[MrMajestic2]

Looking good there Tomb. Wish I had more boards/money to experiment with BJT buffers and different caps. Money is running dry from my Opus and B22 builds

and I still dont have the cases done for my Maxs.


Quote:

Originally Posted by tomb /img/forum/go_quote.gif My first foray into Black Gates on a Millett MAX: http://www.paintedpostcards.com/misc1/BG-MAX-1a-sm.jpg [size=xx-small](click for a larger pic)[/size] Panasonic FM's all around, Wimas, etc. 1" or less parts all around 1000uf 25V BG NX's in CA2 680uf 35V BG NX's in CA7 BJT's - 2SC2238/2SA968's 27VDC, for the PS, DB biasing was straightforward - started out at about 40mV and worked up to 110mV (50ma) in about an hour - everything stable. LM317 is running at about 42deg.C. on the plastic, about 50deg.C on the metal tab. The sink is very hot - almost burning hot to the hand, but more like automobile-door-in-the-sun hot. It should run like that indefinitely. DB's are considerably cooler - 37deg. C. on the plastic, about 42deg.C. on the metal tabs (the 2SC2238's have metal tabs). It sounds OK so far - it's obvious that the frequency response on the BG's is very extended - both extended highs and bass. However, things sound a bit metal and harsh for the time being. I haven't decided whether to put film caps on the output - have some Russian PIO's I'm itching to try, but will probably let it run this way for a week, at least. EDIT: The relay is sounding pretty harsh in the left channel, but I've measured both turn-on and turn-off and the offset is within previously seen values - it just sounds bad. Maybe the caps need to break-in some more. If it continues like that, I may replace the smaller electrolytic with a 10uf 25V tantalum - something Amb recommended over on Headwize in the E12 thread - and see if that helps. I may have a not-so-good relay. EDIT2: Wow - it doesn't take too long for those things to start smoothing out - sounds very good!

 

[pabbi1]

Quote:

Originally Posted by tomb /img/forum/go_quote.gif It sounds OK so far - it's obvious that the frequency response on the BG's is very extended - both extended highs and bass. However, things sound a bit metal and harsh for the time being. I haven't decided whether to put film caps on the output - have some Russian PIO's I'm itching to try, but will probably let it run this way for a week, at least. EDIT: The relay is sounding pretty harsh in the left channel, but I've measured both turn-on and turn-off and the offset is within previously seen values - it just sounds bad. Maybe the caps need to break-in some more. If it continues like that, I may replace the smaller electrolytic with a 10uf 25V tantalum - something Amb recommended over on Headwize in the E12 thread - and see if that helps. I may have a not-so-good relay. EDIT2: Wow - it doesn't take too long for those things to start smoothing out - sounds very good!

My experience is that the BG will 'bloom' at a couple of stages - about 100 hours, then somewhere around 300 hours... oh and some rebloom if it ever stays off for a week or so... all casual observation, of course. I wouldn't even listen for the first 24 hours, or so. Definitely give it a month or so to really burn in before making changes, as it really will take that long to fully settle out.

 

[tomb]

Quote:

Originally Posted by pabbi1 /img/forum/go_quote.gif My experience is that the BG will 'bloom' at a couple of stages - about 100 hours, then somewhere around 300 hours... oh and some rebloom if it ever stays off for a week or so... all casual observation, of course. I wouldn't even listen for the first 24 hours, or so. Definitely give it a month or so to really burn in before making changes, as it really will take that long to fully settle out.

Thanks, pabbi - you, too, MrMajestic2. Unfortunately, I probably won't be able to keep it that long since I'm building it for someone else. I'm going to keep it long enough to determine if it's better with some PIO caps than without, though. To tell the truth, it sounds awfully good right now - it lost a lot of the metallic harshness after about 2 hrs. Right now, it's slightly but pleasantly fuzzy. It helps that those RCA 12AE6's have about 500 hrs on them.

 

[ruZZ.il]

I liked them.. I loved them... I wondered, cause they seemed just good.. then, I loved them again. they really do need their time.

Meanwhile, I'm finalizing a digikey+mouser order tomorrow. Mostly my second max bits. I've also got a better transformer for my steps so I can finally turn up the bias on my M^3. a bunch of knobs too.. and some spares for minis, etc. anyone have any parts suggestions that could come in handy? nice knobs? got knob? I'll have to do some pcx and tube shopping too.. wheee.

 

[tomb]

Quote:

Originally Posted by ruZZ.il /img/forum/go_quote.gif I liked them.. I loved them... I wondered, cause they seemed just good.. then, I loved them again. they really do need their time. Meanwhile, I'm finalizing a digikey+mouser order tomorrow. Mostly my second max bits. I've also got a better transformer for my steps so I can finally turn up the bias on my M^3. a bunch of knobs too.. and some spares for minis, etc. anyone have any parts suggestions that could come in handy? nice knobs? got knob? I'll have to do some pcx and tube shopping too.. wheee.

Get those 1800uf 35V FM's at DigiKey (for CA4 and CA5) - very nice caps and at 1" if you're building a MAX for the 3rd slot.

You might try that 10uf 25V tantalum for the relay, too. Amb recommended it over on Headwize after Nate had similar trouble. I haven't had an issue with the relays until this one, but it may be more common than I thought. It really has an un-nerving punch on turn-on and turn-off. Measurements seem OK - only a volt for a split second or so with KSC75's on turn-off. Still, the tantalum should speed things up a lot.

We had a 20uf 25V tantalum on the 1st proto and it worked very well. Trouble was, it was $2.20 ea.(!) or thereabouts. The 10uf should put us in a lot cheaper territory. I'll try to make the change tomorrow night and see if makes a big difference.

Again, nothing we've measured so far would indicate that there's anything that could be damaging and Amb confirmed as much. I'd just rather that we had as sharp a turn-off with all of the relays, but the quality control may not be that rigid on those things. (could be the cheap Xicon's I'm using, too)

 

[joneeboi]

Question:

I know the BJTs can bias up to 50mA, but I got 1.5" heat sinks for my BJT setup as opposed to the recommended 1" heat sinks. Can I safely bias them higher in that case? Is the current limit more a matter of heat sink size or of which transistor I chose? (Running both MJEs, the middle two of which are considerably hotter than the outer pair.)

 

[ruZZ.il]

ah, thanks

I'll try that cap and I'll scope my relays behavior out sometime soon too. Got the 1800's down already

I'm going to use 1.5" sinks and have them stick out the top.. kind of like vixr's I guess, just not quiet like it probably some cheap plywood again, or something better if I find it.

So anyone have a good recommendation on rca jacks from pcx, not on the looney expensive side?

hmm.. I should get some cable too.. need to make new i.connects.

hey adam how'd all your new ones turn out? looks like you're wired. maybe you can send me some ~?

what wire is that? pics if you're done!


edit: jon, you could bias a bit more, just make sure they don't get much hotter. remember that they have a negative temperature coefficient, ie.. as they heat up, they conduct better, current rises, more heat, etc.., unlike the MOSFETS that as they heat up, the current drops. So thermal runaway is more likely with the BJTs (unikely with mosfets?). something to watch out for. The question also is, how much more would be beneficial and how beneficial, if it all. I'd take it slow and stay away from 'too hot'. I'd also rather get better transistors, than to try push those (there are substantially better, imho). but hey, if you burn them.. you could get better ones

 

[tomb]

Quote:

Originally Posted by joneeboi /img/forum/go_quote.gif Question: I know the BJTs can bias up to 50mA, but I got 1.5" heat sinks for my BJT setup as opposed to the recommended 1" heat sinks. Can I safely bias them higher in that case? Is the current limit more a matter of heat sink size or of which transistor I chose? (Running both MJEs, the middle two of which are considerably hotter than the outer pair.)

Power is the important parameter. In the case of the output transistors, that's V+ times the bias current. That's the same whether it's a BJT or a MOSFET. We know the MOSFETs can easily run 100ma, or 2.7W at 27VDC. However, we also know the MOSFETs require the 1.5" sinks. So, one would think that there's no reason the BJT's can't run the same - 100ma with a 1.5" sink. However, as ruZZ.il says, the BJT's have a positive temp coefficient. They also have different cases in some instances. So, you're in trail-blazing territory to a certain extent and I can't guarantee anything. The best thing to do is look at the table that Amb devised for biasing the M3 relative to supply voltage and the height of the heat sinks:

PSU voltage Maximum recommended quiescent current (w/1.5" sinks)
24V 160mA
26V 140mA
28V 130mA
30V 120mA
32V 115mA
34V 110mA
36V 105mA
38V 100mA
40V 95mA.

So, you can take from that you might be able to attain 135ma maximum with 1.5" sinks. However, this is with a genuine metal-backed, metal-tabbed TO-220. It will be slightly less for the MJE's and less even more for the totally plastic 2SC3422, 2SC3421 pairs, etc. So, as stated in the beginning - 100ma, no question with 1.5" sinks. After that, it's up to you.

 

[thomaskuhn]

Tomb,

I am curious as to what you have enjoyed more so far? The SE - wima combination or the BG combination (based on your current break-in period and possibly an extrapolation)?

Tom Kuhn

 

[tomb]

Quote:

Originally Posted by thomaskuhn /img/forum/go_quote.gif Tomb, I am curious as to what you have enjoyed more so far? The SE - wima combination or the BG combination (based on your current break-in period and possibly an extrapolation)? Tom Kuhn

Well, this is a timely question because I'm close to modifying the website with what's been found so far.. I'm still in the thick of trying many things out and have yet to really try the PIO caps. I've had a bunch of Vitamin Q's for awhile, but they're so big - even if I could find a way to make them work, I think they would be unrealistic to recommend to everyone. However, fordgtlover sent me some of the green Russian PIO caps (0.22uf 160V) that will fit. So, they may be a realistic alternative if they sound good and fordgtlover seems to think they do already. I'm also still testing quite a few of the transistors. I'm pretty familiar with several of them, but not others.

So, keep in mind that my opinion may be modified over the next week or two.

1. 1st config - budget conscious, realiable spectacular results:
ES caps at CA2 and CA7, WIMAs all around, hefty* caps for CA4 and CA5
2SC3422/2SA1359 BJT's

This combo simply rocks. Bass slams the doors off and highs are reasonably effective. It's a very rocking combination that goes well with everything from Senns to KSC75's and inbetween. The WIMA's only have a slight graininess in the mids and a bit of harshness on the top end, but are very tolerable and competitive with amps easily costing twice as much (more than that if commercial).

Frankly, despite that slight deficiencie with the WIMA's, I have heard nothing to compare with the bass in one of these short of a cranked up subwoofer in the car trunk. It can almost be intimidating with certain headphones (Beyer-padded Sony V6's, for instance).

2. Same as 1st config, except add Sonicap GEN II's to byppas CA7 instead of the WIMAs - WIMAs everywhere else. This is sort of the minimum upgrade to #1. Bass is slightly less, but highs are superlative and mids are soothing. Highs are the sort that my poor ears often hear only as a semi-constant fluttering or dog whistle in the ears, but it's something that adds airiness and an incredible sense of hearing every detail that exists. The Sonicaps are only $3.30 each, about $2 more than the WIMAs. So for about $4 more, you can supercharge the highs of #1. The only down side is that some of that subwoofer slam is missed.

3. 3rd config - high cost, audiophile special:
Black Gate NX's at CA2 and CA7, hefty caps at CA4 and CA5
2SC2238/2SA968 BJT's

This seems to be the Audiophile special. To my ears, it's as flat and detailed as a solid state amp - nothing exagerrated, but nothing lost. I think the BJT's add to this effect a lot, since they have always been listed as the ultimate neutral and extended response BJT's. The only downside that I can see to this combo so far is the slight fuzziness in the Black Gates. That may decrease with further running - as pabbi has indicated. Or, it might be corrected immediately with a pair of those PIO caps. I will try them by next weekend and let you know.

The Sonicaps are out of the question for this and so would be just about every other popular cap such as Auricaps - IMHO, it would ruin the wonderful neutrality that this combination seems to have (already). Bass is very, very deep and firm, yet highs are splendid. They don't seem quite as extended yet as the effect gained with the Sonicaps on the ES's, but either further run-in or those PIO caps may address that. The slight fuzziness and tendency toward a bit of metallic sound makes guitar work sound glorious.

So, there's three combos that work very, very well - from inexpensive to expensive. Time will tell if the PIO caps add to this combo, but everyone who's tried them seems to like them. Other than that and perhaps the Sonicaps on the output ES's (or the Millett-standard WIMAs), I would be very, very stingy with any mid-range bloom "boutique" film cap. They seem to do more harm than good - especially in bypassing the CA2 electrolytics, perhaps less so on the output.


* This goes back to some of my recent recommendations. I have always used giant caps in the CA4 and CA5 positions, even though Millett tradition has called for 470uf's in these positions. It's only pennies more when using FM's up to 1200uf and 1800uf (what I use). Other caps are more expensive, so this route may not be as beneficial to some. One day we'll be paying premium prices for those FM's, but until then - get 'em while you can.

EDIT: Didn't answer the question as to which was more enjoyable, did I? Well, I will always have a soft spot for those ES's and 2SC3422 BJT's and their rockin' slam. Are the BG's better? Yes. Are the ES's and 2SC3422's more fun? To me, yes.

EDIT: As I type this, the BG's are sounding worse and worse. That's sort of typical - what pabbi and others have reported and the same thing that happened with my BG Alien DAC. It'll probably be a week's worth of running before they sound good again and continue to improve.

 

[amphead]

Nice work on the BG Maxes TomB!

I was tempted to install a 2sc3421 in place of the 2sc3422, to test your "bad" amp Pabbi. But I think it would be prudent to wait for that 2sc3422 shipment.

 

[pabbi1]

Quote:

Originally Posted by amphead /img/forum/go_quote.gif Nice work on the BG Maxes TomB! I was tempted to install a 2sc3421 in place of the 2sc3422, to test your "bad" amp Pabbi. But I think it would be prudent to wait for that 2sc3422 shipment.

Yes, should be Tuesday or Wednesday, as I opted for an expedited shipment... mainly to respect your time, and I am all but insensitive to cost penalties.

Tom, the Black Gate bloom really is a wonderous thing - you wonder what all the fuss is about, they turn dull, then downright icky, then wham - where did that come from? It is a truly mysterious experience, but something you just have to experience to join the believers. In fact, I didn't go BG mainly because this is an office rig, and I just travel too many weeks to live through all the bloom cycles I would be faced with. So, I thought the Cerrafines would suit my application better. This was not an economic decision for me.

 

[amphead]

Pabbi, Well I couldn't wait and I went ahead and installed 2sc3421 for purposes of troubleshooting. The right channel tube, heats up properly, but there is no bias voltage at TA2R. Measuring the plate voltage at pin 7 of the right tube, gives .9V . Therefore I'll replace QA1R and QA2R, looking at what is the likely culprit upstream on the schematic.

 

[Listen2this1]

Quote:

Originally Posted by tomb /img/forum/go_quote.gif You should not be complaining about a lack of bass with the MAX. Tell us more about the specs of what you put in. I can see that you put ES caps in CA4 and CA5 - are those the 470uf's? Did you say what output transistors you were using? If you're using the BD's, you might get the impression of less bass. Anyway, we can probably help this with a few changes.

Tomb In position CA4/5 they are Muse 470uf 35v, In position CA2 it is a Muse 1000uf 16v. The output trannies are the ones direct linked from the BOM. I was wanting to get some of the Toshiba's but then you had that shipping problem, so I went with the ones from the link. I am currently using Rca 12fk6 tubes. Bias is staying really stable.

Should the mje253's run warmer than the mje243's?

Also I am using the Grado 325i's, I know they are slightly on the bright side, this is why I am wanting a little more bottom end.

 

[joneeboi]

I'm all for biasing up to 150mA, but I just don't have any clue how many mVs I need to hit that mark. I'm running ~24.5Vs so I can easily hit that 150mA mark, I just don't know how carefully I should tread. There's got to be something I can consult wrt to this matter. The MAX website says 264mV for MOSFETs will give me 120mA. Does this transfer over for BJTs? I checked the MAX site, Steinchen's dB site and the M^3 site but I couldn't find anything.

edit: Also, I'm running with the top of the Hammond enclosure removed, so the greenhouse effect will be avoided.