hmm, looking at the graph of the 6080's grid current/voltage it looks like for on output independence of 500ohms the grid voltage is supposed to go from 0 to -80 volts for the full output swing?
I took one more look in the junk box, and lord behold there was a little dual triode; a 12ax7a (ecc83)...
So I if I were going to use the 12ax7a, what values would I have to use for the preamp? I can't find any data sheet that lists data for the tube being used with a 130v supply, only 250v... Will it work that low, or do I need to use a voltage doubler to get b+ up that high?
Thanks!
I took one more look in the junk box, and lord behold there was a little dual triode; a 12ax7a (ecc83)...
So I if I were going to use the 12ax7a, what values would I have to use for the preamp? I can't find any data sheet that lists data for the tube being used with a 130v supply, only 250v... Will it work that low, or do I need to use a voltage doubler to get b+ up that high?
Thanks!
The ECC83 is a high gain tube which is usually used in microphone or phono amplifier duty. It may not have enough current to drive the miller capacitance of the 6080 grids. It might just have the gain you need.
It has a very high output impedance at 80K which will limit its output current capabilities. Still the 6080 datasheet says that the maximum grid resistance for cathode bias is 1megohm which suggests it doesn't need a lot of current to drive the grid so it might just work. Someone else would have to supply a definitive answer.
I will try to see a data sheet an express an opinion.
Shoog
It has a very high output impedance at 80K which will limit its output current capabilities. Still the 6080 datasheet says that the maximum grid resistance for cathode bias is 1megohm which suggests it doesn't need a lot of current to drive the grid so it might just work. Someone else would have to supply a definitive answer.
I will try to see a data sheet an express an opinion.
Shoog
My gut feeling is that the ECC83 wont work in this application. This is for two reasons;
1 - to much gain (especially if CCS loaded). This thing has a gain of 100 and you need at most 50.
2 -when running at 100V and -1V bias it will only pass 0.5mA current, I would be much happier in the 10mA range.
I would say that if you can hold out for a ECC88 variant you will be a lot happier with the result.
Shoog
1 - to much gain (especially if CCS loaded). This thing has a gain of 100 and you need at most 50.
2 -when running at 100V and -1V bias it will only pass 0.5mA current, I would be much happier in the 10mA range.
I would say that if you can hold out for a ECC88 variant you will be a lot happier with the result.
Shoog
Maybe there is another reason, we are old enough to have lived when you could get some old TV's and HiFi sets from the trash and have parts to build a good stereo or guitar amp. I have been building tube amps since the 60's and have only started spending mode than $100 on them a few years ago.
This is possible but not necessarilly true. A transformer that was designed for a guitar amplifier may (and probably does) have restricted frequency response, expecially at the low end. The Triode Electronics transformer that you mention does fall pretty short at the low end. My 300Beast, a push pull amplifier using 300B tubes has guitar amp output transformers. It sounds absolutely awesome, (a design by Kevinkr) but was not cheap.
There was a thread that I took part in where some low cost single ended output transformers were tested.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=72654
The Edcors sound great and cost $18 each. They would work well for SE amps up to about 10 watts per channel. There were two different schematics for simple tube amps based on 6EW7's or 6DN7's or similar vertical output tubes mentioned in this thread. I have built similar amps using the 6EA7/6EM7. They sound great, but only deliver 3 watts per channel.
I used the Allied electronics 6K56VG power transformer ($37) for a total iron cost of $73. Add two 6EM7's for $6 each, and a Chinese 5AR4 for $10 and you are at $95. If you can scrounge the caps and resistors you are done. This a simple design that sounds excellent.
It is possible to use power transformers for output transformers, but you need to keep the DC out of them. That is what the "parafeed" design is all about. The 10M45 chip makes an excellent CCS but it only is good to 100mA.
"It is possible to use power transformers for output transformers, but you need to keep the DC out of them. That is what the "parafeed" design is all about. The 10M45 chip makes an excellent CCS but it only is good to 100mA."
Parafeed is the design the man is working on. The design calls for a CCS of 200mA, can you suggest a setup using these chips which will work at that current. If not he can go SEPP with the two halves of the 6080 for a current of 100mA. Will the 10M45 survive a constant current of 100mA without failing.
Shoog
Parafeed is the design the man is working on. The design calls for a CCS of 200mA, can you suggest a setup using these chips which will work at that current. If not he can go SEPP with the two halves of the 6080 for a current of 100mA. Will the 10M45 survive a constant current of 100mA without failing.
Shoog
I haven't tried paralleling 10M45's, but it should be possible. The 10M45 can put out 100mA for ever but you need to keep the dissipation below the maximum spec. A heat sink is required. For high current CCS circuits, I have the used LM317 wired as a CCS and used to set the cathode current of a pass tube. The usual two transistor or transistor LED circuit can be used also.
Driver tube?
I tend to seperate the voltage gain and current sourcing aspects of a driver circuit by using a mosfet follower. Many die hard tubeheads dismiss this " sandy nonsense" but those who have tried it have liked it, and often send me email saying so. The circuit is on my web site.
http://www.tubelab.com/powerdrive.htm
If you use this technique the mosfet provides the current drive, so that you can choose the driver tube to have the right amount of voltage gain, and set its bias for the most linear operating point. There are dozens of tubes that are pin compatible with your 12ax7. This gives you plenty of options.
Make a list of the tubes that you have available, and we will find something.
Driver tube?
I tend to seperate the voltage gain and current sourcing aspects of a driver circuit by using a mosfet follower. Many die hard tubeheads dismiss this " sandy nonsense" but those who have tried it have liked it, and often send me email saying so. The circuit is on my web site.
http://www.tubelab.com/powerdrive.htm
If you use this technique the mosfet provides the current drive, so that you can choose the driver tube to have the right amount of voltage gain, and set its bias for the most linear operating point. There are dozens of tubes that are pin compatible with your 12ax7. This gives you plenty of options.
Make a list of the tubes that you have available, and we will find something.
I am not worried about power dissipation for the 10m45 because there will never be more than 30w (150v@.2a) on it, which is will within the power dissipation rating for a to-220 case (40w if I remember correctly). Of course it will need heat sinking...
Hmm, your powerdrive looks interesting... However, I don't really see the need to complicate things for this project, it seems like the easiest/cheapest way for me to get this amp off the ground would be to buy some $3 dual triode... I just need to know which one...
As far as the tubes I have relating to amps...
12ax7 (1)
0a2 (2)
6080 (2)
6146 (2)
some unmarked dual triode. It came out of an old military radio, but I have no idea as to what it's operating charistics are so it is useless...
I also have some larger tubes, that I acquired as a single lot when buying the 811a's for my coil that I have plans to sell... There are 8 829b's, 4 CV415's, and a QQVO6-40A, which is the is the extent of the tubes that would be even in the ballpark of a bedroom amp...
Really for this amp I just want to keep it with the lowest part count of the cheapest parts, so that I can get a working amp for minimum investment. If I so feel inclined I can mess with better parts/design, but for now I just want a working amp... Anyone can go out and buy $500 in parts and solder them together to make an decent amp, but in my opinion it takes some skill to buy $50 in parts and make a decent amp.
Thanks!
Hmm, your powerdrive looks interesting... However, I don't really see the need to complicate things for this project, it seems like the easiest/cheapest way for me to get this amp off the ground would be to buy some $3 dual triode... I just need to know which one...
As far as the tubes I have relating to amps...
12ax7 (1)
0a2 (2)
6080 (2)
6146 (2)
some unmarked dual triode. It came out of an old military radio, but I have no idea as to what it's operating charistics are so it is useless...
I also have some larger tubes, that I acquired as a single lot when buying the 811a's for my coil that I have plans to sell... There are 8 829b's, 4 CV415's, and a QQVO6-40A, which is the is the extent of the tubes that would be even in the ballpark of a bedroom amp...
Really for this amp I just want to keep it with the lowest part count of the cheapest parts, so that I can get a working amp for minimum investment. If I so feel inclined I can mess with better parts/design, but for now I just want a working amp... Anyone can go out and buy $500 in parts and solder them together to make an decent amp, but in my opinion it takes some skill to buy $50 in parts and make a decent amp.
Thanks!
Tubelabs can you suggest component values for using the 10M45 at 100mA.
What voltage drop should he expect to get 100V on the plate at 100mA.
I would suggest that for simplicity you go with parallel 10M45's. The other alternative would be to set up the 6146's as current sources. The power supply would be twice as big and the heat would be twice as much. Unfortunately this would mean a less stiff current source which would need at least 100V to function. Not a big issue, but the 10M45 seems simpler in the circumstances.
On the matter of the driver - your just going to have to get another tube.
Shoog
What voltage drop should he expect to get 100V on the plate at 100mA.
I would suggest that for simplicity you go with parallel 10M45's. The other alternative would be to set up the 6146's as current sources. The power supply would be twice as big and the heat would be twice as much. Unfortunately this would mean a less stiff current source which would need at least 100V to function. Not a big issue, but the 10M45 seems simpler in the circumstances.
On the matter of the driver - your just going to have to get another tube.
Shoog
The subject of a chokeless parafeed design has been discussed before in this thread:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=67437&highlight=
This thread lead to at least two prototype amplifier designs, one of which was mine. The point is that when you use a CCS instead of a choke for the plate load, you loose the property of an inductor that allows the plate voltage to rise above the power supply value. I have verified this by experimentation. If you use a CCS load the power supply needs to have enough headroom to allow the plate voltage move upward as the grid goes negative. If you intend to operate the tube at 100 volts on the plate you will need at least 175 volts of B+, probably closer to 200 to avoid distortion.
The 6AS7/6080 was never intended to be used as an audio amplifier tube (although it has been used sucessfully as an audio tube). It was designed as a voltage regulator. In regulator service it operates inside a feedback loop, so thermal stability is not an issue. I experimented with the 6AS7/6080 tube several years ago and found that they are prone to instabillity and destructive runaway if operated in fixed bias situations. This means that you need to use cathode bias to avoid this. This has also been mentioned in a previous thread, I just don't remember which one. Now your B+ needs to be at least 250 volts.
I have been building cheap amplifiers for a long time. I have used all sorts of unusual transformers for output transformers, and regular filament transformers just don't sound good. The toroidial ones do work pretty good though, but they cost as much as real output transformers. What good is a $50 amp if you don't want to listen to it. The reason that I mentioned the 6EM7 - 6DN7 - 6EW7 design is that it can be built for just over $100 and it sounds great, lots of them have been built, and it is a proven design that was the basis for the S.E.X. amplifier from Bottlehead.
If you want to build a minimum dollar amp that uses the 6080 tube, I would suggest using 1/2 of the 12AX7 to drive 1/2 of the 6080 in a conventional circuit design similar to this one. The resistor values will need to be tweaked to set the bias for the different tubes. You can use the filament transformers that you already have, and a voltage doubler off of an isolation transformer to generate about 280 volts of B+. I would run the 6080 at about 60 to 80 mA which would put about 80 to 100 volts at the cathode. The resistor would need to be rated for at least 10 watts, or you could put a 10M45 here. With careful design direct coupling could be used. Output power would be about 2 watts. I would use The $18 SE output transformer from Edcor. This would be a good sounding amp. A 50 VA isolation transformer is about $15, and the Edcor transformers are $18 each. $51 plus shipping.
http://www.audiohobbyist.com/images/6ew7/6EW7sch.JPG
More power? Then build a Push Pull amp using the 6080. One tube for each channel. You would need a phase splitter which would require another dual triode tube. It could be a 12AX7 or a 12AU7. Edcor has a $18 Push Pull output transformer that should work well. Power output, about 10 watts. Same cost as before, but you will need more small parts, and a few more tube sockets.
Back to the original question, the 10M45's vary quite a bit from lot to lot, so you need to measure the current in the actual circuit. 100mA should be around 22 ohms.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=67437&highlight=
This thread lead to at least two prototype amplifier designs, one of which was mine. The point is that when you use a CCS instead of a choke for the plate load, you loose the property of an inductor that allows the plate voltage to rise above the power supply value. I have verified this by experimentation. If you use a CCS load the power supply needs to have enough headroom to allow the plate voltage move upward as the grid goes negative. If you intend to operate the tube at 100 volts on the plate you will need at least 175 volts of B+, probably closer to 200 to avoid distortion.
The 6AS7/6080 was never intended to be used as an audio amplifier tube (although it has been used sucessfully as an audio tube). It was designed as a voltage regulator. In regulator service it operates inside a feedback loop, so thermal stability is not an issue. I experimented with the 6AS7/6080 tube several years ago and found that they are prone to instabillity and destructive runaway if operated in fixed bias situations. This means that you need to use cathode bias to avoid this. This has also been mentioned in a previous thread, I just don't remember which one. Now your B+ needs to be at least 250 volts.
I have been building cheap amplifiers for a long time. I have used all sorts of unusual transformers for output transformers, and regular filament transformers just don't sound good. The toroidial ones do work pretty good though, but they cost as much as real output transformers. What good is a $50 amp if you don't want to listen to it. The reason that I mentioned the 6EM7 - 6DN7 - 6EW7 design is that it can be built for just over $100 and it sounds great, lots of them have been built, and it is a proven design that was the basis for the S.E.X. amplifier from Bottlehead.
If you want to build a minimum dollar amp that uses the 6080 tube, I would suggest using 1/2 of the 12AX7 to drive 1/2 of the 6080 in a conventional circuit design similar to this one. The resistor values will need to be tweaked to set the bias for the different tubes. You can use the filament transformers that you already have, and a voltage doubler off of an isolation transformer to generate about 280 volts of B+. I would run the 6080 at about 60 to 80 mA which would put about 80 to 100 volts at the cathode. The resistor would need to be rated for at least 10 watts, or you could put a 10M45 here. With careful design direct coupling could be used. Output power would be about 2 watts. I would use The $18 SE output transformer from Edcor. This would be a good sounding amp. A 50 VA isolation transformer is about $15, and the Edcor transformers are $18 each. $51 plus shipping.
http://www.audiohobbyist.com/images/6ew7/6EW7sch.JPG
More power? Then build a Push Pull amp using the 6080. One tube for each channel. You would need a phase splitter which would require another dual triode tube. It could be a 12AX7 or a 12AU7. Edcor has a $18 Push Pull output transformer that should work well. Power output, about 10 watts. Same cost as before, but you will need more small parts, and a few more tube sockets.
Back to the original question, the 10M45's vary quite a bit from lot to lot, so you need to measure the current in the actual circuit. 100mA should be around 22 ohms.
If you say that normal plate transformers are not going to sound good then there is not much point in buying parts around them...
Lets see if I were to go pp I would need
2 more 12ax7's ($5/piece here)
2 o/p xormers ($16 here)
A 110->220v transformer... ($20 shipped )
If I am going to buy a transformer it might as well have the right output voltage...
ttl=$60 + some shipping... And I will still need the sockets and other parts... A little high but reasonable... Is there anyone arround here that has an old isolation transformer and the 8/9 pin tube sockets for sale?
Is there any other things I would have to lookout for when going pp? Since it is not going to be parafeed any more, can I eliminate the css? And what kind of powersupply considerations do I need to make?
Thanks!
Lets see if I were to go pp I would need
2 more 12ax7's ($5/piece here)
2 o/p xormers ($16 here)
A 110->220v transformer... ($20 shipped )
If I am going to buy a transformer it might as well have the right output voltage...
ttl=$60 + some shipping... And I will still need the sockets and other parts... A little high but reasonable... Is there anyone arround here that has an old isolation transformer and the 8/9 pin tube sockets for sale?
Is there any other things I would have to lookout for when going pp? Since it is not going to be parafeed any more, can I eliminate the css? And what kind of powersupply considerations do I need to make?
Thanks!
I am sorry about the double post, but I cant figure out how to modify the old one...
I won a mixed lot of 12au7's and some other tubes on ebay ($5 shipped ), and was starting try to figure out the values for all of the other compoents I need...
The first thing I did was try to figure out the plate voltage...
With a 5k transformer (as previously noted as a good being a choice), and a 10w output power, I calcuated 220v (v=sqt[pr]) and 44ma (r=srt[p/r], and plotted it on the characteristics chart for the 8060; but it was up in the corner of the graph where it would take like -100v on the plate to make it work (if even that, the grid voltage lines did not exend that far up).
Is this because for a push pull design the tube see's only 1/2 of indepedance of the o/p transformer, so I need to set the operating point for 2.5k? In that case I need 66ma at 150v... but it still looks like about -70v is necessary on the grid
I won a mixed lot of 12au7's and some other tubes on ebay ($5 shipped
), and was starting try to figure out the values for all of the other compoents I need...The first thing I did was try to figure out the plate voltage...
With a 5k transformer (as previously noted as a good being a choice), and a 10w output power, I calcuated 220v (v=sqt[pr]) and 44ma (r=srt[p/r], and plotted it on the characteristics chart for the 8060; but it was up in the corner of the graph where it would take like -100v on the plate to make it work (if even that, the grid voltage lines did not exend that far up).
Is this because for a push pull design the tube see's only 1/2 of indepedance of the o/p transformer, so I need to set the operating point for 2.5k? In that case I need 66ma at 150v... but it still looks like about -70v is necessary on the grid
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.