[tomb]

 

 
There were really four tweaks that solved the hum problem:

• Use of the choke
• Zener diode tweak
• Replacing the heater supply schottky rectifiers with general purpose diodes.
• Installing snubber caps on the diodes in #3 above.

 
The last two were the latest and solved the hum issue from all reports. 

 

  We're now shipping with a V1.02 PCB that Dsavitsk designed.  Changes are as follows:

• Trace moved under the PCB that was impinging on one of the standoffs
• Allowance for zener tweak with normal through-hole mounting.
• SMD pads on the bottom of the PCB for snubber caps on the heater rectifier (general purpose diodes).

 

 
Wow, thanks for the great information.  A few more questions if you don't mind.  How many of these tweaks are included in a complete kit (I'm pretty sure I read the Zener diode and a Hammond choke are part of it)?  Regarding the scope of the project, aside from the SMD pads for the snubber caps, is the rest of the project through hole?
 
This seems like an exciting project for me, but I may need to tackle a smaller one or two (maybe a CMoy or Mini3) before I take it on, just to feel a little more comfortable.

All of the tweaks are included in the kit.  The new PCB recognizes this by incorporating all of them.  I'm not sure why there would be a question of doing otherwise if it makes the amp better, but maybe making that strong statement will make it clear.

 
The SMD for the snubber caps are trivial.  You can still solder leaded axial capacitors across the diode leads instead, if you wish - making it 100% through-hole.  If you haven't soldered parallel leaded parts before, trust me when I say the SMD capacitors are much simpler.  They're large pads, so 1206 SMD will easily work - or smaller, if you desire.  I'll spec some parts shortly and update the BOM when I get a chance.  For now, I've been including the leaded snubber caps in the kits.

 

[AnalogSavior]

 
All of the tweaks are included in the kit.  The new PCB recognizes this by incorporating all of them.  I'm not sure why there would be a question of doing otherwise if it makes the amp better, but maybe making that strong statement will make it clear.

 
The SMD for the snubber caps are trivial.  You can still solder leaded axial capacitors across the diode leads instead, if you wish - making it 100% through-hole.  If you haven't soldered parallel leaded parts before, trust me when I say the SMD capacitors are much simpler.  They're large pads, so 1206 SMD will easily work - or smaller, if you desire.  I'll spec some parts shortly and update the BOM when I get a chance.  For now, I've been including the leaded snubber caps in the kits.

 
Ha! I guess I was over thinking it a bit, but still thanks for the answer.  You strong statement has indeed put me at ease.

 

[dpump]

Wanted to post some comments about my experience with the Torpedo amps. I had one of the earlier amps with the Schottky diodes and no snubber caps. I did make the diode mod and replace the resistor with the choke. This amp, as has been stated by other users many times before, did have an audible hum on all phones. This is the hum that TomB and Dsavitsk struggled to eliminate, along with other builders who tried all sorts of shielding schemes. Nothing really seemed to work, so I ended up selling this amp.
 
When I read about the fix for the hum-using general purpose diodes and snubber caps-along with the discovery of the E90CC tubes, I knew I had to have another Torpedo. I had loved the sound of my first amp, just not the background hum. So, last December I ordered and built the newest kit and bought some E90CC tubes. Now I have the Torpedo 'sound' back, but even better. There is now either no hum or a very, very small amount of hum with all the phones I have tried. The only phone that I even really notice any hum on is the Beyer T90, and that is due to the high efficiency of that phone. There is a barely audible hum with the Sennheiser HD-800, but nowhere close to being audible if any music is playing. (Yes TomB, the new Torpedo is beautiful with the HD-800 with Charleston Cable Companies' top cable. The Torpedo is that good). Keep in mind that almost all tube amps have some background hum and/or noise-it's the nature of the beast. The new Torpedo is as good as, if not better than, most other tube amps in this regard. And as TomB has said, the E90CC tube takes the Torpedo to an even higher level than the 6J6 tube.
 
If you are thinking about buying a Torpedo but you're not sure you will like it, I can only say do not hesitate. The sound is sublime. It's like relaxing with your feet up with your favorite pair of well broken in slippers. The sound is so smooth and mellow but also very detailed without being 'in your face'.  

 

[AnalogSavior]

Thank you very much for your impressions! I'm pretty much decided on taking on the project, and was more looking into what to expect when I asked about the hum.  All the info I have gotten has been great. 
 
As soon as I have some time to put one together, I will definitely be ordering one,

 

[Punnisher]

I agree that the E90CC has reduced the hum to a nearly inaudible level. Any background noise in your natural environment will almost certainly be louder than the hum.
 
I still say that the Torpedo is one of the best amps (if not the best) I've ever tried, especially with dynamic headphones. You can't go wrong.

 

[dpump]

  I agree that the E90CC has reduced the hum to a nearly inaudible level. Any background noise in your natural environment will almost certainly be louder than the hum.
 
I still say that the Torpedo is one of the best amps (if not the best) I've ever tried, especially with dynamic headphones. You can't go wrong.

To be accurate, it isn't the E90CC tubes that reduced the hum. It is the replacement of the original Schottky heater diodes with general purpose diodes with a shunt capacitor across each diode. Anyone who has an older amp can do this mod to reduce hum, along with the diode mod.
 
When I built my current Torpedo amp I used ClarityCap SA coupling caps. Last week I replaced the SA's with ClarityCaps top of the line MR's. There was an obvious change for the better with a noticeable improvement in detail and a more top end. I had read reviews about ClarityCaps saying that the SA and SE were warmer sounding than the MR, and now I believe it. It's an expensive upgrade but I feel well worth it.

 

[Punnisher]

I noticed a large reduction of the volume of hum when I installed the E90CC over the stock 6J6. I can test it again to confirm but it was noticeable.

 

[tomb]

Yep. It's quite possible that a tube change can be responsible for a reduction in noise.
 
The E90CC is inherently a much, much quieter tube that runs at lower distortion. It also runs at a lower current than the 6J6 tube family.  This may put less stress on the power transformer and surrounding circuit, potentially reducing noise even further.   There are 6J6 tubes that are greater offenders than others, though, so the differences might be more severe in some instances.  I would agree that the primary noise reductions have been through the tweaks that we implemented.
 
That said, the amp simply sounds better in every way with an E90CC.  Noise reduction would be part of that equation, regardless of the specific cause.
 
I guess I'm agreeing with you both, if that's acceptable.

 

 

[ChrisX]

I just finished assembling my Torpedo PCB. At first startup I have seen that the DC voltage at C11/C12 varies between 130 V at one channel and 145 V at the other channel. After exchanging the tubes the voltages followed with the tubes. Would this measurement be a good indicator for tube matching or is this voltage reading unimportant?

 

[dpump]

Here are pictures of my Torpedo with the ClarityCap MR 4.7 mFd capacitors I recently installed.
 

 

[jboehle]

You should install parts with the printed values facing up, so when someone has to repair / mod it in the future, it's easy to see what's what.

 

[dpump]

  You should install parts with the printed values facing up, so when someone has to repair / mod it in the future, it's easy to see what's what.

I agree. However, the leads were already bent from the factory and I didn't want to twist them around to keep from fatiguing the metal. I read of one person who did that and one of the leads broke off. Didn't want to take a chance on that.

 

[tomb]

 

  You should install parts with the printed values facing up, so when someone has to repair / mod it in the future, it's easy to see what's what.

I agree. However, the leads were already bent from the factory and I didn't want to twist them around to keep from fatiguing the metal. I read of one person who did that and one of the leads broke off. Didn't want to take a chance on that.

I normally always recommend the same thing, but with parts that big - I don't think it matters.  One other thing to consider - even more important - is that you can ruin a film cap like these if you try to rotate already bent leads to get the labeling to face up.  It's possible to torque the leads right out of the film capacitor spiral inside, effectively ruining the film cap.
 
Here's what one looks like on the inside:


The leads are soldered directly the foil/dielectric roll.

 

[tomb]

  I just finished assembling my Torpedo PCB. At first startup I have seen that the DC voltage at C11/C12 varies between 130 V at one channel and 145 V at the other channel. After exchanging the tubes the voltages followed with the tubes. Would this measurement be a good indicator for tube matching or is this voltage reading unimportant?

Dsavitsk would have to answer that, but my guess is yes.  If they're my tubes, please let me know and I'll replace them with something better matched.

 

[dsavitsk]

To be accurate, it isn't the E90CC tubes that reduced the hum. It is the replacement of the original Schottky heater diodes ...

With the E90CC the heater draw is lower which puts less stress on the power transformer which then emits less noise. But you are right that the different diodes matter, too.

I just finished assembling my Torpedo PCB. At first startup I have seen that the DC voltage at C11/C12 varies between 130 V at one channel and 145 V at the other channel. After exchanging the tubes the voltages followed with the tubes. Would this measurement be a good indicator for tube matching or is this voltage reading unimportant?

We tend to use DC matching as a proxy for AC matching, but who knows whether this is a reasonable thing to do for any given tube. The easiest way to check yours would be to play a 60Hz tone and measure the AC output from both channels with a multimeter (even cheap meters can measure 60Hz reasonably accurately.) That should give you an idea how mismatched your tubes are. And truth be told, tube matching is not that important here. They'd have to be way way off be before you'd hear much of a difference.

I agree. However, the leads were already bent from the factory and I didn't want to twist them around to keep from fatiguing the metal. I read of one person who did that and one of the leads broke off. Didn't want to take a chance on that.

I ripped a lead out of an Auricap once doing that.