Yup!
So far I only had input jacks on the case. I just bought a pair of insulated ones! (I have to test them)
I also got a fuse holder 😀 (It was about time...)
and some coax cable.
I think I'll also try to run the circuit with the transformers wired for ±36v (Both transformers have two 13VAC 2A windings, and a 7VAC 0.6A one)
Not the voltage, may be the biasing point... got a 9V headphone amp with same tube...will sleep on it.. spent all day faultfinding on my first solidstate amp (diy)...
Keep the isolated sockets.. without you have one definate groundloop.
Do you have another tube and transformer... those can also sometimes create their own hum.
Keep the isolated sockets.. without you have one definate groundloop.
Do you have another tube and transformer... those can also sometimes create their own hum.
A metal case will act as an aerial! It is not the point to connect all your grounds!
Connect them all to some other point that is in turn connected to the case via a resistor and capacitor in parallel.
I also suggest isolating all your input and output connectors from the metal chassis.
Connect them all to some other point that is in turn connected to the case via a resistor and capacitor in parallel.
I also suggest isolating all your input and output connectors from the metal chassis.
I don't, but definitely, without the tube buffer, there are much less hum problems. (without it, moving the grounding points is actually doing something significant)
Doh! That means that there's still some MAJOR work to do! (As you can see in the picture...)
To save time I could just snap off the mounting tab, but then, I'll have to find an other way to screw the thing on.
OK... I setup the amp for ±40vDC (The 13v secondaries are running at almost 15v)
I connected my four 25V caps to get two series pair to run the amp(Is it safe?). All I can say is that the tube distorts much less, but there's much more hum, which leaves the transformers as the last possible hum source.
I have an impression that I'll have loads of fun moving the grounding point.(Or I could try some cheesy setup insulating the bakelite board with o-rings and electrical tape...)
I connected my four 25V caps to get two series pair to run the amp(Is it safe?). All I can say is that the tube distorts much less, but there's much more hum, which leaves the transformers as the last possible hum source.
I have an impression that I'll have loads of fun moving the grounding point.(Or I could try some cheesy setup insulating the bakelite board with o-rings and electrical tape...)
Hey, putting capacitors in series is the same as putting resistors in parallel... to get the effect of two caps, they must be paralleled....
You have not confirmed trying to follow the earh instructions as per the JLTI site... and I mean exactly.
You have not confirmed trying to follow the earh instructions as per the JLTI site... and I mean exactly.
I put them in series because they are rated 25v each and I now have ±40v. They were in parallel when I used a ±20v line.
Not really, my chassis is the ground, while JLTi talks about a wired ground. I'll have to move the grounding point on the bakelite boards.
OK... Well, the wired ground just makes things worse. Same hum, but 4-5Hz constant speaker wobble included.
Before, I only had wobbling when the volume control was set to a certain point, but now, it's happening all the time. If I connect the 10k resistor normally between -40v and tube's output between ground and tube's output, there's still sound but no more wobble.
EDIT: I always got less hum when I connected the speaker negative to the RCA jack. I get no wobbling this way.(But much less gain)
I decided to connect the amp directly in the mains rather than passing by my ISOBAR. Now, when stuff is turned off, it makes loud pops in the speaker. It sounds like there's a filtering problem somewhere...
Before, I only had wobbling when the volume control was set to a certain point, but now, it's happening all the time. If I connect the 10k resistor normally between -40v and tube's output between ground and tube's output, there's still sound but no more wobble.
EDIT: I always got less hum when I connected the speaker negative to the RCA jack. I get no wobbling this way.(But much less gain)
I decided to connect the amp directly in the mains rather than passing by my ISOBAR. Now, when stuff is turned off, it makes loud pops in the speaker. It sounds like there's a filtering problem somewhere...
Negative... mine were housed in seperate housings with interconnects between em...
Can you try to move the gridstopper resistor as close as you can to the tube... it is there to prevent oscillation...
Can you try to move the gridstopper resistor as close as you can to the tube... it is there to prevent oscillation...
Well, I slapped the 3.9kOhms directly on the tube pins, if that's near enough ;-)
But mines are stuck between two toroids.
I had fun browsing the 36 pages of the JLTi GC topic, and a part of my problem is the same as Tor Martin's : He has no grounding or transformer proximity problems apparently, so I'll stop making the grounding scheme more as-per-instructions as it just seems to make things worse for now.
http://www.diyaudio.com/forums/showthread.php?postid=197369#post197369
I isolated my problem to the tube buffer just like Tor since there's almost no hum when I bypass it. His problem is apparently due to PSU ripple going in the tube's output, he measured a few AC millivolts, which I can't do because my DMM can see a minimum of 16mV only.
Maybe the separate rectifiers solution with 220ohms in series for the tube PSU is a good idea to try finally... (It partially fixed Tor's problem - who never came to a final solution
)I heard moamp's GC had no problems, the only thing is that he used some filtering that I can't fit in my box for his version of the JLTi DIY GC.
http://www.moxtone.com/Tube-buffered Gainclone.htm
I have a project list at Mouser with 8 Nichicon 2200µF 50v HE caps and 8 schottky diodes, I was thinking of replacing the current 3900µF 25v cans in placed in series pairs with them.
BTW, I almost have an oscilloscope(since it could help find problems) I received Friday the two oscilloscope/spectrum analyzer/logic analyzer ISA cards I bought on eBay, but I still need to buy/build probes.
OK!
I'm emailing/annoying a few persons from the main JLTi topic inviting them to post, since they probably didn't get any e-mail notification after not having posted for a long time.
I've been running the amp a bit without any buffer(And the 10k to negative disconnected because I don't want to blow the my speakers and the amp) and there IS a little EMI, which is fixed by moving the test lead away from the transformers(which means that I'll have to use a little bit of coax in the amp if I ever get to a point where the EMI hum gets louder than the buffer's)
I'm emailing/annoying a few persons from the main JLTi topic inviting them to post, since they probably didn't get any e-mail notification after not having posted for a long time.
I've been running the amp a bit without any buffer(And the 10k to negative disconnected because I don't want to blow the my speakers and the amp) and there IS a little EMI, which is fixed by moving the test lead away from the transformers(which means that I'll have to use a little bit of coax in the amp if I ever get to a point where the EMI hum gets louder than the buffer's)
I would slap a nice big cap on the output for now to keep the DC off the speakers until we get this sorted..., I am thinking adjustable bias like the Soha headphone amps... should not be to hard to figure out... are you running dual rails i.e. + 0 and - rails or just + and gnd?
Are you giving the tube enough time to warm up or do you freak out and power down before then...? You realy do need a relay to short the output of the tube until it is warm, or it will present the input of your amp with large DC offset. which is then multiplied by the gain of the amp. Maybe install a manual switch for that function... give it 30 seconds to a minute because of your low voltage and open the short...
Are you giving the tube enough time to warm up or do you freak out and power down before then...? You realy do need a relay to short the output of the tube until it is warm, or it will present the input of your amp with large DC offset. which is then multiplied by the gain of the amp. Maybe install a manual switch for that function... give it 30 seconds to a minute because of your low voltage and open the short...
±40v now. I have been using ±20v in the beginning, there was half the amount of hum, but the tube distorted a little at that voltage.
At ±20v secondaries and 14v feeding the LM317 filament regulator I would let it warm up, it was not that hard on the speaker, but now, with ±40v and 8v feeding the filament regulator I prefer to just disconnect the mute wire from Vee until the tube warms up(15 secs when I leave it unmuted).
Should I short the tube output or disconnect the 10k resistor from the negative rail(produces the same effect)?
I could use a SPST switch I have lying around that way, since I don't want to mix both channels.
I just put two 8µF NPs I have in parallel at the speaker output, no more wobble, hum, and push-pull at startup, but no bass ;-)
I guess there's no push-pull because the transformers and caps don't get overloaded by the sudden amp's power request due to the -35v not being compensated by the tube when the speakers have a DC cap?
There is no wobble because there is no DC... even low frequencies say sub 10hz gets close to being DC...
Try largest electrolytic you have, maybe 2200uf or even bigger on the output to filter out less of the bass...
I scratched my head a bit... my circuit used a greenacher voltage doubler... basicaly caps and diodes... made over 100V from a small PCB mount 24V transformer...
Try largest electrolytic you have, maybe 2200uf or even bigger on the output to filter out less of the bass...
I scratched my head a bit... my circuit used a greenacher voltage doubler... basicaly caps and diodes... made over 100V from a small PCB mount 24V transformer...
I don't have this right now though.
The current heatsink becomes barely warm with my 8ohms, I rarely use over 1w when listening to music.
At least with DC caps the transformers don't make almost as much noise as the speaker until the tube warms up...
I see... It'd fix my wobble and startup problems though.
do you have a 3k3 resistor you could put in series with the 10k cathode resistor... or a resistor in the 12k to 15 range you can swap it with... and then try swapping with higher value closer to 33k... just want to see how biasing affects us
You would always have the startup problem when interfaceing a properly running tube with a chip, as the tupe puts out DC until it reaches its operational point... it is not realy a problem, as it is easily dealt with
You would always have the startup problem when interfaceing a properly running tube with a chip, as the tupe puts out DC until it reaches its operational point... it is not realy a problem, as it is easily dealt with
- Status
- Not open for further replies.