[sourced]

Woo hoo! I appear to be somewhat successful (finally)! First, here's the money shot (well, it's mirrored but whatever):

Now, I'm not perfect, so there appear to be a couple issues/notes:
- It hums very quietly (it's not disturbing, but it is audible when no music is playing). Hum goes up when I put my finger near the tube, but silent when I touch the case. Judging from other posts in the thread, I should check all my grounds. I don't have a pic yet, but I have a ground for all the left channel circuit, and a ground for all the right channel circuit (two star grounds, more or less). Then the two grounds are connected through the case (aluminum). The RCA jacks have their 'cases' touching the case, so I figured that would be ground enough (appears to be similar to how pmillett did his). Should I run wires to the ground lugs?

- Right channel is louder than left, since my $3 pot isn't very good. I know turned off, resistance reads ~48.5kΩ through right, ~50.4kΩ through left. Would it be ok to solder in a 2kΩ resistor at the right RCA jack? Obviously when right is around ~20kΩ left might be ~21kΩ, and as I turn it up I'd actually have the left channel louder. I'll see if I have a better pot lying around, but what is the general fix? I panned my computer output left but I'm not sure I can pan left with digital out (when I use a DAC).

- I wired the pot backwards, so the audio taper is all messed up (very sensitive at the low end), which I will fix.

- The mosfets get _really_ hot, and the whole case warms up (good thing my case is an aluminum heatsink). Awesome.

- Watching the tubes heat up together just about makes my day

Regardless, this is pretty amazing. I've been listening for about an hour total (not burned in yet) and it doesn't sound horrid, so... success!

 

[Llama16]

Quote:

Originally Posted by tomb /img/forum/go_quote.gif Over time, ABSOLUTELY. Clean it off. All it takes is an old toothbrush, some paper towels and some 90% or better isopropyl alcohol that you can obtain (very cheaply) at any corner drugstore or Walmart. Use the toothbrush by dipping it into a bowl filled with the alcohol. Scrub the entire board until it's completely coated with alcohol and you've scrubbed down the biggest lumps. Pat up the liquid with paper towels. Repeat as necessary (usually takes 4-6 times). Use several rinses until all of the white powder is gone and your paper towels no longer pick up yellow stains.

:S, I won't argue with you and I'm sure it's for the best. And I wish I had some of that isopropyl alcohol at hand, allthough my fear of breaking a joint when cleaning the board makes me want to be very sure that I really have to as well.

FELDER GmbH | Solder wires for the electronic | Flux cored soft solder wires ISO-Core® This is the solder I'm using: It's the RA version, 60tin and 40lead with 2.5% flux.

At the bottom of that info page it says: "The fluxes have a high temperature resistance! The bright, solid flux residues of these solder wires do not result in corrosion on non-iron metals. Therefore, they can remain on the soldering joint"

 

[tomb]

Quote:

Originally Posted by Llama16 /img/forum/go_quote.gif :S, I won't argue with you and I'm sure it's for the best. And I wish I had some of that isopropyl alcohol at hand, allthough my fear of breaking a joint when cleaning the board makes me want to be very sure that I really have to as well. FELDER GmbH | Solder wires for the electronic | Flux cored soft solder wires ISO-Core® This is the solder I'm using: It's the RA version, 60tin and 40lead with 2.5% flux. At the bottom of that info page it says: "The fluxes have a high temperature resistance! The bright, solid flux residues of these solder wires do not result in corrosion on non-iron metals. Therefore, they can remain on the soldering joint"

Yes, well - corrosion is different. You asked if it would conduct - over time, yes.

Just another note, but if one if your solder joints comes undone from simple scrubbing with a toothbrush and alcohol, then it was too weak to make good electrical contact anyway. You'll be discovering problems that are easier to fix now rather than later.

 

[Llama16]

You're right, I'll stop my search for excuses and see if I can get my hands on some.
Thanks for letting me see the light

, I sometimes need a slap so I don't 'stray' off.

 

[the_equalizer]

Quote:

Originally Posted by sourced /img/forum/go_quote.gif Woo hoo! I appear to be somewhat successful (finally)! First, here's the money shot (well, it's mirrored but whatever): <snip> Now, I'm not perfect, so there appear to be a couple issues/notes: - It hums very quietly (it's not disturbing, but it is audible when no music is playing). Hum goes up when I put my finger near the tube, but silent when I touch the case. Judging from other posts in the thread, I should check all my grounds. I don't have a pic yet, but I have a ground for all the left channel circuit, and a ground for all the right channel circuit (two star grounds, more or less). Then the two grounds are connected through the case (aluminum). The RCA jacks have their 'cases' touching the case, so I figured that would be ground enough (appears to be similar to how pmillett did his). Should I run wires to the ground lugs? - Right channel is louder than left, since my $3 pot isn't very good. I know turned off, resistance reads ~48.5kΩ through right, ~50.4kΩ through left. Would it be ok to solder in a 2kΩ resistor at the right RCA jack? Obviously when right is around ~20kΩ left might be ~21kΩ, and as I turn it up I'd actually have the left channel louder. I'll see if I have a better pot lying around, but what is the general fix? I panned my computer output left but I'm not sure I can pan left with digital out (when I use a DAC). - I wired the pot backwards, so the audio taper is all messed up (very sensitive at the low end), which I will fix. - The mosfets get _really_ hot, and the whole case warms up (good thing my case is an aluminum heatsink). Awesome. - Watching the tubes heat up together just about makes my day Regardless, this is pretty amazing. I've been listening for about an hour total (not burned in yet) and it doesn't sound horrid, so... success!

Congratulations ! It is indeed a warm fuzzy feeling to watch those tubes light up and hear sweet music coming out of the 'phones!!

About the hum, I've found some 19J6's to be much more susceptible to hum than others; specially if you use high sensitivity 'phones. Swap the tubes and see if the hum follows the tube. If it does, well: it's the tube that it's at fault.

On the channel imbalance, indeed, the most 'general' solution would be to use a better pot

. Short of that you can try adding a couple input resistors as done in the PCB version. Take a look at the schematic. This solution won't make the imbalance dissapear, but it'll make it much less noticeable.

cheers !

 

[Llama16]

Sorry to disrupt again, but I want to be sure before I go on.
I went to the farmacy, where I bought myself about 100ml of isopropyl alcohol. The farmacist wanted to know what I am gonna use it for (don't blame her ofcourse

).

I just cleaned of the back of the board for the first time and yes, I noticed the flux dissolving, allthough the board is now covered with a rather mat, dull residue that sticks a little. I'm guessing that's the dissovled flux.
What's the best way to get rid of it?

Also, i'd like to see if I got a connection between the upperside of the RLED and the main ground whole in the middle, I have the caps installed but I get a resistance readin that fluctuates. Is this due to the capacitor that is also connected to the LED?

Thanks
Dries

 

[tomb]

Quote:

Originally Posted by Llama16 /img/forum/go_quote.gif Sorry to disrupt again, but I want to be sure before I go on. I went to the farmacy, where I bought myself about 100ml of isopropyl alcohol. The farmacist wanted to know what I am gonna use it for (don't blame her ofcourse ). I just cleaned of the back of the board for the first time and yes, I noticed the flux dissolving, allthough the board is now covered with a rather mat, dull residue that sticks a little. I'm guessing that's the dissovled flux. What's the best way to get rid of it?

Go back to my post where I described the procedure.

It's going to take 4-6 rinses. As you noted, the alcohol dissolves the flux. What not many people will tell you though, is hitting it with alcohol and a toothbrush again - with no other action - will not accomplish anything but to continue to spread the flux around. What you need to do is dab up the dissolved flux and alcohol with a paper towel while it's still wet. Otherwise, where can it go? the alcohol will dry, but the flux will still be there.*

When you dab while the alcohol is still wet and the flux dissolved, you should see the flux on the paper towel as dirty yellow stains. Repeat the rinsing and dabbing up with the paper towel for as many times as it takes, allowing the board to mostly dry inbetween. After 4-6 times, the PCB should be pretty clean.
Quote:

Also, i'd like to see if I got a connection between the upperside of the RLED and the main ground whole in the middle, I have the caps installed but I get a resistance readin that fluctuates. Is this due to the capacitor that is also connected to the LED? Thanks Dries

I can't tell what you're measuring from what you describe. The best way is to put one probe on the wire lead (I'm assuming you can reach the tip.) That's supposed to be ground. Then place the other probe far removed from that area, but on a ground point. You should get no variation and it should measure 0 ohms. Keep in mind that if you're not an awful lot experienced in measuring with a DMM, the process itself is flawed by how you hold the probes and will cause variation. In other words, you may not be able to keep the target centered at all times with your hands. As long as you read zero sometimes and can repeat it with some assurance while you're touching the wire tip and not the surrounding pad, then you have a successful ground connection.

Note that you can still leave the amp case dissassembled and turn the power on. As long as the tubes are in and you have the MOSFET heat sinks attached, it should be OK. Certainly, you can tell within a second or less if the LEDs will light. That might be the easier way to do it.

* One other method is to turn the board on its side as if you're scraping the dissolved flux down off of the edge of the board, allowing it to drip off. This is much more messy and not as positive as the paper towels, IMHO.

 

[Llama16]

Thanks, I read the procedure (alot!

)and I was just doubting what rinses meant. I went ahead and cleaned it a second, third, forth time or so and yes, you're right now it's alot better

.

About the measuring: That's about what I did with all my the connections, I was wrong when mentioning 'connection' I meant more of a 'solder bridge' allthough there isn't any visible solder (just making sure

)

 

[Llama16]

Ok guys, bad news :S

I assembled it and fired it up for the first time.
I saw the Tube LEds shine rather dim, then they fainted out, the tubes didn't light up...

I know i'll take some measurements but does anyone maybe already now how this happened? The second time nothing happened.

 

[the_equalizer]

Quote:

Originally Posted by Llama16 /img/forum/go_quote.gif Ok guys, bad news :S I assembled it and fired it up for the first time. I saw the Tube LEds shine rather dim, then they fainted out, the tubes didn't light up... I know i'll take some measurements but does anyone maybe already now how this happened? The second time nothing happened.

Sounds like there's a short to ground. The faint light up in the LED's was the power supply just before it shut down due to it's short protection.

Check your MOSFET's. See that you mounted them correctly (i.e. there MUST be a plastic shoulder between the MOSFET tab and the nut, there MUST be a thermal pad isolating the MOSFET from the case, etc.)

cheers!

 

[Llama16]

yeah the mosfets should be fine: shoulder in place abd thermal pads in place, just like tomb instructed I checked the resistance between the MOSFETs iron 'backplate' (dunno the name) and their nuts, not the resistance is higher then 2000k (my meter can't go higher, so I'm guessing there isn't a short.

I'd also like to say that I'm not sure that it's the short protection of the PSU, as it doesn't cycle. it happened once when I booted it up for the first time, then when I shut it down again (even when I pulled out the plug for a min or two) and then fired ti up again, nothing happened, no fainting of the LEDs, nothing.
The tubes don't fire up, the LEDS don't light (and there isn't a smell of burning stuff nor do the heatsinks get warm (I don't know how fast they get warm though)).

Thanks for the help. Really don't know what to do

 

[the_equalizer]

Quote:

Originally Posted by Llama16 /img/forum/go_quote.gif yeah the mosfets should be fine: shoulder in place abd thermal pads in place, just like tomb instructed I checked the resistance between the MOSFETs iron 'backplate' (dunno the name) and their nuts, not the resistance is higher then 2000k (my meter can't go higher, so I'm guessing there isn't a short. I'd also like to say that I'm not sure that it's the short protection of the PSU, as it doesn't cycle. it happened once when I booted it up for the first time, then when I shut it down again (even when I pulled out the plug for a min or two) and then fired ti up again, nothing happened, no fainting of the LEDs, nothing. The tubes don't fire up, the LEDS don't light (and there isn't a smell of burning stuff nor do the heatsinks get warm (I don't know how fast they get warm though)). Thanks for the help. Really don't know what to do

Voltages please

at the power supply connector, at the MOSFET pins. Also measure the voltage at the PSU when it is NOT connected to the amp.

cheers !

 

[Llama16]

Ok

You're right I'll be measuring some voltages, I'm quite new at this and I know that those voltages and even pictures are necessary to be able to help. But I first have a little discovery:

I got the amp back dissassembled (I got it out of the case) and now the leds work fine, they start up in 2 seconds or so and when turned off they darken in abotu 3 seconds.
I see the tubes lighting up (I didn't expect that they don't light up so bright though, but that's maybe my inexperience) and the heatsinks getting warm.
I'm not able to test whether it works as an amp or not (I can't find my analog cable and probably have to make one)

It's really a mistery here :S probably a short but I really can't find it right away (all my cable are insulted, I am 10percent sure there is not single cable left open for even a mm)

I already checked the voltage of the PSU, and the strange thing is: it's 60V (I'm on the belgian 'european' network) while I think it needs to be 48. This could be my inexperience of using a MM, so i'll quickly explain what I did. I placed the black wire of the MM (which is connected to the 'COM' of the device) on the outer side of the PSU connecter (below the yellow insulater) and put the red wire (which is set on the 1000VDC/750VAC/200mAmax conenctor of the MM) inside of the PSU connector, then I got a positive reading of 60V (the scale I used was 200DCV)

I got to go catch some sleep, it's been a long day, but I'll see if i can wake up for a sec to see the responses, allthough the measurements need to be for tomorrow.

Thanks alot equalizer (I think I've added you on MSN, but didn't have the time to say hello yet)

 

[the_equalizer]

Quote:

Originally Posted by Llama16 /img/forum/go_quote.gif Ok You're right I'll be measuring some voltages, I'm quite new at this and I know that those voltages and even pictures are necessary to be able to help. But I first have a little discovery: I got the amp back dissassembled (I got it out of the case) and now the leds work fine, they start up in 2 seconds or so and when turned off they darken in abotu 3 seconds. I see the tubes lighting up (I didn't expect that they don't light up so bright though, but that's maybe my inexperience) and the heatsinks getting warm. I'm not able to test whether it works as an amp or not (I can't find my analog cable and probably have to make one) It's really a mistery here :S probably a short but I really can't find it right away (all my cable are insulted, I am 10percent sure there is not single cable left open for even a mm) I already checked the voltage of the PSU, and the strange thing is: it's 60V (I'm on the belgian 'european' network) while I think it needs to be 48. This could be my inexperience of using a MM, so i'll quickly explain what I did. I placed the black wire of the MM (which is connected to the 'COM' of the device) on the outer side of the PSU connecter (below the yellow insulater) and put the red wire (which is set on the 1000VDC/750VAC/200mAmax conenctor of the MM) inside of the PSU connector, then I got a positive reading of 60V (the scale I used was 200DCV) I got to go catch some sleep, it's been a long day, but I'll see if i can wake up for a sec to see the responses, allthough the measurements need to be for tomorrow. Thanks alot equalizer (I think I've added you on MSN, but didn't have the time to say hello yet)

Glad it's working ! ... Now you just have to find that short when the circuit is in the case

The voltage from the PSU does not seem normal at all; my PSU even gives slightly less than 48 volts. The MOSFETs can take up to 100 volts drain-to-source so, no problem there, but I'm not sure about the gate-to-source..

Anyway, have a good night and let us know the voltage measurements tomorrow.

EDIT: BY THE WAY. I think the most common source of shorts when the circuit is in-case, is the standoff and the LED resistor. Check out TomB's step-by-step build thread.

cheers !

 

[tomb]

Thanks, equalizer!

Yes, casing an amp up and finding a short is a common problem. However, I'm wondering if you dissassembled it, maybe the center standoff was still connected? If so, it might be an issue with the connectors/wiring on the back plate. Do you have any long leads and when the slack wire is bunched up when you bolt up the case, is there a possibility that some exposed leads can come into contact with something? Could the leads bind, etc., etc.?