alco,
I set up my se just as you ended up except for the mosfets. I used 2sk3565
data sheet http://www.toshiba.com/taec/components2/Datasheet_Sync//77/10779.pdf
in light of this thread they seems less then ideal. I bought the parts before this thread developed. none the less my amp still sounds great. I wonder if it is worth removing the parts to change them.
Anyway have fun with your new toy...Evan
I set up my se just as you ended up except for the mosfets. I used 2sk3565
data sheet http://www.toshiba.com/taec/components2/Datasheet_Sync//77/10779.pdf
in light of this thread they seems less then ideal. I bought the parts before this thread developed. none the less my amp still sounds great. I wonder if it is worth removing the parts to change them.
Anyway have fun with your new toy...Evan
Evan,
Congratulations on your new amp !
It's too bad we're on opposite coasts, it would be interesting to hear our amps side by side.
Perhaps someone more knowledgeable can comment on what sorts of differences we should look for in the sound from using different MOSFETs. I think someone brought up phase distortion earlier in the thread.
Congratulations on your new amp !
It's too bad we're on opposite coasts, it would be interesting to hear our amps side by side.
Perhaps someone more knowledgeable can comment on what sorts of differences we should look for in the sound from using different MOSFETs. I think someone brought up phase distortion earlier in the thread.
I doubt it. THe Toshiba spec you referenced only goes to 100 volts. The reverse transfer capacitance is 10 pF at 100 volts, The Tubelab SE runs the mosfets well above 100 volts so the capacitance will be low.
I have had several theoretical discussions with the "solid state is evil" crowd about Phase Intermodulation Distortion. It is definitely possible for a device whose capacitance varies with the applied signal to cause a varying phase shift with applied signal creating PIM.
ANY solid state device can cause PIM. SO CAN MOST TUBES!
It is convienient that PIM can not be easilly measured. It is also convienient that none of the SS=evil crowd can tell me what PIM sounds like. I would think that PIM would cause a loss of high frequency detail. A capacitance that varies a few pF with the signal voltage will only affect the high frequencies.
Back when I was developing the Tubelab SE design I stuffed several different mosfets into the board. I tried some rather large ones just to see what happened. The largest fets seemed to cause a general HF rolloff, but I never heard anything that I could call PIM. My hearing was a lot better then than it is now too. The TSE does not suffer from lack of detail!
If it works and you like it don't mess it up. Even if (more like when) you get the urge to tinker, wait until you have plenty of listening experience. After a while you will pick out some musical favorites to use for auditioning changes.
You want find music that "thrills you", stuff that sounds amazing and couldn't possibly sound better. You also want to find stuff that could sound better. Then audition that music on several other systems to see if it indeed can be better, or the good stuff can be messed up. I don't know how much time I have spent trying to clean up bad sound that is in the source material.
Some music sounds best on a SE amp and the TSE is one of the best, some music sounds best on a P-P amp. I have not found one amp that fits all music.
You want find music that "thrills you", stuff that sounds amazing and couldn't possibly sound better. You also want to find stuff that could sound better. Then audition that music on several other systems to see if it indeed can be better, or the good stuff can be messed up. I don't know how much time I have spent trying to clean up bad sound that is in the source material.
Some music sounds best on a SE amp and the TSE is one of the best, some music sounds best on a P-P amp. I have not found one amp that fits all music.
Bad News !
Well, I got seven harmonious months with my new TSE.....and then it blew up. Or at least it's blowing fuses. And I suspect my MOSFETS.
I haven't done much troubleshooting yet, but the amp powers up, warms up the filaments, and then just when the B+ starts to come on line the fuse blows, and on one occasion I saw a tiny bit of lightning inside the rectifier. I've been using 2 amp fuses. The rectifier test ok but makes me skeptical of any future use.
I've pulled the tubes and checked the PT; it seems fine.
How does one diagnose a blown MOSFET ?
And I'm back to deciding on which mosfet will work best. Perhaps I'll try the fairchild next......
Or maybe I'll just repurpose the transformers and build a simple SE. No mosfets to worry about there.
Anyways, I wanted to let people know before anyone else tries the vishays.
Well, I got seven harmonious months with my new TSE.....and then it blew up. Or at least it's blowing fuses. And I suspect my MOSFETS.
I haven't done much troubleshooting yet, but the amp powers up, warms up the filaments, and then just when the B+ starts to come on line the fuse blows, and on one occasion I saw a tiny bit of lightning inside the rectifier. I've been using 2 amp fuses. The rectifier test ok but makes me skeptical of any future use.
I've pulled the tubes and checked the PT; it seems fine.
How does one diagnose a blown MOSFET ?
And I'm back to deciding on which mosfet will work best. Perhaps I'll try the fairchild next......
Or maybe I'll just repurpose the transformers and build a simple SE. No mosfets to worry about there.
Anyways, I wanted to let people know before anyone else tries the vishays.
thanks for taking an interest, Russ.
So here's as far as I got before the fear set in:
I pulled all the tubes,
plugged the amp into my little 300 watt variac,
and measured the voltages from the power xformer at 110 volts.
They were right on spec.
The power transformer is an Edcor xpwr163-120 B+=550V 150ma
I inserted the rectifier, the JJ GZ34 that had been in the amp when it was last working. I put a 1.5 amp fuse in the amp and brought the variac up to 110 volts. The filament began to glow and warm. Just at the point where I expect the B+ to come on line the fuse burns and I see the briefest small spark in the rectifier.
Thus chastened, I tried another rectifier, this time a chinese 5ar4 known good. With a fresh fuse in the amp I cranked the variac back up and again the filament warmed and as the B+ came on line the fuse burned, this time not just the 1.5 amp fuse in the amp, but also the 3 amp fuse in the variac.
No lightning in the tube, but it felt very hot for the short time it was on.
Thus chastened, I pulled the rectifier and went back to measuring the xformer voltages. Still right on spec. I've got a EMC 215 tester and both rectifiers still test good. It looks like I haven't destroyed anything yet, but I'm game to keep trying.
I think my power transformer circuit is working but thats about it right now.
What do you think I should try next ?
perhaps selling it in times square ?
So here's as far as I got before the fear set in:
I pulled all the tubes,
plugged the amp into my little 300 watt variac,
and measured the voltages from the power xformer at 110 volts.
They were right on spec.
The power transformer is an Edcor xpwr163-120 B+=550V 150ma
I inserted the rectifier, the JJ GZ34 that had been in the amp when it was last working. I put a 1.5 amp fuse in the amp and brought the variac up to 110 volts. The filament began to glow and warm. Just at the point where I expect the B+ to come on line the fuse burns and I see the briefest small spark in the rectifier.
Thus chastened, I tried another rectifier, this time a chinese 5ar4 known good. With a fresh fuse in the amp I cranked the variac back up and again the filament warmed and as the B+ came on line the fuse burned, this time not just the 1.5 amp fuse in the amp, but also the 3 amp fuse in the variac.
No lightning in the tube, but it felt very hot for the short time it was on.
Thus chastened, I pulled the rectifier and went back to measuring the xformer voltages. Still right on spec. I've got a EMC 215 tester and both rectifiers still test good. It looks like I haven't destroyed anything yet, but I'm game to keep trying.
I think my power transformer circuit is working but thats about it right now.
What do you think I should try next ?
perhaps selling it in times square ?
When you are blowing fuses, look for parts that, if shorted, connect a B+ or B- line to ground. One of those will usually be the culprit. The filter caps, and bleeder resistor should be checked before you keep putting rectifier tubes in the amp.
Or, you could just have two bad rectifer tubes.
The mosfets set the negative bias. One can always pull an output tube and the rectifier tube and measure the bias voltage at the output tube socket to determine if the mosfets, and indeed, the entire B- supply are working. The MOSFET's are pretty tough, and harder to burn out than you might think. The bias voltage should be present even without the rectifier tube, as the bias line uses separate solid state diode rectifiers, and does not utilize the 5AR4.
Win W5JAG
Or, you could just have two bad rectifer tubes.
The mosfets set the negative bias. One can always pull an output tube and the rectifier tube and measure the bias voltage at the output tube socket to determine if the mosfets, and indeed, the entire B- supply are working. The MOSFET's are pretty tough, and harder to burn out than you might think. The bias voltage should be present even without the rectifier tube, as the bias line uses separate solid state diode rectifiers, and does not utilize the 5AR4.
Win W5JAG
thanks Russ,
I am using the Triode choke that everyone seems to like so I'll pull it give over the weekend and report back.
I'm also using a 50 mfd motor run cap. I put in a smaller cap for C4 when I set the B+ voltage, although I don't remember tha value of it off hand. I'll check that too. That's starting to seem like a potential culprit as well.
Looking closely at the board I saw no obvious indications of shorting, but R6 shows a tiny bit of cracking in its sandy exterior, as if it might have gotten a bit too hot, or received too much heat from the adjacent rectifier. I measured it without pulling it from the board and got 9.6k ohms. I'm thinking it's less likely to be causing the problem since it's in the bias circuit.
I'm really grateful for your help. This amp was sounding sweet and working without problem for 7 months and then just started blowing fuses. I immeadiately suspected the mosfets since I went off road in selecting them, but what you're suggesting makes all the sense in the world: the fuse pops when the B+ tries to come online, so first check the parts in the B+ circuit......
I am using the Triode choke that everyone seems to like so I'll pull it give over the weekend and report back.
I'm also using a 50 mfd motor run cap. I put in a smaller cap for C4 when I set the B+ voltage, although I don't remember tha value of it off hand. I'll check that too. That's starting to seem like a potential culprit as well.
Looking closely at the board I saw no obvious indications of shorting, but R6 shows a tiny bit of cracking in its sandy exterior, as if it might have gotten a bit too hot, or received too much heat from the adjacent rectifier. I measured it without pulling it from the board and got 9.6k ohms. I'm thinking it's less likely to be causing the problem since it's in the bias circuit.
I'm really grateful for your help. This amp was sounding sweet and working without problem for 7 months and then just started blowing fuses. I immeadiately suspected the mosfets since I went off road in selecting them, but what you're suggesting makes all the sense in the world: the fuse pops when the B+ tries to come online, so first check the parts in the B+ circuit......
Yep, just go through the circuit step by step. Also, you can bring up the B+ more slowly with the variac. At 50% or so, the heater in the 5AR4 will get hot enough for the diode to start conducting. If you have a good DMM that can measure high AC currents (usually a separate jack or setting for "10A"), put it in series with the center tap on the secondary and the PCB. A current spike at the start is normal as the caps charge, but it should drop-off immediately. That might save you some money on fuses. 
It doesn't take much of a short to arc-over the rectifier and blow the fuse. It only takes a few extra 100mA on the secondary side to put an extra 1A over the fuse and blow it. What sort of cap did you end up with in the C4 spot?
The spec'ed part for R6 is over stressed in this design. Somewhere on George's front page or maybe in the TSE description he suggests up-rating this to a 7W part. I used Digikey p/n PPC7W10.0KCT-ND. It fit in the location well enough and seems to handle the load OK. I'll be using a 10W chassis mount in the final amp.
It doesn't take much of a short to arc-over the rectifier and blow the fuse. It only takes a few extra 100mA on the secondary side to put an extra 1A over the fuse and blow it. What sort of cap did you end up with in the C4 spot?
The spec'ed part for R6 is over stressed in this design. Somewhere on George's front page or maybe in the TSE description he suggests up-rating this to a 7W part. I used Digikey p/n PPC7W10.0KCT-ND. It fit in the location well enough and seems to handle the load OK. I'll be using a 10W chassis mount in the final amp.
Good idea on the chasis mount resistor.
So I spent a few minutes with the TSE over the past few days and found that C4 had shorted out. It was a 4.7 uf 450 volt jamico brand.
It was the first component I checked & since it was bad, I didn't yet check the choke.
I wonder if the cap was simply bad, or if it's failure is the symptom of another problem. Are jamico caps known to be crap ?
I spent most of my tube/hobby time over the weekend finishing my SPP. And just as a quick aside, what a wonderful sounding amp it is (currently set up in UL with CFB).....
I tried both of the rectifiers in the SPP and they work well. No damage done.
Next session I'll put in another cap for C4 and go through the check out and set up process for the TSE to see if I can't get it back online. I'll use parts out of my bin for now but I think with my next parts order I'll get a 500 volt cap if I can find one for C4.
Thank you again for pointing me in the right direction...
So I spent a few minutes with the TSE over the past few days and found that C4 had shorted out. It was a 4.7 uf 450 volt jamico brand.
It was the first component I checked & since it was bad, I didn't yet check the choke.
I wonder if the cap was simply bad, or if it's failure is the symptom of another problem. Are jamico caps known to be crap ?
I spent most of my tube/hobby time over the weekend finishing my SPP. And just as a quick aside, what a wonderful sounding amp it is (currently set up in UL with CFB).....
I tried both of the rectifiers in the SPP and they work well. No damage done.
Next session I'll put in another cap for C4 and go through the check out and set up process for the TSE to see if I can't get it back online. I'll use parts out of my bin for now but I think with my next parts order I'll get a 500 volt cap if I can find one for C4.
Thank you again for pointing me in the right direction...
I have a different question about the Mosfets...
I was able to find 4 2SK3563 mosfets, so I ordered them all. When they arrived, they were in a standard tiny bag and not in an ESD bag. I've heard that mosfets are quite static sensitive; should I be concerned? To make matters worse, I'm on the slow boat plan for building my TubeLab amp so if the mosfets are bad I won't know it until it's too late to do anything about it.
Should I question the seller now, while I still can? Is there a way to test mosfets with simple tools (my DMM has a diode test mode)?
Thanks!
I was able to find 4 2SK3563 mosfets, so I ordered them all. When they arrived, they were in a standard tiny bag and not in an ESD bag. I've heard that mosfets are quite static sensitive; should I be concerned? To make matters worse, I'm on the slow boat plan for building my TubeLab amp so if the mosfets are bad I won't know it until it's too late to do anything about it.
Should I question the seller now, while I still can? Is there a way to test mosfets with simple tools (my DMM has a diode test mode)?
Thanks!
You can test them fairly easily using a simple rig using a small load like a flashlight bulb and flashlight battery and a 9V battery for the gate. You can also test them with a DMM in diode test mode, but it doesn't always work because the DMM may not provide enough gate voltage. Google "testing MOSFETs" or similar and you should get plenty of hits and maybe even some YouTube videos.
Success !
It appears the trouble with my amp was not the MOSFETs, after all.
It was indeed a bad C4. Temp rating is 85C; as Russ suggests I'll look for a more heat tolerant replacement. For now I put in another 85C rated 10 uf 450 volt cap. That gives me a B+ of about 290v with around 28.5ma going through the 45 tubes.
In the check out process, before I put in the rectifier, I looked for the negative bias voltage on the output screens and found it. Thank you for that suggestion, w5, made me feel better about my MOSFETs.
I haven't had a chance to turn it up over late night apartment levels, but all the readingss look good so I will soon ! Time to give the new cap a nice warm burn test...
Thank you for the guidance, Russ.
It appears the trouble with my amp was not the MOSFETs, after all.
It was indeed a bad C4. Temp rating is 85C; as Russ suggests I'll look for a more heat tolerant replacement. For now I put in another 85C rated 10 uf 450 volt cap. That gives me a B+ of about 290v with around 28.5ma going through the 45 tubes.
In the check out process, before I put in the rectifier, I looked for the negative bias voltage on the output screens and found it. Thank you for that suggestion, w5, made me feel better about my MOSFETs.
I haven't had a chance to turn it up over late night apartment levels, but all the readingss look good so I will soon ! Time to give the new cap a nice warm burn test...
Thank you for the guidance, Russ.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.