6S19P budget SE.

the thermionic watercooler

6S19P budget SE.

Postby sorenj07 » Wed Oct 03, 2007 11:46 pm

Should be OK for a bit over 2W. No scanner so I'll describe the circuit, from the OPT towards the back.

OPT: Edcor XSE10-16-8K, run into an 8 ohm output for a 4K primary Z.

B+: 310V (allow 5V for OPT loss)

6S19P. Bias at 85Vk, 220Va-k, 40mA with one 100R/1W + 2K/10W, bypassed with 220uF/160V and .22uF/400V Solen. I'm counting on a voltage swing of 150V pk-pk to bring this puppy to full output power.

5687 Driver: RC coupled to 6S19P grid with .47uF 400V + 220K 1/2W. It itself is a standard grounded-cathode stage, direct-coupled from the previous stage, so Rk (unbypassed) will have to be high enough to bypass past the grid voltage, which biased is the high cathode voltage of the top tube in the SRPP. I assume a 5687 has enough grunt to handle a 6S19P.

6N2P (12AX7) Input. SRPP layout. Haven't decided the identical resistor values, but the bottom resistor IS bypassed by a healthy-sized electrolytic and 1/1000 value film cap for higher gain. I'm considering implementing feedback, so I'd probably want to split the bottom resistor into two resistances in series, the one closer to the cathode unbypassed, so I can run feedback to that spot. Input to the 6N2P is straight from the RCA jack, except for a 10K 1/2W carbon composition grid stopper right at the tube socket.

The layout of this amp will probably be a single 5687 in the middle, with a 6N2P to the left and right and a 6S19P at the extreme ends.

The PSU should only need to be around 150mA 350V (choke-input). A 273BX at 350V 175mA should do nicely. A further 1.5H 56R LC filter will be used for each channel's B+, with the capacitor bypassed by a Solen film type. A further LC will feed the 5687 driver, with probably around 15H 1K and 100uF. A last RC tacked off of that will feed the 6N2P SRPP.

All PSU caps will be generic electrolytics.

This amp shouldn't cost too much. I have the 5687's and 6S19P's and just need two 12AX7-types. The OPT's are the cheap Edcors. It'd be nice to splurge on a DIY Enclosures 1712
( http://www.diyenclosures.com/products/chassis/diy1712/ )
BUT it's $115 and might go against the budget-2A3 mentality of this amp - Hammond aluminum seems more my speed. I love symmetry also so might want to just use one 5687 per channel paralleled - quite a lot of current available. Heater voltage should just work out. The 6S19P's will take 2 amps, and if I use two 5687's, that's 1.8 more for a total of 3.8. That leaves just .6A for each 6N2P, which is fine since they take only 340mA each to heat.

How does this sound? Let me know if I ought to sketch this out. I would come up with better values but I left TubeCAD at home...
User avatar
sorenj07
KT88
 
Posts: 925
Joined: Tue May 02, 2006 2:39 pm
Location: Berlin

Postby EWBrown » Thu Oct 04, 2007 11:52 am

The SRPP VA stage, typically the plate and cathode resistors are the same value. If you want to use NFB, a good mix would be 1100 ohms for RK and RA. For Rk (the lower triode) use 100 ohms to ground, then a 1K with bypass cap, to the cathode. Inject the NFB at the junction of the 100 ohm and 1K resistors (the 100 ohm resistor should not be bypassed).

The 6N2P / 12AX7 SRPP is pretty much self-balancing so that the plate of the "lower" triode should be half of the B+ voltage present on the plate of the "upper" triode.

The 5687 should have more than enough "grunt" to drive the 6S19P, it could probably push several of them in parallel... Just remember, their pinout is different that that of 12AX7 or 6DJ8. IICR< filament is 4 and 5, and the center tap is 8 and not 9. THeir filaments are definitely power-hungry... Yellow_Light_Colorz_PDT_07

Your voltage and current choies for the 6S19P agree nearly 100% with what I determined by using the published tube curves. (They do say that great minds think alike). 2100 ohms RK for 40 mA should be perfect - I think I calculated 2125 mathematically. 2K and 100 ohms in series, bypassed with a 220-470 uF 150V cap(s) should be just about absolutely perfect.

The 2K should be at least 10 watts, use at least 3X the actual power dissipation, for safety) The 100 ohm resistor, if it is on the ground end, makes a handy current checking test point. 40 mA would equate to 4VDC. The 100 ohm resistor can be lower power, 1W should be OK, it sees only 160 mA dissipation. Even a 1/2 watt should be OK, and having a "weaker" resistor can act as a safetry fuse if something "shorts out".

/ed B in NH
Real Radios Glow in the Dark
User avatar
EWBrown
Insulator & Iron Magnate
 
Posts: 6389
Joined: Wed Mar 19, 2003 6:03 am
Location: Now located in Clay County, NC !

Postby sorenj07 » Thu Oct 04, 2007 12:58 pm

great minds _do_ think alike - I guess I forgot to post the power ratings of the bias resistors, but that's exactly what I came up with. I'd love to also try PP with these tubes and at the moment am really struggling between the two. I'd need a more robust power transformer (maybe a PA060S) to deal with the filament current but it might be nice to have a bit of power. I'd need 7A of 6.3V to duplicate the circuit, and the PT cost would be just about doubled. Not to mention the iron - I'd probably go with Z565's or maybe the CXPP25-8-7.6K's, either way that's more than twice the cost. Maybe over Christmas break we'll see something materialize, either way. I might also need to work on a regular 6P14P amplifier, I think my brother might like to have a nice little stereo setup :)
User avatar
sorenj07
KT88
 
Posts: 925
Joined: Tue May 02, 2006 2:39 pm
Location: Berlin

Postby EWBrown » Fri Oct 05, 2007 5:46 am

It might be more cost-effective to simply add an extra filament trannie, instead of dropping huge funds into bigger-than-necessary power trannies...

A couple of 6.3V 4 or 5 amp filament trannies should be a good fit. Also check out AnTEk, he has various LV Hi current toroids for pretty cheap, these usually have dual secondaries as well as dual primaries.

/ed B in NH
Real Radios Glow in the Dark
User avatar
EWBrown
Insulator & Iron Magnate
 
Posts: 6389
Joined: Wed Mar 19, 2003 6:03 am
Location: Now located in Clay County, NC !

Postby sorenj07 » Sun Oct 07, 2007 6:30 pm

I'd rather stay with an EI style transformer, and the filament transformer is a good call. I don't suppose a GSG board would work to feed the 150V swing to the 6S19P grids?
User avatar
sorenj07
KT88
 
Posts: 925
Joined: Tue May 02, 2006 2:39 pm
Location: Berlin

Postby EWBrown » Tue Oct 09, 2007 6:41 am

A little more TubeCAD fun:

AN SRPP (schematic D) using a 12AX7 and 1K RK and RA, will have a voltage gain around 69-70, and will deliver up to +/- 85V signal swing, so this looks perfect for feeding the 6C19P-Vs. The SRPP stage draws about 1.2 mA, figure 2.5 for stereo, the DC B+ could be derived by doubling off the main power trannie, or use one of the PA-211 PAS-2 power tannies. THis is the exact same driver that I am using on the "300B on a Dynaco ST70 chassis" project.

I got the delivery of the ten 6S19Ps from Odessa, seven tested out good, and three had no emission. They looked OK, no cracks or visible damage, I suppose that 30-40 years of warehouse storage takes its toll. or perhaps the "OTK" quality control wasn't so stringent after all... Yellow_Light_Colorz_PDT_21 Even so, the tubes came out to less that 3 bucks apiece, which ain't too shabby... Yellow_Light_Colorz_PDT_06

I haven't had other "dud" tubes from over there, except for one 6C33C which was a shipping & handling "casualty" a couple years ago.

We'll soon see if the GU-50s fare any better...

/ed B in NH
Real Radios Glow in the Dark
User avatar
EWBrown
Insulator & Iron Magnate
 
Posts: 6389
Joined: Wed Mar 19, 2003 6:03 am
Location: Now located in Clay County, NC !

7233 is almost 6S19P-V

Postby EWBrown » Wed Oct 10, 2007 5:54 am

Just found this one, the GE 7233. Very similar to the 6S19P-V, slightly different ratings (probably more conservatively rated), and pins 8 and 9 are swapped.

http://www.mif.pg.gda.pl/homepages/fran ... 7/7233.pdf


"Close enough" for Horseshoes, Hand Grenades and Atom Bombs... Yellow_Light_Colorz_PDT_04 Yellow_Light_Colorz_PDT_01 Yellow_Light_Colorz_PDT_08

/ed B in NH
Real Radios Glow in the Dark
User avatar
EWBrown
Insulator & Iron Magnate
 
Posts: 6389
Joined: Wed Mar 19, 2003 6:03 am
Location: Now located in Clay County, NC !

Postby sorenj07 » Sat Oct 13, 2007 8:54 am

I assume that a 5K OPT impedance would also work? Power output should remain pretty much the same right, maybe 2W or so? One other thing: if I build it really nicely, do you think there might be a pretty decent market for an amp such as this? In case it's too little power for me, I mean. I bet it won't be, though, since my speakers are quite efficient. If I ever changed speakers, though, it might be problematic :)
User avatar
sorenj07
KT88
 
Posts: 925
Joined: Tue May 02, 2006 2:39 pm
Location: Berlin

Postby EWBrown » Mon Oct 15, 2007 9:37 am

I paper-worked it out, and came out with about 4300 ohms as being the optimal impedance for SET, anything from 4K to 5K should be perfectly good, the higher impedance will yield a dB or so less power, but at a lower 2nd harmonic and THD content (a good trade-off).

/ed B in NH
Real Radios Glow in the Dark
User avatar
EWBrown
Insulator & Iron Magnate
 
Posts: 6389
Joined: Wed Mar 19, 2003 6:03 am
Location: Now located in Clay County, NC !

Postby sorenj07 » Wed Oct 17, 2007 5:24 pm

Sounds good. Over at AA, they seem to imply that third harmonic distortion rises with primary impedance. I didn't know what to make of that - I'd assumed that all forms of distortion decresae as primary impedance increases. Hmm...

http://www.audioasylum.com/forums/tubed ... 29075.html
User avatar
sorenj07
KT88
 
Posts: 925
Joined: Tue May 02, 2006 2:39 pm
Location: Berlin


Return to diy hifi

Who is online

Users browsing this forum: No registered users and 13 guests