New Millett Hybrid Maxed Amp - Page 105

I am using the je243g's, Afer removing the left channel qb8 It read 840, So I thought that was good. So I re-installed it and tried to bias the left. It also has problems. Is it possible that I put the trimmers to far clockwise. Can they create a dead short? All 5088-5087 are in their right positions. I think on Monday I will order new qb8's Rb14/10's and new trim pots. Exchange those and see where that gets me. Should I order anything else?

Thanks again guys.

oh, btw, the Murata trimmers i have don't click. or don't click so much that i can hear or feel it. So, stick with bourns or whoever. I find it pretty annoying to have no clue save for counting turns whether I've reached the end or not.

The nice thing about the Muratas is that you can't over-turn them and tear them up. So, in effect - you don't have to count. They're less than half the price of the Bourns, too.

MJE243/253 HFE = 180 max

You didn't mention checking the resistors - I'd give them a pretty good look, too. A wrong value here and there can really screw up the currents in the DB's. Also the PN4392 - it's a very critical part. IMHO, bad Murata trimmers are pretty unlikely. It's easy enough to check them with your DMM.

Listen2this1, don't forget that a cold solder joint can increase resistance significantly wherever it occurs. Depending on where that is, it could drive a transistor past its maximum rating. I usually give the most critical areas(what isn't) one more quick application of heat from my soldering iron. ;-)

It lives.

Correcting those three mistakes cleared up my issues. Thanks for all the help!

Now i just have to enclose it. i'd been thinking of doing something creative and original, but now I'm thinking that that's overrated. 'cause i have no good ideas.

Way to go, Eric!

So, how well does it drive those K340's in your pic there? I've heard those things are infamous in their demand on an amp.

Pretty well. 12fm6 didn't have enough gain, so i installed some RCA 12ae6a - which is enough.

Plenty of bass, as others have mentioned. But crystal clear treble, too.

Output is Nichicon KZ 470uf 25v. Even with v+ at 29v and 12ae6a tubes, i actually can't set the tube bias higher than just under 26v, so i have no concerns about their lifespan. Admittedly things would be different w/o tubes or with something wrong in the circuit.

btw, 20-ohm R1 at 29v is running at about 125f. As you can see from my earlier pics, I went up one watt in size, too. It's the hottest thing on the board.

Probably won't be an issue, but if someone was concerned they might be able to get a to220 resistor and heatsink it. Probably have to air-wire the resistor and drill extra holes in the board to mount it.

CA2 are nichicon upw 1000uf, CA9 are CDE polycarbonate 0.68uf. CA3 and CA6 are Nitsuko stacked metalized polyester 4.7uf. They're big and fast - which is good 'cause they're power caps, right? I wouldn't use them in the signal path, but they seem to be doing well on the power rails. CA8 are ERO boxed mylar 0.22uf and nothing special. I might replace these if at some point for some reason I'm dissatisfied with the sound, but by now you know i'm not really entirely a capacitor guy.

RB14 are big honkin ferrites. I have no specs for them - got something like 200 of them, leaded, on tape, on ebay a few years ago.

CA4 and CA5 are relatively cheap Nichicon RS 1000uf 35v. They're good but nothing special. I figure the giant film caps paralleling them make up for any shortcomings they have.

Oh, and the to220 diodes in the power supply are Unitrode SES-5404. They're old-ish. Precursor to the MUR420 iirc. Ultrafast diodes of the early 90's. The voltage drop across them is only about .4v each. Paralleled with unknown-brand .1uf boxed mylar.

I'm driving them with a modified Skynet power brick. Used to be rated 30vdc 1A. Cracked open and pulled out the rectifier board. There's enough room in there that i'll piece together a STEPS-alike AC line filter w/ common mode choke and the works.

With everything biased up and music playing i have almost 33v after rectification, so I'm dropping 4 volts and the regulator doesn't get particularly hot.

Excellent ericj! Another MAX Lives!!!!

Your determination showed on this build. Kudos to the Millett Max Teamwork!

Now, on to the not so fun ground plane, (top mask) for the experimental MAX mask. The registration marks have to line up perfectly to get that right. I may even have to alter the "final" mask if things get rough. One "cheat" is to leave ground plane around the pad areas slightly wide, which will still keep the board quiet. Remember this is the "experimental" MAX board, not the "uber" MAX that many of us have built.

A quick trimmer question

I have seen the other advice regarding starting with the RB12L/R trimmers set at the minimum (full clockwise) setting to start with. I have not been able to find comments on reasonable starting points for the other trimmers.

Is it reasonable to set RR3 and RA1L/R to the half way point of their adjustment as a starting (installation) setting?

it doesn't really matter. The tubes don't care, and RR3 is only a fine-adjust: RR1 and RR2 actually set the range. If you look at Tangent's STEPS schematic, when RR1 and RR2 are 120 and 2K, respectively, then the voltage range is approximately 22 - 27VDC. None of the voltages in that range are going to hurt anything in the short term.

So - not to worry.

EDIT: Tangent's STEPS is slightly different - it uses a 500 ohm trimmer, but the principle is the same: an interval of voltage is set by the resistors, depending on the incoming voltage, and the trimmer is only a fine adjust within that range.

Here are some neater pics of my MH Max :





and more here:

http://i72.photobucket.com/albums/i1...H/P1010063.jpg
http://i72.photobucket.com/albums/i1...H/P1010064.jpg
http://i72.photobucket.com/albums/i1...H/P1010067.jpg
http://i72.photobucket.com/albums/i1...H/P1010068.jpg
http://i72.photobucket.com/albums/i1...H/P1010070.jpg

The BJT version still needs to be cased up and it will be totally differerent .

Very, very nice, Ferarri! I'll post these on the website when I get a chance.

Speaking of the website, ruZZ.il suggested that I put a warning on the Setup and Bias page stating to bias the DB's first. Seemed like a good idea based on what we've been seeing and the recent question from fordgtlover. That change has been added.

as stated before, the trimmers themselves may be unaffected. You did mention both RB10 and RB11 getting excessively hot, so there is a possibility that QB9 is damaged too. You could go about removing the BJTs that you already have, checking what's functional, trying to use the good ones, etc.. if you find a good pair you'll be able to make sure the rest of the circit is good, but all the inserting/resoldering could pay its toll on the board, so you may not be too bad off just ordering spare QB9s too, specially for their price. I suggest getting a pair of nicer transistors, but you may like to make sure everything else is fully functional before putting nicer ones in. Once you're interested in getting some, mention it around here, or better yet, at the DIYForums. You may find someone to share orders with. If you do manage to find a good pair in the meantime and can make sure everything is good, maybe just go straight for a good bunch. since you'll be removing quiet a few transistors, may I suggest looking at this tip. It was also noted to me that QB1 is in a sensitive position, or at least in one that could be effected by excessively high currents in the DBs, since most of it would go through it.

Rest assured, you'll get it all solved soon enough. I was recently in exactly your situation but the support here keeps you going (thanks guys) Anyway, you're bound to learn more this way

My Millet w/ the Triad 1A 24VDC wallwart won't get past 26.7V. This is with MOSFETs biased at around 230 mV each. Is this normal?