and see the hand drilled part i was talking about connecting to the other mosfets heatsink ... the heatsinks are insulated by washer/pad though (at least i thought so, if not would that be a possibility ?) I have no idea on what to check now .. might be the mosfet thats fried ? I have some spare tubes but wanted to ask first .. 
greets,
Marcus
 

Users who are viewing this thread

Back
Top