Help Debugging bias (?) problem.

knowledge base for the classic Dynaco ST70

Help Debugging bias (?) problem.

Postby DoubleAA » Fri Jul 31, 2009 1:43 pm

Hello,

I have a st-70 that has been pretty much completely rebuilt. I originally got it off of ebay, but have replaced the power supply stage (curcio kit), the transformers (triode - both opt and power), tube sockets, and of course the driver board - I am using the one from here.

The amp was always kind of intermittant and it essentially has sat for about a year because it kept blowing fuses. Finally I've dug into it. I've determined I've had a number of bad tubes - EL-34s and KT-77s and the rectifier. Since I am waiting for a new rectifier tube in the mail, I've wired things up to use solid state rectification.

My problem appears to be isolated to one channel and I think the driver board. On the right channel, using KT-77 power tubes (these are the only 100% sure brand new tubes I have on hand that I know are good), I am able to bias things just fine. With full ccw on the wire wound, I get 46.5 mA, and I can set to 50mA without issue and it will stay there. This is without any power tubes on the left side.

The left side, the bias current just climbs up and up, slowly, but in less than a minute, I'm turning things off as it creeps past 100mA. This is without the right channel power tubes. If I pull the coupling caps on this (bad) channel, I can bias that side and life is good - full ccw is around 46.5 mA. Note, all of this testing is using the same 2 KT-77s, so I know the tubes are fine. I've swapped the 12AU7s from left to right and eliminated those as the issue. I've swapped out the 12ax7 and that works the same regardless.

Here's the one thing I am able to see. Without any power tubes in the amp, across J4/J5 and J15/J16, I see the 6.3 VAC; however, on J4 and J5 there is a DC component of about 1.3 VDC (measured from J4 to gnd and J5 to gnd). On J15/J16 (the working channel) I see a much smaller DC component - 0.150 VDC.

I've gone so far as to pull the transformer leads on this channel (green) and I see the 6.3VAC and a negligable DC component. I've pulled the wires at J4 & J5 and I do not see any significant DC component - only when things are all connected.

This leads me to think the issue is somewhere on the driver board.

Sorry for the book above, but ANY help on this issue would be beyond appreciated. The ONLY original items left in this amp are the wire wound pots (those appear to be identical when I compare resistances, voltages) and the 3 wire cap (this too appears fine). Everything else is new and wired as the schematics in the diytube manual.

Thoughts on where this DC component is coming from and causing me my troubles?

Thanks!!!

Aaron
DoubleAA
 
Posts: 8
Joined: Fri Jul 31, 2009 1:22 pm

Postby TomMcNally » Fri Jul 31, 2009 2:35 pm

Did you jumper J4 and J5 on the driver board over to the 6.3VAC ?
Check the center tap of the 6.3 winding and make sure the capacitor is
properly connected to ground ...
It sounds like you have an issue with the filament voltage on
that 12AU7.
User avatar
TomMcNally
Darling du Jour
 
Posts: 2729
Joined: Sat Nov 19, 2005 2:19 pm
Location: Northfield, NJ

Postby DoubleAA » Fri Jul 31, 2009 11:03 pm

Thanks for the reply. I have J4 & J5 jumpered to pins 7 & 2 (power tube sockets) on the left side, which ultimately leads back to the transformer (the green leads). I checked the center tap and everything looks fine there - I even resoldered the line and have ensured a proper ground connection. The 3 wire cap also appears to be fine - I removed it and measured the values on the right and left halves (with a capacitance meter) and things are identical (~0.02uF).

I've swapped 12au7s - since the right channel is stable, I moved that 12au7 to the left side (and put the left side's on the right) - to no avail. The right side continued to perform as it should and the left continued to exhibit the same behaviors.

I did notice that the 1.2VDC I see at J4 & J5 is also present on the centertap (green/yellow)

Any other suggestions I might try? I went over things with my ohm meter and as far as I can tell, the left and right channels are the same (measuring various points to ground and comparing with the other channel), but obviously I must be missing something.

Thanks!
DoubleAA
 
Posts: 8
Joined: Fri Jul 31, 2009 1:22 pm

Re: Help Debugging bias (?) problem.

Postby Brik » Sat Aug 01, 2009 8:19 am

Hi Aaron,
Welcome to the forum.

"Abandon all disposable income and spare time ye who enter here." (666) ;)

DoubleAA wrote:The left side, the bias current just climbs up and up, slowly, but in less than a minute, I'm turning things off as it creeps past 100mA. This is without the right channel power tubes. If I pull the coupling caps on this (bad) channel, I can bias that side and life is good - full ccw is around 46.5 mA.

Assuming the output valves are OK, one possible cause of this "slow runaway" is a slow drift of the bias voltage upwards. Have you measured the bias voltages at the power valves' grids (J1 and J2)? Reconnect the the coupling caps, remove the output valves, and measure the voltages at J1 and J2. They should be stable at -35V ~ -40V.

If the voltages drift upwards (less negative), you may have bad coupling caps.

Regarding the DC component on the heater circuit, it is theoretically possible for the heater to act as the directly heated cathode especially if there is an exposed filament segment that is in the line-of-sight of the anode. Basically, the filament becomes the positive terminal of a directly heated rectifier.

/Brik
User avatar
Brik
 
Posts: 204
Joined: Fri Apr 07, 2006 5:30 pm
Location: West of Boston

Postby DoubleAA » Sat Aug 01, 2009 8:53 am

Hi Brik - Thanks for the welcome!

I did as you suggested but I don't think that's the issue. Here's what I found.

J1: -41, -42.1, -42.2 vdc
J2: -40.8, -41.9, -42 vdc
(for comparison on the right)
J22: -41.3, -42.2, -41.9 vdc
J23: -41.1, -41.9, -42 vdc

The 3 values for each test point are at just at turn on, after 5 minutes, and then after another 10 minutes. With the left channel tubes in, the runaway would have long blown the fuse - I get no more than maybe 2 minutes at the absolute longest.

This does bring up a point I noticed initially. At full ccw on the bias pots, the values above are what I get. Obviously a bit higher than spec (though not by much). I put it off to having the new power transformer. With some of the tubes I have that I suspect might be bad, I noted them bad as I couldn't get the bias point anywhere near 50 mA at full ccw - it was always higher. Since I have tubes that work (~46mA at full ccw) I thought the tubes might be suspect. Regardless, I thought about changing the 10k resistor into the bias pots to a lower value to compensate, but wanted to get this working right first. Still, -42 vdc doesn't seem that far off.

Thanks!
DoubleAA
 
Posts: 8
Joined: Fri Jul 31, 2009 1:22 pm

Postby Brik » Sat Aug 01, 2009 10:40 am

Hi,

Perhaps output valves are not getting good, solid contacts at screen/suppressor grid pins?

Right now, the bias circuit is not designed to generate much beyond -45V because too much voltage is wasted across Ra.

Image
Image

Replacing Ra with 5K will give you the voltage down to -57V.

Image
Image

That should be good enough for all but the most "enthusiastic" KT88s. (lol)

Cheers,

/Brik
Last edited by Brik on Sun Feb 28, 2010 10:09 pm, edited 1 time in total.
User avatar
Brik
 
Posts: 204
Joined: Fri Apr 07, 2006 5:30 pm
Location: West of Boston

Postby DoubleAA » Sat Aug 01, 2009 12:42 pm

Nice explanation. I think once I get things working, that's on the (neverending) agenda to give me some extra space when biasing.

I made some additional observations. Without output tubes, and removing the 12ax7, both the right and left J4/J5 & J15/J16 are about the same. I swapped out a handful of different 12ax7s I have around and then went back to observing the disparity in regards to the DC component - so I'm thinking not all of those 12ax7s are bad (if any, of course with the luck I've had...).

I'm not even sure if this is really an issue - it's just that this is the only observable disparity I have right now. I resoldered the left 12au7 and the 12ax7 sockets in case of a cold solder joint. No difference.

I resoldered all of the output tube pins on the left side in case a good connection was not being made as you suggested. This also yielded no difference - within moments of powering up, I see the voltage at the bias points start to sky rocket and I power down as it approaches 1VDC (100 mA).

Frustrating to say the least. Many thanks for the suggestions!!!
DoubleAA
 
Posts: 8
Joined: Fri Jul 31, 2009 1:22 pm

Solved! (??)

Postby DoubleAA » Sat Aug 01, 2009 1:41 pm

Well, I seem to have solved the 'problem.'

:$

The issue was how I was shorting the inputs - I simply had a small wire with two alligator clips that I clipped one to each input. Mostly as I don't have a 'proper' set of shorting plugs. While measuring resistances, I saw that between J7 and J17, what should have been a short, was not. I grabbed an extra rca-rca cable and plugged both ends into the inputs (I had not done this initially as I didn't want a 3 foot antenna hanging off...). Upon doing this, the left side biases just as the right was. Sigh.

If this is what it took to get a working amp, so be it, but man...

I'm letting the tubes cool down a bit and will insert all 4 and see where that takes me. I'm feeling a bit more confident now. If that works, in goes the tube rectifier (which conveniently came in the mail today).

I'm still a bit mystified as why one channel was fine and the other not - but considering I was running out of things to check...

Thanks for all the assistance and ideas and help!
DoubleAA
 
Posts: 8
Joined: Fri Jul 31, 2009 1:22 pm

Postby Brik » Sat Aug 01, 2009 8:40 pm

Hi,

Good find!

The left channel input J7 and left channel bias wiper J6 are right next to each other...

For taking voltage measurements I prefer shorting each input to the ground.
In the absence of shorting plugs take an RCA-to-RCA cable, short out one end by connecting the center pin to the ground shield, plug the other end into the amplifier.

Good luck!

/Brik
User avatar
Brik
 
Posts: 204
Joined: Fri Apr 07, 2006 5:30 pm
Location: West of Boston


Return to st70

Who is online

Users browsing this forum: No registered users and 22 guests