Tubelab Se Bias current jumping

[jamicomon]

I was going though the Tubelab Se checkout.

Everything is well now that i've got the diodes replaced and in the right polarity. (I was the R5 going in fusion guy)

But as soon as I turn the bias pots on either channel for more than 1/8 of a turn, the output tube bias jumps to over 150mA. Very sudden.

The only change I did was lower C4 to 2.2uf to go from 450v to 403v


Here are my stats:

Hammond 376x / JJ 300b

B+: 403v

negative bias: -283v

5842 voltage: 169v each

output tube grid voltage at lowest bias: -141v


Case shot:
https://www.dropbox.com/s/dbg5ycsvij5fp62/2014-12-14 13.11.56.jpg?dl=0


Board:
https://www.dropbox.com/s/50okupf2cnvvhyi/2014-12-14 17.32.28.jpg?dl=0

https://www.dropbox.com/s/y9erxjy0corghya/2014-12-14 17.32.21.jpg?dl=0

 

[Evenharmonics]

Bias adjustment for TSE is very sensitive. Only a tiny turn will either shoot up or down the bias. I turn mine in fraction of millimeter at a time while monitoring the voltmeter.

 

[jamicomon]

Thank you for replying Evenharmonics, but I cannot go more than 20mA before it jumps to 200mA. Even with the slightest adjustment.

I will try to replace the mosfets tonight as well as install a bigger heatsink on them and then report back.

 

[jamicomon]

Just done replacing the mosfets. They were probably still good.

Here is the new heatsink (old thermalright ram cooler to stay with the PC theme) :

https://www.dropbox.com/s/f4vfjd5b50hhz2i/2014-12-15 22.16.16.jpg?dl=0


My issue it seems, is that there has to be music playing for the bias to be stable. Whenever my cell phone stops playing and goes in standby, the bias shoots up like crazy. (Same when I power off the amp, but that's probably normal)

Any ideas?

 

[jamicomon]

So my right channel bias is adjustable and stable but as soon as the left goes past 5mA, both channels shoot up to about 150mA. Lowering the left channel back to its lowest bias returns the right channel to correct bias.

Swapped all semiconductors in the circuit except both CCS.

New bias pots installed. Tried to reflow most soldering pads. Still no luck.

Used resistor legs to make direct connections to components around the R5 area to make sure pcb damage is not the cause.

Here is a shot of the top of the board:

https://www.dropbox.com/s/09emrm8s6zw9fms/2014-12-16 19.14.10.jpg?dl=0


Please help.

 

[jamicomon]

Tried swapping output tubes from left channel to right. Same issue.

I can bias only one channel correctly but as soon as I start to bias the second one both bias currents jump to 150mA.

Will try to make negative bias voltage more positive tonight by adjusting C6. (Currently -296v)

 

[Evenharmonics]

Did your "check out" process go the same way as posted on Tubelab website?

 

[jamicomon]

Yes, the rest of the checkout did go as planned.

I'm thinking out loud..

My mosfets Q1 & Q2 (2sk3563) are not rated to go above 500v.

With my B+ at 410v and the output tube grid being at -95v at around 75mA bias current.

The voltage potential difference between its drain-source is 505v. Could this be the issue?

 

[Evenharmonics]

Can you list the voltages per each "check out" step as shown here? I'm trying to see how yours would compare to mine.

 

[jamicomon]

I didn't have enough tube time to go through checkout again but i did try to remove both mosfets to check and see if I can still adjust negative grid voltage as suggested by George to you here (post 5):

http://www.diyaudio.com/forums/tube...ab-se-finished-but-need-help.html#post1458429

Negative grid voltage won't budge with both bias pots using no mosfets, so they should be good.

I will try to lower B+ a little more by adding a resistor in series with the choke R4 (currently 400v) on saturday and go through the checkout again.

Thanks again for helping out Evenharmonics.

 

[JP]

Where did you get the MOSFETs? 3563 have been generally unobtanium for some time.

 

[jamicomon]

I bought two batches. Both from ebay.

One located in the U.S. and one from china.

I figured at least one of them would be good.

I tried some from both batches. They do work but I can't say if they meet their 500v Source-Drain spec.

 

[JP]

Three years ago I bought some 3563 off eBay - there was only one listing for them from a US seller and they were all fake. Took him better than three weeks to ship, so my guess was he didn't even have them on-hand and sourced them from China.

My bet would be your MOSFETs are fakes. I'd get a current part from a trusted distributor.

 

[jamicomon]

Changed coupling caps, Lowered loaded B+ by 10v, still can't set bias.

Here is how the checkout procedure went:

1) w/ no tubes in,
B+ = -0.68v
B- = -305.8v
Output tube filament supply left = 4.98v, right = 4.98v

2) w/ rectifier tube in,
B+ = 440v
B- = -303.2v
Output tube grid to ground left =-156.2v , right = -147.8v

3) w/ rectifier tube and input tubes in,
B+ = 394.5v
B- = -301v
Input tube plate to ground left = 174.9v, right = 175.1v

4) w/ all tubes in and set the output tube bias,
B+ = 390v
B- = -295v
Input tube plate to ground left = cannot set bias, right = cannot set bias

Like jpjones3318 said, my mosfets are probably fakes. Anyone know a current and suitable model I can source from mouser?

 

[jamicomon]

I ended up going for some toshiba 2sk3565 from mouser.

They are really stable and adjust perfectly.

The amp sound very powerful and is insanely detailed.

The 5k tap is VERY sufficient for driving my klipsch RF-82 to LOUD volume.


My B+ is sagging now.

I can only get 330v loaded.

C4 is at 47uf (5AR4) and no dropping resistors in series with R4.

Anything I can try to bring up my B+

 

[Evenharmonics]

My B+ is sagging now.

I can only get 330v loaded.

C4 is at 47uf (5AR4) and no dropping resistors in series with R4.

Anything I can try to bring up my B+

Is 47uf the max recommended for C4? I thought it was higher.
Do you use choke for R4?

 

[jamicomon]

60uf is the max recommended for C4 and an 5AR4 rectifier.

I use the CX-14 triad choke with 150ohm dc resistance.

My hammond 376x maxes out 173mA so it might be out of juice.

I played with PSUDII last night and I don't think I can do much to help besides changing the transformer.

I will remove a 500uf cap that I had on C5 to help ease the burden on my rectifier.

I do have a 90-140v adjustable step-up isolation transformer hooked up on the amp so I can boost it up a bit. (120v instead of 115 for an added 10v B+)


Also, once I'm done messing around with B+ I will post back here for a comparison between the sound of 2sk3565 and 2sk3566. (If there is one)

 

[JP]

How much current are you running through the output tubes? How much B+ are you after?

 

[jamicomon]

Right now 80mA @ 320v B+

I was looking to get 80mA @ 380-400v B+

If I increase bias current then my B+ drops even more, so it's pointless..

 

[JP]

IIRC, with the Edcor XPWR-131 PT and all BOM values in the PS I could run 80mA @ ~390vdc. No free lunch - seems the best path is to get an appropriately sized PT.