brainstorming building my first single ended tube amp (well, more like a test bed) with Antek MS-15W60 toroidal opt's

[nerdorama]

@pac1085 be sure to check the polarity of the Tek windings before you parallel connect those secondaries. Some of their data sheets indicate which terminals are positive polarity with respect to the primary but they are sometimes a bit funky and it isn't always obvious. I've used several Tek power transformers. They are built like a tank but you have to be careful of details.

What is the audio analyzer you are using. Looks pretty neat.

 

[pac1085]

@pac1085 be sure to check the polarity of the Tek windings before you parallel connect those secondaries. Some of their data sheets indicate which terminals are positive polarity with respect to the primary but they are sometimes a bit funky and it isn't always obvious. I've used several Tek power transformers. They are built like a tank but you have to be careful of details.

What is the audio analyzer you are using. Looks pretty neat.

Thanks, yeah I need to validate that for sure. I might actually keep them separate and build a separate regulated supply for each channel after getting a better understanding of the max cathode current of the series pass tubes I was gonna use. Nothing like a 3 watt amp with a dozen tubes in it, lol. I think I'm going to have more fun on the power supply than the amp itself..

Some of the pics are from the Levear VP-7725D audio analyzer, but the plots are from a Rohde & Schwarz UPL: https://www.kenrockwell.com/audio/rohde-schwarz/upl.htm

I love my Fluke 407D, but man, ya gotta be careful and ensure the set-ups are well connected with well insulated wire. Oopsies are spectacular without programmable current limiting. Found that out the hard way..

Yea... I have some Octal relay breakout sockets showing up today so I can get a much less sketchy test bench setup going. Although that's not going to help with the Novar's. I'm using all decent Pomona cabling, but the pre-made ones are only rated 300v hands free. The ones I made myself are using Pomona 10kv silicone wire. I usually tape the B+ connection down somewhere that I can't bump it but there's still plenty of opportunity for things to go wrong the way it was.

I have a Kepco BHK500-80mg that I've been using too, since it has a current limit but the damn fan is soooo loud.

 

[nerdorama]

Also, maybe check the direction of the coils so you can orient it in the most quiet direction relative to your output transformers. Or just make a test with the primary energized and move an output transformer around it with the primary on a scope while looking for the best orientation.

 

[gadget73]

Although that's not going to help with the Novar's.

surprises me that those don't have anyone making breadboard sockets. They exist for everything else including 12 pin. Only thing I can think of is either design a board, or get a chassis mount socket mounted on short standoffs and run wires to barrier strips on a small chunk of wood.

 

[pac1085]

surprises me that those don't have anyone making breadboard sockets. They exist for everything else including 12 pin. Only thing I can think of is either design a board, or get a chassis mount socket mounted on short standoffs and run wires to barrier strips on a small chunk of wood.

I think some guys on diyaudio designed one using mill-max pin sockets on a PCB. I can also 3d print a box to hold a novar socket and break it out into terminals or whatever makes sense too

these octal relay breakout sockets were $8 for 10 on amazon, can't beat that lol https://www.amazon.com/dp/B0BYRY285N

 

[gadget73]

yeah I have several octal relay sockets salvaged from demo equipment at work or picked up stupid cheap. They are extremely common in industrial controls.

 

[AlexHempel]

I bought a pair of Novar breadboard sockets from the guy selling them on DIYaudio (assuming there's only one), but honestly given the price and hassle (he made me pay in Litecoin...) I wish I had just made and ordered my own PCBs. Wouldn't take long to design the PCB, although sadly shipping costs for JLCPCB have increased substantially.

I actually tried to make a run of chassis mountable Novar sockets using laser cut Delrin and press fit Mill-Max receptacles, but unfortunately I didn't really get the tolerances right and the holes were too large for the press fit. I still have all the receptacles -- maybe I should take another stab at it. I have like 300 6KY8s, hence my interest in Novar sockets.

 

[nerdorama]

Here is a thread detailing a DIY pcb design for Novar tubes using Keystone female pins, Keystone Part #1434.

DIY Novar Socket

I found a source for PC solder pins that work with Novar tubes so I designed a circuit board for a Novar damper diode power supply. It can be used in a hybrid bridge configuration or with a center-tapped xfmr by just leaving out the diodes. Also, the heaters can be powered by a separate or...

www.diyaudio.com

 

[nerdorama]

It would be useful to have a PCB design that we could have fabbed as required for our projects. The designer in the above post included a zip file which I think are design files but I'm not familiar with them or if they require a certain application to use. Would be cool if someone that knows how to do it would post Gerber files that we could use.

 

[pac1085]

I think I am going to order a regular hammond transformer before I got too crazy with this. I've tried so many things in the past few days but that low frequency distortion is still there.

 

[AlexHempel]

If there's serious interest in some PCB fab files for novar sockets, I could post some. But don't want to derail this thread from the subject of @pac1085's SE amp, so perhaps a new thread if so. And a word of caution: the Mill-Max receptacles aren't cheap, so the cost per socket ends up being fairly high from the receptacles alone. That's why I ended up simply buying available Novar sockets on Ebay when they popped up. Although they're expensive by normal standards, it's still cheaper than the Mill-Max approach unless one needs larger volumes than can be found periodically on Ebay.

 

[nerdorama]

The Keystone sockets/jacks aren't too expensive. In quantity 25 they are about 72 cents each at Mouser. RS Electronics has them for 18 cents each. If making a socket for damper diodes you don't need to fully populate the socket pins.

 

[gadget73]

I've tried so many things in the past few days but that low frequency distortion is still there.

the peak at 60? Does this have a shield lead by chance? Wondering if its just picking up 60hz in the air. Less likely if its a no-feedback design but currents induced in the output transformer of a feedback amp go right back to the input so they can be amplified.

 

[nerdorama]

I was able to load the Gerber zip file from the above mentioned post at JLCPCB and it looks ok. It is a 50mm x 100mm board so not tiny but holds two Novar tubes and provision for SS diodes if wanting to make a hybrid bridge.

 

[pac1085]

the peak at 60? Does this have a shield lead by chance? Wondering if its just picking up 60hz in the air. Less likely if its a no-feedback design but currents induced in the output transformer of a feedback amp go right back to the input so they can be amplified.

The OPT does not have a shield lead. I tried adding some global negative feedback but it didn't change the spike at all. The last few sweeps I did with more resolution it actually seemed to shift around a bit, with 5hz resolution it was starting around 80hz. But it doesn't recover right after, it doesn't seem to drop back down to normal until around 500hz.

I'm going to get this thing built a bit more cleanly and documented better before sending you guys on any more wild goose chases. I noticed at some point I must have bumped the switch on my dummy load (4 ohm, 8 ohm, off) position to 'off' so that probably didn't help with some of the earlier experiments (although to be honest, I think I just did that a moment ago when I brought the DC power supply back onto the bench)

I was thinking about buying the Hammond 125ESE since the local guy who's system I heard was based around that, but it seems that the advertised frequency response on that one isn't very good (Sounded great though). So, may not be the best for me and wanting to understand this low frequency distortion issue.