Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › New Millett Hybrid MiniMAX (what happened to this thread?)
New Posts  All Forums:Forum Nav:

New Millett Hybrid MiniMAX (what happened to this thread?) - Page 8

post #106 of 1944
Finished my MiniMAX (click for bigger pics):









For you prototypers out there, a jumper at RR4 results in about 0.215mVAC ripple. So, a jumper is not the optimum component to put in there. Right now, I have the old Millett Hybrid Polyfuse (RXEF050) in there - it has an R of about 1.5ohms - and the ripple went down to 0.038mVAC! That is under a load, given the tube heaters and buffer bias at somewhere between 250 -300ma. Under music the ripple varies some, but not enough to be detrimental. In fact, it sounds much, much better than the jumper.

I'll be doing additional testing with different resistances the rest of the weekend and try to find an optimum value/device to put in RR4, but for now, I'm convinced we have a winner in the design. I've got some ideas on some minor changes for cooling that I'll be discussing with Colin, but most likely, the first run of production boards will be ordered Monday.

Heat is a bit of an issue with the middle sinks and the PS sink, but not outside of what's acceptable. At 27VDC and 50 ma bias in the buffers, I'm measuring the following when equilibrium is reached after about an hour or so:
Outside sinks - 33 deg.C. (91.4 deg.F.)
Center sinks - 43 deg.C. (109.4 deg.F.)
PS sink - 45 deg.C. (113 deg.F.)
The center sinks feel quite hot to the touch, but as you can see, this is not much more than the metal on a car during a hot sunny day.
post #107 of 1944
Looks gorgeous! I really like the Vit-Q orientation!
post #108 of 1944
Awesome. Any news from Lansing?
post #109 of 1944
Quote:
Originally Posted by FallenAngel View Post
Awesome. Any news from Lansing?
They're just waiting for me to give them the word ... and the money. Hopefully, I can organize the group buys this weekend. I have a couple more posts to do to help everyone visualize the case, but here's a teaser:


post #110 of 1944
Don't do that to me at work Tom! I'm drooling over the notepad in front of me!

Yay! Group-buy, group-buy!

Hell, I'll build another amp just so I can have 2 of those cases.
post #111 of 1944
OMGOMGOMGOMGOMG SEXY!!!! I was planning on building two.. those look so good I might have to do three!
post #112 of 1944
Wow!! I already know what I'm doing with mine: Super bright LED in the center of the case so that all of those grill openings glow! That will look so hot!

EDIT - Would there be any issues with running several LEDs in parallel from LEDC?
post #113 of 1944
I don't think I will even care how they sound...... that case is a work of art
post #114 of 1944
Thread Starter 
Will it also be available in Black?
post #115 of 1944
Quote:
Originally Posted by m0b1liz3 View Post
Will it also be available in Black?
Yes. It's just that AutoCAD (translated = operator error) is having trouble rendering in black and mapping white text to the endplates.

Lansing doesn't care about that, of course.
post #116 of 1944
Quote:
Originally Posted by patton713MW View Post
Wow!! I already know what I'm doing with mine: Super bright LED in the center of the case so that all of those grill openings glow! That will look so hot!

EDIT - Would there be any issues with running several LEDs in parallel from LEDC?
I would test it first with a breadboard - or a bunch of alligator clips and wire. There's a variance in some of these LEDs and they may not all glow equally (at least when I've tried it). Plus, you could quickly end up with a huge power resistor to handle all the current - 4 or 5 LEDs at 10ma a piece and you'd have some real power usage. That's the same as the bias on the transistors - and they have 1" extruded heat sinks.
post #117 of 1944
Black -




Also, this is something SPECIAL, if your Adobe Acrobat Reader version is up-to-date, try clicking on the image in the following files. You should be able to spin, twist, zoom in and out - all in real time:



(These are linked to PDF's - either click to load Adobe Acrobat in your browser, or right-click and download the PDF to call up in Adobe Acrobat.)

Watch for the Group Buys shortly ...
post #118 of 1944
Tom = Awesome.
post #119 of 1944
Ooooooh, 3D-tastic!
post #120 of 1944
Based on the prototype builds, here's the updated layout and schematic (click for larger files):




Here are some of the issues found out by the prototypers and how they've been addressed:

1. Axial Schottky's too difficult to fit into the rectifier positions.
I believe the holes have been made a bit bigger, but a very tight bend allows them to fit without issue.

In addition, we are changing the ceramic discs to axial ceramics. This should get that back end squared away with a really low profile.

2. PS trimmer operates backwards from previous MAXes.
The silkscreen for the PS trimmer has been turned around.

3. LED resistors are too large for the pads.
The traces are really tight on the MiniMAX, as you prototypers can attest. There isn't any room to make these pads larger. However, we have adjusted the BOM for an inexpensive 1W metal film that will fit the pads. I have yet to look for the same alternative at DigiKey, but will followup with that when I can. Also, cetoole won't like me saying this, but kneeling 1/2W carbon films works just fine.

4. The 0.018uf VitQ's won't fit easily between the front heat sinks.
We knew this was tight from the beginning, but I thought perhaps they would fit horizontally. Unfortunately, those stupid VitaminQ nipples get in the way. We want everyone to be able to fit these in without resorting to expensive teflon tubing or the hassle of heat shrink. A creative Z-bend worked on mine to allow the cap to kneel to one side without touching any of the sinks. However, cetoole put some additional pads further out toward the sinks that should make this task easier.

I will draw up a detail and post it on the website, as was done for the regular MAX, that will show how to bend the leads to make these fit - without touching the heat sinks and remaining below 1" height.

5. The heat sink temperatures are ~10 degrees higher on the PS sink and the middle two sinks, compared to the outboard sinks.
The silkscreen pattern used in Eagle is slightly off from the actual extruded heat sink footprint. This resulted in some of the cooling holes actually being blocked.

Cetoole and I worked to increase the number of holes where possible and to move them further out toward the tips of the heat sink fins for better cooling. We don't anticipate any issues with the machined case design (Group Buys in progress), but we still want the option for a manually drilled case as with the original MAX. Either a CT1 or DT1 Lansing case will work and are economical, but using simple drilled cooling holes will not be as efficient as the machined slots on the Group Buy case. So, hopefully the extra cooling holes in the board will benefit a manually drilled case.

Of course, the original prototype was a manually drilled Lansing CT1 (has been featured on the temporary MiniMAX website page for several months), but the vinyl covering with that case option makes things pretty toasty. A manually drilled Lansing DT1 case should be completely analogous to the original MAX. I will provide drawings and drilling templates on the website for these manual options as well.

6. The LNMP tests resulted in two changes to the Power Supply in order to reduce ripple close to the target values (<0.060mVAC).
The result is a measured ripple/noise in the neighborhood of 0.038 to 0.54 mVAC, depending on components and line voltage. These two changes are:

RR4 is now PF1, a polyfuse - part #RXEF050. This polyfuse was original Millett Hybrid and revMH Millett Hybrid equipment. The resistance is anywhere from about 1.5 to 0.5 ohms, decreasing with higher current rating. This works perfectly for our purposes and results in the very low ripple noted above. In addition, the polyfuse will serve to protect the buffers - separate from the PS fuse, which actually serves to protect the walwart to a large extent.

Note that the part# is for a 1/2A rated polyfuse. This has been tested (historically, too) and works fine for BJT buffers. Those who are willing to try a MOSFET-MiniMAX may need to adjust this rating upward, however. Again, we are not really recommending a MOSFET version except for experimentation by advanced builders.

CR4 will most likely change to an electrolytic. This will reduce the ripple by about ~0.010mVAC (one of the reasons I cited a range above). An added benefit is that an electrolytic is low cost compared to a tantalum. A 100uf 35V will be the recommendation.

One final note - a DIP-16 or two DIP-8 sockets are pretty convenient for plugging in the relay. However, be sure to snip the leads that are not used - there are not pads for all 16 positions of a socket. Also, if you don't snip these very close to the plasic socket frame, you'll end up with a very tall relay installation that may interfere with using a center front-panel LED bezel.
New Posts  All Forums:Forum Nav:
  Return Home
Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › New Millett Hybrid MiniMAX (what happened to this thread?)