the CTH (Compact Tube Hybrid) Rev A thread...

[Mullet]

We're finally here! Here's a little teaser...
 

 
Thus far the build is going well. The only non-standard situation that I've run into is for the C13P and C14P film caps. I had to bend the leads outwards to fit them into position. Not a big deal though. Overall, it seems like the new and modified positions for everything in the power and heater sections is making the parts fit a bit easier. Also, as you can see, C8P is kind of hiding part of the 4 heater positions. Next step is to wire everything up and test this puppy out tomorrow sometime. I'll be sure to let everyone know my progress and any trials and tribulations of the process.
 
When everything is verified and working a new BOM will be posted on cavalliaudio.com. Then I'll proceed to the next phase of shipping out everyone's boards from the group buy.

 

[cfcubed]

We owe much thanks to Mullet for doing most of the BoM revision, associated parts subbing/etc. and building out the 1st of the new rev.   For completeness:
GB for this: "Group buy for CTH" - http://www.head-fi.org/forum/thread/536098/closed-group-buy-for-cth-compact-tube-hybrid-pcb
Earlier  "A Very Compact Hybrid Amp" spec & build thread - http://www.head-fi.org/forum/thread/398839/a-very-compact-hybrid-amp
 
>  The only non-standard situation that I've run into is for the C13P and C14P film caps. I had to bend the leads outwards to fit them into position
 
This is because places were out of the 7.5mm pitch cap spec'd there so the rev'd BoM must have a 5mm pitch part #.
No biggie but we should note this in BoM comment or ID a better sub. 
BTW Mouser seems to be out of common parts lately (e.g. many Wima films, some common BC transistors, and the "long" hammond cases until recently, etc.)
 
The changes in this new rev are largely ones that make building easier & should not result in sound or performance difference compared to proper earlier CTHs.   
 

• Space to better fit C3H.  Though new BoM 2026A+ 330uf part has better specs and 330uf has proved plenty to fix early CTH hum issues, a 470uf can be fitted there now at the builder's option.
   
• C1H, C2H increased to 150u taking advantage of a new series of caps (thanks to user rds).
   
• A TO225 (2N49922) Q1P that can handle much larger currents like cases where there is a short on the output of the regulator.  Though the earlier re-spec'd BC337 has proven sufficient for expected loads, this change provides even more current "headroom".
   
• More room around C4L and C4R to give you more space to tombstone other caps in that area.
   
• More room to fit the heatsink on the 24V regulator.
   
• PCB support for up to 4 heater voltages.  Base 12.6V & 6.3V plus up to two other voltages if you wish & provide a way to switch amongst them.  This was accomplished through off-board resistors in the past, and you can still support as many heater Vs as you wish using that method.
   
• Misc organization tweaks WRT some diode positions, comp spacing, etc.

 
As Mullet indicated, runeight will update the CTH site with the various revision-related files & info once Mullet's build is proofed.
 
For the curious, here's the last PCB spec image I have from runeight:

 

[drclaw]

Were you able to get any work done over the weekend?  I'm excited to get this project rolling!

 

[Mullet]

I know the feeling!
 
Had some early hiccups with reversed resistors -- R1H and R2H and as a result ended up frying what looks to be the right side of the buffer output when trying to troubleshoot why I wasn't getting proper heater voltages. New parts are on the way and if everything goes smoothly with the transplant I think everything will be good to go by this weekend. Good thing we're vetting this BOM because I'm sure others would have had similar results with the heater voltages being so off.
 
I'm pretty stoked as a whole. Here is a list of tubes that I've collected in the past few weeks in anticipation for this amp:
 
Amperex Bugle Boy ECC88
Amperex Bugle Boy ECC82
Amperex USA Gold Pin JAN-6922
Pro Comm Miniwatt PCC88
Raytheon 6GU7
RCA Clear Top 6CG7
RCA Holland Gold Pin 7308
RCA 5963 Black Plate D-Getter
Siemens ECC82 Chrome Plate
Siemens ECC802S
Siemens PCC88
Sylvania Gold Brand 5814A
Telefunken ECC82 Smooth Plate
...and a variety of 4BQ7 tubes from Tung-Sol, Sylvania, DuMont, and Westinghouse.
 
I know a few of these are generally well liked by many using the CTH. Any comments or suggestions regarding other tubes or these ones in particular? I'm trying to locate an Amperex Holland 8416 PQ if I can find one. I hear it's quite nice.
 
 

 

[cfcubed]

>  New parts are on the way and if everything goes smoothly with the transplant I think everything will be good to go by this weekend. Good thing we're vetting this BOM because I'm sure others would have had similar results with the heater voltages being so off.
 
Yes, we think that the new CTH rev will work w/R1H/R2H correction in the new BoM.  The errant heater V caused by the swap was precisely what would be correct for the swap.  Before I could diagnose R1H/R2H swap as the cause for low heater (1.37v) & high plate Vs, a short may have occurred trying to research things.   These occurrences, a BoM error (esp. in modified heater area) & short while diagnosing a CTH PCB, are not unexpected.  I blew an OB once taking measurements to make sure everything was OK (it was prior to me trying to confirm so
 
In an effort to expedite things I've a new CTH PCB in hand & my parts arrive tomorrow (Wed) evening.  I should have it built & tested no later than this Saturday.
So one way or another we'll have answers this weekend.  

 

[cfcubed]

Well we have a working build now (mine), so I think we can call the new CTH PCB & newest BoM proofed.
 
Been listening to it for a while & it's working well, voltage measurements are fine and it sounds good.
 

 
Notes:

1. I did not have to raise 24V reg & flip it heatsink upside down to fit it, the combo will fit snugly where it belongs if mounted properly.  But I've something strange in mind for casing so my 24V reg heatsinking is not final.
2. Yes those are Gen II Sonicaps even though the BoM box caps are fine.  The "F" indicates the foil side/lead end going to the tube plates (input).  Used this to ID the foil: http://mailman.io.com/pipermail/sound/2006-October/014043.html
3. Here are the 4 heater voltage/resistors I included & my results (within reason given the available resistor values):
   Voltage / parallel resistor value / my measured voltage under tube load w/25 yr old, uncalibrated DMM:
   4.2v / 3.3k / 4.2v
   6.3v / 7.15k / 6.2v
   7v / 9.31k / 6.9v
   12.6v / <none> / 12.7v
   
   So I've not yet confirmed these but it seems they should work & wiatrob noted no issues w/them:
   5.6v / 5.8k
   8.4v / 15.1k
   9.4v / 23.3k

 

[Mullet]

Awesome cfcubed! Congrats on your successful build.

I get my new parts on Monday. I'm possibly planning on just redoing my whole buffer sand all in fell swoop. At a minimum I'll do at least the right side, which was giving issues reading 24v instead of 12. That way I don't have to pull my hair out -- only a few.

Can't wait for the CTH goodness!

 

[Mullet]

So I finally got my parts and have proceeded to replace a few BJTs on the ride side of my OB. Everything is testing fine with these readings...
 
 
99.6v SG>TB+
11.97v SG>OG
77.8v SG>Pin 1
77.5v SG>Pin 6
11.97v SG>OL
11.96v SG>OR
 
My pin 4 to 5 readings for four different heater voltages are as follows using the 2-12 position NKK rotary switch that cfcubed recommended earlier on in the GB post...
 
Position C - 6v = 6.04
Position 1 - 7v = 6.87
Position 2 - 4v = 4.058
Position 3 - 12v = 12.58
 
I'm not sure why they aren't in the same order as the BOM would suggest ie 12v, 6v, 7v, 4v, but it must have something to do with the way the switch works. I could reverse some wires to put it in order but I think that would just complicate things. The next step for me is to wire this puppy up and test out the audio. Then I'm on to panel drilling fun!
 

 

 

[cfcubed]

Sournds good Mullet - Yep, heater voltage WRT switch position can only be the installed order of resistors, their wiring to the rotary switch or the switch mechanics. 
Of course you may move the wires on the switch's tabs to get the order you want.
A couple voltages are a tad low, differing from mine, but are within reason I suppose.  It could be we are seeing variance in our DMMs or the resistor values, even something like a 0.2k variance in the actual switcher feedback resistor values could account for this.  We'll see if we need to better target the resistor values if others report their actuals.  Note sometimes the precise resistor we'd like is not available (e.g. 7.25k).

 

[Mullet]

I have a quick question about wiring going to to IG and G+ on my RK27 pot.  This diagram looks like it's calling for G+ coming from my RCA jacks to be both connected to G+ and IG on the pot. Also it looks like the IG wire going from the pot to the IG pad on the PCB needs to be connected to both contacts on the pot as well. Is this correct? Can someone clarify what this diagram is calling for?
 

 
 

 

[wiatrob]

Mullet,
 
All the input grounds get tied together. So, Left and Right input Ground can go to their respective (I might have labeled them IGL and IGR on the pot, but +G and IG in this diagram). Tie the pot ground pins together, then run a lead to IG on the PCB. DO NOT tie them to the output ground or case.
 
 
Bill
 

 

[Mullet]

Cool, makes a little more sense now.

Just figured that I'd post an updated pic of the current status of my Rev A CTH... Almost there... Perhaps, I'll be there tomorrow... Just need to wire up the pot...

 

[Mullet]

Well we finally made it... FTW! For the size and cost to build, this amp really packs a punch. My K701s aren't having a tad bit of trouble. Thus far I've popped in a few select tubes and I'm enjoying every minute of it. I'm liking my Amperex ECC82 Bugle Boy the best. Furthermore, I'm already thinking of putting my La Figaro 332C up on the chopping block. Thanks to everyone who made this possible -- runeight, cfcubed, and a few select others.
 

 

[cfcubed]

Looking good... Great its up & running and you're enjoying it...
 
>  Thanks to everyone who made this possible 
 
And you are one of them... Handling the GB, finding subs for the BoM, and almost having the first working Rev A (we'll remember that pesky R1H/R2H thing)

 

[wiatrob]

Great to hear! And my thanks as well for your participation!