[unbalanced/balanced to balanced tube input]

For reference, this design can take the E288CC (uber 6922 basically) with just minimal or no modifications. (Note they are 1/2" taller than a 6922) I had to slightly adjust the fixed cathode resistor value to get enough range out of the cathode trimpot, but that's about it. Dialed in to effectively zero offset/balance. A couple of the mirror transistors may get a bit warmer, but I found nothing alarmingly warm after many hours of use. 

It's controversial, and I'm quite a skeptic usually, but I found a significant improvement going from very nice but "normal" Amperex A-frame 6DJ8/ECC88 (essentially a 6922) to the Siemens E288CC. Got mine from Brent Jessee in Chicago. Detail, width, "3D". Tighter less boomy, but still very strong/punchy low end. My .02c

[and now for something completely different part 3]

Protector board noise... I still have noise on whichever "side" (dual mono) I connect the protector board power to. If I use low sensitivity cans, like my HE6SE, I can't hear anything, but high sensitivity cans, like my LCD-X, there is still a persistent and annoying very high frequency noise. I'm more and more sure the protector board is causing it. I'm currently running no shield at all from outputs to headphones, just 4 wires. I've tried connecting the protector board ground to the star ground as well as "chassis". No change.

At this point I think I need a separate supply all together, or possibly a RC filter between the grlv and protector board, to both filter and drop the 30v down closer to 12v a bit. Bad idea?

EDIT: I added 1k resistors on the +30v and -30v inputs. Noise is gone! Getting about 13v on the + input after the added 1k, before the regulator, and about 25v on the - input. Everything seems to be functioning properly, looks like that did it, unless anyone can think of a reason this is a bad idea?

Here's the wiring scheme and pics.

Messy wiring at top by transformers is temporary, waiting on a proper dist block. 

And no, my speaker jacks are wired yet, my headphone/speakers switch is just a smidge too short to allow a nut on the inside, so I can't tighten it down. Prob going to 3d print a bracket of some sort to hold it.

All wire is silver plated copper/teflon, not that it prob makes any difference.

[balanced headphone protector]

Alright, the vast majority of my noise issue is resolved, but I still have noise on whichever "side" I connect the protector board power to. If I use low sensitivity cans, like my HE6SE, I can't hear anything, but high sensitivity cans, like my LCD-X, there is still a persistent and annoying very high frequency noise. I'm more and more sure the protector board is causing it. I'm currently running no shield at all from outputs to headphones, just 4 wires. I've tried connecting the protector board ground to the star ground as well as "chassis". No change.

At this point I think I need a separate supply all together, or possibly a RC filter between the grlv and protector board, to both filter and drop the 30v down closer to 12v a bit. Bad idea?

EDIT: I added 1k resistors on the +30v and -30v inputs. Noise is gone! Getting about 13v on the + input after the added 1k, before the regulator, and about 25v on the - input. Probably only needed the resistor on the + side that drives the relay, but can't hurt I guess. Everything seems to be functioning properly, looks like that did it, unless anyone can think of a reason this is a bad idea? If anyone else has this problem with 30v rails, give it a shot!

Here's the wiring scheme and pics.

[and now for something completely different part 3]

On 12/26/2023 at 12:20 AM, Pars said:

I'm just driving ortho's (T50RP modified, and now a pair of DCA Ether C Flows). I wouldn't even consider trying to drive speakers with it. Caps in the GRLVs are Panasonic TSHA 4700uf 63V.

And Merry Christmas to you!

Nice cans! I just nabbed a pair of HE6SEs, but also have LCD-2 and LCD-X. I surely won't be driving anything large with the CFA, just some small, efficient near field monitors on my desk, supported with a powered sub. Dukei said that his customers had good things to say about it with small nearfield speakers. We'll see I guess. I def have other amps if it doesn't work out. (various modded adcoms, GFA-535, 545, and 555)

This build is taking it's toll, sooooo much work, although I'm admittedly chose hard mode with the dual mono, dual chassis, tube input version. 

[and now for something completely different part 3]

4 hours ago, Pars said:

I wouldn't think that large transformers would equal less heat; maybe more since the unloaded voltage will be higher. You can remote mount the pass devices for the GRLVs; I'm doing that on my DynaFET. You are doing dual mono, so you have 2 CFP2 boards per GRLV. The DynaFET is similar, with a pair of DynaFET boards powered by one GRLV. I have that in a 300mm deep Pesante case, with the pair of DynaFETs on one heatsink, and the GRLV on the other. The GRLV heatsink gets almost as warm as the DynaFET side does. In your case, I might consider rotating the GRLVs 90 degrees to point the pass devices at the sides of the case and mount the pass devices there, either directly to the side or on your L bracket?

Also, are you sure you have the rectifier boards in right on the GRLVs? I guess one side wouldn't work if it was backwards, and you are using the 3 bridge boards, which I'm not.

 

Thanks Pars! I’ll give that configuration a look! Are you driving speakers or just headphones with the dynafet? Weird that mine were running so cool, even with a susvara. 

As for the bridge boards, It must be because we have different bridge boards, as these have running great for quite a while, just finally stuffing it all into the chassis. Early on I actually put my bridge boards in like yours and got no output on one side because of it, so must be a board difference. Good eye!

Side question: what value are you using for main filter caps on your grlv’s? They look large. I was contemplating if I could benefit from bigger caps on mine, I have plenty of vertical space. I also see some putting larger local filter caps on each cfa board. 
 

Thank You! Happy Holidays!

[and now for something completely different part 3]

On 12/2/2023 at 9:06 AM, Pars said:

I think you might need more heatsinking for the GRLVs, particularly if you are going to be running some speakers judging from the binding posts I see.  How big in VA are the low voltage transformers for the GRLVs?

Hi Pars, could you (or anyone!) elaborate on the extra heatsinking for the grlvs? I admittedly have only tested with 50-600ohm headphone loads, but the small sinks I had on the grlv's then barely got warm at all. Maybe 105*F max. I have 100VA 30v transformers. I seem to recall someone mentioning that big transformers (and the 30vac secondaries) mean less reg heat from the grlvs? Below is what I was thinking of doing. I could also rotate them 90* left, bolting the L bracket to the side plate but that's even thinner steel than the bottom plate, and would move the DC outputs closer to the 90vac input on the grhv to it's right. I could also rotate the grlv 180* and remote mount the regs to the ~1.5mm aluminum back plate, but that would bring the grlv 30vac inputs closer to the DC jacks as well. 

Second, here's what I can get away with for IEC - grhv proximity. Think it will be a problem?

It would give me enough room for the power switch at the top, plus room for distribution blocks for mains/grounds.

Big thanks to anyone willing to share their wisdom.

[and now for something completely different part 3]

3 hours ago, Pars said:

For the amp portion, I like the layout on the right better, with the tube input boards right along the center.

You really are going dual mono, aren't you? For the PSU, that does look tight. I think you might need more heatsinking for the GRLVs, particularly if you are going to be running some speakers judging from the binding posts I see. The GRHV100 is too close to the IEC as it is. How big in VA are the low voltage transformers for the GRLVs?

And being team crimped connectors, I'll ignore those phoenix blocks

Hi Pars,

Maybe I need to find a new power switch. This one sticks out about 1.5”. That’d let me shift everything back. I could also remote-mount the grlv regs to side panels with an added alum/copper bar. 

The lv transformers are 100VA Antek

About the phoenix blocks, I think I may agree with you. The amount of times these became loose on me makes one question their reliability.

 

[and now for something completely different part 3]

Can anyone offer their input on my layout before I start drilling?

Amp: Left or right layout? Input comes in the middle, so figured I'd go right down the middle to the vol pot. DC comes in on each side. Which layout is best?

Power Supply: The center transformer will interfere with the rather deep power switch if I push it back any further. Too close to the GRHV100 board? Is the GRHV too close to IEC input at the rear/bottom? Note the heatsink on the GRHV won't be there, sand will be sinked to bottom plate via a block I think. Flip position of GRHV?

Thanks!

[and now for something completely different part 3]

2 hours ago, audiostar said:

The only question that remains for me is why wouldn't you want to convert SE headphones to balanced? Just recabling necessary and using 4 instead of 3 wires!? 

I surely am using balanced headphones and balanced speaker outs. The issue just lies in that the speaker level inputs on most subs have negative terminals that are common, which would pop the cfa3 -outs if wired traditionally. Maybe that question was more aimed towards Pars' comment on how Kerry wires SE headphones to a differential amp, and not me specifically, but thought I'd clarify. Thanks!

[and now for something completely different part 3]

1 hour ago, Pars said:

I would think for that application it would be the same as how you would wire an SE headphone jack. L+, R+ and PSU gnd for the - for each channel (L and R). For example, look at how Kerry did it for the SS mini dynalo.

 

49 minutes ago, audiostar said:

Oh, yes, right in this case - the same as how you would wire single-ended headphones, as suggest by Pars. 

Thanks Guys! I’ll give it a shot! 

[and now for something completely different part 3]

5 hours ago, audiostar said:

No, chassis (and ground as well) do not touch the signal. You have two halves + and - per channel, from input to output. Speakers connect the same way to the amp as headphones. The only difference is the impedance. 

Sorry if I mis-wrote. The sub is a powered sub, the speaker level inputs are just a way to get signal to it. Said speaker level inputs have an opamp buffer/mixer to combine the L and R signals and feed it to the built-in plate amp of the sub. So the CFA should see no load from the speaker level inputs on the sub. This method is how REL tells customers to connect their powered subs to a differential amp. According to them the only necessary connections are the L+ and R+. They suggest the cfa chassis to sub high level - connection “if there is hum”.

https://rel.net/blog/2015-07-23/how-to/how-to-connect-a-rel-to-a-balanced-amp/

[and now for something completely different part 3]

Can anyone verify this is the proper way to connect a subwoofer with speaker level inputs to a differential amp like the CFA3:

CFA3       >   SUB
R+           >   R+ SPK LVL IN
L+           >    L+ SPK LVL IN
Chassis  >    R- and L- SPK LVL IN

I'm excited to use the CFA3 with high efficiency monitors, but with "only" 15W/chan will definitely want a small sub. Sub is a small velodyne.

[and now for something completely different part 3]

6 hours ago, n_maher said:

Suggestion - instead of trimming the lids to get the IEC to fit, why not 3d print a spacer?  Be a shame to risk mangling an otherwise pristine set of cases. Looks like you only need a mm or two.

Good idea! I was planning on using a nibbler to just cut a notch in the lips, which should look decent, but a spacer would definitely work too.

[and now for something completely different part 3]

Chassis Mocked up. Still mostly empty. Definitely will need to trim the lips of the top and bottom plates of the PSU chassis to fit the IEC block's trim, but otherwise extremely happy. Not sure if I've mentioned but this is just a one-off for personal use. Definitely a huge setup, and doesn't come close to the refinement of many of your amazing builds, but for the first headphone amp I've built from scratch, (RIGHT to the deep end, haha!) I'm pretty content so far. haha! 

Not pictured, JAE DC connectors, Neutrik XLR inputs, CMC Rhodium 5-ways.

[balanced headphone protector]

On 3/12/2015 at 6:05 PM, kevin gilmore said:

the 15 second startup delay is intentional.

you can change the resistor string to make it more or less sensitive

Could anyone offer a starting point to lengthen delay? First 100k to 200k too big of a jump?

[balanced headphone protector]

1 hour ago, audiostar said:

No, balanced headphones do not touch ground. That "ground" pin on the 4P XLR connector goes actually to chassis; that is if everything is implemented properly, and if not, to nowhere.

Exactly. Or properly to chassis and it goes then to the star ground in the PSU.

You had introduced a ground loop.

Yep, apparently. Didn't think just a floating shield could do that...  Thank you for all of your help here and elsewhere, I seriously appreciate it! 

[balanced headphone protector]

Well, right as I was about to give up and buy a pre-made board I figured my issue out. 4P XLR shield shouldn't be connected to protector board gnd apparently? If I disconnect the ground between 4pin XLR jack shield and the protector board, no noise. Also no noise if I ground the 4P XLR connector shield to star ground. Protector board itself is also connected to star ground of course. So, ground loop/shared ground return current issue? Sure was a high pitched noise...

What gave me the idea to try this was when I touched the barrel of the XLR jack, the noise lessened and then slowly went  completely away. Let go, noise came back. That told me "ground issue" and here we are. Fixed.

[and now for something completely different part 3]

I got lucky with them, usually they charge if you don't supply CNC files directly. I made vector drawings to-scale, but don't know solidworks or such, so they definitely had to do some conversion and recreation to some degree, but I didn't get charged for the additional file creation fee. Gianluca and I went back and forth probably 50+ times over a number of months. Definitely great to work with, and like you mentioned, easily cheaper than sourcing locally or taking a risk with aliexpress/etc. 

If anyone with a CFA hasn't tried the ubal2baltube boards, they add just the right amount of "tube" flavor, and still keeps the CFA detail/slam. I splurged on nice Siemens E288CC tubes (not quite the same as a 6922, double heater and plate current) which meant some additional tweaking of the tube boards, but I think it was worth it. Compared to basic EH 6922 they are a clear improvement. They ARE a bit taller than a 6922 though, so they may end up sticking out the top of the chassis, we'll see, ha! Could look pretty neat with some tube cages, and make swapping easier. Like this: 

[and now for something completely different part 3]

1 hour ago, audiostar said:

Nice. How much did you pay for them?

About $650 shipped, with all the customizations and upgraded alum top/rear panels and 10mm front panels. Rear panels are machined and labeled as well. Pricey, but fair I think, I wanted to go all-out. The tube input boards are fantastic sounding.

[and now for something completely different part 3]

Just wanted to share a sneak peak of the machined panels for my split chassis tube input CFA3. Just got pics from modushop. 10mm thick. Very pleased with how they turned out! Will be 1x 2U Dissipante and 1x 2U Pessante. Gianluca was amazing to work with, top notch customer service.

 

[balanced headphone protector]

18 hours ago, simmconn said:

I bought this finished protector board on Taobao for about $12. The parts look reasonably good. It could use better cleaning off the flux residue on the bottom side. Totally worth the time I would otherwise have to spend soldering it.

 

Thanks! Yeah, impossible to beat that price! I had my boards made at oshpark. While I've had great luck with them, maybe something is up with the boards themselves.

[balanced headphone protector]

7 hours ago, jamesmking said:

its possible that Q1 is only just providing enough current for the replay contacts to just about close resulting in poor switch contact - that might lead to noise.

Also does the relay have a built in diode for transient protection when the relay switches state? If not the transients could be damaging Q1 and or the opamps.

Thanks for the reply! In my case the noise is on the 30v rails, not the protector board ins/outs. I have a dual mono CFA3. All I need to do is connect the board to a grlv also powering one channel of the CFA and it injects the noise into that side of the CFA3, audible in headphones that are connected directly to the CFA3 outs, nothing connected to protector board ins/outs. So it's adding the noise to the grlv, which is then audible in anything also connected to that grlv.

And yes, there's a number of caps and diodes on the board that aren't in the drawing. There's a diode across the relay coil.

Thank You!

[balanced headphone protector]

Well I made progress on my noise issue. There is no noise present if I remove Q1, the mosfet that turns on the relay and instead connect the relay to ground directly, effectively hardwiring the relay on. That tells me that the noise likely (?) isn't from an over-taxed 7812 regulator being fed 30v. Narrows it down to Q1 or the comparator circuit driving it... right?

[unbalanced/balanced to balanced tube input]

Edit: I'm considering not running a HV delay at all, and relying on headphone protector board for delay and DC protection. A "standby switch" necessity is widely overstated, especially at the rather low 100v B+ we're dealing with in this circuit. Some even argue that hv delay (for long periods) can cause cathode poisoning. In my build, getting the delay working would be mean an additional 30vin to 12vout reg board as well as a 12v 555 delay board to operate the cpc relay. (can't find a 30v delay board without a noisy buck converter). KISS principle I figure. Any thoughts?

[and now for something completely different part 3]

Please don't kill me for asking this: Can anyone suggest the best method to connect a subwoofer with SE or speaker-level inputs only to a CFA3? (specifically ubal2baltube boards into 4x cfa2) I imagine the sub inputs grounds are not floating, which obviously presents a problem. Maybe a 15k 1:1 audio transformer tapping signal after volume control? aka a "DI box"? Or maybe something like this paralleled to CFA inputs post-volume?

Am I missing an easier method?