ECP Torpedo III [indexed in first post]

Two things to add to @bazelio's excellent writeup:

1) The sound differences between the best sounding circuit (T3(C), in my opinion) and the least awesome sounding one were more noticeable than the sound differences between the group's top 3 tube choice. That said, once I really zeroed in on what to listen for, there were significant differences between the tubes, and it's nice to have so many good options for a single amp.

2) The T3 is a hell of an amp. I was secretly hoping to walk away 100% satisfied with my Night Blues Mini, but that did not happen. The NBM is a really nice amp, and an incredible value, but the T3 is in a different league in terms of technicalities like resolution, dynamics (a big one for me), and engagement factor. Again, the NBM is quite good in all these categories, but the T3 is flat out great.

I was unsurprised that we didn't favor the 800s over the 650s, but I dislike a lean-sounding headphone no matter how capable it might be in other departments.

 

I have mostly GE's in the 12AZ7's. I've tried a pair of Dumonts, which are usually re-branded RCAs. They sounded pretty good, too, but were terribly microphonic. I tend to pull tubes out that are like that, because I'm never quite sure whether I'm hearing the tubes or some sort of mechanical effect.

Looking at those Rike caps, I agree with Morten that drilling holes in the PCB would be best. That is pretty drastic in and of itself, but as I noted in a conversation offline, box caps that big are designed to sit flush on a PCB. Most come with tabs on the bottom to facilitate this (look at the Epcos for an example). Caps like the Mundorfs are different, even if we've all shoe-horned them into the T3. Axial, cylindrical caps are built to be supported by a small contact area or simply suspended by the leads in a point-to-point arrangement (maybe with a cable tie).

 

@TomB would you typically mount box caps totally flush on a PCB, or leave a small air gap (or shim with something) to provide room for solder to wick through? Tabs probably accomplish this but as I recall, there are no tabs on these like Epcos. I was also thinking they are best suited to point to point in larger chassis's where they can be secured to a surface off the PCB or turret board within the length of the leads. I didn't get the impression these are PCB-mount caps.

I finished up this afternoon. I had some neoprene (I think) sheet material that I used as shims between the PCB and bottom cap surfaces. I had left a little slack in the leads when soldering to facilitate the shimming. And now there's no play whatever after removing the electrical tape jig that held them in place during soldering. So now, essentially, the teflon insulated leads run in "channels" under the caps. I also couldn't totally avoid filling the gaps in order to make them wiggle-free, but the neoprene shim in the gap is only at the very top corners, and again, might not matter anyhow. Drilling for bringing the leads through the under side is indeed a good idea, and doesn't sound too difficult.

 

Yes, that's it... Big box caps or any other big and heavy component should be flush on the PCB and securely installed mecanically. This also reduces resonances, and when I contacted Rike Audio they were quite specific on resonance control for best sound with these.

In my young days I worked in R&D in a company that manufactured electronics. We learned this the hard way by failure rate going up, up, up when our designs did not ensure proper mechanical coupling between the components and the PCB. Never air gaps, never something soft under the components that could leave them vibrating, and use cable ties or glue to hold the bigger components.

 

Yeah, insecure installations are to be avoided, but we aren't saying that because of a box shape, it was therefore necessarily intended as a PCB flush mount are we? Understood on typical radial vs axial, but my Rike caps have no support tabs underneath. PCB "flush" mounting of other types of box caps would create PCB contact only at the raised perimeter corners on the under side. On those, there is in fact a supported gap such on the Epcos. This supported gap can also be created with perimeter shimming using material of appropriate hardness which also provides space for the lead wires. Rubber damping material, for example, is available in various durometers and need not be soft.

Edit: BTW, I find "best practices" discussion both useful and appropriate in DIY amp threads such as this... Hope others do as well.

 

I don't know how mechanical stability is tested in R&D and producion these days, but 25 years ago when I was involved in R&D of electonics we had this machine to shake a product. A machine quite similar to the machine they have to mix paint at (what's such a shop called in English?) the place you buy paint for painting your house. It shakes like crazy...

We inserted the product into this machinend and it had to ''survive'' for XX seconds or minutes without components breaking loose. Any heavy component that did not have a positive strong and flush contact with the PCB would be ratteling around inside the product after a short time. This positive contact can be made in several way: With caps like the Epox that has a raised perimeter, but still a positive and strong flush connection. It can be made with stand offs like with the transformers in the T3. It can also be done with gluing larger areas of a component to the PCB and/or to other components. Like when we see glue between electrolytic caps and/or between cap and PCB. Or in this case when gluing the Rike box cap for a positive, solid and flush contact on one of it's sides. The same should be done with the Mundorfs as it's also shown by Tom and Dsavitsk: Glue between them, and glue between cap and PCB. Different ways to get a positiv, solid and flush installation. Such installations would survive the shake test (the Mundorfs would not without also securing them with cable ties). Anything with air gaps or soft materials between the components and the PCB would rattle loose almost instantly. We thought that we could use such softer materials and a less direct connection to the PCB to prevent stress on the components. Reality proved us wrong.

Why is this relevant for a DIY amp sitting safely on a shelve at home?

Because the same principle still apply and we tend to get bad solder joints and component legs breaking over time if the components can vibrate or move around. It's also a safety hazard where high voltages are involved. Actually some also believe, that vibrating components should be avoided in audio because it can degrade the sound.

I always try and follow these relatively simple principles when doing DIY projects.

What I wanted with my installation of the Rike caps was a positive and flush mechanical coupling between the caps and the PCB on a large area to make it mechanically stable, and I wanted the legs of the caps not holding anything to prevent stress on the legs/solderings. I consider this the best possible installation of such a component. So yes, there is a large area of mechanical coupling, but not between the caps (this is a soft contact), the mechanical contact is between the cap and the PCB.

 

@TomB: Now that more positive feedback comes in on the Amtrans plate load resistors, maybe it's time to try them out, even if you guys are skeptic? I feel they are so important for the potential of the T3 that they might be worth for you to consider as options when purchasing the T3, similar to the options with Mundorfs, CCS, Cinemags etc.

Just a thought

 

MortenB had sent me a private preview of the Rike caps before having much listening time, and I haven't read his later impressions that were posted a few pages back yet. I will. And since I've let the cat out of the bag locally, will give my $0.02 on the Rike S-Cap 5.0 uF here....

I forgot to weigh them but it seemed in-hand, the Rike 5 uF caps were noticeably heavier than the 10 uF Mundorfs. Installation notes are obviously contained in the past several posts. Secure the caps well, by some method, and be done. LOL. And the sound: I feel these are simply a more transparent capacitor than the Mundorfs, meaning tube characteristics are more clearly expressed and the amp is a bit more responsive to or revealing of changes in the chain. E.g., the "bright tubes" are a bit more clearly differentiated from the "warm tubes" than before, etc. If that makes sense. And I think, as I mentioned already, I might have a different favorite tube preference if we were to redo the tubeup after the cap upgrade. But I need a long break from tube rolling now fo' sho'. The tonal balance of the blue Valvo 6201 in my T3 at this moment is, to my ears, fantastic. BTW big thanks to @shaizada for turning me on to this tube some time back. He probably doesn't even remember doing so. Now, price-to-performance of Rike vs Mundorf: I'd say it's lacking. On the one hand, I wouldn't revert back to the Mundorfs. But on the other hand, I also can't suggest everyone run out and do this to their Mundorf-based T3s. I have just really enjoyed tinkering with this little amazing amp, and can't seem to leave well enough alone! But Mundorf to Rike is definitely not low hanging fruit.

Edit: one more note. If you are in the USA and purchase these caps, a VAT surcharge is deducted at checkout if you use the Latvia dealer. This brings the cost down quite a bit.

 

The Blue Valvo 6201 with the Windmill Getter is probably the best 12AT7 type in existence. You found a pair? PM Coming!

 

Great review of the Rike S-Caps Bazelio, and it seems your impressons on them are much in line with what I hear. They are more like no capacitor at all...

Here is the website again to the seller in Estonia (I have linked to it earlier). It seems though, that his stock is running dry...
https://www.audiohobby.eu/en/5524-rike-audio-s-cap-paper-in-oil

For buyers from outside the euro zone 20% VAT is deducted from the 47 Euro price.

 

I realize now how lucky I was... I don't see them anywhere at the moment!

 

I found one for 300 euro's, fuuuuuuuuuuuck that shit.

 

As far as I know, it's simply called a paint shaker.

 

There you go. Thanks..! This is good practice of my English skills

 

I'm not totally sold, but just in case - I'm going to try some Kiwame resistors and see how they sound. They can be had for 46 cents each. They're a favorite of Pete Millett's and Doug has used them many times, too. Maybe they can be a cheaper, easy-to-source alternative to the metal or Amtrans.

 

Whether they work or not, my skepticism of the Amtrans is that they are an off the shelf $0.15 resistor in a tube being sold for $6.50. The investment necessary to design and produce a new resistor is extraordinary. I have my doubts there could possibly be a large enough return in the audio market, even at $6.50, that anyone would make that investment. If anything, my suspicion, assuming they are better in a way detectable in a blind test, would be that the epoxy keeps them from resonating. In which case, I'd suggest buying some epoxy and coating whatever other resistors you have laying about. On the other hand, if you are comfortable paying someone $6.50 for a metal tube and some epoxy, then maybe it is a fine investment.

 

In my opinion there are different perspectives on this...

I'm also an electronic engineer, and from that perspective I understand and accept the perspective that any resistor is a cheap component, and that it's unacceptable to put this component into a metal tube, add gold plating on the legs and sell it for $6.50. For this alone such a product is doomed. Years back I would not even consider such a component, and if I listened to it my skepticism and negative approach would mean I would hear what I wanted to hear = I would want to hear, that it sounds like any other off the shelf $0.15 resistors.

But there is also another perspective. With shipping, import taxes, upgrade transformers, Mundorf caps, NOS tubes etc I have invested almost $1500 in my T3, it's no small investment, but I was still comfortable about it. If a component costing $6.50 improves the sound more than anything else, it's a very good investment in itself, and a protection of my $1500 investment in the product. Especially since the improvement in sound is bigger than tubes or caps costing much more. To me, this makes these resistors the best possible investment for the T3 and I would share this find with other T3 owners.

I actually started my plate load resistor experiments with cheaper and well respected carbon resistors. I was sure they would be as good as anything else (the engineer in me). To my surprise they did very little for the sound, and I was equally surprised that the Amtrans made such a difference. I have shared openly in the thread, that I found this hard to believe. Others hear something similar, latest in a direct comparison between different spec T3's. I found that latest comparison quite convincing and I wish I could have been part of that comparison.

But I hear what you're saying TomB and Dsavitsk, and as mentioned from the beginning of this post I understand and respect ''the engineers skepticism'' and why it rules these resistors out, also if you actually try them. It was a mistake of me to suggest you offering them as options when purchasing the T3.

 

I wouldn't take my skepticism personally. I just have not experienced these parts myself, and I am not ready to believe in their value until I do. I have certainly experienced lots of seemingly odd things make a difference. I will say that the $6.50 price troubles me as being, as mentioned, too high for an off the shelf part, but also way too low for a newly developed part. I find it puzzling in that regard.

Anyhow, offering parts as a part of a kit upgrade is an entirely different proposition. It has to pay for itself. So unless people are willing to pay us a markup to stock these, or we are able to secure them at a steep discount, it doesn't make sense on our end.

 

I should add that we appreciate the level of experimentation going on. It is great to see just how far people have taken this amp, and we hope to see even more going forward. (So far as I know, nobody has played with changing the output transistors yet.)

 

You focus on the part Dsavitsk, I focus on the sound quality.

In an ''improvement of sound per $ invested'' these over priced resistors are exceptional value to those of us who have tried them and listened. This spot in the circuit seems to be really critical for best sound.

I also found the Amtrans a much better investment than other well respected (and cheaper) carbon resistors I also tried.

I realized now, that these are ruled out for you and TomB without listening. It's your amp, and your decision. All good, since you know what's best for you...