[Mullet]

BK~ that is indeed a nice looking build. Your bro is gonna be stoked.

Im tempted to do something similar to civilmonkey with regards to switching the R18 resistors. The only difference is that I'd most likely not use SIP sockets because I'd be only using 100 and 150 ohm resistors. I still haven't decided on whether I'm doing this or not.

The other thing I'd want to do is use higher rated L1H and L2H so I can use a 6n6p tube among others. If anyone has spares or wants to trade for the BoM versions please PM me.

 

[Halvor]

Finally I'm getting around to build my CTH, so I have a couple of questions.
 
- How large transformer can be used before you can't tell the difference?
 
- I will probably have plenty of room left in my case. Will this amp benefit from bigger/additional heatsinking?
 
Thanks

 

[civilmonkey]

No one has weighed in yet, so here are my thoughts.
 

- How large transformer can be used before you can't tell the difference?

Use the BOM 24VAC adapter.  The price is reasonable, and it's sized to work with the CTH.  I don't think a larger transformer will yield any gains.  Most people report (myself included) that with most headphones there is no hum or hiss.  Sensitive phones (e.g IEMs) do have some hiss.  With my DT880's (very hard to drive headphones), there is more punch and volume than I personally need.
 

- I will probably have plenty of room left in my case. Will this amp benefit from bigger/additional heatsinking?

My understanding is that typically, the cooler some components run, the longer the life.  How much difference there is I don't know.  I'm pretty sure I have seen on some (not all) capacitor electrolytic datasheets estimates of life and discussion of reduced life with elevated temperature.  Think laptops, they burn out much more frequently than a desktop computer which typically has better cooling.  Still, saying that, I don't think I have come across a post of anyone with a failed CTH due to temperatures, but I could be wrong, the original thread is looooong, and I might have missed bits here and there.

 
 

 

[cfcubed]

If you stick with target tube draws (< 600ma) then IIRC the amp draws 0.6A -> 0.75A (lower side for usual 300ma heaters), and you need not heatsink the LM2595.  The 24V reg does get hot & must be heat-sinked. And there should be adequate case ventilation.  So 0.8A+ 24VAC adapters are fine, unless you are shooting for higher-draw tube heaters...  In that case 1.0A+ 24VAC adapter (BoM one) is better & a small heatsink on the LM2595 makes sense.
 
The amp's draw is relatively constant so higher-current adapters should not yield a difference/benefit (but will not hurt anything either).
 
If you had plenty of room & felt like doing it a larger-than-spec'd heatsink on 24V reg (or two BoM ones back-to-back on the reg) could be considered.  As well as using the small heatsink on the LM2595.  But again, for the recommended use cases this is not necessary.

 

[Mullet]

I agree that the BOM 24 VAC adapter is the one to go for. There are a few tweaks that can be found here as well as the "Cavalli Compact Tube Hybrid (CTH) Tube & Tweak thread" that enhance the amp. The first and foremost are different C4 bypass caps. There are several out there that people have tried. I'm going to use those green Russian paper in oil caps for my next build. Then there is the option of using a switch and possibly SIP sockets to use different R18 resistors for headphones of varying impedance. Civilmonkey did this tweak. Finally there is the option of using higher rated amperage on your L1H and L2H to allow for the capability of using a Russian 6N6P or 6N30, which draws more than the recommended current (600mA) at 750mA+. It looks like cfcubed is saying you can get by with the BOM ac adapter for higher current tubes. The question is do you really even need to upgrade L1H and L2H for this? I've read that some builders in the past upgraded and I think some didn't and still used the 6N6P without a problem. The definitive answer is that you have to have ICH heat sinked when doing this.
 
With regards to heat I'd say just drill a ton of holes in the case. A few on top near ICP and ICH. Then maybe 6 to 8 on each side and finally one or two on the bottom. Of course, if you can source some perforated metal then you don't need to drill any holes other than the tube hole. I found a metal hole punch works best for this. The only caveat is they're quite expensive and you need to have the right size for your hole.

 

[Halvor]

Thanks for the answers guys! Seems like there is no need for "overkill". But I would still like to stick with a toroid and have everything integrated, rather than a wallwart. So one rated 1A and 24VAC would be fine?

 

[cfcubed]

>  It looks like cfcubed is saying you can get by with the BOM ac adapter for higher current tubes.
 
Yes, again IIRC from my CTH current draw measurements from 2 yrs ago or so.  It is a simple matter for builders to use a panel mount jack & measure their CTH's adapter draw under various conditions.  Just make sure your ammeter can deal with 2A+ to handle any instant/power-on draw spike.
 
>  The question is do you really even need to upgrade L1H and L2H for this? I've read that some builders in the past upgraded and I think some didn't and still used the 6N6P without a problem. The definitive answer is that you have to have ICH heat sinked when doing this.
 
IIRC BoM coils are rated for 800ma and its always good to have headroom.  That said my CTHs had no issues/troubles driving 750ma heater loads (w/little ICH heatsink as Mullet says), but doing so w/BoM parts is still not recommended.
 
>  But I would still like to stick with a toroid and have everything integrated..   So one rated 1A and 24VAC would be fine?
 
From above that rating should be fine esp. sticking w/target tube heater loads.  But experience & paranoia has me making all my builds two-box ones, providing separation between transformers and input stages/low-level stuff.  A/C adapters are a nice cheap way to get this two-box build advantage.  Anyway, if doing one-box build for this I'd be ready to allow 6" or more between transformer & input stage, use good wiring practices and test/experiment listening with transformer position/orientation/rotation prior to pining things down.

 

[Halvor]

The plans have changed in regards to my choice of case, so a wallwart might be considered. But i have yet to find one with the recommended specs in my country. Does anyone know where to find one with 230V primary?


> Anyway, if doing one-box build for this I'd be ready to allow 6" or more between transformer & input stage, use good wiring practices and test/experiment listening with transformer position/orientation/rotation prior to pining things down.

If I don't find a cheap wallwart in reasonable time, I will use a separate case for the transformer.

 

[Mullet]

So like my first CTH build what fun is it all without any problems?
 
I stuffed all my parts, checked soldering joints, and then cleaned the board with isopropyl alcohol. Then I proceeded to wire up my power to check voltages et al. First thing I noticed is that the tube wasn't working. I tested pins 4 to 5 -- no voltage what-so-ever. Tested everything else. SG to OG 11.92v, SG to OL and OR same readings as SG to OG. SG to Pin 1 -- 107v, SG to Pin 6 on the tube -- again 107v. SG to TB+ -- 107v. Then I checked the LM2595 and I'm getting the normal readings of 33v in and 12.7v out. The only things after this in the circuit that aren't making sense are the L1H is getting 12.7v between SG and it's input, but no reading between SG and its output. Could the inductor have been bad from the get go? I tried to reflow some solder to no avail.
 
Also, I think I might have semi-fried Q3P. When I was poking around a ring on my finger touched the board (this was two minutes after unplugging the thing) and I heard a spark pop. Now I'm getting a 97v reading on Pin 6 of the tube instead of 107. Is there supposed to be a 10v voltage drop from Q3P? Looks like I'm going to have to do another Mouser order. Any one have a spare Q3P and L1H on hand?
 
What else should I be looking at readings wise to make sure I'm covering all bases?

 

[civilmonkey]

Has anyone tried the knock off 50K log DACTs from ebay?  I have one while waiting for a 50K Alps pot.  My experience so far is that they sound fine, the attenuation curve seems pretty reasonable, channels seemed to be matched well, and build quality is so/so  (for $10 it's just fine).  What I am experiencing though is that as I switch volumes, the click often causes the amp to unlatch for 1-2 seconds.  turning the know slower to eliminate the click reduces this, but does not eliminate it.
 
Can anyone confirm?
 
EDIT:  The amp has warmed up (10 minutes of music) and I can no  longer recreate the above.

 

[civilmonkey]

Quote:

 Any one have a spare Q3P and L1H on hand?

A little late but, I have one QP3 but no L1H.  PM you address if you want, I can mail it out tomorrow.

 

[wiatrob]

Quote:

Has anyone tried the knock off 50K log DACTs from ebay?  

The Gigaworks? I plan to use one in my build, I have the case finished with the stepped attenuator mounted (but I haven't built my rev'd CTH yet).   Any further symptoms?
 
 
 

 

[cfcubed]

Quote:

  What I am experiencing though is that as I switch volumes, the click often causes the amp to unlatch for 1-2 seconds.  turning the know slower to eliminate the click reduces this, but does not eliminate it. EDIT:  The amp has warmed up (10 minutes of music) and I can no  longer recreate the above.

I have a guess.  I've found that if a lineout-connected iPod is turned on (wakened causing Apple logo) with CTH on & volume set it often causes e12 to cycle.  Turning CTH volume down avoids this.
Thinking CTH is very sensitive to input offset / input offset transients.  Don't know the make/break behavior of that DACT but perhaps if you scoped what is happening when you turn it you might see spikes.  With input connected/etc so you are simulating what is being seen by the CTH input stage.
Also think the particular C4s used may have influence on this, BoM/box caps exhibiting it less / getting earlier, more resilient lock than large, fancy caps.
 
Having never owned / used a stepped attenuator, I'd be worried about granularity when most never go past 11am on std pots in CTH.  E.g. will I get the precise volume I want - something I'm quite picky about.  Thinking amps w/adjustable/configurable gain may be better suited as you can set/use more vol range, but again don't know much about stepped attenuators.

 

[civilmonkey]

 

The Gigaworks? I plan to use one in my build, I have the case finished with the stepped attenuator mounted (but I haven't built my rev'd CTH yet).   Any further symptoms?
 
 
 

No other symptoms.  When the amp is warmer, it seems to affect the latching less.  It still occurs occasionally but I think last night I decided I could live with it.  I'm using non bom caps.  Unfortunately ebay links die quickly, but the dact I used was $10 from Hong Kong, 21 step gold plated.
http://cgi.ebay.ca/DACT-Type-21-Stepped-Attenuator-Potentiometer-50K-300B-/200430354774?pt=LH_DefaultDomain_0&hash=item2eaa948156 

 

I have a guess.  I've found that if a lineout-connected iPod is turned on (wakened causing Apple logo) with CTH on & volume set it often causes e12 to cycle.  Turning CTH volume down avoids this.
.

My source is an IPOD with homemade line out, but I have only connected pins 1 to 4 (not the 21 pin to ground with a 68kohm resistor).  I don't experience this luckily.
 
 

Having never owned / used a stepped attenuator, I'd be worried about granularity when most never go past 11am on std pots in CTH.  E.g. will I get the precise volume I want - something I'm quite picky about.  Thinking amps w/adjustable/configurable gain may be better suited as you can set/use more vol range, but again don't know much about stepped attenuators.

All good points.  I personally find I have enough control with the stepped attenuator but everyone is different of course.  With my DT880's I get to around 1pm on the volume knob.  Anything higher risks the amp unlatching from the load and is also pretty darn loud.  Aside,  I've found that it's not necessarily bass heavy music which will cause unlatching, rather music which is very 'full', i.e., lots of mids as well as base.  If you are familiar Alexi Murdoch - Orange Sky, that song will cause unlatching more than any electronic, heavy bass, etc.
 
One nice thing I could say is that I can use a very low volume and still have the volume balanced in both channels.  I have never tried the alps 50k log which I understand is supposed to be pretty awesome, but other pots I've tried (the small green 10k alps for instance), usually don't balance channels at low volumes as well.
 

 

[rds]

Quote:

Having never owned / used a stepped attenuator, I'd be worried about granularity when most never go past 11am on std pots in CTH.  E.g. will I get the precise volume I want - something I'm quite picky about.  Thinking amps w/adjustable/configurable gain may be better suited as you can set/use more vol range, but again don't know much about stepped attenuators.

@civilmonkey
Another thing to consider is whether a DACT makes sense in this application.  Considering the CTH uses a single dual triode tube you will inherently have some degree of mismatch in gain between the two channels.  Using a volume control that's intended to track two channels very precisely might not have any advantage in such a circuit.
I'd suggest a nice pot as the best choice in this application.