Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › A lo-cost HT-supply true transformer-coupled tube headphone amplifier
New Posts  All Forums:Forum Nav:

A lo-cost HT-supply true transformer-coupled tube headphone amplifier - Page 5

post #61 of 108

popcorn.gif This is a really long movie with no plot...  I hope you're OK Waki.

post #62 of 108

I think its pretty clear that he isnt OK. Over 3 weeks and nothing.  :(

post #63 of 108
Thread Starter 
Sorry for the long delay folks, but I have not felt very well for the past few weeks. Immediately following treatment with radioactive iodine I was put on a high dose of beta-blocker. My thyroxine levels then fell very low and the combined effect was that I started to feel very tired all the time, having barely enugh energy to walk the dogs once a day. I recently had a consultation following blood tests and I was recommended to ramp up the amount of thyroxine I am taking from 25 micrograms to 125 micrograms, a five-fold increase, so you can see that the amount I was taking was way off. Half-way through this process I already feel much better. I hope to be taken off the beta-blocker soon too, which hopefully will cause me to return to my usual waspish energetic self.
Seriously though, the overactive thyroid which was the initial cause of these problems is known to result in sufferers becoming very short-tempered, so if I have been rude to anybody in the recent past, I do apologise. I should be a bit less inclined to jump down people's throats in future.
The good news is that the amp has a clean, satisfactory sound with no hum or perceptible hiss with plenty of gain when turned up full. Perhaps a little light on bass, but I haven't had a chance to evaluate it fully.
There is, unfortunately a fault, a regular ticking on both channels, which is very noticeable with no program material. I will have to get to the bottom of what the cause of this is. It's about at 1Hz, or perhaps a bit quicker, not very loud, it sounds like the right frequency for a relaxation oscillator, something charging to a potential and undergoing discharge or something like that. It's completely masked when the volume is turned up full, but there is a bit too much volume with the phones (cheap) and source I chose for the first trial to listen continuously at full volume.
The amp does work though, and not that differently from what I predicted.
I have taken Frank Cooter's advice regarding DC coupling, and I have worked up a simulation with an active (CCS) anode load. This should perform slightly better than the amp in it's current configuration. You can see the circuit and results here. Treble is somewhat constrained, but bass is only a couple of db down @ 20 Hz, although these simulations are done with a fairly primitive transformer model, and while useful for development purposes, have to be taken with a pinch of salt.





This shows the amp just below clipping, all harmonics above the second disappear to all intents and purposes when the drive level is reduced.
All the best to everybody...
post #64 of 108

Great to hear you're back and that your health is improving! We were beginning to wonder what happened. Hopefully your improvement is permanent and there'll be more amps in the future. There are very few of us DIY tube builders here and last thing we want is to lose one!


It's been a pleasure watching the evolution of your design and I'm happy to have played a small part in the process. Happy to hear it works.


The "clicking" you're hearing sounds almost like a decoupling problem, although this shouldn't be an issue with a two stage design. Can't hurt to try something like a 10uf electrolytic in the C9 position. I'd also be looking for a bad solder joint in the power supply or grounds.

post #65 of 108

Hey, there you are! It's great to see you back in action.


I once had a 'tick, tick, tick' in a circuit that turned out to be a hairline crack in a power trace on a PCB that kept arcing across. Discovered it with the lights out! You may have an arc in you HT circuit somewhere, turn off the lights and have a look; it won't cost a dime.


What was your solution on the filament supply?

Edited by elliottstudio - 11/24/12 at 9:17am
post #66 of 108

This is pretty slick.


Hope you feel better soon

post #67 of 108

Good to see you back on deck, Waki. 

post #68 of 108
Thread Starter 
Hi, people, thanks for the concern.
To answer the question about the LT, first I tried wiring it up to AC. This proved a pain to achieve physically, given that I had to contrive a resistive splitter to connect the ground to and get the soldering iron in through the later wiring to get at the valve bases. So I went back to the DC I had originally intended. I changed the cap in the filament supply to 22,000 uF. This improved the ripple by a factor of 10 to under a volt, where it should have been in the first place. The peak voltage was ~8.3V after rectification. ~0.7V ripple took this down to 7.6V. I designed and built a discrete low-dropout regulator, but I had some issues getting this to work, by this time I was pretty groggy due to the medication. Anyway, I started looking at the LM317 again, the dropout voltage is well below 2V at most temperatures, so I wired up 2 more LM317's with 1k2 and 5k1 resistors, which adds up to 6.3V and decided to let the dropout look out for itself.
Now I knew this part of the circuit was going to work OK, unlike the discrete regulator, so I wired it up on some perfboard and wired it in.
Then I left it for nearly 3 days until I got the nerve to turn it on. I knew I had put some dents in one of the HT caps when I took out the original cap for the LT supply, but I couldn't face stripping the heatshrink and substituting it, even though I had a spare. So eventually I turned it on and watched the tubes for any signs of serious problems.
When both heaters fired up, with no signs of distress on the plates, I measured the voltage on both the input and output jacks. When it turned out there was nothing much I decided to connect up some old 'phones and a source. The source is a Sony NW-E003 flash Walkman, known to be a bit quiet, and a pair of Koss K/20's which are quite robust.
Warm it up, turn on the Walkman, crank the volume, and there it is. Loud. But good. Although with an annoying tick, tick when turned down.
Here is the actual circuit as built:-



The big mistake I made with this build was the chassis.
In order to keep the cost down I looked at the first chassis on the site I was buying from, the chassis was 8 * 4  inches and 2 inches deep. I saw that the prices increased down the page, I though that the 8 * 4 would probably be acceptable and I went with it.
I ended up with a really tight build that was very hard to modify and fix when I had a few minor problems. Not really very smart considering it was a prototype.
It's only on going back to the site where I bought the chassis that I realize that the general trend on the page was rising prices going down the page, but it was only superficially true. I paid 11.80 euros but further down the page was a 10 * 6 inch chassis, 2 inches deep, for 11.20 euros. Not only cheaper, but with 60 square inches of floorspace as opposed to 32 square inches.
Anyway, now I want to rebuild the amp with a slightly different configuration I'm going to order up the larger chassis.
Although this will mean a bit of effort in marking up and drilling and punching another chassis (particularly the hole for the power connector), it will result in a much less cramped build with room for turret boards or tag strips, and it will be much easier to correct any problems or modify the design if I decide on a third iteration.
Here's the circuit as I'm currently thinking I will build it. I couldn't resist the symmetry of the 2 current sources.



It'll probably take me a few days to check the BOM, order up the parts and punch and drill the chassis. I'll get back to you when I've got some real progress to show.




post #69 of 108

Thanks Waki - very professional circuit diagrams. I'll leave it to those with a few more clues to comment further !   wink.gif

post #70 of 108

very professional circuit diagrams.

...except for the .jpg - never post tech doc, line drawing in jpeg - use gif or png


clik to see 4x zoom of actual bitmap





just a suggestion

post #71 of 108
Thread Starter 

Turns out I couldn't re-use the CCS bias on the second channel because it makes the current drain assymetric, so it needs to be duplicated at the cost of another 10mA or so...


Good job that the PT and choke are overspecced.


You can see the most recent version of the circuit (arranged for PCB layout) in high detail and save an easy-to-manipulate copy (by right-clicking and selecting 'save image as...') by visiting this link:- http://wakibaki.com/audio.php#t11







post #72 of 108


post #73 of 108

Why is it -3db at ~5khz? 

post #74 of 108
Thread Starter 
Originally Posted by nikongod View Post

Why is it -3db at ~5khz? 


If you change the coupling coefficient of the transformer to 1, the -3dB frequency response goes up to > ~100kHz. The transformer model is primitive, and the 0.95 coupling coefficient is arbitrary. I checked the low frequency rolloff by substituting in progressively smaller caps, you can see things going a bit awry when you put in about a tenth of the capacitance, so I have a bit more confidence that the low end simulation is meaningful, but I don't see much point in worrying about the significance of the top end.


Let's wait for the RMAA results for the rebuilt amp. I ordered some transistors for the CCS's, but I haven't ordered the new chassis yet.



post #75 of 108

Are you going to build it out on that PCB you designed? That would be a treat.


Looking at your first stage tube circuit, I've tried using a diode (and an LED too) for the cathode bias and measured more THD than with a plain resistor. Does the CCS load on the plate help alleviate some of the distortion?

New Posts  All Forums:Forum Nav:
  Return Home
Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › A lo-cost HT-supply true transformer-coupled tube headphone amplifier