Satyrnine

Fired up first two cfa2's last night. A couple issues: Board 1) Without servo, I'm seeing around 400mv (!) offset on output with input shorted. If I plug in the opamp, it slowly comes down to 1-2mv over 30sec or so. I started with bias pot all the way counter clockwise. Not seeing anything across the 1R resistors after a good 8 turns clockwise, do I just need to keep going, or should I be seeing some idle current by that point? Board 2) LED's not lighting. It was late, and I gave up before investigating further.

Thanks for the input guys. As a testament to their great customer service, Goldpoint is letting me exchange my (unused) 25k quad for a 10k quad, for cost of shipping only. Their shipping is prompt too. I haven't used one of their pots yet, but so far so great. My config will be just 4x cfa2 at first, and then the tube ubaltobal input once I get boards/bom/iron ordered up and together. Since it's a free exchange for the 10k, might as well take advantage I figure. As mentioned earlier my legit dealer sourced 15031's range from 70 to 120hfe, approx evenly distributed. If it even matters, what would be the best way to distribute this range across my 4x cfa2 boards? If I recall correctly, one of the 7 sinked devices on each board doesn't get as hot as the main 3/3 set, should I put outliers in that position, or...???

Just rereading some of this thread. So uh... lets say someone ordered a Goldpoint in 25k. How much of a problem is that. Asking for a friend. 😑

So is this still the balanced/current mirror design of kevins but cascoded 6922's? I love it. Cascoded 6SN7's is one of my fav recipes. I have a SE pre setup that way, and it sounds great. I'm shooting for balanced all the way through for this build though. 6SN7's ideally should get more B+ vs 6922 obv, not sure how far down a that rabbit hole I want to go. Tempting as heck.

Here's one of Dukei's he sent me. Primo work as usual:

So you you suggest three different enclosures? Even with the tube pre in the power amp chassis, it would be only DC coming into the power amp chassis. Would put the tube pre powersupply in the PSU chassis. Are you concerned more about the 6v and 100v or some sort of other interference/issue I'm overlooking? So far, everything should fit in two 2U chassis, (one a pessante, one a dissipante) with plenty of room. Thank you!

Sounds like something I don't want to mess with. Thanks Kevin. Just chatted with Miroslav/Dukei, and he had great things to say about the sound/performance of the tube bal/ubal input. I think I'm going to go for it. This will likely be my endgame amp, so why not go all the way.

I wonder how easy it would be to adjust the reg output above 100v on the grhv100. I definitely don't understand all of what's going on there with the hv regulation. I was playing with the idea of using 6sn7's vs 6922's, which can handle quite a bit more voltage, and imo, sound a bit better. I don't want to make too much extra work for myself though.

To make a CFA3. bal/unbal + CFA2 = CFA3 right? ignoring that I'm going w the tube version, Not much more components in path than a base CFA3 I figure? It seems people really enjoy the added ss/zf (zf only in the tube version of course) part of the CFA3, and if I'm going through all this trouble I figure why not go the full distance. Plus, having a little tube flavor on the input is always appreciated, since I typically prefer tube amps/designs. A balanced tube pre isn't common, so when I saw a tube bal/unbal, that seemed like a perfect match vs a standalone tube pre. Last, seems like not many people have built the tube bal/unbal, so figure I can be of some use by guinea pigging it. I figure this is the best of both worlds. Am I off? If yall think there's nothing to be gained, I'll consider skipping it. Thanks!

Was looking at this. 115v. Nice size, and toroid. https://www.mouser.com/ProductDetail/Triad-Magnetics/VPT230-110?qs=wkKrz7WmEgNNNgyT8w4YqA%3D%3D

I was planning on sinking directly to the 10mm 2U front plate (along with 2x 30v grlvs) with good thermal compound. Think that'd cut it? Obviously "try it and find out" is the best answer.

That looks like the one! I would definitely be interested in buying one. I'm in Madison, WI USA. Thanks for suggestions on connectors. Excellent! I like switchcraft. I'll check all the options out. Many thanks! Since you ran those voltage, you don't happen to have a recommendation for a 90vac trafo around .5A or so? Seems to be a uncommon range. That is, unless I can use higher voltage ac than 90vac, 100-115 or so would be easy to source. Edit: nvmd on trafo rec. found some good options.

Excellent! What about the 90vac, and 8vac are they flexible too? 100vac seems a bit more prevalent in a reasonable VA. Or a 1:1 115v trafo would be great too. Will it reg down to 100v ok still? Same with the 8vac heater supply, I find a lot of 9vac trafos in a good size. Here's a very preliminary mockup of what I was thinking. To anyone: Will bundling all these DC rails (6vdc, +/-30dc, and 100dc) together in a single multi-conductor connection work? Best connector for that? A bit new to external power like this.

Can anyone confirm if I can use a 100V secondary trafo for the grhv100? Will it still regulate down to 100vdc without issue? Having a hard time finding a suitable 90V unit that isn't 1A+/huge.

Kevin, is there any way the ubaltobaltubeschem2 can be made to take +/-30 vs the +/-15? Otherwise I'd need a third grlv I think. Would be running 4x cfa2 from 2x grlv30. That is, unless a separate supply is recommended for the ubaltobaltube +/-15v requirement regardless. Also, suggestion for VA on the 90vac and 8vac transformers? I'd like to have 1.2A min available for heaters and 40ma plate current available if possible, to try 2x 6sn7's. (pinout is dif, but compatible)

Welp after all that work trying to stuff a quad cfa into a 3U, I think I'm going to add a ubaltobaltubeschem2 and grhv100, and go split chassis. Anyone have any ubaltobaltube or grhv100 boards?

Excellent, thanks everyone. Wow, The man himself responded to one of my questions.

Reading EL_Ken's post about grounding, (even though I know the incorrect grounding didn't cause a device failure in the end) I assume plastic shoulder washers are best to make sure boards don't ground to chassis through standoffs. Are brass or alum standoffs the best move, or would a nylon standoff/screw be a better option? I know it "depends" but does anyone have a suggestion for best size/height/etc for those?

Haha, I guess not.

A bit off topic, but I just have to say that this is the most knowledgeable, helpful, and BS free forum I've ever been a part of. Bravo everyone!

Oh wow, that must have been recent stocking! I just ordered from Farnell/Newark. Thanks! Definitely not something to risk.

I started populating my grlv's tonight. I tested my MJF15030 and 31's and they vary quite a bit. I recall it said that matching isn't too terribly important, but want to double check. I know I'll get scolded, but I DID source the 30's from a seemingly legit and well reviewed ebay seller, since they are out of stock everywhere. I'm wondering if I got a batch of rejects. Here's what I found: MJF15031 from mouser: All around 280hfe +/- 5. MJF15030 from THIS ebay seller: 80-150hfe. Oof. Visually they look identical, even with a loupe. I know that doesn't nec. mean much. Is this normal? If these are questionable, anyone have a source for 30's? Also, the larger tantalums are 36v, not sure if they see full 30v or not, that's a bit too close to 30v if so. I ordered from an existing bom, and didn't catch this.

Oh I did mostly, or I definitely wouldn't have made it this far, haha! I must have missed this one or overlooked thinking I needed to heatsink mount them with 30v.

Well hell, wish I would have come across that photo earlier, haha! That sounds like a great solution. I'll set it up just like that. Thanks Guys!

Here's another idea. Mount the grlv sand offboard, and flip is on its side. Mounting grlvs on top of trafo covers and remote mounting sand to sink.