I want to build a PP amp with either KT88 or EL34 (or possibly 6L6GC?) tubes.
The main 'candidates' on my list right now are:
Pointz (AudioTropic) Musical Machine with 6GK5 and EL34 tubes.
http://usr.audioasylum.com/images/2/20130/Musical_Machine_6gk5_Audio_Shematic.jpg
EDIT: this is the wrong schematic, sorry.
Allen Wright KT88 PP2C but building it without the 'SuperReg' - using the same type of power supply as in the PP1C:
PP2C
http://basaudio.net/gm70/index.php?action=dlattach;topic=89.0;attach=205
PP1C Power Supply
http://www.vacuumstate.com/images_upload/gross/pp_1c_ps_s.gif
or Allen Wright EL34 PP1C
http://www.vacuumstate.com/images_upload/gross/pp_1c_s.gif
I've read a couple of cryptic comments that the Musical Machine has 'design problems', so I'm wondering about that.
I'm not interested in very complex power supplies that need PCBs and a lot of SS regulation (like the SuperReg), though I have no problems with wiring up something on terminal strips or perforated board- like the CCS in those amps.
I won't be buying 'audiophile grade' capacitors and resistors.
Output transformers will be coming from my parts shelf (Fisher and similar) or ordered from Edcor.
Speakers will be Troels Gravesen DTQWT or AudioNirvana 10s - both in the 95dB sensitivity range.
I already have an EL34 SE amp, and some other SE projects on my list, so concentrating on PP designs for now.
What other designs do you think I should add to my list for consideration?
Thanks!
The main 'candidates' on my list right now are:
Pointz (AudioTropic) Musical Machine with 6GK5 and EL34 tubes.
http://usr.audioasylum.com/images/2/20130/Musical_Machine_6gk5_Audio_Shematic.jpg
EDIT: this is the wrong schematic, sorry.
Allen Wright KT88 PP2C but building it without the 'SuperReg' - using the same type of power supply as in the PP1C:
PP2C
http://basaudio.net/gm70/index.php?action=dlattach;topic=89.0;attach=205
PP1C Power Supply
http://www.vacuumstate.com/images_upload/gross/pp_1c_ps_s.gif
or Allen Wright EL34 PP1C
http://www.vacuumstate.com/images_upload/gross/pp_1c_s.gif
I've read a couple of cryptic comments that the Musical Machine has 'design problems', so I'm wondering about that.
I'm not interested in very complex power supplies that need PCBs and a lot of SS regulation (like the SuperReg), though I have no problems with wiring up something on terminal strips or perforated board- like the CCS in those amps.
I won't be buying 'audiophile grade' capacitors and resistors.
Output transformers will be coming from my parts shelf (Fisher and similar) or ordered from Edcor.
Speakers will be Troels Gravesen DTQWT or AudioNirvana 10s - both in the 95dB sensitivity range.
I already have an EL34 SE amp, and some other SE projects on my list, so concentrating on PP designs for now.
What other designs do you think I should add to my list for consideration?
Thanks!
Last edited:
8 WPC is plenty for 95 dB. sensitive speakers. KT88s are overkill. 7591s are as easy to drive as "12" W. tubes. Triode wired 7591s should prove ample, for your purposes.
Poinz's "Musical Machine" would do nicely. A "taller" B+ PSU and different O/P "iron" are the changes need to switch from the "12" W. 6V6 to the "19" W. 7591.
Poinz's "Musical Machine" would do nicely. A "taller" B+ PSU and different O/P "iron" are the changes need to switch from the "12" W. 6V6 to the "19" W. 7591.
That one is SE not PP. They do have a SPP but I don't see it being built as much on here so there isn't as much experience to feed off of.
Thanks for all the quick replies. Love this place!
I found the comment about "problems" with the Musical Machine - it was SY who said:
Eli-
I was thinking that I was looking at too high power (KT88) for my speakers but I keep running across comments that the 'extra' is needed for bass, transients, etc.. For example, Troels Gravesen says:
I don't want to end up with an amp where I adjust the volume only between zero and 2 (on a 0-10 scale) . I have a SS amp for that!
I thought Williamson designs needed good/very large/specifically wound output transformers?
I found the comment about "problems" with the Musical Machine - it was SY who said:
Eli-
I was thinking that I was looking at too high power (KT88) for my speakers but I keep running across comments that the 'extra' is needed for bass, transients, etc.. For example, Troels Gravesen says:
I don't want to end up with an amp where I adjust the volume only between zero and 2 (on a 0-10 scale) . I have a SS amp for that!
I thought Williamson designs needed good/very large/specifically wound output transformers?
From what I have read here and elsewhere, SE is better for High efficiency speakers, more so than PP, YMMV.
Is that what Williamson used in 1949? Dynaco was built on an improved transformer for Williamson amps. You don't have to get any more esoteric than that type of modern quality. They also explained how to improve the stability and reduce the stages. Lots of good 2-channel mid to late '50's circuits used of the day trannies for Williamson type circuits. But the real improvements came from better driver stage design for less phase shifting allowing more GNF.
I've heard the same but I was just going off of this line: "I already have an EL34 SE amp, and some other SE projects on my list, so concentrating on PP designs for now."
I am building DTQWTII and I will feed it with a PRIMA LUNA ( 38 watts/channel )
I am currently using it with vintage JBL LE14C from around 95/96 DB, and it is not so.
Power control is done by the potentiometers that are logarithmic, so that our ears, which are also, are compensated ....
You will NOT have any problems at all with a PP KT88. And you can hear high SPLs without the typical cross distortion of amplifiers SS class AB.
Everyone has been typing faster than me. But here goes some points:
Amp # 1
At turn on, the 6GK5 can have large grid current equal to the current source value, if the B-supply powers up before the B+ plate supply.
Or, on a ‘hot start’ after a very brief power outage (filaments still warm, but B- comes up before B+ comes up).
Wow! way to much grid current (can potentially damage the tube).
You will have to be sure about the order that the 2 power supplies power up.
The grid current can come out the volume 100k input pot, especially if the pot is turned all the way up. Might be a disaster for some solid state preamps, or CD output ICs.
If the power supplies do come up in the wrong order, higher resistance grid stopper resistors could limit the grid current.
But it is not practical to limit the grid current by changing the grid stopper resistors. For example, 64k Ohms would limit the grid current to 1mA. But that would severely roll off the high frequency response.
The 6V6 are using an 11k plate to plate transformer.
Those are not likely on old Fisher amps, and probably not in your parts box.
If those diodes are 100V zener, that is the wrong symbol for a zener.
And if they are a zener, they are drawn in the wrong polarity. The cathode always goes to the more positive voltage (towards the 330 Ohm resistor).
If those are ‘regular’ diodes on the 6V6 screen, don’t use them; just connect directly across where they would have been.
Pentode mode of this amp has a very low damping factor.
The frequency response will be almost entirely dependent on the loudspeaker impedance versus frequency.
Negative feedback, if properly implemented could fix this, but not that easy to do unless you know how to calculate and or measure, etc.
And do not let the loudspeaker ever be disconnected, it may destroy the output transformer and 6V6s, because of the super high gain of unloaded pentodes.
And this amp uses 6V6, not EL34, KT88, etc. That would require lots of extra re-design.
Amp # 2
These are much more complex amps and power supplies.
Again, the amp can draw grid current under a wrong order of B+ and B-power supply turn-on,
and/or ‘hot’ start B+ and B- power recovery times and order.
But the current source supply could rise to 1.4 times -12V.
So, grid stoppers that are large enough to limit the current may also cause high frequency roll off.
You have to calculate the resistance necessary to limit the current to a value you decide on.
Then you have to find the -3 dB bandwidth of that resistance versus the grid to plate capacitance times the gain of the plate in the bottom of the cascode.
Cascode connected ECC88s will almost certainly have to have elevated filament supplies.
The same might be true for 6H30s. It depends on the tube cathode to filament voltage ratings, and what voltage the cathode of the ‘top’ tube of the cascode is at (including the small signal voltage swing).
The triode wiring of these KT88 and EL34 amps provides a moderate damping factor.
The output transformer primary impedance, the tube type, and the amount of standing current all determine how good it is.
Amp # X
Amps with input stages that are cathode coupled phase splitters can work too.
I have used these myself, but careful design is required to have limited grid current,
good high frequency response, and enough signal swing for the rest of the amp.
Other Amps
As noted by others in this thread, there are many push pull amps out there that could do a good job, with little or no modifications.
Most have favorites, and each one should have good points, but may also have tradeoffs.
Amp # 1
At turn on, the 6GK5 can have large grid current equal to the current source value, if the B-supply powers up before the B+ plate supply.
Or, on a ‘hot start’ after a very brief power outage (filaments still warm, but B- comes up before B+ comes up).
Wow! way to much grid current (can potentially damage the tube).
You will have to be sure about the order that the 2 power supplies power up.
The grid current can come out the volume 100k input pot, especially if the pot is turned all the way up. Might be a disaster for some solid state preamps, or CD output ICs.
If the power supplies do come up in the wrong order, higher resistance grid stopper resistors could limit the grid current.
But it is not practical to limit the grid current by changing the grid stopper resistors. For example, 64k Ohms would limit the grid current to 1mA. But that would severely roll off the high frequency response.
The 6V6 are using an 11k plate to plate transformer.
Those are not likely on old Fisher amps, and probably not in your parts box.
If those diodes are 100V zener, that is the wrong symbol for a zener.
And if they are a zener, they are drawn in the wrong polarity. The cathode always goes to the more positive voltage (towards the 330 Ohm resistor).
If those are ‘regular’ diodes on the 6V6 screen, don’t use them; just connect directly across where they would have been.
Pentode mode of this amp has a very low damping factor.
The frequency response will be almost entirely dependent on the loudspeaker impedance versus frequency.
Negative feedback, if properly implemented could fix this, but not that easy to do unless you know how to calculate and or measure, etc.
And do not let the loudspeaker ever be disconnected, it may destroy the output transformer and 6V6s, because of the super high gain of unloaded pentodes.
And this amp uses 6V6, not EL34, KT88, etc. That would require lots of extra re-design.
Amp # 2
These are much more complex amps and power supplies.
Again, the amp can draw grid current under a wrong order of B+ and B-power supply turn-on,
and/or ‘hot’ start B+ and B- power recovery times and order.
But the current source supply could rise to 1.4 times -12V.
So, grid stoppers that are large enough to limit the current may also cause high frequency roll off.
You have to calculate the resistance necessary to limit the current to a value you decide on.
Then you have to find the -3 dB bandwidth of that resistance versus the grid to plate capacitance times the gain of the plate in the bottom of the cascode.
Cascode connected ECC88s will almost certainly have to have elevated filament supplies.
The same might be true for 6H30s. It depends on the tube cathode to filament voltage ratings, and what voltage the cathode of the ‘top’ tube of the cascode is at (including the small signal voltage swing).
The triode wiring of these KT88 and EL34 amps provides a moderate damping factor.
The output transformer primary impedance, the tube type, and the amount of standing current all determine how good it is.
Amp # X
Amps with input stages that are cathode coupled phase splitters can work too.
I have used these myself, but careful design is required to have limited grid current,
good high frequency response, and enough signal swing for the rest of the amp.
Other Amps
As noted by others in this thread, there are many push pull amps out there that could do a good job, with little or no modifications.
Most have favorites, and each one should have good points, but may also have tradeoffs.
The "El Cheapo" I'm associated with is similar to Poinz's "Musical Machine". I took slowing B- rise down into account. A hybrid bridge rectifier using an indirectly heated type achieves the need. 
The OP is recycling Fisher 7591 "iron" and seems to be leaning towards that tube. He has to decide whether or not regarding GNFB. He definitely has plenty of gain from the 6GK5 LTP, to make either option viable.
The OP is recycling Fisher 7591 "iron" and seems to be leaning towards that tube. He has to decide whether or not regarding GNFB. He definitely has plenty of gain from the 6GK5 LTP, to make either option viable.
Attachments
Last edited by a moderator:
In Allen Wright's PP1C amplifier the ECC88 filaments are not elevated, and checking the schematic it is not necessary.
I'd say the filaments of "upper" and "lower" tube are some 60V apart.
Please note that the ECC88 was a tube specially designed for cascode circuits.
Another to consider is "Gingertubes" Baby Huey EL84 pp amp. Has an extensive thread on this forum. Less power than the other amps but still plenty for your speakers. Inexpensive.
I have built both the Baby Huey and PP1C/PP2C and would strongly recommend either. On my less sensitive speakers the Allen Wright designs were maybe more dynamic.
I have built both the Baby Huey and PP1C/PP2C and would strongly recommend either. On my less sensitive speakers the Allen Wright designs were maybe more dynamic.
I don't mind buying (cheap Edcor or similar) iron and I have a few other options from the parts box, so no decisions yet! But leaning, yes!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.