I was interested in Eli Duttman's suggestion in post 5 of the Looking to build a tube integrated thread, which he enlarged on in post10.
It got me thinking that I could replace a SS amp at home with such a thing; but I wanted more (what's new?)..:
25 watts
A cool running amp
Instant on from standby
Inactivity power-down to standby
These features are needed because of the expectations and bad habits engrained by using the SS amp for over 12 years.
Well, Eli has made some further suggestions and I thought I'd post my interpretation of them (errors accepted as mine). There are borrowings from others too.
I'm currently thinking about the PSU..
Here's the power amp bare bones:
It got me thinking that I could replace a SS amp at home with such a thing; but I wanted more (what's new?)..:
25 watts
A cool running amp
Instant on from standby
Inactivity power-down to standby
These features are needed because of the expectations and bad habits engrained by using the SS amp for over 12 years.
Well, Eli has made some further suggestions and I thought I'd post my interpretation of them (errors accepted as mine). There are borrowings from others too.
I'm currently thinking about the PSU..
Here's the power amp bare bones:
The 7591 has some interesting characteristics, low heater requirements and higher than usual transconductance which reduces drive requirements. It is not widely known and therefore has a novelty factor, however it is very likely that it will become hard to get in the future.
The first consideration IMO is to select tubes and an OT that fit the requirements, preferably tubes that will be around for some time and at a reasonable cost.
Have people checked the going price for 7591s with all the talk about them?
The EL34 also has high transconductance and higher plate dissipation, it's only disadvantage is higher heater current which is a minor point. It is widely available and much less expensive.
I would not use the 7591 in new designs, the EL34 is a good choice because it will be produced for all those sockets already out there.
Pete B.
The first consideration IMO is to select tubes and an OT that fit the requirements, preferably tubes that will be around for some time and at a reasonable cost.
Have people checked the going price for 7591s with all the talk about them?
The EL34 also has high transconductance and higher plate dissipation, it's only disadvantage is higher heater current which is a minor point. It is widely available and much less expensive.
I would not use the 7591 in new designs, the EL34 is a good choice because it will be produced for all those sockets already out there.
Pete B.
The 7591's are in production again from 2 sources: JJ & EH
They are even available in the UK for £10-16 each.
The source follower FET's do not have voltage gain, but act as current amplifiers to allow the driver stage 12AT7 to run with large value anode resistors; yes equvalent to cathode followers. Otherwise it would not be possible to drive the 7591 maximum grid leak of 300K in fixed bias without distortion.
They are even available in the UK for £10-16 each.The source follower FET's do not have voltage gain, but act as current amplifiers to allow the driver stage 12AT7 to run with large value anode resistors; yes equvalent to cathode followers. Otherwise it would not be possible to drive the 7591 maximum grid leak of 300K in fixed bias without distortion.
dhaen,
Don't know if you've seen this before, but here's a link to my DIY PP 7591 amp:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=64164&highlight=
I think the 7591 is a great tube-easy to drive, clean transparent sound-think SS with that great tube mid-range. I'm no expert by any stretch of the imagination, but if you've got a question about using this tube in a DIY amp that my feeble skills can answer, just give me a shout...
I use both the JJ and the EH tubes. Slightly different sound between each-JJ has the advantage of less adjustment of bias over time. Both have been reliable, good sounding tubes and run @ $60 US for a matched quad and are readily available.
Best,
mr mojo
Don't know if you've seen this before, but here's a link to my DIY PP 7591 amp:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=64164&highlight=
I think the 7591 is a great tube-easy to drive, clean transparent sound-think SS with that great tube mid-range. I'm no expert by any stretch of the imagination, but if you've got a question about using this tube in a DIY amp that my feeble skills can answer, just give me a shout...
I use both the JJ and the EH tubes. Slightly different sound between each-JJ has the advantage of less adjustment of bias over time. Both have been reliable, good sounding tubes and run @ $60 US for a matched quad and are readily available.
Best,
mr mojo
anatech said:
Hi Chris,
Not to get into a debate over this but the first price I came up with was $17.80 each ($53 for 4) for EL34EH and $20.20 each for the 7591EH ($76 for 4). Just my opinion to go with the EL34s and also have more plate dissipation headroom. Could always build with enough heater power for other tubes.
I like the 6L6GC and 5881 also but they don't provide as much power in a ST70 but obvioulsy work well in the old VTL type arrangement which I notice is very similar to the Fender Twin Reverb. Fender is smart enough to use low value grid leak resistors on the outputs, VTLs were over 300K in some older models as I remember.
I notice 100K in the design in this thread, might be better at half that.
Pete B.
dhaen said:The 7591's are in production again from 2 sources: JJ & EH
The source follower FET's do not have voltage gain, but act as current amplifiers to allow the driver stage 12AT7 to run with large value anode resistors; yes equvalent to cathode followers. Otherwise it would not be possible to drive the 7591 maximum grid leak of 300K in fixed bias without distortion.
Just my opinion about the tubes, no big deal.
The follower buffering looks similar to Modafferi's patent 4,918,394 from 1989:
http://patft.uspto.gov/netacgi/nph-...L&S1=4918394.WKU.&OS=PN/4918394&RS=PN/4918394
Pete B.
Hi Pete,
No intention to debate. I was just drawing attention to an easy source of good tubes. Price differential is not that large either. Not unless you go for inflated NOS sources. This was your initial point and a valid one at that.
7591 has a different sound than EL34, so the choice is really about taste, although I agree I'd use something more common for a new design. My preference is more towards a 6CA7 rather than EL34's, but that's just me.
mr mojo,
Excellent workmanship! Have you considered an angled mirror below. This would be for ventillation, of course.
-Chris
No intention to debate. I was just drawing attention to an easy source of good tubes. Price differential is not that large either. Not unless you go for inflated NOS sources. This was your initial point and a valid one at that.
7591 has a different sound than EL34, so the choice is really about taste, although I agree I'd use something more common for a new design. My preference is more towards a 6CA7 rather than EL34's, but that's just me.
mr mojo,
Excellent workmanship! Have you considered an angled mirror below. This would be for ventillation, of course.
-Chris
Hi Pete,
Being lazy, and not willing to hunt down reasonably priced NOS tubes ...
For the past three years, I have been buying both Electroharmonix and the JAN tubes from New Sensor. I've had good luck with them. About ten years ago I started with Sovtek. At least they weren't as noisy as the National brand and my NOS sources were drying up. The prices were getting nutty too.
I still get other tubes from time to time, but the Electroharmonix have been reliable and sound better than some NOS tubes (now that's sad). When it comes to 6L6GC time, I have to buy either 7581A or 6L6EH. Otherwise I have to worry about hum and other tube problems.
I don't have the time for that any more. I can not afford recalls on serviced equipment. Why do the job twice?
-Chris
Being lazy, and not willing to hunt down reasonably priced NOS tubes ...
For the past three years, I have been buying both Electroharmonix and the JAN tubes from New Sensor. I've had good luck with them. About ten years ago I started with Sovtek. At least they weren't as noisy as the National brand and my NOS sources were drying up. The prices were getting nutty too.
I still get other tubes from time to time, but the Electroharmonix have been reliable and sound better than some NOS tubes (now that's sad). When it comes to 6L6GC time, I have to buy either 7581A or 6L6EH. Otherwise I have to worry about hum and other tube problems.
I don't have the time for that any more. I can not afford recalls on serviced equipment. Why do the job twice?
-Chris
Perhaps it'll be a full-wave rectifier next.....
I wasn't suggesting any conflict hope it didn't come across that way. Thinking you might get ideas from the patent and that great minds think alike. Improving on existing patent ideas and experimenting with them is encouraged.
Chris, your comment on buying tubes makes perfect sense, thanks.
Pete B.
Chris, your comment on buying tubes makes perfect sense, thanks.
Pete B.
With a pair of 7591 you ought to do 20 W without getting too far from a true Class A bias. You'll be AB1, but it will be closer than most that claim the 'A'.
Stick with pentode connection, and run an E-Linear front end for the NFB. Try a good high-gm twin triode with a MOSFET upper element. I'll refer you to the Merlin V2 schematic in Audio Roundtable's Projects folder.
If you pick the load properly, and put the load line at half the a-a impedance, and run it out the g1=0V line to the left of the flat part, you'll be just fine. At maximum swing, you'll be looking into a quarter of a-a load and the line will be steeper at the upper end.
You'll get a minimum of odd order stuff for the NFB to flatten out, and you'll get max power for your load and bias scheme.
cheers,
Douglas
Stick with pentode connection, and run an E-Linear front end for the NFB. Try a good high-gm twin triode with a MOSFET upper element. I'll refer you to the Merlin V2 schematic in Audio Roundtable's Projects folder.
If you pick the load properly, and put the load line at half the a-a impedance, and run it out the g1=0V line to the left of the flat part, you'll be just fine. At maximum swing, you'll be looking into a quarter of a-a load and the line will be steeper at the upper end.
You'll get a minimum of odd order stuff for the NFB to flatten out, and you'll get max power for your load and bias scheme.
cheers,
Douglas
John,
R5 and R6 are way too big. A 1.5 KV. B+ rail is needed for 470 KOhms.
Ib is 3 mA. per 'T7 section. 68 KOhm INDUCTIVE wirewound parts seem right. The FET buffer allows "rational" values for the B+ rail and load resistor.
Speaking of FET buffers, I see you selected the topology that works for an AB2 amp. I was thinking in terms of the simpler DC coupled gate and cap. coupled source. With a DC coupled gate and cap. coupled source, PSU requirements are B+, B-, and C-. I don't think a 7591 will be "happy" if control grid current is present. The FET I have in mind is the International Rectifier IRFBG20.
IRFBG20 Info.
Might as well take advantage of the LOW O/P impedance of the source follower. Make the 7591 grid leak resistors 33 KOhms. The IRFBG20's transconductance (gfs) is a "mere" 1 A./V.
R5 and R6 are way too big. A 1.5 KV. B+ rail is needed for 470 KOhms.
Ib is 3 mA. per 'T7 section. 68 KOhm INDUCTIVE wirewound parts seem right. The FET buffer allows "rational" values for the B+ rail and load resistor.Speaking of FET buffers, I see you selected the topology that works for an AB2 amp. I was thinking in terms of the simpler DC coupled gate and cap. coupled source. With a DC coupled gate and cap. coupled source, PSU requirements are B+, B-, and C-. I don't think a 7591 will be "happy" if control grid current is present. The FET I have in mind is the International Rectifier IRFBG20.
IRFBG20 Info.
Might as well take advantage of the LOW O/P impedance of the source follower. Make the 7591 grid leak resistors 33 KOhms. The IRFBG20's transconductance (gfs) is a "mere" 1 A./V.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.