Build coming together nicely - Page 6

jrobson · 21st January 2012, 04:02 PM

Thanks again, when I switch it on it trails B+ into negative, but about 10v lower, once B+ starts to reverse the plates go with, but stuck at 0 while b+ goes to 450v.

jrobson · 21st January 2012, 04:18 PM

If I switch off it goes positive to about 5V or so for a second or two.

jrobson · 21st January 2012, 07:05 PM

I can't find any shorts anywhere, if I adjust the output tube grid the voltage on the plate of the 5482's go to 12-17 while I'm adjusting, then settles back to 0.

jrenkin · 21st January 2012, 07:58 PM

Hmmm...I am not an expert, I hope one will chime in. I suggest trying the resistor instead of the ccs. This won't work as well in the final circuit, but might help you determine where the problem lies. Also, if at all possible, completely disconnect the output stage, even the mosfets, so you can see if it is causing a problem.
Pain in the...

jrobson · 22nd January 2012, 08:46 AM

Thanks, it's not clear exactly what size is required and how to wire it in. Do I need to wire in between B+ and plate?

I'm using the online calculators, the previous one I used was a little too simplified. It appears that 45K is required to for the 450V @ 10ma.

Anyway from here: Voltage Divider Calculator

27.5K/17.5k = 45k

Is this correct?

jrenkin · 22nd January 2012, 12:11 PM

You don't need a voltage divider, just a resistive load. The ccs is a more ideal load, many les complicated amps use simple resistive loads. So all you need is a single resistor between B+ and the plate. The value of this resistor is determined by the B+, the desired plate voltage, and cathode bias, which then determines the current in the tube. Since the current is not constant, the load line on the tube curves is sloped rather than a more ideal horizontal as with a ccs. With a pure resistor load, the tube will operate in a less linear fashion and not achieve the full MU of the tube, which in this case may be inadequate, although may be fine and certainly will for testing.
I ran a simulation on tubecad, and I thought less than 45 K was needed, but my B+ was lower.
It is 6am here and the baby has a fever, so I can only try again a little later.
But if you put in a 45k and set the bias low, you may just get too low a plate voltage. Play with the bias a bit and measure it all. That may be adequate to determine if the problem is the ccs or something downstream.
Be sure the resistor you use is rated enough to handle the voltage drop at the current of the tube. That could require a 3 or even 5 watt resistor.
I have took my of these ccs, I could mail some out, but don't know how that is to S.A. From Denver...

jrobson · 22nd January 2012, 12:43 PM

Thank you again, good luck with the baby.
I have around 450v B+, I'll pick up some big 45K resistors tomorrow and see what happens... Is there a way to test a CCS off board?

Thanks for the offer, let me see if I can confirm where the problem lies and I can purchase the CCS's off you if you have some spare.

jrenkin · 22nd January 2012, 12:59 PM

I'll try and run a simulation again later and let you know some more accurate values to pick up. If I can get the right number, you could try it in the entire circuit to drive the output and see if you can hear music! It wouldn't be optimal in this circuit, but it would be encouraging to be able to hear it.

When I had the problem with my build, by not grounding the 5842 grid leak, I thought is was the ccs so I bought a bunch. I will use some in other circuits, but...they are definitely not $30 each. I paid more like $4, maybe less. And I have too many.

I would measure some voltages on mine, but i didn't use the pcb, rather made a semi- point to point and modular build. Then I jumpered it all together to test and now have the modules apart to build back into the chassis that my wood working skills are slow (and poor) at building and my daughters fevers have stymied me working on. She comes first after all. But side time is good to see if I can help you troubleshoot.

I am curious on your measures on the MOSFET. They seem a bit odd, but solid state devices defy me. That is why I suggested also trying the driver stage completely separate from the MOSFET and output stage.

Oh, since I didn't build on the pcb, I also have two unused pcb's around, since I believe in paying george for his intellectual property, but if somehow you found a fault in a board, or kill it unsoldering things,I could send one of those too.

jrobson · 22nd January 2012, 01:20 PM

I have spare mosfets that I can try out as well but I would like to keep them in case of a rebuild, I've disconnected the coupling caps before and I couldn't get it to work then either, but a few things have changed since then so I'll give it another shot, I quite fancy the idea of a resistor to bypass the CSS just to see if anything else is wrong and to pinpoint the problem.

It looks like I might have to rebuild this if I want it working, a bit of a pain in the *** but maybe less hassle than to restart something else completely from scratch considering the investment in the power transformer, tubes, OPT's and caps, the lounge can look like a warzone for a few more weeks I suppose...

jrenkin · 22nd January 2012, 01:31 PM

can you disconnect the mosfets and OPTs from B+ as well to isolate the driver stage? I doubt there is a problem from the output stage,but that would prove it.
Baby and mom went back to sleep,I will try and run the simulation, but it means rebooting the old computer...

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Dolby AC3 coming on nicely.	Doz	Tubes / Valves	13	21st September 2011 05:41 AM
Coming soon to a thread near you...	Cal Weldon	Multi-Way	57	20th May 2011 12:03 AM
Delta3 coming along ...	David Gatti	Multi-Way	18	11th July 2007 06:06 AM

New To Site?	Need Help?
Register to Participate Search Privacy Statement Contact Us	Frequently Asked Questions Did you forget your password? Mark Forums Read

21st January 2012, 04:02 PM	#51
jrobson diyAudio Member Join Date: Aug 2011	Thanks again, when I switch it on it trails B+ into negative, but about 10v lower, once B+ starts to reverse the plates go with, but stuck at 0 while b+ goes to 450v.

21st January 2012, 04:18 PM	#52
jrobson diyAudio Member Join Date: Aug 2011	If I switch off it goes positive to about 5V or so for a second or two.

21st January 2012, 07:05 PM	#53
jrobson diyAudio Member Join Date: Aug 2011	I can't find any shorts anywhere, if I adjust the output tube grid the voltage on the plate of the 5482's go to 12-17 while I'm adjusting, then settles back to 0.

21st January 2012, 07:58 PM	#54
jrenkin diyAudio Member Join Date: Nov 2007 Location: Denver, Colorado	Hmmm...I am not an expert, I hope one will chime in. I suggest trying the resistor instead of the ccs. This won't work as well in the final circuit, but might help you determine where the problem lies. Also, if at all possible, completely disconnect the output stage, even the mosfets, so you can see if it is causing a problem. Pain in the...

22nd January 2012, 08:46 AM	#55
jrobson diyAudio Member Join Date: Aug 2011	Thanks, it's not clear exactly what size is required and how to wire it in. Do I need to wire in between B+ and plate? I'm using the online calculators, the previous one I used was a little too simplified. It appears that 45K is required to for the 450V @ 10ma. Anyway from here: Voltage Divider Calculator 27.5K/17.5k = 45k Is this correct?

22nd January 2012, 12:11 PM	#56
jrenkin diyAudio Member Join Date: Nov 2007 Location: Denver, Colorado	You don't need a voltage divider, just a resistive load. The ccs is a more ideal load, many les complicated amps use simple resistive loads. So all you need is a single resistor between B+ and the plate. The value of this resistor is determined by the B+, the desired plate voltage, and cathode bias, which then determines the current in the tube. Since the current is not constant, the load line on the tube curves is sloped rather than a more ideal horizontal as with a ccs. With a pure resistor load, the tube will operate in a less linear fashion and not achieve the full MU of the tube, which in this case may be inadequate, although may be fine and certainly will for testing. I ran a simulation on tubecad, and I thought less than 45 K was needed, but my B+ was lower. It is 6am here and the baby has a fever, so I can only try again a little later. But if you put in a 45k and set the bias low, you may just get too low a plate voltage. Play with the bias a bit and measure it all. That may be adequate to determine if the problem is the ccs or something downstream. Be sure the resistor you use is rated enough to handle the voltage drop at the current of the tube. That could require a 3 or even 5 watt resistor. I have took my of these ccs, I could mail some out, but don't know how that is to S.A. From Denver...

22nd January 2012, 12:43 PM	#57
jrobson diyAudio Member Join Date: Aug 2011	Thank you again, good luck with the baby. I have around 450v B+, I'll pick up some big 45K resistors tomorrow and see what happens... Is there a way to test a CCS off board? Thanks for the offer, let me see if I can confirm where the problem lies and I can purchase the CCS's off you if you have some spare.

22nd January 2012, 12:59 PM	#58
jrenkin diyAudio Member Join Date: Nov 2007 Location: Denver, Colorado	I'll try and run a simulation again later and let you know some more accurate values to pick up. If I can get the right number, you could try it in the entire circuit to drive the output and see if you can hear music! It wouldn't be optimal in this circuit, but it would be encouraging to be able to hear it. When I had the problem with my build, by not grounding the 5842 grid leak, I thought is was the ccs so I bought a bunch. I will use some in other circuits, but...they are definitely not $30 each. I paid more like $4, maybe less. And I have too many. I would measure some voltages on mine, but i didn't use the pcb, rather made a semi- point to point and modular build. Then I jumpered it all together to test and now have the modules apart to build back into the chassis that my wood working skills are slow (and poor) at building and my daughters fevers have stymied me working on. She comes first after all. But side time is good to see if I can help you troubleshoot. I am curious on your measures on the MOSFET. They seem a bit odd, but solid state devices defy me. That is why I suggested also trying the driver stage completely separate from the MOSFET and output stage. Oh, since I didn't build on the pcb, I also have two unused pcb's around, since I believe in paying george for his intellectual property, but if somehow you found a fault in a board, or kill it unsoldering things,I could send one of those too.

22nd January 2012, 01:20 PM	#59
jrobson diyAudio Member Join Date: Aug 2011	I have spare mosfets that I can try out as well but I would like to keep them in case of a rebuild, I've disconnected the coupling caps before and I couldn't get it to work then either, but a few things have changed since then so I'll give it another shot, I quite fancy the idea of a resistor to bypass the CSS just to see if anything else is wrong and to pinpoint the problem. It looks like I might have to rebuild this if I want it working, a bit of a pain in the *** but maybe less hassle than to restart something else completely from scratch considering the investment in the power transformer, tubes, OPT's and caps, the lounge can look like a warzone for a few more weeks I suppose...

22nd January 2012, 01:31 PM	#60
jrenkin diyAudio Member Join Date: Nov 2007 Location: Denver, Colorado	can you disconnect the mosfets and OPTs from B+ as well to isolate the driver stage? I doubt there is a problem from the output stage,but that would prove it. Baby and mom went back to sleep,I will try and run the simulation, but it means rebooting the old computer...