- 189 Posts. Joined 2/2010
- Location: Den Haag, Netherlands
- Select All Posts By This User
Millett "Starving Student" hybrid amp - Page 388
Actually, the more I think of it, the fact that the hum goes away when you touch the metal plate suggests your issue is not so much with grounding but with shielding, so I'm pretty sure that a conductive enclosure will help.
No need to get a fancy Hammond extruded aluminum case, you can either:
a) use glue and alumminun foil to cover the interior of the radio shack project box, then ground it (with a nut and bolt, since you can't solder to alumminum). I did this for my first build of RJM audio's "Very Simple Phono Stage". It definitely worked in removing some buzzing the circuit was picking up.
b) get a cheap metal box for other product, like Mr Pete Millet's build, done on a wallet box. I've been considering using the tin boxes from some dominoes games I've seen sold at Wal-Mart and Office Depot for around $10.00 USD
Excellent. The tin foil worked. I cut tin foil and overlapped/taped the edges together on all sides of the box, then grounded it with a nut and bolt to the ground bus you said. To not make a mess with glue, I stuck the foil to the sides with glue dots. In case you're not familiar, they're little sticky blob things of glue.
They don't have to dry like regular glue (and don't make a mess). You just stick them onto the side, and stick the tin foil to them. They're popular for sticking things in scrapbooks.
That was the suggestion I was looking for. Thanks equalizer. No more buzz/hum/anything. I hope someone reading this thread can learn from this.
Great! I'm glad you solved it! Enjoy your amp!
- 141 Posts. Joined 1/2010
- Location: Houston, Texas
- Select All Posts By This User
I have read for about an hour now but really cannot invest any more time until I get a whole night to read through everything.
I would like to make an order through mouser and bezzar soon. I was planning on using the BOM from diyforums.com that was poster here a while back.
Something that I was wanting input in was:
- seems there are multiple tube options and it is not simply changing out tubes for different ones (i guess the term tube rolling would apply)
- with all the different tubes i assume different tones and sounds go with them?
- the decision on which tubes to be used seems to change the BOM other than the tubes themselves. seems other parts must be changed in order to handle tubes other than the 19J6. would like to look at different companies and options. would prefer telefunkens if possible.
- if i choose based on the different sound qualities how do i know what all needs to change for the BOM in order to build accordingly?
- not sure if it matters but I listen to rock, harder music, folk, indie, slow stuff, etc so a big variety
Sorry for the questions that are most likely redundant. I just would like to order soon so I can work with a friend in a few weeks on it. The only thing hanging me up from ordering everything is the tubes and what changes to the BOM the tube decision makes...
EDIT: talked with my dad a little kinda busy at work on my end and trying to hurry up and type this out but told him i read about the 12au7 and 19j6 and he said he would go with the 12au7 since tons of companies made them and it was a standard tube for stereo amps. he has two mc225s so he has spent his fair time with tubes. so now to update the BOM to use a 12au7 and go about ordering some parts!
Thanks for any advice and help,
I am surprised, looks like 12AU7's have a large range in cost..
Which is a good thing.
Really, IIRC, there is just one resistor change between the 12AU7 and the 17EW8, so you could do either/box
They use the same sockets (again, IIRC).
One thing to keep in mind is that the power capacitors in the BOM were chosen because of space issues.
If don't have the same space issues, you can get panasonic FM caps that are longer, etc. for cheaper (I think)
Yeah it seems like you only need to change the resistor(s) and have the appropriate sockets. I had to go further back in my reading then i did initially, i had started reading around page 300. As far as space issues I plan to make my own box out of wood or metal. my dad has a shop and we can fab something real nice. Not sure if wood is ok to use or metal is the standard....
I am going to order all my stuff tonight. I guess I just need the appropriate sockets and resistors.
Thanks for the advice. Any one else is welcome to chime in.
Also as previously stated any differentiation between 12au7, 12ax7, 19j6 tubes would be appreciated. Does it really matter? Do they all sound different / have certain traits? Have some questions about the tubes themselves and pins, plates, etc but my dad knows all that so I will have to get him to explain it all so I know what I am buying. He has like 3 testers as well so should be good on that end.
EDIT: not sure if i will get a reply to the above but I would like to add....if I am going to buy a few extras what are the things I should for sure order a few backups of?
Edited by eertelppa - 4/4/11 at 9:36am
- 1,758 Posts. Joined 10/2008
- Select All Posts By This User
So, this means that the MOSFET's are dropping 29*150 = 4.350 watts each?
Is there a real benefit to doing this?
What would happen if you powered the amp using a 26V source?
Or was this done just because a 48V power supply is super cheap?
Not exactly. You have to realize that the MOSFET and the heater are in series, from the 48 volt supply line to ground. So whatever voltage the heater 'uses' the rest is taken by the MOSFET; they being in series, the same current flows through them. So it's 150ma @ 48 volts per channel, as Beftus pointed out.
Yup, that's why the heatsinks get so hot. If you take a lower voltage PSU the MOSFETs will dissipate less heat and you will need to turn up the central heating in your home. The Cisco PSU's are cheap (mine were), hence the name: Starving Student.
True. Silly me, I forgot that. Less plate voltage in essence means a starved plate which leads to the tube operating in a non-linear area. This will give interesting distortion if you're into that. An area usually only interesting to rock guitar players.
Yeah, I'd say the choice of PS was a combination of those two factors: reasonably high plate voltage for the tubes and drain voltage for the MOSFETs as well as a compact, plentiful and cheap supply brick.
Well, If I remember correctly, from a simulation in TINA-TI, using a 1 KHz sine wave as input, the amp was putting out 200 mili Watts into a 62 ohm resistor (the impedance of the K701) with .5% harmonic distortion. It's a simulation but it gives you a reference. Does it seem low compared to the theoretical dissipation? That's class A operation for you !
You can get some reference measurements in real life: get a 1 KHz tone mp3 file, and use your DVM to measure AC volts at the headphone jack. Then, knowing the impedance of your heaphones (which is usually rated @ 1 KHz) you can calculate the power output by the amp; maybe at the volume knob position you usually listen to.
EDIT: Take a look at http://www.pmillett.com/starving.htm Almost at the bottom of the page Mr. Millet states that the 5% Total Harmonic Distortion point is 7 volts RMS @ 100 ohms and 3 volts RMS @ 32 ohms. That means 280 mW and 490 mW respectively.
Edited by the_equalizer - 4/4/11 at 6:45pm
sorry if this has been answered, but i looked through a fair bit of this thread before giving up and deciding just to ask this.
Did anyone ever come up with a simple way to install a "safety" to stop potential damage to headphones during start up / turn off?
I only came across one (besides using AMB's epsilon12) - it was briefly detailed by ArtemF in one of his first posts.
I sent him a PM clarifying how he has it set up, he responded with this:
the idea of output diode protection is the following: when the caps are charging (through
the headphones) we'd like to bypass the headphones. On the other hand, we don't want
to loose the signal when the amplifier is working.
Correspondingly, if we put a diode (or two in series) parallel to the headphone then
the current will flow through the diode when the voltage is higher than the opening voltage
of the diode (0.7-1.2V). When the caps are charged and the amplitude of the output signal
is lower than the diode's opening voltage (that almost never happens) the diode will be closed and
the headphone will get the full signal.
Correspondingly, I took 8 diodes = 2 channels x 2 diodes in series x forward/reverse directions and
put the set of 4 (two in series forward, two in series reverse) in parallel to the output jack,
I mean, parallel to headphones. As a result, I do not overload the headphones by turning the amp
on and off. 2.4V opening voltage limit is never reached when the music is playing.
I guess, any diodes will be fast enough for this application. I used blue LEDs just for visual control.
Hope this will help,
I considered doing this protection with zener diodes, but I decided I wasn't strong enough electronically to stray from what Artem did, so I ended up using some LEDs.
When the caps are charged the upper points of R6 and R12 are at zero potential with respect to ground.
It's fine to use the diodes you named. One in forward, one in reverse for each channel.
The scheme for one channel is like that:
----| |--------*-------------------*------------*---------> output to headphones
| | |
----- | |
| | ------- -------
| | R6 \ / /\
| | \ / / \ diodes
----- \/ / \
| ----- -------
| | |
-------------- *------------------*-----------*----- ground
I can't say how good of a job they do, and I still get some pop when turning on (and especially turning off) the amp. I'm not sure if it's damaging to my headphones but I've done it a few times early on and my headphones were okay. I unplug my headphones anyway now that the ones I'm using are more expensive.
awesome, thanks for the fast reply, i think i might just copy what ArtemF did, im not that great at DIY so i will play it safe.
I was mainly looking for a fail safe just in case i forget to unplug before i turn on / off (i can be a real air head at times :P ) rather than risk blowing up my headphones. this fits the bill pretty well.
I just put a double pole switch on the outputs. So when I want to use it, I turn the main power switch on, wait a little while, then turn on the headphone switch (a sort of reverse mute switch). To turn it off, flip the mute, flip the power. Works every time, no pop.