Soooo,
I'm still wrapping my brains around the scratch-built thing. I can read a schematic, modify an exisiting amp, and have a 6V6 DIY amp that *almost* works (shoulda breadboarded first -power supply issues - read: attention to detail lacking
).
My question is this:
I have some decent 50's era OPT's and a few PT's kickign around which will fit the bill for a 45 amp, and a decent stash of 45's, 26's, 27's, and 80's etc, etc . . .
I'm leaning towards building the JE Labs Simple 45/2A3 cause it's you know - simple, BUT, in digging around for info I've seen two things:
1) 45 tubes are loved for their transparency and midrange
2) some of the more impressive amps I've seen use the same tube for the driver and the output tube.
Could you do this with a 45 ? is it just a matter of getting the operating points right on the 'driver' 45. I've searched all over, and can see guys doign this with the other power triodes (845, 211, 300b), but why not the 45 ?
I know the short answer is 'time to break out the 45 data sheet and get intimate', but I thought I'd see what the experts think . . .
thanks !
I'm still wrapping my brains around the scratch-built thing. I can read a schematic, modify an exisiting amp, and have a 6V6 DIY amp that *almost* works (shoulda breadboarded first -power supply issues - read: attention to detail lacking
).My question is this:
I have some decent 50's era OPT's and a few PT's kickign around which will fit the bill for a 45 amp, and a decent stash of 45's, 26's, 27's, and 80's etc, etc . . .
I'm leaning towards building the JE Labs Simple 45/2A3 cause it's you know - simple, BUT, in digging around for info I've seen two things:
1) 45 tubes are loved for their transparency and midrange
2) some of the more impressive amps I've seen use the same tube for the driver and the output tube.
Could you do this with a 45 ? is it just a matter of getting the operating points right on the 'driver' 45. I've searched all over, and can see guys doign this with the other power triodes (845, 211, 300b), but why not the 45 ?
I know the short answer is 'time to break out the 45 data sheet and get intimate', but I thought I'd see what the experts think . . .
thanks !
I'd be hesitant to do this for the following reasons:
1. NOS 45s are evermore rare and hard to find. Preserving the existing stock is important. Why burn up a tube in the driver role when you can burn up a tube in the output stage role?
2. The mission of driving a 45 with 45, or 845 with 845 is to cancel
distortion. The 45 already has such low distortion that this is pretty much fruitless. I dont think you will hear anything but it will "look"
impressive to use more rare tubes I guess.
3. Using a a 45 to drive an 845 makes sense. Because you are using 1.5W of the 845 to drive the grid of the 845 to get 20W out. In this case, the 45 is doing a great job swinging lots of volts at low distortion. This is a case where burning up a 45 as a driver begins to make sense.
-- Jim
1. NOS 45s are evermore rare and hard to find. Preserving the existing stock is important. Why burn up a tube in the driver role when you can burn up a tube in the output stage role?
2. The mission of driving a 45 with 45, or 845 with 845 is to cancel
distortion. The 45 already has such low distortion that this is pretty much fruitless. I dont think you will hear anything but it will "look"
impressive to use more rare tubes I guess.
3. Using a a 45 to drive an 845 makes sense. Because you are using 1.5W of the 845 to drive the grid of the 845 to get 20W out. In this case, the 45 is doing a great job swinging lots of volts at low distortion. This is a case where burning up a 45 as a driver begins to make sense.
-- Jim
hey-Hey!!!,
gain might be an issue. A 5k:16 output TX is ~17:1 step down. Two 45's in cascade will have a combined gain of ~1:13. Your amp will have a gain less than 1.
It gets worse when you go 5k:8R which is a 25:1 step down. Consider perhaps a 26 driving a 45. Still not going to have a lot of gain, but power output won't need a lot of output voltage. A 45 SE won't do 4V rms into 8R, more like 3.5 V rms. A reasonable pre-amp/linestage should be able to deliver a proper signal to the amps.
cheers,
Douglas
gain might be an issue. A 5k:16 output TX is ~17:1 step down. Two 45's in cascade will have a combined gain of ~1:13. Your amp will have a gain less than 1.
It gets worse when you go 5k:8R which is a 25:1 step down. Consider perhaps a 26 driving a 45. Still not going to have a lot of gain, but power output won't need a lot of output voltage. A 45 SE won't do 4V rms into 8R, more like 3.5 V rms. A reasonable pre-amp/linestage should be able to deliver a proper signal to the amps.
cheers,
Douglas
jrdmedford said:
It's not a given in any case, effective cancellation doesn't result from simply cascading identical tubes. The driver must be set up to generate exactly the correct level of even harmonic and maintain it across a wide input variance. Since the two 45's will typically be under significantly different operating conditions it's by no means assured.
bottom_feeder said:
this and the power supply, will be the most important aspects, in the final sound.
i vote yes
from here-
http://www10.big.or.jp/~dh/sp/DHWS_2.html
- Through listening, I discovered that the driver's tone contributes the greater part of the amplifier's voice.-
An excellent power tube demands a similar level of quality from its driver. Few driver tubes can sing along in beautiful harmony with a good power tube ."
http://www10.big.or.jp/~dh/index.html#koko1
58
food for thought . . .
I appreciate all the input guys : )
to Jim's point, all my tubes mentioned in this thread are pulls from old radios, so not NOS, but I definitely see the point of not burning up a dwindling resource unnecessarily. . .
Doug: my opt's into an 8 Ohm load will be 2.2K:1 - I see the gain issues, and a preamp / linestage would definitely be used. The details there are yet to be decided, but ya gotta start somewhere
tomtt: the 'Sakuma Philosophy' was where I got the hairbrained idea
the power supply will be using a TON of oil caps, and I have some great old chokes which will be put to use also.
Loudspeakers to-be-driven for starters would be a very simple OB design; I have a few kicking around, and a few more 'in the works'. I'm not really expecting earth shaking volume, I basically want to explore the sound of the 45 tube. . .
Regardless, this will be a two-chassis design, so down the road I can experiment with different output stages . . . time to hit the books . . .
I appreciate all the input guys : )
to Jim's point, all my tubes mentioned in this thread are pulls from old radios, so not NOS, but I definitely see the point of not burning up a dwindling resource unnecessarily. . .
Doug: my opt's into an 8 Ohm load will be 2.2K:1 - I see the gain issues, and a preamp / linestage would definitely be used. The details there are yet to be decided, but ya gotta start somewhere
tomtt: the 'Sakuma Philosophy' was where I got the hairbrained idea
the power supply will be using a TON of oil caps, and I have some great old chokes which will be put to use also. Loudspeakers to-be-driven for starters would be a very simple OB design; I have a few kicking around, and a few more 'in the works'. I'm not really expecting earth shaking volume, I basically want to explore the sound of the 45 tube. . .
Regardless, this will be a two-chassis design, so down the road I can experiment with different output stages . . . time to hit the books . . .
Hi
2.2k :1 ? you meant 2.2K:8ohm? If so, this will not give you optimum power delivery and also poor damping factor... rp of 45 is 1700ohm, from the data sheet, recomand
plate load of
1) 4600 gets 2W
2) 3900 gets 1.6W
3) 2700 gets 0.85W
You may need to parallel 2 tubes to yse the 2.2k OPT to get 4W and/or high efficiency speakers...
Cheers
2.2k :1 ? you meant 2.2K:8ohm? If so, this will not give you optimum power delivery and also poor damping factor... rp of 45 is 1700ohm, from the data sheet, recomand
plate load of
1) 4600 gets 2W
2) 3900 gets 1.6W
3) 2700 gets 0.85W
You may need to parallel 2 tubes to yse the 2.2k OPT to get 4W and/or high efficiency speakers...
Cheers
OK, Some hints on how to get really good results with the 45
as well as the 6V6. It's cheapo parafeed and it sounds very good!
Go shopping around various WMass Radio Shacks and try to pickup
a couple of their 70V 10W line audio transformers. (P/N 32-1031B).
They are in a Red package. RS discontinued these a few years ago, but some stores still have them. I paid $2.00/ea. These guys have
4/8/16 ohm taps as well as multiple primary taps. Use the "2W" tap
for an equivalent 5K plate load.
You need a parafeed cap... Any oil cap around 4-8uF 400V will do fine.
For a plate load choke on the 45 or 6V6, you can use anything that
have about 30H inductance... A CCS would also work like a champ!
The RS transformers when done up in Parafeed are amazing... You
can get 15Hz-30kHz and beyond no problem. You will discover some other benefits of parafeed too - like having the PSU out of your signal loop.
-- Jim
as well as the 6V6. It's cheapo parafeed and it sounds very good!
Go shopping around various WMass Radio Shacks and try to pickup
a couple of their 70V 10W line audio transformers. (P/N 32-1031B).
They are in a Red package. RS discontinued these a few years ago, but some stores still have them. I paid $2.00/ea. These guys have
4/8/16 ohm taps as well as multiple primary taps. Use the "2W" tap
for an equivalent 5K plate load.
You need a parafeed cap... Any oil cap around 4-8uF 400V will do fine.
For a plate load choke on the 45 or 6V6, you can use anything that
have about 30H inductance... A CCS would also work like a champ!
The RS transformers when done up in Parafeed are amazing... You
can get 15Hz-30kHz and beyond no problem. You will discover some other benefits of parafeed too - like having the PSU out of your signal loop.
-- Jim
Audio_idiot: yup - I meant 2.2K:8ohm . . . looks like I'll use some 16 Ohm speakers simple enough; just have to dig into the 'console speaker stockpile'
The parafeed idea sounds REALLY interesting. There is a shop in the next town that was a Radio Shack until the mid-90's. He still had lots of the raw speaker drivers which I got for a song (HE tweeters, FR drivers, Lineaum dipole tweeters). My guess is he has some of those transformers kicking around too . . .
simple enough; just have to dig into the 'console speaker stockpile'The parafeed idea sounds REALLY interesting. There is a shop in the next town that was a Radio Shack until the mid-90's. He still had lots of the raw speaker drivers which I got for a song (HE tweeters, FR drivers, Lineaum dipole tweeters). My guess is he has some of those transformers kicking around too . . .
The Radio Shack parafeed option opens up many many possibilities.
Whats more, the Rat Shackers that I bought (mfgd in the 70s) had
nickel cores!
The higher wattage primary taps have lower impedance. So, to chase
more power out of your rat shackers you have to start finding driver
tubes with ever small plate resistance. The 45 and 6V6 are abit on
thie high impedance side. With 2A3's , or , triode wired 6/12/25AV5GAs, you're looking at Rp=800 ohms. Going lower gets
harder. The 6AS7 and various other shunt regulator tubes get you
Rp=300 ohms. With the 6AS7, you could get some nice power and
there are ST (bottle) shaped variants out there.
I used them with a SE 845 amp for awhile and was suprised at how
much better they sounded. Measurement clearly proved what I heard:
http://world.std.com/~doyle/845seBode.html
If you compare the "flatness" of the Bode plot in the HF, you see that the Rat Shacker is kicking the pants off the $90 and 13 pound Hammond 1628SE. This is because the former 1628SE part had
a design flaw that results in cruddy dropoff of HF around 13kHz.
The 26kHz limit, OTOH, is miller capacitance of the driver tube.
However, the 1628 was giving up long before miller capacitance began clobbering the gain.
The Rat Shackers would be ideal with some HE fullrangers or
horns esp. if you are on a brutal budget.
-- Jim
Whats more, the Rat Shackers that I bought (mfgd in the 70s) had
nickel cores!
The higher wattage primary taps have lower impedance. So, to chase
more power out of your rat shackers you have to start finding driver
tubes with ever small plate resistance. The 45 and 6V6 are abit on
thie high impedance side. With 2A3's , or , triode wired 6/12/25AV5GAs, you're looking at Rp=800 ohms. Going lower gets
harder. The 6AS7 and various other shunt regulator tubes get you
Rp=300 ohms. With the 6AS7, you could get some nice power and
there are ST (bottle) shaped variants out there.
I used them with a SE 845 amp for awhile and was suprised at how
much better they sounded. Measurement clearly proved what I heard:
http://world.std.com/~doyle/845seBode.html
If you compare the "flatness" of the Bode plot in the HF, you see that the Rat Shacker is kicking the pants off the $90 and 13 pound Hammond 1628SE. This is because the former 1628SE part had
a design flaw that results in cruddy dropoff of HF around 13kHz.
The 26kHz limit, OTOH, is miller capacitance of the driver tube.
However, the 1628 was giving up long before miller capacitance began clobbering the gain.
The Rat Shackers would be ideal with some HE fullrangers or
horns esp. if you are on a brutal budget.
-- Jim
Sure thing... It's easy to explain. Transformers are power transfer
devices. So, power in = power out. "70 V line" means that 70V is
the maximum AC level allowed on the line.
P = V^2 / Z. So, if 2 W is going through the transformer, and
70V is the maximum AC signal on the primary then Zpri = 70*70/2 =
2450 ohms. On the 0.5W tap, Zpri = 4900 ohms.
You can now figure out the turns ratio. With 2W coming out the 8 ohm
tap and 2 W going in the primary 2W tap, n = sqrt(4900 / 8) = 25
With the RS #32-1031B the primary taps are: 5, 10, 2.5, 1.25, 0.62 W. Build a table Zpri and work the formula for Zpri = 70*70/Pin. The 5 W tap is about 1 K...
There are also 100V line transformers out there too. You can
work the math. However, keep in mind the whole point of this
is budget. If you start wanting to buy a $20 line transformer, maybe its best to just get the real thing.
-- Jim
devices. So, power in = power out. "70 V line" means that 70V is
the maximum AC level allowed on the line.
P = V^2 / Z. So, if 2 W is going through the transformer, and
70V is the maximum AC signal on the primary then Zpri = 70*70/2 =
2450 ohms. On the 0.5W tap, Zpri = 4900 ohms.
You can now figure out the turns ratio. With 2W coming out the 8 ohm
tap and 2 W going in the primary 2W tap, n = sqrt(4900 / 8) = 25
With the RS #32-1031B the primary taps are: 5, 10, 2.5, 1.25, 0.62 W. Build a table Zpri and work the formula for Zpri = 70*70/Pin. The 5 W tap is about 1 K...
There are also 100V line transformers out there too. You can
work the math. However, keep in mind the whole point of this
is budget. If you start wanting to buy a $20 line transformer, maybe its best to just get the real thing.
-- Jim
Hi Jim,
Thanks for the clear explaination.
Always wanted to play with parafeed, but suitable OPT can cost me an arm or leg. I did see some matching transformers collecting dust in the local shops, and had been wondering how they can be useful. Will surely have a closer look at them soon.
I'll most likely use active load to replace the plate choke, also because premium quality plate choke are expensive too.
Cheers
Ken
Thanks for the clear explaination.
Always wanted to play with parafeed, but suitable OPT can cost me an arm or leg. I did see some matching transformers collecting dust in the local shops, and had been wondering how they can be useful. Will surely have a closer look at them soon.
I'll most likely use active load to replace the plate choke, also because premium quality plate choke are expensive too.
Cheers
Ken
jrdmedford said:
Hi Jim,
I bought a line matching tranny of the shelf for ~USD5 in local shop. But there isn't any specification on the max AC voltage alloable. So I can't really use your calculations.
Is it through that Matching Output Trans are autoformers? If so we can measure the DCR or L of the individual coils and multiply the load Z to botain the Zpri?
Cheers
Ken
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.