D-AOC PCBs - the building and help Thread

[scott2000]

R105 ,R107 the right value?

 

[L091c]

Ah good call, checked, they are 47K and 10K resp, like the BOM :/
Just in case, I will trace back the values using the schematics, double check any divider that is in play.
Hmm all resistor values look ok.
Could it be because the board is not (yet) connected to any potmeter?

 

[scott2000]

I'm sure you'll find something. Mabe try a different tube if you haven't and checking cathode resistor would be good as well.

 

[radiance]

How's the plate resistance tolerance on these ECC81/12AT7 tubes?
If everything else checks out fine the difference in voltage could be due to difference in plate resistance between tubes....

 

[L091c]

How's the plate resistance tolerance on these ECC81/12AT7 tubes?
If everything else checks out fine the difference in voltage could be due to difference in plate resistance between tubes....

Can I measure the plate resistance somehow sensibly? (This is my first tube project, I might ask silly questions)

 

[audiophreak]

Ah good call, checked, they are 47K and 10K resp, like the BOM :/
Just in case, I will trace back the values using the schematics, double check any divider that is in play.
Hmm all resistor values look ok.
Could it be because the board is not (yet) connected to any potmeter?

I seem to remember I got antsy to check voltages and found discrepancies , posted here ( not sure when ) and made the same statement about not having the pots in yet , when I put the pots in the circuit and retested - all voltages checked out good.

 

[audiophreak]

Ah good call, checked, they are 47K and 10K resp, like the BOM :/
Just in case, I will trace back the values using the schematics, double check any divider that is in play.
Hmm all resistor values look ok.
Could it be because the board is not (yet) connected to any potmeter?

Yes, I posted on pageb69 of this thread , look at my voltages , that was without pots , after I installed pots , all voltages were good.

 

[radiance]

Can I measure the plate resistance somehow sensibly? (This is my first tube project, I might ask silly questions)

Sorry, I made a mistake while looking at the PCB pic you posted earlier. I don't think it's the plate resistance.

 

[L091c]

I seem to remember I got antsy to check voltages and found discrepancies , posted here ( not sure when ) and made the same statement about not having the pots in yet , when I put the pots in the circuit and retested - all voltages checked out good.

Thanks! When I searched for "voltages" (plural) your post didn't match, my bad.

I went into the rabbit hole and followed the advice to try other tubes, so I bought a matched pair of Miniwatts. The voltages on pin 1 of these EEC81s are now: 225V (L) and 218V (R). Doesn't measure matched, just not sure if this difference (hear-ably) affects the performance, the previous set of Siemens measured both the same voltage. What would your preference be?

 

[scott2000]

Are pots installed?

 

[L091c]

Are pots installed?

Thanks for your comment Scott, no not yet, we found out that the missing pots are the cause of our 130V difference from spec. Voltages will drop when the pots are installed.

But my next question was: how important is having equal plate voltages on each side? With the new tubes they're 3% off. Radiance mentioned plate resistance tolerances so hence my question. With my limited knowledge: when Mu and Gm are matched, I don't think it matters that one plate runs hotter than the other. Is that an acceptable assumption?

 

[Script]

Planning on adding Lundahl LL1540s to input. Is it as simple as implementing what the datasheeet says (schematic on right with 2 x 12K) or is it even simpler and could leave out the resistors ?

 

[MisterCMRR]

Those resistors are necessary to avoid a resonant peak in frequency response (flatten HF response) and minimize overshoot and ringing for square-wave input signals. See the section under "self-resonance". Such networks are generally required for any input transformer that contains a true Faraday shield.

 

[Script]

Thank you. I understand Zobel networks and that the datasheet says that RC is not required with R<25K. So it just the two 12Ks for CMRR of >60dB, which seems very good.. Had just wondered and wanted to confirm as I recall it being mentioned somewhere above that the LL1540 could be implemented without any additional resistors. It'll be hooked to input of circuitry which looks like this with a 15K.



In general, I think this circuit doesn't really call for input transformers at all. Can feed unbalanced as is or -- and this has crossed my mind -- add IC debalancing instead, which might be sacrilege for some, but definitely cheaper and probably also 'cleaner'.

 

[L091c]

In general, I think this circuit doesn't really call for input transformers at all. Can feed unbalanced as is or -- and this has crossed my mind -- add IC debalancing instead, which might be sacrilege for some, but definitely cheaper and probably also 'cleaner'.

Ohayo! Yes, since most of us do not use it in live concert situation with cable runs of >100m, it can run unbalanced fine (see: What are you using your D-AOC on? Switchable transformers?).

The point of using 'iron' is its sound when it saturates. I am building mine not only for its compression function but also for the sound & color it brings into my chain. I had Lundahls planned for it, but after thinking hard I actually will put the Edcors and Lundahls in a separate transformer box with some Triads as well for switchable flavours.

 

[fragletrollet]

Just turned on power for the first time on this thing. After settling in I'm reading 243.6V on first test point on the left channel and 242.9V on the right, so so far so good. Pretty stoked, my first tube project. The accumulation of a couple of years of thoroughly studying this thread and whatever else "noob-friendly" tube design/safety information I could get ahold of.

Still haven't sent my front panel to production, as I am still deciding what (and possible extra) functions I want to include. These couple of years building this thing has resulted in some major feature-creep in my (various) front panel design(s).

 

[fragletrollet]

Attached are measured voltages after the unit is left on for a while to stabilize/warm up. I am presuming that rx27 is lower since I used 1n5370 that has 56V zener voltage instead of the 60V that the 1n5371 is specc'ed at. Not sure how this will affect my circuit, is it merely affecting Meter behavior?

The top resistors are also measuring a bit high, but again I am presuming that it's well within an acceptable range. The two channels are also showing very similar values to each other.

Time to try out some sound....

 

[fragletrollet]

Well, we all like nudies right

Actually a little surprised I got everything right before power up. Only had to swap around an output XLR Had some very weird behavior on one of the channels at first, vastly different gain and behavior on the makeup gain potmeter, somehow affected by the compression control (jumpy and uneven levels, low gain and other weirdness) that was all cleared up when mounting the pots to a temporary aluminium front panel. Guess the pots didn't like lying around with their body not connected to the chassis. Although I've changed the leftmost pot as you can see on the picture, I'm pretty confident I could put it back in, as it reads like it should out of circuit. Grounding problems are weird.

Now it's performing tip top; from some quick measurements in REW the noisefloor is way down in the -70-80dBu area (still pretty new in REW and all it's functionality). Sounds very clear and natural. I've only audited the 4c2 so far, looking forward to checking out the 2c2's. I've used trimmers instead of input pots, channels are surprisingly matched without much component sorting.
Compression action is very pleasant. I have not hooked up the meters yet, and am pretty sure I'm doing some insane amounts of compression but it all sounds great. Many different timing actions depending on the compression "threshold" control.

Kubi mods, double (quadruple, really) vactrols, toroid mounted on the back of the NRG case, JJ "Gold Selected" tubes (whatever that means), Lundahl 1540's on inputs and VTB2291's on output.
As you can see there's a bunch of stuff in there; as rack-real-estate is hard to come by I would really love to add some more features to this thing. So the plan is to add a MS board, after reading all these hardships people are having with the stereo link. I also would love to add some simple EQ functionality to it, I have some Helios69 boards from Ian that have a very suitable fixed low and high band (all there is room for on the panel, so skipping the mid band). Contemplating a 2520 style solution for the eq makeup, hence the SMPS (gotta do some current measurement, I'm guessing I might have headroom for a couple more valves as eq makeup amps). The JLM relay board is for the vactrol switching, and further bypassing of said anticipated modules.

Have a bunch of rotary switches I'm trying to choose from too, so I guess there's a fair bit of work left. But really happy that my first valve project is alive (and with so little headscratching!)

 

[Script]

Like

Why insert MS board here ? Have you considered putting MS into a dedicated case ? More flexible that way.

How are your 2W 2k2 (R211) resistors doing ? They too were running quite hot on mine so I used 25W for them also. (Actually all those 2W better be checked for heat.)

I added a fan at the back. -- This unit is great for winter --

 

[fragletrollet]

Well I sort of went on a bit of a shopping spree over at KA-electronics, so now it's time to M/S'ify everything!
Jokes aside, I am planning to build a mastering style console with extended MS capabilities, but it's very practical to have a simple matrix right in the unit itself. I find M/S processing so much more logical and easy to use on stereo signals and buses than L/R. As a musician/engineer who makes electronic music I deal mostly with stereo sources from my synthesizers and effects, where L/R has been previously used to define pan positions and it's modulation. Slapping an L/R at the end of the chain seems counter intuitive to me, the stereo signal is already been made to my liking with individual l/r gains. But with M/S you can now have control over the ambience/reverberation of the bus (if using stereo reverbs), compress the mids (often bass heavy which dominates the sidechain and following compression) to a different degree than the sides, apply locut filter to the side for a more centered low end (anyone say vinyl mastering)... or just to boost the side for that wonderful stereo widening, or do it with a high shelf band....

Aaand I just finished populating a MS76, so hey, why not make a theme of'em

We'll see what happens in the end

Regarding the 2w resistors, I dont have any means to measure the temp on them like an IR gun or what they call'em. I lifted them all up from the pcb to allow for some airflow and protecting the pcb in case they would get very hot. With my increased film caps there ain't any more room width wise, so I'd have to install them under the pcb if I wanted bigger ones. I have a 25w for the first 470r which gets comfortably warm. The heatsink with the ht and heater regulators share doesn't seem to get more than a little warm. It all seems to run pretty cool. We'll see when it's in the rack on a hot summerday