AMI Model AA jukebox monoblocks?
- Thread starter ruraldrew
- Start date
bimasta
Addicted Member
What do you mean by "very hi-fi sounding"? Just curious.
A 6L6GC won't do anything the 6L6 won't. It might last longer since its running way below max ratings but it will not give more output power. You'd have to re-design a few things to make use of more voltage to try and get more power, and then you might run into limitations of the output transformer.
I'll just run them as is then, unless 6L6s are particularly cheaper. I've never shopped them since other amps I've had required 6L6gc.
The metal 6L6's can often be had for fairly low prices considering what other NOS 6L6 family tubes go for. I have 2 pairs that I got for stupid cheap because I own one old radio that uses a pair. The one set I think i paid 20 bucks for, the other pair I think came in a box of other stuff.
maxhifi
Addicted Member
The cheapest tube for 6L6G applications, is the Russian 6P3. You can commonly find it re-badged as a 6L6GC, which has lead to reliability issues, because it isn't one and will fail if subjected to 6L6GC operating conditions. That said, it is a reliable and very affordable 6L6G replacement.
Finally revisiting this project, trying to sort out how to remove that relay from the 6L6 heaters. It looks like disconnect an 11 volt wire coming from the transformer and cap it off, then directly connect the 6.3 volt wire to the one 6L6 pin 2 that also feeds the other heater.
What's up with the resistor coming off the cathodes and going to the relay? Is that just something the relay needed to work or is it working with that 20u/25v capacitor as a filter and the relay was just borrowing a signal there as a "sensor"?
What's up with the resistor coming off the cathodes and going to the relay? Is that just something the relay needed to work or is it working with that 20u/25v capacitor as a filter and the relay was just borrowing a signal there as a "sensor"?
nasadowsk
Bow:good, hunt:needs work
The relay is there to turn the turntable on, in case the mech pulls and loads the record before the amp's warmed up, which happens, esp as the output tubes get old. You can actualy hear the sound quality change as the thing warms up for the first few records.
Frankly the early AMI jukebox amps weren't anything special in terms of sound quality, and like the others at the time put out more light and heat than sound.
Frankly the early AMI jukebox amps weren't anything special in terms of sound quality, and like the others at the time put out more light and heat than sound.
We might be talking about different relays. The one I'm looking at sends 11 volts to the 6L6's until they're warmed up then the relay trips and drops it back to 6.3 volts.
heater booster for faster warmup then? Interesting. Probably doesn't do the tube life any favors at all.
Looks like you can ditch the relay and move the heater supply to the Green / Yellow (I think thats G/Y) wire. Whatever is on the normally open relay contact.
I would probably dump the relay coil and sub it for a resistor to bias the output tubes too. I can't quite read what it says the coil measures at so you might need to experiment a bit to get the cathode voltage right. The resistor is there to get the total resistance correct and to limit how much current flows through that relay. The resistor and the coil in parallel form the cathode resistor so it creates bias for the output tube. Since no current flows until the tube warms up some, its a reasonably effective way of kicking the relay only after the tubes are hot enough to conduct.
Looks like you can ditch the relay and move the heater supply to the Green / Yellow (I think thats G/Y) wire. Whatever is on the normally open relay contact.
I would probably dump the relay coil and sub it for a resistor to bias the output tubes too. I can't quite read what it says the coil measures at so you might need to experiment a bit to get the cathode voltage right. The resistor is there to get the total resistance correct and to limit how much current flows through that relay. The resistor and the coil in parallel form the cathode resistor so it creates bias for the output tube. Since no current flows until the tube warms up some, its a reasonably effective way of kicking the relay only after the tubes are hot enough to conduct.
Thanks, any chance I could beg you to explain the napkin math? I'm trying to learn some theory as I go. Since I was looking for resistance I'm guessing I would need to already know a current and a voltage. It looks like "k" stands for cathode in the pinout on the schematic and that specifies 18.4 volts? How do I know what the current should be?
its the heater voltage bump that I'd be most keen to ditch. Could be wired around and the coil left though. I wouldn't want my output tubes being bumped with near double the heater supply just to get it hot faster. Re-purpose the relay to run a two color LED that will switch from a red "warmup" indicator to a green "ready" indicator.
So napkin math. On the schematic, there is a 62 ohm resistor. It also spells out DC resistance of the relay coil as 62 ohms, so 124 ohms. 18.4v / 124 ohms = 148 milliamps plate+screen current for the pair.
For a little more math and a sanity check, the schematic calls for what looks like 286v at the plate. Less the 18.4v for the cathode, lets call that 268 volts from cathode to plate. 6L6G datasheet gives us operating conditions for a class A pentode push-pull stage at 270 volts. It specifies 17.5v of bias, 134ma for the plates, and 11ma for the screens for a total of 145ma. 17.5v / 145ma = 120.7 ohms, pretty darn close to what the schematic values for the resistor + coil come out to be. If the output transformer is somewhere near the 5k specified on the datasheet, that should give 17.5 watts at the plate, or ~15 to the speaker.
If this were in front of me I'd be comfortable with a cathode resistor in the 120-130 range. Its going to be cooking off a bit shy of 3 watts. 5w part at minimum for de-rating, but a 7-10 watt part would be a better fit.
So napkin math. On the schematic, there is a 62 ohm resistor. It also spells out DC resistance of the relay coil as 62 ohms, so 124 ohms. 18.4v / 124 ohms = 148 milliamps plate+screen current for the pair.
For a little more math and a sanity check, the schematic calls for what looks like 286v at the plate. Less the 18.4v for the cathode, lets call that 268 volts from cathode to plate. 6L6G datasheet gives us operating conditions for a class A pentode push-pull stage at 270 volts. It specifies 17.5v of bias, 134ma for the plates, and 11ma for the screens for a total of 145ma. 17.5v / 145ma = 120.7 ohms, pretty darn close to what the schematic values for the resistor + coil come out to be. If the output transformer is somewhere near the 5k specified on the datasheet, that should give 17.5 watts at the plate, or ~15 to the speaker.
If this were in front of me I'd be comfortable with a cathode resistor in the 120-130 range. Its going to be cooking off a bit shy of 3 watts. 5w part at minimum for de-rating, but a 7-10 watt part would be a better fit.
That would be 17.5 at the plate of both tubes right? With this being push-pull does that mean each tube is only working half the time such that the speaker is only ever driven by one tube at a time meaning it's effectively 15 watts total? Or is each tube applying it's power to only half the signal but all the time, so it would be more like 30? I'm having trouble finding a good explanation of push-pull...
it calls for 17.5v of bias, which can be either negative V at the grid, or positive V at the cathode. Bias is measured from cathode to grid, and the grid wants to be negative relative to the cathode. Both tubes would see the same voltage.
Unless you meant power. If you were to measure AC voltage from plate to plate and use that number and the specified 5k load, you'd get 17.5. Transformers are ballpark 85% or so efficient, so that leaves you with near enough 15 watts at the speaker. Might be a bit one way or the other if you actually measure it.
Unless you meant power. If you were to measure AC voltage from plate to plate and use that number and the specified 5k load, you'd get 17.5. Transformers are ballpark 85% or so efficient, so that leaves you with near enough 15 watts at the speaker. Might be a bit one way or the other if you actually measure it.
Are you using the 295 from the socket diagram to get that 17.5? Isn't 295 the screen voltage? If P is plate it's kind of blurry but looks like 285, so would that be
285^2 / 5,000 = 16.245
Concept wise apart from the math I'm having trouble understanding exactly how the tubes work together in push-pull configuration. It seems kind of similar to balanced line level inputs with the way the signal is split out of phase and put back together.