[cetoole]

Quote:

Originally Posted by xmokshax /img/forum/go_quote.gif colin: i hear scratchiness only when i turn the shaft of the pot. i wasn't terribly specific about this in my initial post. is this still a symptom that would be consistent with DC on the input? i've only used an iPod as the source thus far, and i've measured very little DC on its output in the past. i'm thinking of ordering some 12FK6 tubes for the next MAX i build (this one is a gift, so now i have to take care of "number one"). where do you recommend buying them online? i'm sure this information can be found readily on the forums, and i'm doing a search as we speak, but i thought i might as well ask since your mention of your own 12FK6es reminded me.

DC offset on the pot is noticeable when turning the pot. Measuring the DC offset is a real quick way of determining if this is a problem with your amp. Also, is your pot grounded? Some funny sounds can ensue if it isnt. You can test for this if you are unsure with your DMM, by measuring resistance from some metal portion of the pot's case to ground.

No idea where to buy Max tubes these days, I bought mine from vacuumtubes.net and radiodaze. Have GE, Tung Sol, Raytheon, RCA, DuMont, and Sylvania, in various different models. Probably have between 2-3 dozen Millett appropriate tubes in total, so I havnt had need to look for any more lately.

 

[amphead]

One small point that probably wont be affecting Max builders but is something to keep in mind, is........The heater voltage on the tubes should not be allowed to drop too low, say 10-20% lower than the tubes spec. The heaters which are a part of the filament, provide thermionic emmission. Sooooo they boil off electrons into the vacuum of the tube. This is essentially the electrons used to produce current flow from the cathode to the anode/plate. If the heaters voltage is too low, there is not enough available electrons sooooo.......the cathode slowly becomes pitted as electrons are stripped from its metallic surfaces, and eventually fails. This phenomenon is called cathode-stripping. Just some tube 101...... I don't know how much of a problem this is with low voltage tubes like the ones we use however. Dsavitsk?

 

[tomb]

Quote:

Originally Posted by cetoole /img/forum/go_quote.gif <snip>DC offset on the pot is noticeable when turning the pot. Measuring the DC offset is a real quick way of determining if this is a problem with your amp. Also, is your pot grounded? Some funny sounds can ensue if it isnt. You can test for this if you are unsure with your DMM, by measuring resistance from some metal portion of the pot's case to ground. </snip>

Xmoshakx,
Cetoole has a good point with the pot. If it really only happens when you touch and turn it, then it could very well be grounding as cetoole suggests. Did you install that wire from the pot ground pad to a screw on the pot?

Amphead,
I think most of us run the MAX at 27V. For those that don't and run lower voltages, your post is a good admonition. The Millett tubes all have definite warnings that heater voltage should never be below 10V. Even with a heater resistor of 20 ohms (most use 10 ohms, I think), 24VDC in the PS would still result in about 10.7V per heater. So as you say, MAX builders shouldn't be affected.

 

[xmokshax]

cetoole and tomb: i haven't grounded the pot yet. i know that the buzzing sound i hear when i touch the pot's shaft is from lack of grounding, but didn't realize that the scratchiness might be caused by the same thing. i'll install the ground wire and see if it goes away.

EDIT: ok, minimal DC on the inputs w/ my iPod - <1mV on each channel. soldered in a ground wire, which eliminated the buzzing (of course), but the scratchiness is still there when i turn up the volume. sounds like a (VERY) softened version of the sound you hear when you clean a record player's needle while the amp is on.

 

[tomb]

Xmokshax,
I think you have a dirty pot. Buzz when you touch without grounding, but as you've found out - scratch stays. Most likely, you've got some dried flux in there.

Same thing with my latest here

-


[size=xx-small](click for bigger pics)[/size]

I've got a scratching on the left channel when I adjust the pot. I haven't attached the ground strap yet, but just for a test: I attached a test jumper from ground to the pot. No go - the scratch is still there. It probably happened when I had to replace a couple of transistors a few hours ago.

I should be working on drilling the holes in the MiniMAX case, but hey - it's a holiday.

Anyway, this is another Black Gate MAX I had built, but had never fired it up. I decided to work on it last night so that I could fire it up this morning.

So first thing this morning - carefully, I screwed down the DB trimmers, stuck my probes in the left DB test points and fired it up. Hmm ... the meter says 0.0 - that can't be right - checked the meter setting, that was OK - checked the probes to see if they were making contact - that was OK. That's when I smelled it - burning plastic!

I cut everything off and began the troubleshooting. First, I suspected that maybe the Vitamin Q's that Colin gave me were contacting the opposite pads in their tombstone position, but no - measurements checked out. Everything else looked great, so I started checking the transistor numbers vs. their silkscreen locations. Even-numbered transistor, odd numbered silkscreen and vice-versa. Then I saw it: the two small signal transistors QB2L and QB3L were reversed. I had a 2N5088 in QB2L and a 2N5087 in QB3L. That was even-even and odd-odd - not right!

About an hour's worth of surgery later (those little pins are tough to get out of the board holes!), I fired it up again and it sounds great. I can't tell whether it's Colin's Vitamin Q's or the 2SC2238/2SA968 transistor pairs, but it sounds better than my other Black Gate MAX did. The Vitamin Q's add a tiny bit of sparkle over the K42's perhaps, but they wouldn't sparkle if the sparkle wasn't there to begin with. I guess I'd have to say that the Chinese-Toshiba thing mentioned by 04BluMach is real. The other BG-MAX had Chinese versions. I'm still not convinced that these pairs are any better than the 2SC3422/2SC3421/etc. transistors - just different. However, I don't think there's any doubt the Chinese versions are not as good - noticeably. I'm starting to wonder if the "conjestion" I reported with BG's was simply the Chinese transistors.

Anyway, specs:
Black Gate NX's for CA2 and CA7, Panasonic FM's all around - 1000uf in PS, 1800uf elsewhere, Wimas in the unimportant positions, Philco 0.47uf Vitamin Q's (I think that's what Colin said they were) - about 1-1/4" tall, 10ohm Kiwames on the output. I have a couple of Tung-Sol 12AE6A's in there right now, but the one on the right is irritatingly microphonic. It's only biased to 40ma so far - being extra specially careful after smoking those two trannies. Despite that, it sounds great already. What BG break-in?

 

[Negatron]

You might also check the continuity of RA7. It's also easy enough to paralell another 1M R between the grid pin of the tube socket and gnd.

It's not uncommon for the neg V on the grid to cause noise from a pot if the grid R is not adequate.

 

[tomb]

Quote:

Originally Posted by Negatron /img/forum/go_quote.gif You might also check the continuity of RA7. It's also easy enough to paralell another 1M R between the grid pin of the tube socket and gnd. It's not uncommon for the neg V on the grid to cause noise from a pot if the grid R is not adequate.

Well, you may have something there.

After basically drenching the pot in flux remover and contact cleaner, it had no effect.

After populating 7 of these boards so far, I'm fairly certain that RA7 is both sufficient and connected well on this board. However, the pins on the left tube are pretty bad. I've been straightening them all day on and off because the slightest twist in the socket bends them. They have some corrosion near the glass base. So, guessing that you were right - only maybe it's the pin connection to the grid - I switched tubes with their sockets. The scratching is almost gone except at very high volume travel and at that point, the very slight scratching is in both channels. So, maybe a little steel wool on the pins is in order - assuming that one tube doesn't drop its pins altogether, which I suspect will happen soon.

Gotta put me some good GE's in there. Meanwhile, problem solved - thanks.

 

[Negatron]

Quote:

Originally Posted by tomb /img/forum/go_quote.gif Well, you may have something there. After basically drenching the pot in flux remover and contact cleaner, it had no effect. After populating 7 of these boards so far, I'm fairly certain that RA7 is both sufficient and connected well on this board. However, the pins on the left tube are pretty bad. I've been straightening them all day on and off because the slightest twist in the socket bends them. They have some corrosion near the glass base. So, guessing that you were right - only maybe it's the pin connection to the grid - I switched tubes with their sockets. The scratching is almost gone except at very high volume travel and at that point, the very slight scratching is in both channels. So, maybe a little steel wool on the pins is in order - assuming that one tube doesn't drop its pins altogether, which I suspect will happen soon. Gotta put me some good GE's in there. Meanwhile, problem solved - thanks.

I get a lucky guess sometimes. I realize you know, but for the benefit of others, steel wool for cleaning tube pins should be 4/0 or at least no more than 2/0. I mention this because I cringe when I hear people say they will 'wire brush' the pins to clean 'em. BTW, for bad corrosion, CLR or even 409 and a toothbrush works OK with a GOOD after rinse of H2O and then pure Isopropyl alky.

 

[tomb]

Agreed. It was a poor choice of phrase (key words:"maybe" and "little") - I don't own any steel wool, anyway.

More than likely, it will be a little alcohol and a soft toothbrush.

Biased up to 50ma now and it's burning in nicely.

BTW, your solution helps explain another issue that pops up from time-to-time that I couldn't explain: cutoff noise in the delay circuit. The cutoff on the delay circuit almost always has a small pop or click. However, every once in awhile I get one with quite a "whoosh" in one channel on cutoff. Up till now, I assumed it was a problem with the relay.

Well, this one had it in the left channel. After switching the tubes, now it doesn't.

I would've never guessed a poor connection with the tube grid was causing it.

EDIT: Argggh - Head-Fi's changing smileys on me while posting this. Now "allteeth" is "big grin" - ":" and "D" instead of "

".

 

[Negatron]

Quote:

Originally Posted by tomb /img/forum/go_quote.gif Agreed. It was a poor choice of phrase (key words:"maybe" and "little") - I don't own any steel wool, anyway. More than likely, it will be a little alcohol and a soft toothbrush. Biased up to 50ma now and it's burning in nicely. BTW, your solution helps explain another issue that pops up from time-to-time that I couldn't explain: cutoff noise in the delay circuit. The cutoff on the delay circuit almost always has a small pop or click. However, every once in awhile I get one with quite a "whoosh" in one channel on cutoff. Up till now, I assumed it was a problem with the relay. Well, this one had it in the left channel. After switching the tubes, now it doesn't. I would've never guessed a poor connection with the tube grid was causing it. EDIT: Argggh - Head-Fi's changing smileys on me while posting this. Now "allteeth" is "big grin" - ":" and "D" instead of " ".

The 'Smileys' have been interesting, but the site reconstruction was quite an impressive undertaking. I'm starting to write my own in brackets <g>

I think my 4 most used tools with tubes has come down to Toothbrush, Alchohol, Caig DeOxit and most called upon of the lot - patience. (large grin)

 

[regal]

With film bypass caps degrading the bass on the cathode caps I wonder if it would make sense to try no CA2 cap at all. I know many times designers just use a resistor on the cathode. Has anyone tried this?

 

[tomb]

Quote:

Originally Posted by regal /img/forum/go_quote.gif With film bypass caps degrading the bass on the cathode caps I wonder if it would make sense to try no CA2 cap at all. I know many times designers just use a resistor on the cathode. Has anyone tried this?

Regal,
What the MAX reveals is that many, if not most boutique film caps roll off in the bass and highs. That's not a damaging scenario unless paired as a bypass to the electrolytic cap in the cathode bypass. If the film cap in that scenario is not flat and neutral, then frequency response of the entire amplifier suffers.

This can be gleaned from places other than experience with the MAX. Dsavitsk first mentioned it in this thread and his excellent Notes on Output Coupling Capacitors reference gives some inkling to this.

So, bottom line - a boutique is not as good as a plain Wima in this scenario, where the Wima at least provides some highs resolution in contrast to the electrolytic it bypasses. Go further than that, and the blooming mids - tuned "boutique" film cap does almost nothing but damage the response.

PIO's such as the Russian K42 or Vitamin Q's - contrary to conventional wisdom (also confirmed by Dsavitsk) - are actually very neutral. These are fine bypassing the electrolytic in the CA2 position, but offer little over the plain Wima - except when it comes to Black Gates. Black Gates are good enough that the slight squeekiness and grain in the Wima's degrades the BG's. In that case, the K42's or the VitQ's will do a good job.

IMHO and YMMV, of course - but this is what we have found with the MAX.

P.S. Colin actually experimented with his MAX just using a resistor (an adjustable one, though) in the cathode bypass position. He eventually went to ES electrolytics and VitaminQ's - for a better sound, I think.

 

[Negatron]

A few comments regarding cathode caps in the MH or Max. As a rule over time I have come to avoid the use of cathode caps as an detriment to detail and further, I regard electrolytics as no more than a sometimes necessary evil. That being said, I can make a case for EL K bypass in the MH or Max.

One of the effects of a cathode cap is reduction of effective plate resistance or rp and the increase of transconductance or gm. In the case of the automotive tube used in the Max the gm is already extremely low. The omission of the cathode cap lowers the gm of the tube to very low levels as well as a pretty good apparent raise in rp and the tube already exhibits a pretty strange distortion spectrum with unusually high odd order distortion products to a level usually seen only in solid state devices. The large drop in gm and a major increase of rp by not using the K cap don't help the condition at all. The tube requires an almost zero load to operate under these conditions.

The original MH used an opamp buffer presenting a load to the tube greater than 100Meg, and the lack of load was a condition the tube benefited from. The buffer BJT input in the max presents a considerably lower input impedance and much greater load for the tube to drive. And the tube is not up to the task without a (large) cap on K. To my ears the original MH sounded best without a K cap. When I replaced the 551's with a DB I found the sound to degrade quite a bit. I'm not a fan of solid state but the change was hard to accept even with SS prejudice. The problem was fixed by the addition of a K cap. Drat, foiled again! (frowning icon)

I don't know if this clouds the waters worse or makes sense, I can't write worth a darn. Once it leaves my head and hits type it gets muddled up. But I am an avowed enemy of the cathode cap and yet recommend one with the Millett. There are contributors to this thread who have a far better understanding than I about this, as well as a lot more experience with the Max. Hopefully they can simplify this better than I can. (confused Icon)

 

[steinchen]

The load on the tubes is definitely a problem that degrades performance. While prototyping buffers about a year ago a tried several buffers with my revMH I owned that time, although the buffer was meant for my SOHA amp. To me the JISBOS (with its high impedance JFET input stage) sounded best. Colin, Tom, you may consider a JISBOS buffer output stage for the next evolution.

Nevertheless, Kudos and thanks to Colin and Tom for the MAX project. I just completed my MAX a week ago during the holidays: pic1 pic2

 

[tomb]

As you both imply, the cathode cap is important, if not critical in the MAX. This is why we went to the 1000uf rating. The MAX suffers if the cap is not that big. The original Millett and revMH Millett used a 220uf cap for C2. That size was thought to be about twice what was needed, according to the corner frequency equation. As far as I know, no revMH Millett with Diamond Buffers had a bigger C2 cap than 220uf - it simply wouldn't fit. (Actually, if memory serves, Bperboy built one with rather large Elna Cerfines that canted out from the edges of the buffer board in a horizontal direction.)

Through extensive prototyping, we found certain aspects of the MAX that were needed to ensure optimum performance. Some of these items didn't follow the perceived theory: sizing the CA2 cap is one of them. CA4 and CA5 are important as well. That's why we included the selection of "Can't Miss MAX Builds" on the MAX website - to emphasize that there are some important differences from what went before. Perhaps Negatron's posts finally explain what's going on with CA2. I'm guessing that some of Amb's recent posts on power rail variance probably explain the importance of CA4 and CA5 as well.

At any rate, each MAX that I've built performs better than the four Millett Hybrids I built before (one was pictured earlier with the aqua LED's). With Steinchen's DB boards, they were no slouches, either, and were my favorites!! I hope you found the same, Steinchen, but with the level of superlative amps in your collection - it may be hard to tell.

That's definitely a great looking MAX to add to your collection!

BTW, your development, testing, and reviews of the DB's and DB transistors has been critical to the MAX's success. Your work has been referenced numerous times in the MAX's documentation. Thanks for the kudo's and all of the work you did that we built on!

P.S. Colin is going to work on further MAX refinement. I'm pretty sure one of the first things he will include will be his Bipolar Discrete Shunt Regulators. They recently got some nice comments from Jan Didden over on DIYAudio. There are some other things that may be pretty surprising, but we'll have to wait!!