Millett "Starving Student" hybrid amp - Page 113

Eagle Library

I threw this together, an Eagle Library for the 19J6 Tube. This was my first real attempt at Eagle, but to the best of my knowledge it should be all correct. I used the template from the original Millet Hybrid, and made the appropriate changes. Maybe this will help someone out.

I am just working on a little project for myself, and I am going to etch a few PCBs.

Oh yeah, Thanks for the Tutorials Tangent!

Not this week but I'll call them again in the morning. If someone can PM me when some show up somewhere I would appreciate it.

It's a great piece equipment for fab work, but so are a lot of other tools. Tools...

Thanks and good luck with the build.

I just done a beasty mouser order.
AU$90 for the components to build one starving student and two cmoys.

I've received the PSUs, I have one extra so might sell that and all I need is some tubes (and mounts).

Maybe I could trade someone a PSU for a few tubes.

I've been reluctant to post on this thread after the Thermafilm debate, but I have another question.

My amp has developed a problem. The right channel's volume is no longer controlled by the pot. Could this be due to the MOSFET dieing?

It's not a bad tube as I swapped them and the problem didn't change channels.

I checked the 2 test voltages and I have 48V and 21V. The schematic says ~19V, is 21 close enough?

I checked the pot and both channels are working correctly. If it matters, I used a 100k pot with the theory that I would be able to use more of the range.

I have 2 friends that want one of these amps after hearing mine, so I'll be placing another parts order soon. Do I need to add a couple extra MOSFETs? I'm not sure how to test a MOSFET as my meter doesn't have a transistor testing function.

In regards to the earlier discussion of the need for grease, while I respect everyones opinions, it seemed the position that grease was needed was based on the description of Thermasil not needing it. So the assumption was that the others do need it. I have looked at everything I can find from Aavid. I sent them an email asking about it. We'll see what they say. If my MOSFETs are bad, then I guess I have the proof I need

Kuroguy, before you chime in, I want to say that I considered apologizing, but you never offered any support to your labeling of my advice as "bad". Then you had to follow me to another thread to try and rehash the debate there. You're so full of your own self importance, you even quoted yourself in your signature!

Michael Knowles: Extras

1) download track #55. Play in a loop.
2) DMM set to AC mode
3) Measure voltages at the amp's input (RCA jacks).
4) Put pot at minimum volume.
5) Measure voltages going into tube.
6) Measure voltages coming out of the tube.
7) Measure voltages at amp output (headphone jack).
8) Put pot at maximum volume.
9) Repeat steps 5, 6, and 7.

Into the tube will be what the pot puts out.

Out of the tube will be the signal with voltage gain applied.

Amp output will be after the MOSFET buffer.

How do they compare to the RCA input?

That will determine if the problem area is the pot or the tube or the mosfet. After that, it is to determine why there is an issue.

I built a hungry hybrid.

Its somewhere between a classic millett hybrid (same gainstage, and direct coupling between gain and power stages) with a mosfet output using the heater as its source resistor like the starving student.



I got some knobs on my way to work today now I can adjust the stepper without pliers.
probably not a bad mosfet.

It sounds like the pot is not wired correctly, I would try hard to figure out how the pot got OUT of the signal path.
Probably close enough. Tube and part inconsistencies will result in small differences.
Are you 100% sure its "in" the circuit? With the amp off: Measure resistance from the grid/input of the tube to the center of the RCA. It should change from 100k ohms at "minimum volume" to about 0 ohms at maximum volume.

You could also measure the resistance at the RCA plug as you rotated the knob too. It should stay the same as you spin.

If you twisted signal and ground wires its easy to get them confused somewhere.

That's a possibility, but check this - is the MOSFET hot? If so, then it's probably operating. Yes. Again, it could be many things.

Exactly how did you determine this? My first suspicion would be loss of a signal wire at the pot or the RCA connections - or even at the output headphone jack. Again, if you're splitting the voltages, then the heaters are working and probably the tubes, too. If the MOSFETs are hot (without burning up), then they're probably working, too. This is easy with a very cheap meter - if the MOSFET is unmounted:

Testing a MosFet, Testing

Note the following lines in this reference:
"When MOSFETS fail they often go short-circuit drain-to-gate. This can put the drain voltage back onto the gate where of course it feeds (via the gate resistors) into the drive circuitry, prossibly blowing that section. It will also get to any other paralleled MosFet gates, blowing them also."

Now, the SSMH doesn't have a driver stage except the tubes. Nevertheless, methinks you would notice a lot more bad stuff if the MOSFET had blown.

I'm not sure exactly how to test inside the circuit except the old tried and true - measure like points for voltage from one channel to the other. You should be able to narrow it down in short order if you compare with a schematic and record your readings. The one channel works, so it gives you a perfect basis for comparison.
You don't need grease with Thermasil and its use is probably detrimental - it could dissolve the Sil pad, etc. Also, grease is not necessarily an insulator, so you have to be careful that the pad is inbetween the greased surfaces. That might be tricky to do with a Thermasil pad that's more or less meant to liquify under heat, but I'm just guessing. We'll see what Aavid tells you.

I haven't assembled my mhss yet, but I did the diode test on my mosfets before hand, and I think one of them may be dead.

On one of them, the reading on my meter starts off low, at around .1v and rises slowly to about 3v, before the reading cuts off.
On the other one, it starts off at about .1v, but rises fast to 2.9v, where it lingers.

I may have overheated one, or both while putting heat-shrink over the pins.. but i'm not sure. =/
The mosfet isn't soldered onto anything, and power hasn't been applied (except from the dmm)

Last week I purchased a pair of old surplus film caps for C2, C3 . The box was labeled 0.15uF 74V . Yesterday I got to looking and the caps are labeled .15K A26 8052. Are these indeed 0.15uF ? How do I check them with my meter ? How do they compare to the WIMA 0.1uF ? Thought I better check here before I make my order to Mouser.
Thanks

hey guys,

I just signed up today. I have no electronic background or anything but I still want to try this DIY project.

Could someone link me to the final schematics and materials list please?

I was looking through the first few pages and there seemed to be a couple of tweaks/corrections so I don't want to build the wrong thing.

Thanks

I'm from Australia too, so if you guys have any recommendations on where to get the parts from that would be great.

Cool. I'll try this when I get home. Thanks

That looks cool! I've been sort of waiting for an enclosure to inspire me. That might be it
I think the pot is in the signal path. This amp was working great for awhile, other then a hum when I tried using my BantamDAC. It sounded great when I fed it from the line out of a modded iPod.

I checked the pot with my meter only at the pot. I'll check it in other places tonight.

I'll take a look at that link tonight and test the right channel mosfet.

I checked all my connections last night. I did notice that the area around one of the pins for the right heatsink had melted the solder away from the pin. I used the 4 points where the heatsink pins connect to the copper board as my ground points. So I reflowed this one and added some more solder to fill in the gap. Here is the picture from the build (the gap isn't in this picture, it's to show how I did my grounds)


I tested the pot just by measuring between ground and the other 2 points and between the 2 point on both channels. It was definitely changing from 0 to 100k ohms (I used a 100k pot as I said above)

After I posted last night, I did give the amp another listen. The left channel seems to be OK, but it's kinda hard to tell with a wonky right channel. I just used my old iBuds to test as I didn't want to damage my good headphones. I did realize that my previous description was a little inaccurate. The right channel is indeed still affected by the pot. Here the best way I can describe the situation.

With the volume all the way down, the left channel is silent, but the right channel is moderately loud. Probably a little lower volume then I might listen to when "rocking out", so kind loud. As I turn up the volume, the left channel increases, but so does the right. If I turn it up loud enough, it seems like they eventually catch up and the left channel eventually gets louder then the right. This is hard to tell because it's well beyond a comfortable listening level so I sorta pulled the ibuds halfway out of my ears.

I'll have to double check when I get home, but I also don't think the right heatsink was getting warm. Of course I don't think the left was either in the couple of minutes I had it on.

Thanks for all the help everyone. I love this community
I'll try the suggestions when I get home. (working a little late tonight)

-----------------------------------------------------------------------
I got a response from Aavid, so hopefully we can put this subject to rest once and for all

First, here is what I sent them:
From: deleted this part
Sent: Wednesday, December 03, 2008 2:42 AM
To: info
Subject: product question

When using part # 4880G, is a thermal grease needed?

In your description of Thermasil products, you specify as one of its
features not needing any. Does the imply that the Thermafilm, such as
the 4880G, products DO need it?

Here is their reply:
RE: product question
Wednesday, December 3, 2008 5:41 AM
From: "Snow, Rick" <snow@aavid.com>
Add sender to Contacts
To: deleted again
Good morning,
It is not necessary to use a thermal grease, however it will increase
thermal transfer as it fills air gaps between the device and the
heatsink.

Rick Snow
Technical Support Specialist

Aavid Thermalloy, LLC
70 Commercial Street, Suite 200
Concord, NH 03301

So it appears it's a case of "can't hurt, might help", but the manufacturer doesn't require it.

Come on guys

holland, hayduke, tomb, n_maher, error401

Can anybody answer my post #1690 ?

This is NOT about heatsink grease !

I didn't have a good answer so I didn't reply, but you can have my answer that doesn't really help you anyway...

Markings on obsolete components are really going to be best guess if you can't find a datasheet. The value for capacitors is typically given in picofarads in a similar (but numeric) fashion to resistors. 104 for example is 10x10^4 pf, 0.1uF. Yours obviously doesn't match this, but if it says they're 0.15uF, it makes sense to assume that value is correct. The K is probably a tolerance marking. Who knows about the rest of it.

If you have a meter with capacitance measurement, just, uhm, attach the leads and read the value... If you don't have a capacitance measurement function, you will need to use a more convoluted technique. For example, attach a known R and charge the cap while measuring the time constant (or measure the discharge time constant, which might be easier to set up), but this won't work well with small caps because the timing accuracy gets very bad (or impossible). Or pass a known AC signal through an RC network with the cap and a known R, and measure the attenuation. There are lots of techniques, google turns up this that looks useful: Measuring Capacitance: Methods (fwd)

How do they compare? Can't comment, I'm not one of those guys that listens to a dozen different caps and takes notes. If it measures well, it sounds good, as far as I'm concerned.