JLTI tube gainclone LM3886, no sound output

[DragonMaster]

My JLTi gainclone doesn't work. Everything powers up, nothing gets hot, but there's no sound output. The PSU for the lot is ±20vDC (Yes, it's low, but the only affordable solution I found)

The tube itself actually outputs something(at a much lower gain than directly taking the input)

The problem is, there is no sound at all in the speakers output. No pop when starting, no hissing, simply nothing.

Is there any place I should look at? I modified the schematic as per the LM3886 datasheet, V+ gets to destination, V- as well, Vin+ and - too. It's pretty strange. By the way, I have a 6DJ8 tube.

[DragonMaster]

I just tried connecting my source(portable radio, adjustable volume) directly to Vin- through a DC blocking cap, and I still get nothing. The only time I get a thump on startup is if I unplug the -20v line. I get scratching on the sound output while I plug/unplug the -20v line. Speaker is plugged on LM3886's pin 3, v- on pin 4, v+ on pin 1, pin 7 to ground, vin+/- pins 9/10.

One thing, I left the mute pin floating, is this a right thing to do?

[Drew Eckhardt]

Originally posted by DragonMaster
I just tried connecting my source(portable radio, adjustable volume) directly to Vin- through a DC blocking cap, and I still get nothing. The only time I get a thump on startup is if I unplug the -20v line. I get scratching on the sound output while I plug/unplug the -20v line. Speaker is plugged on LM3886's pin 3, v- on pin 4, v+ on pin 1, pin 7 to ground, vin+/- pins 9/10.

One thing, I left the mute pin floating, is this a right thing to do?

Leaving the mute pin unconnected causes the LM3886 to mute.

Drawing at least .5ma from the mute pin lets the amp play, where the mute pin is at V+ - 2.6V

This is all in the data sheet.

[jerishi]

If I remember, it's pin 8..A resistor required-But datasheet will give info. I think you're real close to getting sound.

[DragonMaster]

The datasheet talks about a capacitor. Is it necessary?

I'll first try to find a 10kOhm resistor and see what happens.

[Minion]

I personally use a 100uF cap from the Mute pin to ground as it seems to get rid of any poping sound when I enable the mute switch but if you aren"t useing a Mute switch then a resistor should work fine.....


Cheers

[DragonMaster]

I tried a 22k, it works.

I seriously think I'll put a mute switch because it acts crazy until the tube is warmed up. The woofer cone is pushed forward and backwards(transformer hum included) at .5 sec intervals until the tube has warmed up. If there is one single connection to the tube I disconnect, it starts going crazy again. It also hums a lot (speakers output) and the LM317 used for the filaments becomes burning hot. It's rectifying a 20v line to 6.3v (380mA filament) I checked and the voltage and current remain the same even once the lm318's heatsink becomes finger-burning hot.

I just measured the temps and the LM317 goes to 60°C in a matter of minutes.

Should I try to feed the filament PSU's rectifier with 7VAC instead of 14VAC? I'm about to order some parts at Mouser, do you think I should get an other LM317 in a plastic to-220 package this time and stick it to the main heatsink?

Note that there are no decoupling caps at all on any of the DC lines. There are filter caps though: 2 x 3900µF per ±20v lines for the power amps, a 2700µF for the LM317's input, and an other 2700µF for it's output.

[Nordic]

Add the capacitor... it acts as a kind of deadshort until it is charged, forming a kind of power-on delay which may be long enough to wait for your tube to warm up... 15 to 20 seconds should do.. Your tube buffer is just that, a buffer, not a voltage amplifier... it likely runs at unity gain, which explains why the signal sounds a bit weaker through it... the amp will love it tough...

You could add another regulator before the LM317 to drop the voltage some and carry some of the heat... also add a nice little heatsink to the LM317 one.

The heat generated is basicaly directly related to volts dropped x current drawn...

If your transformer can deliver the current, give the lower voltage a shot... provided the rectified voltage is higher than the 6.3V + rectifier and regulator dropouts, it could work, I suspect it may be just too little...

In the old days I guess they would just drop the 7v with a 2R resistor..

[DragonMaster]

The LM317 already has an heatsink ;-)

So far the LM317 seems happy with an 8.5vDC input. The amp ran for about 15 minutes and it's just at 36°C right now. I wonder if it will be able to dissipate heat well into the small, completely closed enclosure once I'll close the cover.

Well, I'll try the capacitor. With an 8v input to the filament regulator, it takes much more time to heat the tube though. (EDIT: A 470µF lasts 2 secs, is it really a good solution?)

Any idea of what I should shield to stop the humming?

EDIT: The temp continues to rise slowly with time. I guess I'll get an LM317P and stick it on the main heatsink, which remains cold.

[Nordic]

are there any biasing trimpots?
With 8.5V you are only dropping 2.2V @ 365mA ...less than 1W... 1 inch square heatsink is enough

60C may be hot to touch, but wont kill a chip overnight.