Hi.
My problem in deciding about the Williamson topology is that i don't need too much voltage output, only 18V peak (36Vpk to pk) so i don't know if the addition of the LTP will help. Also after examining a broken 6SN7 like tube i think i can drill the tube's locator key from the bottom and insert a LED in there.
Chris
My problem in deciding about the Williamson topology is that i don't need too much voltage output, only 18V peak (36Vpk to pk) so i don't know if the addition of the LTP will help. Also after examining a broken 6SN7 like tube i think i can drill the tube's locator key from the bottom and insert a LED in there.
Chris
If you mean the 829 sockets you can find them here B7A PTFE Valve Base Tube Sockets BY HAT For QQVO64A 829B Valves NOS 1 Pc | eBay
About the key under the tube then maybe i can melt a hole with the soldering iron?
Chris
About the key under the tube then maybe i can melt a hole with the soldering iron?
Chris
hendrixgr, looks like you have a nice build there. if you get your amp to a flat frequency response it will sound dull. The response you have adds a presence to the music. This is why I have the tone controls left in. My amp is 70w SS but driven by a two stage 6922 preamp 1/2 before volume pot 1/2 after. Sound is fantastic with 1 ohm series resistor and my s/n is approaching 110db @ 1 watt.
My air conditioner gives me a SNR of about 40db @ 1W LOL
FWIW I don't imclude tone controls in ANY of my builds. Modern 30 band shapers are cheap and cheerful if you really need tone.
If someone can show me how to build 30Hz and 16kHz shelving EQs I'd be interested, as these are the only bands ever lacking IMHO.
FWIW I don't imclude tone controls in ANY of my builds. Modern 30 band shapers are cheap and cheerful if you really need tone.
If someone can show me how to build 30Hz and 16kHz shelving EQs I'd be interested, as these are the only bands ever lacking IMHO.
I think you are right, the response curve tricks me in to believing that NFB sounds dull when in fact it is the frequency response that changes. The SNR is measured at 99db with 20db of global NFB.
I was thinking of not including tone controls but the amplifier does need an extra gain stage with NFB so why not adding a tone stack? Anyway i think i am missing something here, the input sensitivity, for what input sensitivity should i aim 100mv rms, 1v rms ? I mostly listen cd's from a JVC cd juke box and maybe a tuner.
One final problem i need to address is the primary impedance of the output transformer, i have the feeling that there is something wrong with it. Monday i will connect the transformer to a high voltage generator and check it's turns ratio.
Here a photo of the final test build with the CCS leds under the tubes, right now i am constructing the "enclosure" which is actually a wooden frame with an aluminum top plate.
I also tested the 6N8S and i found that it works almost exactly like the 6SN7 (i THINK that the 6SN7 sounds a bit sweeter) and it has a little more resistance to vibrations, i can hear the little fan's (visible in the photo) vibrations in the output sound with both tubes.
Chris
Attachments
Last edited:
best 6sn7 has made by us factory in old days .your 6sn7 in pictures is russian made .difference in mid and hi .
I really, really like the Russian 6N8S tubes, but as far as US production goes, Westinghouse, GE, and National Union JAN types are my preference, but only so far as consistency, as the Russians will perform excellent and nearly identically in the same applications when implemented properly. I also don't tend to use many circuits that require matched sections, so the Russians are an excellent value. Even in situations with a CCS tail on a LTP they work great. The snobs make up all these excuses as to why they don't like them compared to whatever NOS one they run that week. Trust your own measurements and ears, don't just take my word for it
The Ultron 12SN7 are pretty good, I have a handful of 12SN7 types, handy for when using DC heaters.
Last edited:
Hi.
I really like the Russian tube as t is cheap and performs well. Monday i will also try two FU29 i got, with those if someone would see my amplifier he would think that it came out the CCCP with all those "CCCP" on the tubes.
Btw here is the latest tested schematic, for those who may not like where the global feedback is connected the only thing i can say is that after much testing this is where feedback really works better.
BTW R35 value (at the lower right corner that connects the right 829 grid bias trimmer to earth) is wrong, it should be 10Kohm NOT 8.2Kohm. Also R20, the concertina grid leak resistor should be 470K not 1 MΩ although the difference is negligible.
Chris
I really like the Russian tube as t is cheap and performs well. Monday i will also try two FU29 i got, with those if someone would see my amplifier he would think that it came out the CCCP
with all those "CCCP" on the tubes.Btw here is the latest tested schematic, for those who may not like where the global feedback is connected the only thing i can say is that after much testing this is where feedback really works better.
BTW R35 value (at the lower right corner that connects the right 829 grid bias trimmer to earth) is wrong, it should be 10Kohm NOT 8.2Kohm. Also R20, the concertina grid leak resistor should be 470K not 1 MΩ although the difference is negligible.
Chris
Attachments
Not for smoothing, but for peak current capacity. When I use a small cap on that module, the voltage will sag almost 40V at full power peaks. Also my latest amp idles at 22ma/tube but peaks at 250ma or so (Class B).
OTOH you're using tetrode connection which is less sensitive to voltage variations if I remember correctly.
OTOH you're using tetrode connection which is less sensitive to voltage variations if I remember correctly.
During testing i used up to 110 microfarads (two 220uf in series) but i didn't noticed any voltage sagging more than 10 volts, to tell you the truth i didn't measured the supply voltage at full load with the small capacitors something that i will do on Monday.
I went with small capacitors because during power up the power supply draws a lot of current due to capacitor inrush current, something that needs to be addressed with a relay, a charging resistor and some delay. In the schematic i forgot to show that the screens of 829 are fed with 200 volts from a separate smps. I use 60ma per tube (30ma each valve) for idle current, maybe i am a bit high for AB1?
Chris
I went with small capacitors because during power up the power supply draws a lot of current due to capacitor inrush current, something that needs to be addressed with a relay, a charging resistor and some delay. In the schematic i forgot to show that the screens of 829 are fed with 200 volts from a separate smps. I use 60ma per tube (30ma each valve) for idle current, maybe i am a bit high for AB1?
Chris
Last edited:
30ma seems fine to me... As far as inrush is concerned you can use a 1kW supply
I used a 300R resistor on a time delay relay board... One of these: NE555 DC 12V Delay Relay shield Timer Switch Adjustable Module 0 To 10 Second | eBay
I used a 300R resistor on a time delay relay board... One of these: NE555 DC 12V Delay Relay shield Timer Switch Adjustable Module 0 To 10 Second | eBay
@hendrixgr - please stop quoting the post directly above yours when replying.- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.