Page 2 of 3

Sansui 1000A

PostPosted: Thu Aug 26, 2004 5:38 am
by EWBrown
If I remember right, the Sansui 1000A used 7591s, which have a P-P load impedance of 6600 - 9000 Ohms which is within the range for this circuit.
It doesn't have the UL screen grid taps, so the amplifier would have to be run in either pentode or triode mode.

My SWAG is that the Sansui iron is 7200 ohms CT which would actually be 1800-0-1800 ohms, as the impedance is a square of the turns ratio, so a 7200 ohm CT, 8 ohm secondary output transformer would have a 30:1 turns ratio (30^2 =900, 900 x 8 = 7200). For the 4 ohm secondary, the turns ratio would be 42.42:1 and for the 16 ohm tap, the turns ratio would be 21.21:1

/ed b in NH

Diytube ST 50

PostPosted: Fri Dec 17, 2004 8:20 am
by dhuebert
Well, here it is 8 months later and I just got my ST 50 going last night. Apart from a cap that I broke during the chassis assembly it worked first try.

Cathode current is 60 mA, with a B+ 0f 370 V, using a 272 HX. I had to take the 0.1 R resistors out of the heater circuit because of the larger draw of the 6L6s, heater voltage is now 6.38V. I got some 6L6GBs for cheap, but may upgrade to GCs if I blow them up. Maximum ratings are 360 volts on the plate and 270 for G2. They say tubes are fairly robust and I guess we'll find out.

This weekend I will take some pictures, tweak the NFB and hopefully fry some eggs (Southern Culture on the Skids?).

Don

cathode current

PostPosted: Fri Dec 17, 2004 6:48 pm
by tubes4hifi
Hey Don,
if it was me I'd run those tubes between 40 - 50 ma, not 60.
They'll last 5X longer and I doubt you'll hear any difference.
Roy

ST-50

PostPosted: Mon Dec 20, 2004 9:06 am
by dhuebert
I bought a quad of Mesa-Boogie (read Sovtek) 6L6GCs on the weekend for $75CDN (~50US) They were beautifully matched and came to 70 mA without tweaking. As I suspected, gain was not high enough to run the 6L6s so I bypassed the cathode resistors on the 12AX7s with 220uF caps, lots of power now. I haven't bought a pair of dummy loads yet, so I can't tell how much. The data sheets I have seen suggest a cathode current of 100mA at a plate of 400V, I'm running 350V @ 70mA right now. I'll have to make a resistor change to get to 50mA. Just a guess, but I think I may have to find some more gain, with the volume control @ 3:00 I'm hearing distortion, once I get the dummy loads I can track down the source, preamp? Power amp? 12AX7s? 6L6s?

Don

Re: ST-50

PostPosted: Tue Dec 21, 2004 7:50 am
by Shannon Parks
dhuebert wrote:I bought a quad of Mesa-Boogie (read Sovtek) 6L6GCs on the weekend for $75CDN (~50US) They were beautifully matched and came to 70 mA without tweaking. As I suspected, gain was not high enough to run the 6L6s so I bypassed the cathode resistors on the 12AX7s with 220uF caps, lots of power now. I haven't bought a pair of dummy loads yet, so I can't tell how much. The data sheets I have seen suggest a cathode current of 100mA at a plate of 400V, I'm running 350V @ 70mA right now. I'll have to make a resistor change to get to 50mA. Just a guess, but I think I may have to find some more gain, with the volume control @ 3:00 I'm hearing distortion, once I get the dummy loads I can track down the source, preamp? Power amp? 12AX7s? 6L6s?

Don


Hi Don,

Couple questions, so I can help you tweak this:

1) Important: so far, what are your deviations from the regular DIY ST35 schematic?

2) Are the cathodes separate? Are you stating current per tube or per pair?

Bypassing the cathode of the 12AX7 is not a good idea, and I'm sure you're getting some distortion there. Why? The cap feedback circuit from the 12AU7 is acting in the same function (I believe RDH4 talks about this technique) - the .22uF and resistor lower the output impedance of the 12AX7 and increase the gain already. If you need an increase in gain, start with doubling your negative feedback pot resistance.

Once you let me know what your schematic looks like, I'll give you my tips. I'll see if we can't tweak more drive out of the phase splitter.

Shannon

Tim Tailoring the ST-35

PostPosted: Tue Dec 21, 2004 9:27 am
by dhuebert
So far the only deviations from the ST-35 schematic are a pair of 220 uF caps and of course the 6L6s. I also tried a 12AX7 in the phase splitter. I know that 6L6s need about 70V p-p for full output in AB1 so again once I get the dummy loads I can see what I'm really making.

Cathode currents are 70mA per each. The octals are pretty much soldered into the 6BQ5 sockets. The only change I made was to short pins 1 and 8 together on the octals in case I wanted to try EL34s.

I didn't increase Rf because I had already tried turning the pot all the way up which produced what I thought looked like an over-damped squarewave on the output. (a sharp rise, then an exponential increase to about double the initial rise then a sharp fall to zero) Turning the pot all the way down produced a good squarewave but at very low power. I was thinking of defeating the feedback and see what the effect.


This project is a warmup for the bass guitar amp I will build next.
Don

6l6

PostPosted: Tue Dec 21, 2004 9:55 am
by Shannon Parks
I'll see what I can come up with at lunchtime, Don.

Shannon

PostPosted: Tue Dec 21, 2004 12:21 pm
by Shannon Parks
Don, what does your cathode voltage look like? Around 28V?

Also, what output iron are you using? You might try running on the other tap(s).

Without sweeping the unit on my bench, nothing jumps out at me. You could use a little more drive, but I don't think you're getting your distortion there, and it really doesn't look too bad. I would yank the screen feedback caps if those are in there.

Shannon

Watch those cathode resistors' wattage ratings

PostPosted: Tue Dec 21, 2004 1:27 pm
by EWBrown
If you're running 70 mA thru those cathode resistors, you may want to replace them with some with higher wattage, as there will be around 2 watts continuous running through each.

You could drop them down to 40-50 mA, but the resistors would have to be higher resistance (560 - 700 ohms each, 5W or greater) in order to get the biasing closer to levels that 6L6GTs are designed for.

/ed B in NH

Hang on to the 1000A Sansui outputs

PostPosted: Tue Dec 21, 2004 1:32 pm
by Guest
Hey Norris,

The 1000A output transformers should work with 6L6 types if you say it is 7200 Ohm plate to plate. Most 6L6 amps use around 6.6k impedance, but the difference is small and easily in the order of speaker to speaker impedance differences. The Sansui 1000A transformers are reputed to be very good. On the other hand, if you wanted to get rid of them just let me know :wink:

Francois

Cowboy tech

PostPosted: Tue Dec 21, 2004 3:41 pm
by dhuebert
I will check the cathode V tonite.

The iron is standard ST-35 stuff (I believe 272HX and 1650F)

To me
screen feedback caps
means caps on the screen grid in parallel with the UL taps?

Twice now I have been told these tubes are supposed to run @ 40-50 mA, the data sheets I have seen indicate 100mA, whats the deal here? I will try and turn down the current and see if it makes a difference to output.

I sure appreciate the help, I'm the kind of guy who has to know why

Oh yeah, is the ST-35 schematic posted here somewhere?

6L6GC data

PostPosted: Wed Dec 22, 2004 11:07 am
by dhuebert
So the catode V is 27 for a current of 60mA at a plate V of 350. I will try a 5 W 560R resistor here next. I think I'll contact Mesa-Boogie and see what they say about their tubes. I read thru their site today which was pretty educational. So far I have read everything I can find on tube biasing and I must say, I almost think I'm reading about religion based on the amount of contradiction...

Don

Plate Voltage

PostPosted: Mon Dec 27, 2004 1:17 pm
by Casher O'Neill
Isn't 350 V still a bit low for a 6L6?

Re: Cowboy tech

PostPosted: Tue Dec 28, 2004 7:39 am
by Shannon Parks
dhuebert wrote:To me
screen feedback caps
means caps on the screen grid in parallel with the UL taps?


Sorry I got sidetracked last week, Don!

Yes, the screen feedback caps are going from the 'lower screen' to the cathode of the 12AX7. This feedback is in parallel with the negative feedback from the output transformer secondary. This is used in several of Hafler's designs, mostly a tweak due to the way those outputs were wound (my guess).

In other words, I think we can change the whole feedback stability design when using non-Hafler outputs. We can omit, or lessen, the input low pass filter. We can omit the screen cap. At that point we can do open loop testing (to see where the upper frequency pole is at so that we achieve less than unity). Then add a feedback resistor for the amount of dB of NFB we want and close the loop. Then shoot a square wave through the DUT and tweak with our cap in parallel with the feedback resistor.


dhuebert wrote:Twice now I have been told these tubes are supposed to run @ 40-50 mA, the data sheets I have seen indicate 100mA, whats the deal here? I will try and turn down the current and see if it makes a difference to output.

I sure appreciate the help, I'm the kind of guy who has to know why


Don, I'm wondering if that 100mA is the rated current for a pair of PP tubes. Anyhow, now that you are using 6L6GC's, I wouldn't worry much.

dhuebert wrote:Oh yeah, is the ST-35 schematic posted here somewhere?


I just upped a copy here:
http://www.diytube.com/ST35_REVD_circuit.pdf

Shannon

PostPosted: Tue Dec 28, 2004 10:10 am
by EWBrown
A lot of the tube manuals / data specify if the total current is for a single tube (usually class A) or a PP pair (class AB1, B, etc). Some tubes allow for sharing a common cathode resistor, others require separate cathode resistors (If I remember right, KT66s do). 100 mA through a single 6L6GC would make for red hot glowing plates and a very short tube life.

/ed b in NH