Build coming together nicely

[Alexontherocks]

then buy a new board, sit calmly, get every component ready, plenty of beer and start soldering

[jrenkin]

Or just cobble it together to test with some perf board and alligator clips. Just be careful.

Tubecad suggest that a plate resistor of 25-30k with a bias resistance adjusted to just under 400 ohm would suffice, but you will have to balance for the 5842 variation. can't seem to print it...

I will look into shipping Monday. I am happy to help since I haven't been able to get to mine, it keeps me interested. When I got mine running in test set up, it really did sound so very nice. Worth getting it done!

[jrobson]

Okay I got the one side adjusted to 175v as well, didn't think they would be that off each other, anyway, split my resistors up into 43K again, so I'm down to 120-130v and adjusted the current to 70mA, got it working... On one side only, it could be that the other side doesn't have enough input, I'll give it another go again tomorrow when I get some 32K resistors and see if I can get both sides going...

Thanks a lot for all the help... seems like at least some of the problem is with the CCS not sure exactly why the new ones are bust, maybe I overheated them when soldering.

[tubelab.com]

I have not had the time to dig into this thread with my work schedule lately. At first I saw some pictures of the most amazing looking chassis I have seen. Now I see a discussion about a non functioning board. I apologize for not helping but most of my access has been limited to my smartphone at work.

First off, if you can't make the board work, and aren't sure what to do next, you can send me the board and I will make it work. It will only cost you the actual shipping charges both ways unless there are parts needed that I must purchase (unlikely). This has only been neccessary twice in Tubelab's history. If this is a viable option, PM or email me for details.

It seems like the first stage has some real problems.

70 ma is far too much for a 5842 and will fry it in a short time. It needs to run at about 10 mA and that is near the upper end of the safe dissipation range.

Testing should be done with no signal applied to the amp. Your music source must not have any DC offset on its output, any DC will make it impossible to adjust the plate voltage. Disconnecting the source during testing removes this possibility.

Connecting or disconnecting the coupling cap should not affect any DC readings.

The plate voltage should be near the B+ voltage with no 5842 installed. It should drop when the tube is installed and some adjustment should be afforded by the pot in the cathode circuit.

The CCS can be tested by connecting a current meter in series with a 2 watt 15K resistor and connecting this combination from the plate pin of the 5842 to ground with no tube installed. The current meter should read about 10 mA. The 15K resistor is needed to protect the meter in the case of a shorted CCS. It does not need to be exactly 15K.

If 10M45's can't be easilly obtained the Supertex DN2540 can be used. The 330 ohm current set resistor needs to be changed to get 10 mA. The value will be around 200 ohms.

This many problems could be called by oscillation. This was the case in the first returned board. The customer had mounted the 300B tube sockets off board and the wire lengths were making resonant circuits causing massive oscillation resulting in blown mosfets. Ferrite beads and extra stopper resistors were needed to kill the oscillation.

[jrobson]

Hi

Thank you for your response, sorry I wasn't clear enough, I meant I biased the 300b's at 70mA (R18&R29). At the point I was at last night I was only able to get 120/130v at the 5842's with 43K resistors in place of the CCS's, with the 300b's in this dropped some 10v. I will pick up some resistors today and see if I can get it to 175v/10mA tonight. But with the current setup it plays, only one side but it is working well on that side, sounds very good even at high volumes(surprisingly loud as well), no hum or hiss, it does cut out when pushing the volume but I think this will be fixed once I get the voltage and current on the 5842's up. Thanks for the help on testing the CCS's I'll give that a shot as well tonight.

I have the mosfets and the CCS's mounted off board about 6" away, not sure if this could cause any oscillation.

I still would like to rebuild this board anyway, it's looking quite ugly with so many parts already removed and replaced, some delaminations, burn marks from the blown fuse earlier etc. If the new one still has a problem then I'll certainly take you up on your offer, thanks.

[jrobson]

Okay with 15K resistor here is what happens, no 5842's or 300b's installed.

Plate voltage follows B+ negative, B+ only goes to about -5V, current at this point over the resistor is -10.8mA.

Then B+ starts to go towards zero and plate volts follow while current drops. B+ goes positive but plate volts and current is stuck at 0.

[jrobson]

For interest sake with 46K resistor 9.6mA, probably best to go no lower than 45K for testing...

[jrobson]

Okay here we go, version 2.

Big thank you to Mr. Jrenkin for the parts. I got the sockets yesterday and did the rebuilt, took the whole day, was very tricky getting the components off the old board.

Anyway I picked up some connector blocks from the electronics store for the power lines but I got the wrong size, so I decided to break them open, I ended up placing these connectors everywhere a wire is to be connected to the board. Since a lot of my caps are off board I have a whole lot of connectors

[jrobson]

Got it working after the 3rd pair of 5842's, first and second pair had a faulty tube that kept blow CCS's and mosfets. Would not build something again that relies on rare NOS.

[jrenkin]

Wow, bad luck. I tried 2 different pairs of 5842s and none were bad. Of course, my build is completely stalled at the moment.

So how does it sound? Post a picture of the finished amp. Everyone was impressed with the chassis so I am sure there is interest in seeing it. Hopefully a little praise from your peers and a lot of fine listening will soften the trials and tribulations of the build.

When I get back to it and measure a few things, I am going to try D3a's in it. I think it will be easy, only needing a different socket wired as the pinout is different, but it may be fine at the same operating parameters or just need a little tweaking. My main concern would be frequency response, but I think the CCS will help in that regard.

If I ever get around to it, I will post findings. May be a while though...