Maybe this has already been discussed here, but....
I've never liked the sound of 2A3s, 6B4Gs, etc. if the drivers are RC coupled to their grids. For some reason these power triodes like a strong, low impedance driver which RC coupling can't provide, at least not at the drive voltages required by these output triodes.
If you have access to the old Sound Practices magazine, look at an article in Issue #6 called "Hidrotrac". It's about a PP 2A3 amp that the designer used a cathode follower direct coupled to the grids of the triodes as drivers. The author explains that even though 2 2A3s in PP are rated for 15 watts output, he was never able to get more than 7-8 watts output without distortion setting in using an RC coupled driver stage. Direct coupling of the triodes brings the power output to 14-15 watts in a PP pair, where it should be.
Of course, IT coupling will also work great in this application, also.
One more thing... I find that these amps are very dependent on the output transformer for good sound, especially if no GNF is used. I had to try several output transformers before I found a pair that really made this amp sing. I ended up using a pair of UTC transformers.
Daniel
I've never liked the sound of 2A3s, 6B4Gs, etc. if the drivers are RC coupled to their grids. For some reason these power triodes like a strong, low impedance driver which RC coupling can't provide, at least not at the drive voltages required by these output triodes.
If you have access to the old Sound Practices magazine, look at an article in Issue #6 called "Hidrotrac". It's about a PP 2A3 amp that the designer used a cathode follower direct coupled to the grids of the triodes as drivers. The author explains that even though 2 2A3s in PP are rated for 15 watts output, he was never able to get more than 7-8 watts output without distortion setting in using an RC coupled driver stage. Direct coupling of the triodes brings the power output to 14-15 watts in a PP pair, where it should be.
Of course, IT coupling will also work great in this application, also.
One more thing... I find that these amps are very dependent on the output transformer for good sound, especially if no GNF is used. I had to try several output transformers before I found a pair that really made this amp sing. I ended up using a pair of UTC transformers.
Daniel
Last edited:
Hi Daniel --
Completely agreed. Actually, a lot of folks have been using MOSFET followers instead of cathode followers. Tubelab, MOSFET Follies, gingertube's Baby Huey, etc.
Much of the limitation here is the OP's wish list of features. The amp needs to be really simple, inexpensive, and easy to implement. It also needs to use the chassis, transformers and tubes he already has.
PS - Just saw what you added about OPT's. Absolutely, I'm with ya there. The OP is using Dyna A470 OPT's from a Stereo 70. I'm lucky to have a pair of Hirata-era Tango XE-45-5 OPT's for my PP 2A3 amp.
________________________________________________________________
youtubeyou, would interstage transformers be too expensive?
http://www.lundahl.se/pdfs/datash/1660s.pdf ($150 each from K&K Audio)
Electra-Print.com Interstage Transformers
Hammond 124B Audio Interstage Transformer (Must be used in parafeed configuration, which requires a plate choke or CCS-loaded driver stage cap-coupled to the IT. The value of coupling cap needs to be chosen for best bass response without artificial boost.)
--
Last edited:
the amps doesnt need to be super simple, I just want to keep this project under 200$ therefore most likely keep the output tubes and opt and mains to reduce cost
The hammond interstage at 42 each is not out of budget, but I cannot change the opt for 300$, this blows the budget out of the water...
The hammond interstage at 42 each is not out of budget, but I cannot change the opt for 300$, this blows the budget out of the water...
The other transformers I posted about are not OPT's. They are interstage transformers that can take the plate current of the driver tube. You're right that they look like OPT's. You see, high performance interstage transformers are big and expensive. Just like OPT's.
The cheap Hammond interstage will need to be used with a large value film cap, large physically too. That's because the cheap Hammond cannot take the plate current from the driver stage. It needs to be used in parallel feed ("parafeed") configuration, with no DC current on its primary. The large value film cap will be directly in the signal path. Its quality will directly affect the sound quality. That may cost $$ unless you already have something you can use.
How much work are you willing to put into this?
- Would you be willing to add a couple of extra power supplies in each amp?
- Would you be willing to manually adjust a DN2540 based CCS to get the current you need? It cannot be done as easily as the LM317 CCS. Each DN2540 will need to be adjusted individually, using a trimmer. Each DN2540 will need a heatsink, so you'll need to purchase those and install them. You will also need the insulation wafers and other small hardware.
- Would you be willing to install two MOSFET followers for each amp? They will need heatsinks, so you'll need to find space for those.
Please take a look at Tubelab's and Chrish's PP 6L6GC AB2 amp circuit:
http://www.diyaudio.com/forums/atta...-chrish-6l6gc-pp-1-amp-schematic-post-460.pdf
That's the type of driver stage I would recommend for best performance at least monetary cost. Chrish posted about it here:
http://www.diyaudio.com/forums/tubes-valves/242483-mullard-5-20-build-mods-3.html#post3633501
Chrish mentioned that he wound up using the amp with no global NFB loop, and he's using 6L6GC outputs in ultralinear (not triode).
Apparently it can be done!
--
Last edited:
Wow you guys are moving very fast!
Well as you say, anything can be done, but to what direction? I would say that the direction has to be without additional holes on the chassis and without extra iron.
To go to this direction, but to keep the 6B4G , we either need extra power transformers (for the power drive ) or extra holes (for the cathode followers) Well maybe both, for the cathode follower route.
Suddenly, EL34's sounds a better idea than all this...
Well as you say, anything can be done, but to what direction? I would say that the direction has to be without additional holes on the chassis and without extra iron.
To go to this direction, but to keep the 6B4G , we either need extra power transformers (for the power drive ) or extra holes (for the cathode followers) Well maybe both, for the cathode follower route.
Suddenly, EL34's sounds a better idea than all this...
But how can you "add" feedback? If you add it like that, you suddenly need 10x the input voltage (for 20db for example). Well maybe if you can modify your preamp to have 10x the gain (if you have a peramp and if it is possible to modify it),
My point is that it's getting pretty complicated, you'd better design it in from the beginning. And all that without going into all the other considerations about different choices that have to be made when you use it or not.
I meant that 'in general.' You don't take an existing amp that wasn't designed for negative feedback, and then just add feedback to it. You're absolutely right that you first must have enough extra gain to lose by applying negative feedback. I wasn't clear in how I phrased my assertion.
youknowyou, if you really want to use NFB around this amplifier, you need to redesign the amp to have enough extra gain to lose when you add the feedback.
Also, are you using a preamp with gain?
Yes, I'm with you.
That's right. My earlier questions still stand. I'd like them answered before we get farther into this.
It's relatively easy to come up with circuits that will work to this or that degree. It's physical labor to punch and drill holes in chassis, etc.
Step Number One is to decide on a basic circuit. We are still at that stage.
At this point, we don't even know the current capacity of the power transformer. Are we at its limits? Is it rated for 150mA, or 200mA, or ???
That's a basic question that absolutely must be answered before we can go any further with this.
--
Would you be willing to add a couple of extra power supplies in each amp?
yes thats could be possible
Would you be willing to manually adjust a DN2540 based CCS to get the current you need? It cannot be done as easily as the LM317 CCS. Each DN2540 will need to be adjusted individually, using a trimmer. Each DN2540 will need a heatsink, so you'll need to purchase those and install them. You will also need the insulation wafers and other small hardware.
How hard is it to manually adjust a DN2540?
- Would you be willing to install two MOSFET followers for each amp? They will need heatsinks, so you'll need to find space for those.
I think theres enough place in the chassis, so yes thats is possible. I can follow everything as long as someone can add all the segments on the schematic.
As for the power transformers. Id say that when I was pushing the amp pretty loud, the power transformer were hot, I couldnt stand having my hand more then 6 seconds on their sides...
If you guys think im better off without NFB, I will not insist. I really trust what you guys will tell me.
EL34 are not cheap tube, but I'm willing to change the 6B4G for EL34 if that can help.
I'm willing to change no matter what you think must be change as long as its within the budget, 200$ maximum...
yes thats could be possible
Would you be willing to manually adjust a DN2540 based CCS to get the current you need? It cannot be done as easily as the LM317 CCS. Each DN2540 will need to be adjusted individually, using a trimmer. Each DN2540 will need a heatsink, so you'll need to purchase those and install them. You will also need the insulation wafers and other small hardware.
How hard is it to manually adjust a DN2540?
- Would you be willing to install two MOSFET followers for each amp? They will need heatsinks, so you'll need to find space for those.
I think theres enough place in the chassis, so yes thats is possible. I can follow everything as long as someone can add all the segments on the schematic.
As for the power transformers. Id say that when I was pushing the amp pretty loud, the power transformer were hot, I couldnt stand having my hand more then 6 seconds on their sides...
If you guys think im better off without NFB, I will not insist. I really trust what you guys will tell me.
EL34 are not cheap tube, but I'm willing to change the 6B4G for EL34 if that can help.
I'm willing to change no matter what you think must be change as long as its within the budget, 200$ maximum...
Last edited:
Easy: The "Einstein" amp. I don't see what's so wrong with it.
More ambitious: Tubelab's and Chrish's PP 6L6GC AB2 amp circuit:
http://www.diyaudio.com/forums/attac...c-post-460.pdf
More ambitious: Tubelab's and Chrish's PP 6L6GC AB2 amp circuit:
http://www.diyaudio.com/forums/attac...c-post-460.pdf
The Einstein is a good start. You can use your 6GU7 tubes too, with little modifications. Configure it like this and see how you like it.
and if you don't, 4xEL34 start from 30...40 $
and if you don't, 4xEL34 start from 30...40 $
okay great guys. So the einstein schematic is undoubtly better then my actual amp?
what mod I need to do on the einstein schematic to use the 6GU7? Many sugegstion have been made in this thread. Do I finally just follow the einstein schematic?
What do I have to change in the power supply? I keep the psu as is?
what mod I need to do on the einstein schematic to use the 6GU7? Many sugegstion have been made in this thread. Do I finally just follow the einstein schematic?
What do I have to change in the power supply? I keep the psu as is?
Last edited:
Actually, I'm not 100% sure. Maybe someone else can answer this.
I believe the Einstein circuit would be an efficient use of the chassis and tubes you already have. Since your amps are mono blocks, and they have two 9-pin sockets each, and you already have (I assume good, working) 6GU7's, the Einstein amp easily fits what you have.
No, there will be different resistor and capacitor values, for sure. There are little things that can be improved, too. Tweaks.
______________________________
Even better (IMO), you have the opportunity to learn something here.
!!! -- You should definitely replace the 500R cathode resistor for the 6B4G's with a 750R 25W resistor. Do that, take voltage measurements for the entire amp, and listen to the difference. That will tell you what the 6B4G's 'sound like' when more properly biased, rather than at their currently super-hot bias point. The change in voltages you observe will also give a hint of how close to maximum current your transformer is running. If the voltages change a lot, then the transformer is probably close to its maximum current load. If the voltages don't move much, then the transformer probably has more in it.
The power supply is definitely a limiting factor here. That should certainly be improved. But again, we don't even know what the current rating for the transformer is.
--
I do not see the point in doing all those mods if it may not sound better. Am I missing something here?
You should definitely replace the 500R cathode resistor for the 6B4G's with a 750R 25W resistor
will sure do
after the 750R 25W added, will this give a good idea on the mains and so will let us upgrade the psu?
You should definitely replace the 500R cathode resistor for the 6B4G's with a 750R 25W resistor
will sure do
after the 750R 25W added, will this give a good idea on the mains and so will let us upgrade the psu?
i say tweak your amp as you have it already...
take one small step and listen each time a change was made...
remember, your issues are based on your subjective assessments....
no amount is technical discussions here will suffice if you are not happy with your amp...
you may also try connecting your 8ohm speaker to the 16 ohm tap....
pentode plate resistances just like collector resistances of bjt's are way much too high that it is often ignored in computations.....
for pentodes, voltage gains = gm x Rl
take one small step and listen each time a change was made...
remember, your issues are based on your subjective assessments....
no amount is technical discussions here will suffice if you are not happy with your amp...
you may also try connecting your 8ohm speaker to the 16 ohm tap....
pentode plate resistances just like collector resistances of bjt's are way much too high that it is often ignored in computations.....
for pentodes, voltage gains = gm x Rl
going one step at a time is not a option because I have to pay 10$ for any shipping of even 2 resistors and I use a technician to help me do the modification and he lives quie a bit far, so going over there also cost money and time.
Id rather spent more money, change the tubes if that can allow me to definitely build a better schematic....
since 4xEL34 start from 30...40 $, if going with EL34 allow me to use the mains and opt better, maybe that is a better option? I could sell the 6B4G so changing the tube may not be a bad idea after al...
Id rather spent more money, change the tubes if that can allow me to definitely build a better schematic....
since 4xEL34 start from 30...40 $, if going with EL34 allow me to use the mains and opt better, maybe that is a better option? I could sell the 6B4G so changing the tube may not be a bad idea after al...
Last edited:
Power supply mods. Please examine the attached screenshots.
I modeled the psu with a 5H choke. If you decrease the value to 2H, it still works, but ripple goes up. Still not bad, with ripple to the output stage at 700mV (millivolts) and to the input stage 450uV (microvolts).
If you increase the choke to 10H, things look really good. Output stage ripple 127mV, input stage ripple 90uV.
It's obvious that the better the choke, the better the psu will perform. 7H 150mA would be good. The Hammond 159Q meets those specs and has a DCR of 100R.
--
I modeled the psu with a 5H choke. If you decrease the value to 2H, it still works, but ripple goes up. Still not bad, with ripple to the output stage at 700mV (millivolts) and to the input stage 450uV (microvolts).
If you increase the choke to 10H, things look really good. Output stage ripple 127mV, input stage ripple 90uV.
It's obvious that the better the choke, the better the psu will perform. 7H 150mA would be good. The Hammond 159Q meets those specs and has a DCR of 100R.
--
Attachments
Last edited:
The late great Allen Wright of Vacuum State Electronics had some nice differential circuit designs with no global negative feedback that may work nicely for you. There is a PP-2C schematic for 6550 output tubes and 35W/CH that he he gave out also.
Schematics
Schematics
in that case, go to your friend's house and listen to tube amps they might have, having found one that you like, buy a ready made amp so that you do not have problems....😉
hi, yes I plan on buying the new ifi audio retro 30 tube amp but I have those amp and would like to make them as good as possible with help of you guys. This is for my HT system which is a secondary system but would like to get the best sound with what I have. Now I'm willing to go with el34 tube...
In that case, I agree with AJT. Or, be brave and buy a 50 value 1 Watt set of resistors from ebay, a soldering iron, some solder and learn to solder! This is a DIY forum !
It is much more fun that way than having to drag a technician to do 2 changes. We cant tell the technician to twist the heater wire better for example...
Where do you live by the way?
- Status
- Not open for further replies.