ECP Torpedo III [indexed in first post]

The spacers are intended to prevent any possibility whatsoever of the leads coming into contact in the center. This would be the back row of caps contacting the front row of caps in the center between the two rows of caps. There, the voltage is still very high and could potentially cause the most damage.

Doug felt my original attempt at electrical tape was not sufficient. Even if you use heat-shrink or teflon tubing, there is still the possibility that the center intersection where the lead exits the cap cannot be insulated. (The elbow bend stays bare, iow.)

The use of the spacer, along with an initial bend parallel to the PCB, will prevent any possibility of this happening.

Here's a pic illustrating the spacer and its use in the middle where the caps meet, front to back:

A link to a scale (I think) PDF you can print out:
http://www.diyforums.org/Torpedo-III/tweaks/Mundorf/spacer-glue/cap-spacer.pdf

Finally, the Dow Corning 748 sealant was selected by Doug specifically as a non-corrosive caulk used in electronics. You need some of this to "glue" the cap arrangement together, mainly so that the spacer doesn't fall out. This is a pic of an amp that I sent to Doug that he modified:



Those with a keen eye might catch a glimpse of the output CCS boards near the bottom.

Anyway, I apologize for not documenting this sooner, but the only change is the 90 degree bend parallel to the PCB, first. Then the rest continues as before. I would suggest to get your first bends done, put two caps together (rubber bands or tape) with the spacer fitting in between, then continue with the rest of your bends to line up with the holes.

If you can't fit the spacer in between two caps held together, then your bends are wrong. It won't let you do them any other way.

 

Ah, I wondered what the silly putty was for also, but with electrical tape completely covering the lead outlet, things seem pretty safe. Extra insurance never hurts, of course.

Agreed that any sort of heatshrink without tape still leaves the initial bend exposed. Even if you cut a half-round extension which extended above that initial bend, I'd worry. So I hadn't considered heat shrink sufficient here. I had thought about a thin piece of balsa wood or something as a shim between them. It seems any sort of shim that prevents them from ever coming in contact would work regardless of precisely how it's positioned.

 

Just wondering which Z-setting does everyone prefer to use and with which cans? So far with my modded 650's the high-Z setting seems fuller and the low-Z seems more detailed. I find myself switching depending on the recording and mood, though in the past with different amps i had a strong preference for low output impendance with 650's, 600 i prefer with high-Z.

 

hi z for high ohms, low z for low ohms.

 

I think i get it...
Edited - massaged the leads to get it aligned better. Looks nothing like the pdf bending pattern now.

 

You do know it's just an output impedance toggle right? The high-z labeling is just so people don't use efficient planars or other low-z headphones on the high-z setting and get fart bass

 

yeah

 

Not sure if I understand what you mean by this, but just to clarify: The switch toggles between two different tabs on the secondary winding of the OPT, see the schematic. http://www.diyforums.org/Torpedo-III/schem/TorpedoIIIschem-1-Eagle.png

So the switch is not there to add resistors in series with the output to create a high output impedance, if that's what you're referring to.

 

I thought there was a difference in output impedance between the two? That's the way it sounds to me, now how that is achieved i figured it had something to do with the OPT's yeah. With the difference in gain between the two, are there more differences like voltage swing between the two? Was i wrong and should i actually use the high-z setting for my senns?

Can't read schematics yet

 

There are two secondary transformer taps selected by the toggle switch. I don't know what the difference in turns ratios are; couldn't find a data sheet for this Cinemag OPT. But no parallel or series resistance, as MortenB says.

 

I asked that same question before over at head-fi though.

 

Interesting, i think i read Doug mentioning that there was a difference between the two on his L-2 amp, i kind of assumed it was the case here as well.
Thanks Jun, haven't gotten around to reading the HF thread yet. Still goes above my head a bit though, i can solder pretty decently and follow instructions that's where it ends for me haha

 

The Edcor OPTs are much more documented than the Cinemags, but until Doug says otherwise, I would assume the same windings for each (accounting for the nickel core differences in the Cinemags).

The Edcors have a choice of either 7K or 10K input impedance from the tube output (differential buffer output in the T3). However, for all the Torpedo versions, it doesn't look like Doug used anything but the 10K on the input side. On the output, it's a choice between 300 or 32 ohms output impedance. At that point in the circuit, the connection is directly through the Z-switch to the headphone jack, so I don't know any other way to interpret that except to mean either 32 ohms output impedance or 300 ohms output impedance, period.

That said, Doug has stated from the beginning - try out the switch either way you want with whatever headphone. Whichever sounds good to you, forget what the values mean.

 

There is a difference between load impedance and output impedance. The transformers are roughly 10K:300 or 10K:32 depending upon setting. That means that a 300R load on the 300R setting will appear to the circuit as a 10K load (same for the 32, but substitute 32 for 300). A 32R load on the 300 setting will appear as a roughly 1K load to the circuit, and a 300R load on the 32 will appear to be close to 100,000 Ohms. Neither of those is really ideal - in the first case you may have insufficient current to drive the phones, and in the second you'll lose bass.

From the other direction, the output Z of the circuit is very low, and when translated down by the winding ratio is very very low. So only the copper matters. Thus the output Z on the 32R tap is not 32R, it is something less than 10R (I don't have one in front of me to measure), and the Zout on the 300R tap is lower than 300 - probably about 20.

 

Did you get the answer you needed from the explanations..!? Sometimes we tend to get a bit technical. Let me try to make this a bit clearer (hopefully I don't mess it up and make it sound complicated anyway).

Output impedance =
The output impedance of the amp. In this case the output impedance is quite low (10 - 20 ohm) on both the low and the high setting as described by dsavitsk.

Load impedance =
The impedance of the headphones.

For this amp the LO and HI settings of the switch refers to the load impedance = the headphones. When using high impedance cans like your Sennheisers and the HI setting of the switch the circuit is working at it's best. When using low impedance cans and the LO setting of the switch the circuit also work at it's best.

With other combinations like LO setting for high impedance cans, or the HI setting for low impedance cans the circuit is not loaded correctly and is not working at it's best. You get insuficient current to drive the cans or lack of bass as dsavitsk mention above.

Does the switch make sense now?

 

Yeah i get it now thanks guys. This also explains what i was hearing and it makes sense

Think i'm gonna look into some other caps, are you still selling your Mundorf's @MortenB ? They are a bit thinner sounding then the Solens right?

Also anyone know how close the EH 12au7 and 12ay7's are sound wise except for gain ofcourse? Might stick to cheap tubes and just use this amp normally and don't go all ocd over expensive tubes.

 

I have not tried the Solens in the T3, so I can't really comment on the difference. A few pages back TomB talked about them compared with the Mundorf, so look back in the thread.

From experience in other circuits I would say the Mundorfs are leaner/tighter in the bass, more open and a bit more forward. And the Solens fuller sounding but less transparent and might be a bit grainy...

My Mundorfs are retired, so yes I will be happy to send them to Holland Let's work out the details by PM if you decide to want to try them.

 

Thanks for explaining, Doug! I shouldn't have referenced the Edcor data sheets without understanding more, as usual.

 

I would say that's a pretty accurate description of the differences. Referenced to the Epcos, I think the Solens are closer to the Mundorfs than the Epcos box caps. I'm not sure I hear the grain as you describe with the Solens, but I definitely hear it with the Epcos. While slightly lean, the Mundorfs are definitely more transparent and you will hear more detail. Early in the thread, before anyone tried out the Solens, the original comparison was between the Epcos and the Mundorfs. The difference there, IMHO, was immediately and obviously apparent: the Mundorfs were superior in every way.

I've been very familiar with the Solens since the days of the Torpedo I where again, they were the standard supplied for both kits and assembled versions (and still are!). They also work in the Torpedo III, sound pretty d*mn good, and are ridiculously easy to install compared to the Mundorfs. They are also fairly cheap and while not available at places like Mouser/DigiKey/Newark, they are readily available at Parts Express, Madisound, PartsConnexion, Antique Electronic Supply, The Tube Store, etc., etc. So the decision from my perspective in supplying the kits was relative to the Epcos: the Solens became the standard offering.

I still think the Mundorfs are the preferred option for a high-quality cap that can be shoe-horned into the Torpedo III. If that wasn't the case, we wouldn't have gone to such extraordinary lengths in detailing the installation and providing additional safety tools.

 

Cheap tubes for me, too!

JMHO, the T3 is still fundamentally a tube-driven amplifier. Changing the tubes is still going to have the most effect on the sound signature.

Just as a general rule, the EH tubes seem to be well-detailed, but a bit strident. I still think they are a good standard for the Torpedo III, because they immediately demonstrate the detail of the amplifier. They also test well in distortion and flat response. Finally, they are an assured new-manufacture source. One can't say that for NOS tubes - especially in volume. Even with the ready availability of the 12A*7 family of NOS tubes, you need to purchase a few hundred to ensure an adequate, consistently reliable stock for matching, etc. So, supplying the EH tubes as the basic standard was the better choice.

There's always the basic NOS 12AY7 you could try. With a good pair, they should be more refined than the EH's, but probably not as detailed. They're still somewhat pursued as "audiophile" tubes, so you won't see that many price breaks, I think.

IMHO, the 12AT7 family of tubes should be among the first tubes you try besides the EH 12AY7. I think perhaps the gain is so low on the 12AU7's that they end up flat-sounding, even if you match gain correctly. There are a lot of variants within the 12AT7 family - notably, the European E180CC/7062 tubes, which many seem to prefer here on SBAF. However, they're pretty expensive at a minimum of $40/pair, but often much higher. I tried a pair of Amperex 7062's and they sounded a bit thin/lean, especially when paired with the Mundorfs. That could just be my pair, though.

The basic 12AT7 is much more tubey sounding, even in the EH versions. My EH pair are almost comical in their tubey-ness. The EH12AT7 is a pretty wild tube to put in the T3. However, there's the 6829, which is sort of a US version of the E180CC. At least compared to my Amperex 7062's, the 6829's have a much more prominent bass and overall tubey sound. For some reason, the similar 6414 tube does not sound as good and they are often much more expensive because of their use in Manley amps, I think. There again, it may be just because I got cheap ones that aren't so good.

The 12AZ7 continues to be the real sleeper, IMHO, and has been my preferred tube for awhile, now. It's very clean sounding (low distortion) and seems to simply disappear, letting you listen to the music instead. They don't exhibit the "audiophile price growth" of the 12AT7 and 12AY7, but are somewhat scarce and more difficult to find than other 12A*7 tubes. Also - at least with the testing I did - they seem to favor high-impedance loads instead of low ones.

The 12AV7 can be found very, very cheap (especially the 5965). I got a pair, but was not impressed. Mine were not matched well and I don't trust the quality of the tubes I got. So, I would be interested in someone experimenting with more of those. They might be a great low-cost alternative. Or, it could be my pair is representative and they're not good. One pair is not enough to know for sure.