At the risk of making a fool of myself again, I think I will ask another question. I was looking at the data sheet for the 2SK3563 and Crss has a pretty good rise starting at Vds = 15V or so. Is that a problem? It seems like you would only be able to get the power tube grids about 10 volts positive before that capacitance starts to kick in and possibly start causing some distortion. FQPF2N90 seems to have less capacitance at lower voltages but a little voltage headroom seems to be the best solution if you want to be able to pull the grid more than 20 volts positive, and I'm not saying that is really necessary.
I use the FQPF2N90 for the same reason. I like to bolt them to the chassis and I don't feel that comfortable with those rubber pads, even though they are rated for more voltage than I am putting on them.
I really like the FQAF Fairchild FETs. TO-3P isolated case, lots of space between the leads. Unfortunately, they don't seem to have any with as low Crss as the FQPF choices.
I use the FQPF2N90 for the same reason. I like to bolt them to the chassis and I don't feel that comfortable with those rubber pads, even though they are rated for more voltage than I am putting on them.
I really like the FQAF Fairchild FETs. TO-3P isolated case, lots of space between the leads. Unfortunately, they don't seem to have any with as low Crss as the FQPF choices.
In theory this or any voltage variable capacitance (nearly all silicon devices) can lead to "Phase Intermodulation Distortion". PIM is one of the current theories used to support the "silicon is evil" debate. PIM is also nearly impossible to measure with equipment that any of us can afford. Being an engineer capable of understanding this theory, I usually run the positive supply in the 100 volt range which keeps most fets in the constant Crss range. For now there is only so much room on my bench and the old Sorensen 25 volt supply doesn't eat much of that space. It comes in handy for lighting up those odd voltage TV tubes.
When I build the real amp there will be dual (positive and negative) supplies either derived from an isolation transformer or the 70 volt taps on an Antek toroid. They will be at least +- 100 volts.
I have a good collection of mosfets to try in this board including the IRF820. I see no reason why they wouldn't work. I didn't use them in this build because having two metal heat sinks at different voltages right next to the trim pots leads to the use of bad language!
I have some of those too.
I got 6 of them too. Ordered them Friday night at 10PM after reading about them on this forum somewhere. They arrived Monday morning. I am still trying to figure out how that happened. The USPS just isn't that good!
You can use one for C2 also. Just add a fuse..... Understand that these things are big. You will need to fill the entire chassis with them. They also have a rather ugly mounting stud sticking out of the top of the can (the other end from the terminals).
I can share this warning. These caps can store a LOT of energy! Their extremely low ESR will lead to destructive discharge currents. Don't discharge them with a screwdriver or a clip lead. How do I know this? I blew a Radio Shack clip lead in half! It made a very large bang that got my attention and scared Sherri sensless. I would advise adding a fuse to protect the transformers, wiring and circuit boards from damage in the event of a tube arc or short.
Wire a bleeder resistor across the terminals and leave it there. My usual 150K resistor takes a couple of minutes to discharge this cap.
Those GLY513's are a crazy good deal. Get 'em before the seller realizes what he's doing.
I had the data sheet on them (and lost it?), and aside from being HUGE, the specs are good. I used a pair in both the power supply and the output decoupling in a headphone amp. They're nearly as good on paper as 100uF caps I've got that cost 4 times as much.
Slightly off topic, but I think they would be excellent in a power supply in a high powered amp.
Just be sure to keep those terminals shorted while not in use!
I had the data sheet on them (and lost it?), and aside from being HUGE, the specs are good. I used a pair in both the power supply and the output decoupling in a headphone amp. They're nearly as good on paper as 100uF caps I've got that cost 4 times as much.
Slightly off topic, but I think they would be excellent in a power supply in a high powered amp.
Just be sure to keep those terminals shorted while not in use!
Beverly Hills Florida looks to be in the middle of nowhere, half way between Tampa and Gainsville. I live in Sunrise which is a suburb of Fort Lauderdale. I guess that the planets aligned and the USPS truck driver had a lead foot.
I can see the post office from my front door, although it is about 1 mile by road. I have taken invitations to the post office for local delivery, and the arrival dates were anywhere from 2 days to a week.
I have been shipping boards all over the world through that post office using prioity mail. Boards to the east coast take from 1 to 5 days, the west coast is 3 to 7 days. I sent out a stack of them and the guy is boston gets his before the guy that is 10 miles away.
One thing that has yet to fail is that regular UPS ground (brown) shipping is always 1 day anywhere within Florida.
I can see the post office from my front door, although it is about 1 mile by road. I have taken invitations to the post office for local delivery, and the arrival dates were anywhere from 2 days to a week.
I have been shipping boards all over the world through that post office using prioity mail. Boards to the east coast take from 1 to 5 days, the west coast is 3 to 7 days. I sent out a stack of them and the guy is boston gets his before the guy that is 10 miles away.
One thing that has yet to fail is that regular UPS ground (brown) shipping is always 1 day anywhere within Florida.
I try to ship brown when I can because their tracking is better than anyone else's. Sometimes they beat-up heavy items, like car parts, but as long as the vendor packs things properly I don't have issues. FedEx's tracking is better now, but not updated often. USPS tracking is completely useless.
Shipping times with brown are like clockwork. It almost always arrives exactly on the day they say unless I buy from Mouser or New Egg. Very often it comes a day early for those. USPS is totally random, but their Priority is a little more predictable.
Shipping times with brown are like clockwork. It almost always arrives exactly on the day they say unless I buy from Mouser or New Egg. Very often it comes a day early for those. USPS is totally random, but their Priority is a little more predictable.
Don't thank me. I found this quote in the "Futterman style 6AS7G OTL with other tubes" thread posted by "whitelabrat". I then went straight to Ebay and ordered 6.
One is connected across my somewhat unstable Fluke 407D. I need a cap to keep it from motorboating when I peg the current meter. The KT88's at full tilt just peg the 300 mA meter.
One will go into the power supply that I build for the amp that is being designed here, unless I need two because of the supply voltage.
I took the boards to wood shop class last night and trimmed them to the correct size. I will build the second one this weekend. I will right down all of the parts values as I insert them. Some values may need to be adjusted to suit the supply voltage.
I'm beginning to have crazy thoughts like dragging out the monster 400 Watt Plitron OPT's and hooking up a few sweep tubes.........
I now try to only ship UPS. I had a large order from mouser "lost" in the mail and all I recieved was a slip from the postal service saying it was received with no contents. I used priority with delivery confirmation...but there was no insurance so I was SOL. The post office did nothing to help, mouser gave me free shipping on my next order. UPS costs a bit more but if it means I get insurance I am fine with that.
It should be pretty easy to see it on a scope if it is bad enough, though. I'm thinking a small 20kHz signal riding on 2kHz or something. It may be easy to craft a signal with free sofware and burn to a CD. I wouldn't be surprised if things get pretty ugly in the region where that capacitance is high. I will be checking this when I get my amp up and running again.
I think it would have to be really bad. I have tried to capture it with a scope without success, but I didn't try to make PIM. Maybe wiring a bunch of reverse biased diodes across a plate load resistor might do it. I tried using a 10 KHz square wave as the "signal" since it would be easier to see the blurring of the edges. I tried several low frequency sine, square, and triangle waves as the "disruptor".
An FFT analyzer is excellent for capturing distortions that occur in the amplitude domain. THD or Total Harmonic Distortion does show up as unwanted harmonics. THD is simply a "total" of all of the unwanted harmonics. The FFT analyzer will let you see each harmonic. This is good since some harmonics sound much worse than others. It can be used to look at classic IMD or intermodulation distortion. This occurs when one signal mixes with another signal to create two or more new unwanted signals that are not harmonically related to either of the originals. The classic case is when a vocal (or other) signal (at say 1000Hz) gets blasted by a bass guitar signal (at say 60 Hz) creating two new signals at 940 Hz and 1060 Hz. This ALWAYS sounds bad. It isn't always your amplifier either I have found obvious IMD recorded into some CD's. Metallica's S&M comes to mind.
PIM is something different. It is intermodulation in the frequency domain. All electronic circuits have a delay associated with a signal passing through them. The signal comes out of your amplifier a few microseconds later than when it went in. As long as thes delay remains constant everything is OK.
Much of the delay is caused by RC circuits. Consider a mosfet with a gate stopper resistor. The mosfet has considerable input capacitance which forms a low pass filter against the gate stopper which will have a frequency response and a delay associated with it. Now what do you think will happen if the fets input capacitance varies with signal level. A large signal will cause the delay through the circuit to change with the music. In mild cases a heavy bass note could blur the edge of a transient or other signal with fast rise and fall times. In a bad case you are actually applying frequency modulation to a signal.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.