Give them a shot. Help confirm for future builders

I have stuffed my amp boards with them.

I'll let you know how they react.

Give them a shot. Help confirm for future builders

Sounds good. I'm stupid enough to be building a second one in parallel so can do both. 

/

 

Edited November 12, 201510 yr by JoaMat

That's awesome. My plan was to go SMD for the amp section, including the batteries. I was thinking of putting pads for both battery styles.

what FET did you use?

Good idea pads both styles. FET is Toshiba 2sk208. LEDs are TLMS1100.

Edited November 11, 201510 yr by JoaMat

Well I've swapped out the takmans in the battery for xicons, but the noise is still present. George suggested that it might be the input tubes which are noisy. Aside from that I'm not sure where to look.

Perhaps the trimmers in the batteries?

The trimmers are the regular bourns ones. I remember lower ppm trimmers being mentioned in this thread a while back, but that they were much more expensive.

if you have a lower trimmer maybe put a resistor in series? just fishing

Years ago pedefede mentioned that replacing the 10k trimmers with some vishay PTF56 resistors (since they're 5ppm) did the trick. If it isn't the input tubes that's the next thing I'll try.

Anyone happen to have 9kOhm PTF56's lying around?

Edited September 25, 201510 yr by s_r

Check the supply voltage -500.

For me, the current limit has been reached 90mA in rail -500V. One channel worked noiseless with the consumption of 48mA at -500V.

I measured R33 at about 93mA. What would be the best way to measure each channels current usage? Connect one channel at a time and measure R33 for each?

Yup, connect a single channel and measure across the big resistor in series with the MOSFETs in the PSU

Alright, I went ahead and did that. The current usage for each channel was close to identical, at about 47mA.

I also plugged in some headphones and each channel exhibits some weird behavior. Left only results in no audible noise, but the channel itself is much quieter. So maybe the noise is still there, just quieter since the whole channel is? Right only still gives the same noise, and the channel is about the right volume.

I'm starting to think this is where the black magic is required.

Ok. Do the same and swap the input tubes. Compare.

Swapped the input tubes, no change in each channels behavior when powered individually.

Hm. If the battery voltages are on and offset is fine, LEDs are lit, the circuit pretty much has to be working. The reduced volume in one channel to me is very odd. I would perhaps continue to swap the easy stuff, input (assuming this is good) source and output tubes and even the two umbilicals. See if you can get that behavior to swap channels.

Would it be worth adding a bit more cap to cap that was added to remove the oscillation. If it got worse or better

then it might give a clue.

I originally thought the left channel might not have all its LEDs lit or maybe the battery was out of whack, but it checks out fine. I also tried all the combinations under the sun and the channels don't swap their behavior.

Going to order some parts to try and swap in, mainly some PTF56s to replace the 10k trimmers in the battery. Would 10pf work as something to try in place of the 5pf cap?

I think the trim pot is a good place to look.  I remember Inu looking at them as well as some other members needing to replace them to clean up the noise.  I think Inu did a full substitution with a fixed resistor at some point.  It's been a while so I don't completely remember.  For me, the noise was definitely coming from the resistors.  I used PRP and ended up replacing with everything in the batteries with Xicon, which did the trick.

Finicky thing these batteries. 

Speaking of trim pots, JoaMat, what trim pots are you using in your SMD battery above.  I'm just laying out my new boards and I wanted to size them properly.  Given that yours worked, I thought it would be a good source.

Lastly, I finally got to testing my new PS.  +/-500V, +250, -300, -60, +/-12V and 5V supplies.  Everything checked out. 

One note is that the 47nF or .1uF bypassing the resistors going into the 20K resistors and feeding the error amp seem to increase any oscillations coming from the prior stage.  I've taken them out for now.  I'll post something about this in the carbon thread as well, since these new supplies are used there as well.  I'm seeing noise levels under 6mV, but there is likely noise coming from my room that I'm seeing. Also, I'm seeing some random spikes at around 50mV. 

EDIT: I'm going to try adding an EMI filter at the input to see if it cleans this up.

 

Edited September 28, 201510 yr by Kerry

The power supply looks nice and compact.  Congratulations! 

Great work the new T2 power supply.

 

Edited November 11, 201510 yr by JoaMat

Thanks.  I was just watching the eclipse.  Very cool.

left is 450v 220u, right is 450v 100u. inu posted some pictures for counterfeit caps, where the negative strip is misaligned. should I throw out the 220u? I soldered a bunch of them to my psu already...

NEGATIVE BLACK...  Where???