[samplesj]

Can you bundle more than one kit into the same chassis or is interference too big a risk?  Bundling the boxes seems nice because you can do straight hookup wire instead of hookup wire-connector-interconnect-connector-hookup wire. 

I did see the ClariCorn, but it was all on/shared power.  My system is in my office and I have cds on all day where vinyl would only be in the evenings and weekends so I'd want to be able power down the phono when listening to cds just in case it caused EMI issues. 

Looking at the schematic for the Clarinet, Cornet2, and Chime it looks like they all share the same power supply.  Would it be possible to extend that idea for a ClariCornChi (switching extra power between between dac and phono [not all three at once])?  I'm not sure if I want to replace my benchmark dac1, but the Chime does look interesting especially if they could be bundled into a single clean box.  It'd also save a ton on ICs and PCs  .

[hagtech]

Sure, you can fit all three into one box.  Power supply has to change if you want to run just one common supply.  You might need a bigger tranny.  Switching between phono & DAC is also an option.  Only drawback is that tube equipment takes awhile to warm up.  Sonically, anyway.  It plays within 60 seconds, you just don't reach peak performance until after about an hour.

Yes, you save a lot on interconnects.  That's my new approach with the RGP (an upgrade to a CLARICORNCHI).  Route everything inside.

With three of these boards inside, you'll need a very big box.  It would be 22 inches wide minimum.

jh

[samplesj]

Power supply has to change if you want to run just one common supply.  You might need a bigger tranny.  Switching between phono & DAC is also an option.

I was actually planning on adding a switch (as long as I can find a good one rated for the voltage) so that it would only power either phono or DAC.  Given that either or option where I'm only running two of them can I just borrow the previous work you did on the upgrades required for the ClairCorn power supply or does it need more beefing up to handle the Clarinet/Chime?

That's my new approach with the RGP (an upgrade to a CLARICORNCHI).  Route everything inside.

Can you share anything more about this new product?  Will it be available as a half-kit?  How far along in the design process are you?

I've got a few options for a metal case (maybe a sideways lansing full width, 3u high, 22" deep but I'm not sure about the edge lip), but I could also build a fairly nice wooden box.  I know that won't keep stray EMI in or out, but if this is the bulk of my gear in one box I should be able to space it pretty far from the rest.  Any other drawbacks to wooden cases (assuming sufficent ventilation)?

Thank you for your help on both these threads.

[hagtech]

You should be able to just copy what was done for the claricorn.  Difference is the low voltage tranny on the CHIME.  It can be left on all the time.

Info on the RGP can be found at www.haglabs.com.  I also post about it occasionally at my blog.  It will not be a kit.  Assembled only, retailing somewhere between ten and twenty grand.

Wooden cases are fine if you use a metal plate for the top.

jh

[samplesj]

Hmmm, I really wanted to share a power supply as an either/or, but I'm not so sure about switching H+/B+.  It scares me a bit for 300+VDC to be running through a switch.

I looked at running them all three at once, but I ran into a couple of different problems.  First I think it'd need a bit over 4 amps (AC) for the heater winding (7 300mA dc tubes).  The stock transformer product line does have transformers big enough, but the bigger ones all run the rectifier heater at 3 amp instead of 2 like the stock tx.  After more learning/research I found out why.  The 5y3 isn't really supposed to drive 7 tubes (~125mA limit and each tube is ~20mA) so why make a tx with enough 6.3v amps for 7 tubes.  A 5u4 would be happy with the 5 volt 3 amp heater winding, and would even have enough juice to drive all 7 tubes, but I don't know anything about tube design so unless its pretty much just switch out, then I can't go there.

So is three power supplies going to be a killer for EMI?  All those transformers right next to each other seems like asking for trouble.

It looks like I don't get to sell my extra interconnects and power cables after all 

[hagtech]

It's fine to have three power transformers side by side.  Or stacked.  It's no different that placing the machines in a rack.  Proximity effects do not change just because the chassis does, unless it is made of steel.  Indeed, this is the easy solution.  It also makes it possible to change chassis later for a new configuration.

However, you can still run a single tranny to drive all three boards.  Yes, the switch to a bigger rectifier is probably necessary.  Note the CLARICORN design uses the beefier 372BX.  Maybe it has enough juice?  I haven't looked.  If anything, Hammond makes a tranny that does.  I just don't know the number off the top of my head.

jh

[samplesj]

I don't really know what I'm doing, but let me see if this is close then...

Here is what I am using to look up the tx family details.

Maybe something like
370FX - 275-0-275@150mA / 5Vct@3A / 6.3Vct@5A

~300mA heater for 12AX7 and 12AU7.
7 tubes at 300mA each = 2.1A so the 5A is within the 2x safety margin you mentioned with the ClariCorn.

Now for the H+ resistors R223, R313, and R423.

R313 -
stock - 6.3v (tx voltage) / 2a (tx amp) / .6a (heater draw) / 2 (safety margin) = 2.625 which is really close to the 3ohm resistor used
ClariCorn - 6.3v (tx voltage) / 3a (tx amp) / .6a (heater draw) / 2 (safety margin) = 1.75 which is really close to the 2ohm resistor you recommended
ClariCornChi - 6.3v (tx voltage) / 5a (tx amp) / .6a (heater draw) / 2 (safety margin) = 1.05????  Maybe a 1.2 (using the same .875 derating as above)?

R223 -
stock - 6.3v (tx voltage) / 2a (tx amp) / .9a (heater draw) / 2 (safety margin) = 1.75 which is really close to the 1.8 ohm resistor used
ClariCorn - 6.3v (tx voltage) / 3a (tx amp) / .9a (heater draw) / 2 (safety margin) = 1.166... which is really close to the 1.4 ohm resistor you recommended
ClariCornChi - 6.3v (tx voltage) / 5a (tx amp) / .9a (heater draw) / 2 (safety margin) = 0.7??? Maybe a 0.82?

R423 should match R313 since they both use 2 tubes and both are stock 3ohm right?

So is this about right for H+ resistors?
R313 = 1.2ohm
R223 = 0.82ohm
R423 = 1.2ohm

If this is even close and it won't blow up if I start here, then I'll get extra values around these and nail it down even closer via testing.


But the B+ resistor changes lost me.  The ClariCorn suggested changing R220 & R221 to 3.3k.  I don't understand this part (Cornet2's B+ resistors changed, but not the Clarinet's).  Was this just done because the ClariCorn transformer was higher voltage?  If so then I can just leave mine alone.  If not and its also based on mA then I'm not sure how to calculate the proper values for this bigger tx with 3x the mA.


If I switch the rectifier tube to 5u4 do any cap values need to change.  I realize it won't drop the same voltage as the 5y3, but those were variable anyway for different tubes.  Does that mean I am home free with the switching to a 5u4 rectifier as long as I give it the heater juice it wants?  Is there a different rectifier tube that may work better?

[hagtech]

close and it won't blow up if I start here, then I'll get extra values around these and nail it down even closer via testing

Exactly, I think you got the hang of it.  That tranny sounds about right.  I use the same one in the CYMBAL and VACUTRACE.  AES has them in stock, because I keep buying them.

But the B+ resistor changes lost me

Honestly, I don't remember why I did anything.  Nothing will blow up, though.  Just measure and tweak once you're up and running.  The answer depends on which rectifier you choose.  The 5AR4 might also work.  We don't really use that much B+ current.  All of my production designs are tuned on the bench.  Use math to get as close as you can, then adjust for the real world.  Things like secondary winding resistance make a difference.

Only issue is that you can't adjust voltage one board at a time.  Tranny has to be fully loaded.

I think cap values are ok.  What you need to do is run one rectifier for B+ and one set for H+.  You might need bigger schottky diodes for heater?  You have to do this to prevent a nasty ground loop.  Then use multiple series resistors (3.0 ohm type) to split power to each board.  Make sure there is a good return wire for this loop.  The B+ is similarly split between channels and boards with series resistors (3.9k type).  Does that make sense?  Run the diodes and rectifier on one board. 

jh

[samplesj]

Ok, I know its been a while since I've posted here but I'm almost finished up now.  I had a backorder on the transformer and the DACT selector switch.  The switch finally came in yesterday.  I'm down to testing the voltage levels and hooking up the switches.

So that brings me to my next 2 grounding questions.

1: From what I can tell all three of the boards tie input and output grounds to their ground plane which is tied to mains earth ground.  I do have the ground blade on each board tied to my star ground.  As a convenience I'm going to tie the output jacks directly to my star ground point instead of the board.  Is that a problem?  Did I mistrace the board?  If there is a reason to tie from the board to output grounds that is cool.  For that matter what about floating the output grounds altogether?

2: I'm not using the board's switches.  I just ran wires to the front of the case and I've soldered their output into the vol output on the board.  Because I'm not using its switches then I don't have to ground all of my input grounds.  I'd like to switch them.  However I need to feed the dac and phono boards from the jacks and their output to the switch.  If I wire them normally that will always have their signal ground tied to mains ground.  Would it be safe (and sound good still), if I don't tied their signal ground to the board and don't use the boards output signal ground and instead just tie from the input jacks to the switch.  That way if I'm using the phono the dac input ground isn't dirtying anything (or vice-versa - or even further neither phono/dac grounds dirty aux 1-3 inputs).

DAC/phono input jack signal -> DAC/phono board -> switch -> clarinet -> output jack signal
Aux input jack signal -> switch -> clarinet -> output jack signal
ALL input jack ground -> switch -> mains earth ground star point -> output jack ground


BTW if anyone is interested a Lansing 22" deep case will hold three boards only IF the center board is raised or dropped a bit.  My clarinet board is in the middle and only has like .3" drop, but the others have 1.3".  The other board tubes just clear the top enough for a herbie's damper.  I had to order it un-vented of course since I'm using the sides as the fronts.  Boy was drilling vent holes fun.  I've got a cheap drill press, but it was hard to secure the side (their front) panel so my pattern wandered a bit.  I've got 2 rows of .25" holes .5" on center (took FOREVER).

[samplesj]

Well I hoped the above grounding questions would be my last questions, but I've measured something odd and want to ask a few more questions if I can.

Right now I'm using a sovtek 5AR4 for the rectifier.  Everything powers up fine.  The heater voltages are pretty much dead on thanks to your help (especially the Claricorn example).

However I noticed I was getting as much as 60v low for some of the bigger B+ measurements (-20%).  Before I started switching out resistors for the B+ I checked the tube and tx pins.  Both yellows, the yellow/black, and the bottom 2 tube socket pins all read only 320ish.  Do I need to try a different tube?  Do resistors R101/R102 (220) feeding the tube need to change?  I thought since it measures low at the socket upstream of the load splitting resistors they don't come into play, but am I VERY fuzzy on this so is this low measurement just a side effect of not having enough resistance further down the chain?

What exactly does the lower B+ cause?  Is it just a gain reduction or will it change the sound.

[samplesj]

OK, I got brave yesterday at lunch and hooked it up to my cheapo system (sb2 feeding a pot in a box feeding a teac digital amp to a pair onix ELTs) in the game room.

I injected it between the sb2 and the pot in a box (in case the volume control was wired wrong).  I used both a normal input to the clarinet and a digital to the chime.  First thing I noticed is that it had a hum.
It wasn't too bad on that gear, but still something to fix.  Using both digital to the chime then internally to the clarinet and analog to just the clarinet it sounded great.  Once the music was going you couldn't make out the hum at all.  That gear is really pretty junky and usually is fairly bright and thin, but it actually sounded good.  As I took it down I saw I had sat the power cord on top of the output cables and they were snaked back and forth.  I thought maybe that was my hum problem.  When I put it all back to normal it was a dramatic difference.

Encouraged by that I went ahead and hooked it to the good system (sb3 as source with butler 2250 amp feeding Magnepan 3.6).  It was CRAZY loud with hum.  The hum wasn't volume dependant.  At full on volume (with no music of course) I could hear some buzz.  Since it was still in testing mode I hadn't totally cleaned my wiring up very well.  I moved stuff around as much as I could.  Still had a hum.  Then I tried clipping the grounds apart so that 3 boards + chassis were tied to mains earth ground.  Selected input group, vol switch ground and output grounds were tied together, but floating from mains earth ground.  Still didn't really help.  I even took a sheet of ERS cloth I had spare and tried to shield the output wires and the source wires and it didn't seem to help.

I did stagger the boards so they would fit better.  Because I had bigger caps on the coronet2 and chime boards I let them be the lower ones.  In order to keep the tubes from dropping too load I did end up with the clarinet board within .25" of the top.  Is the clarinet board too close to the power transformer?  Could that be part of the problem?

EDIT:  I also realized as I was looking for a longer screw to drop the center down that I am potentially shorting the board mounting points to the chassis and thereby mains earth ground.  Is the board conductive around the mounting screws and would that cause issues?

[hagtech]

Sound like something isn't grounded.  Like chassis.  Can you post a photo?  Don't bother with ERS paper.  Use it after we solve the hum issue.

jh

[samplesj]

I can get a picture this evening.  I've checked the chassis to the ground pin on the power inlet and it reads ultra low ohms so its connected. 

I had several posts in a row and most are on the last page, but I did not connect from the input jack grounds to the boards (IE the phono input grounds aren't tied to the coronet2 board and the dac input ground isn't tied to the chime board).  It looked like the board just tied them to earth ground so I just tied them to the switch and grounded that (its floating now as a later "fix").  Should I go back and ground those jack shields instead?

I also get a 20% low reading out of the 5AR4 rectifier (new sovtek) if that makes a difference.

[hagtech]

The internal signal wiring should be done as if everything was in a separate box.  Each interconnect should be signal + return.  Just like using a coax cable.  In fact, the returns can provide the connection (through other boards) to eventual Earth.

It's not easy to wire up several boards that were not designed specifically for this approach.  But we can get there.  Eventually, we can remove all hum. 

jh

[samplesj]

I had to charge the camera battery so it'll be this evening for sure before I have any pictures.

I did a bit more testing with the cheapie system and to my suprise I found that it causes a hum even when powered off.

sb2 [ungrounded] - big box [grounded] - pot in a box [unpowered] - teac tripath amp [ungrounded]

Lifting the ground on the big box (FOR TESTING ONLY) did reduce the hum a bit, but it was still present.  Removing the input connections to the big box also reduced the hum some.  Obviously removing the output cables removes the hum  Unplugging the power cable also removed the hum totally.

At least the power off tells me its not HV DC buzz.  I am using a switched and fused corcom filter instead of the stock inlets with a switch.  My output jack cable does run right across the top of the power filter (totally wrapped in ers cloth, but I'm not sure it really does anything...).

Evidentally I've done something wrong lately because the hum is much louder up there than it was.  I may try to play with the output jack cable and see if I can find a cleaner route between the power filter and the power feed for the chime (output jacks are on that side)

[samplesj]

Well I've got a project in mind to reuse for the big chassis.

The lost cost of the chassis was all that was keeping me from single boxing after I ran into issues.  Since I can use it for another project then I'm not really out a whole lot for this experiment, but the time.  I suspect that I can dismantle the big box, add power supply sections to the other boards and wire them up in their own chassis faster than I can find the best way to run all those wires (2 sets of 120vac (board power + chime transformer), 2 H+, 2 B+, 5 inputs, 2 outputs, and tons of grounds).  Why wrestle all the routing issues with hum/buzz potential when you've already solved them for me if I just use the board traces which I couldn't do in 1 box, but will for sure in seperate boxes.

I really appreciate all of the help and double checking with finding out what resistors/tx/tubes had to be tweaked and now with the hum.  Hopefully in their own boxes it will run much smoother and I can just post a few success pictures in a week or so.