A Very Compact Hybrid Amp - Page 77

cfcubed, do your heater circuit calculations permit someone to put a larger cap in for C3H? This would be a very interesting test.

Let me take a look at the info I have on heater parts values. It's been a while.
For sure the SM heater is "chopping" more to put out 6V from its 33V or so input vs. 12v.
Another interesting test would be to disable the switcher & use a very clean 6V heater supply like a battery. I did such a test on the P2P POC & did not notice a difference in any noise. But the POC's layout, etc. was very different than the CTH PCB (more spread out, etc.)

Would be nice to get an idea of the frequency of that wave too.

y axis is 5mv per division

x axis is 10u secs, so high frequency.

no obvious change on when subbing in a 6n1p

can you overlay traces of the heater on the second channel? should be pin 4 and 5, voltage relative to real ground.

holland, i can btu not tonight, I was given the scope a few days ago, and dont have the right probes.

here is a pic of thre scope looking at lower frequencies.

Just two quick questions:

With my headphones (MS-1's, 32 ohm), I only have to turn the pot about 1/3rd around to get the maximum volume I'd want. I am using 100 ohm resistors for R18L/R. Would roughly doubling R18 give me more control with the pot? And will this decrease the noise present?

Secondly, would increasing the value of C4 improve the bass response? Or is the input impedence into Q4/Q5 sufficiently high that the low frequency response is not effected?

Finished my 4th one. Here is the CTH parking lot..The middle one is going to stay with me. The one closest is heading over to Japan.

I do with my CTHs.

Excellent. Yes I am hearing different bass with different tubes, I just wanted to check that C4 wasn't holding it back at all. I'll put in a larger R18 soon, thanks.

Nice!! Wish I could have built it myself. Can't wait to test it out

Cool beans sachu. I'm sure it'll enjoy it's stay. I'm kinda jealous of it really, I've never been to Japan

hahha...I am sure you will enjoy it a lot.

..Thanks Smeggy.

You're like a factory. I just mailed off my second to my bro as a gift, 2 more to build. That's not quite the pace other builders are going at.

Basically dedicated 2 days to this ****...tested the EHHA and built two CTHs...productive wouldn't you say

CTH switcher info

Won't have time to tinker for a bit, but wanted to put up info.

At issue here is that it seems common that w/sensitive, lo-Z cans & 6,3v heated tubes, a noise/hum can be heard. For most, this does not intrude on music but is heard between songs. Use of design-center 12au7 variants lessen the noise.

I'm giving design info here, but if/when I get time these are the experiments in order that I'd conduct:

• disable switcher by disconnecting its supply. lift output of 3uH & use clean 6V heater supply (e.g. batt) for pins 4/5. See if noise disappears, just to be sure switcher is involved.
• if this confirmed switcher is causing the noise, restore 3uH & introduce resistance between lifted switcher input leg & connect. Resistor should handle load (e.g. 5W?)
• Perhaps alter cap values.

If switcher is found the culprit (1st experiment), I'd suspect we're asking a lot of it to cut from its supply (about 33V, its max is 40V) all the way down to 6.3V. Its min is stated as 1.2V but I'd suspect an ugly waveform at that voltage.
So, the 2nd experiment would be to alleviate some of the load through a resistor.

Here are National's Simple Switcher results at 6.3V & 12.6V:



WRT above, note that our input (Cin) & output (Cout) caps are higher than required, a good thing I'd suppose.

Note also that ideal main coil values vary between the two (220uH, 330uH), but in this next snip of the datasheet, 220uH appears best compromise value (& much smaller as well):


Now the ideal ripple filter cap may be a bit higher (180uF) than we are using (100uF). BUT we are getting very nice ripple reduction vs switcher w/o ripple reduction (see scope image from spec above). And in P2P proto I used a "worse" switcher (LM2575 - 52kHz) with NO ripple filter & found it to work well. So what may be happening @ 6.3V is the actual chop freq (or a harmonic of it) is entering audible range. More research would need to happen to reach that conclusion. Alleviating some of the switchers job through resistance OR altering assoc cap values might then address this.