• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Three PP KT88 designs. Which one ... (please;-)

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi good folks!

I need help to choose!

I have a question that have been posed a lot of times in here!
The difference is that this question isn't a general one, it's more specific than "the usual" ...
So please show pity on a "new-bee" - and excuse yet another question on "which to choose". I know that I probably would have been happy with any of these. At least I have narrowed my choice down to three designs.

First a short description of "my needs":
I need a good PP amp of approx 30 to 50W power for an active bi-amped loudspeaker.
I allready have a Lectron JH50 that I plan on using for the low end - up to 1000Hz. To make a long story (very) short: This is one of the best PP amps in the bass and mids that I know of. To match it, and possibly better it in the mids/highs, I have found that a very sound PP KT88 may just be right ...

I have looked around, and found these fine designs:

KaT88 from World Audio Designs
- But does World Audio Design still exist?
- Where to find more info.?

Ella Signature with KT88 from DIY Hi Fi, used as a pure poweramp.
This amp. with these modifications?

Or Mono Bills, designed by "Triode Dick"
Here I can get a transformer/choke-kit from AE at a good price also ...

Please give me feedback! All is welcome! I really need your opinion and experience to find wich one of these that will give me the best result!

Looking very much forward to all thoughts, advice and general comments you may have!

All the Best for the New Year to all of you!

Aril.
 
Looks like the KT88's are being run in pentode mode. More power but more output impedance and more NFB needed to run it stably. Not that NFB is bad...

Can't find a schematic!

Personally it all depends on your taste, the type of music you like, your source, your speakers, and the way all these things interact, before you start thinking about mods. Try to "demo" different ones if you can, and if you can't, be prepared to take a hit to your wallet distributing on A-gon... sorry I can't be more specific.
 
No schematic for the Ella, so hard to say.

The WAD is based on a classic Mullard 5-20 circuit (probably less linear), but their choice of an ECC82 for a long-tail phase splitter is particularly unfortunate.

The Bill looks somewhat better, but will need careful choice of speakers- there's no feedback, the source impedance will be pretty high, and the frequency response with most speaker loads will look like a roller coaster.

Just MHO.
 
The WAD is based on a classic Mullard 5-20 circuit (probably less linear), but their choice of an ECC82 for a long-tail phase splitter is particularly unfortunate.


SY,

IMO, the Mullard circuit is not without virtue. If nothing else, pesky phase shifts are reasonably well controlled. Do you think a 6AB4 as the voltage gain tube and an ECC99 as the LTP have merit driving PP KT88s?
 
I hope I didn't imply anything different- the 5-20 is a classic for a reason.

I haven't used the ECC99, but it sure looks good on paper. For an input tube, I might consider the 6AB4 (it's half a 12AT7 if memory serves). A 6AQ4 might be close to optimum; the higher mu will result in higher feedback, and the linearity is excellent. Trioded D3a would be overkill, but if cost were no object...
 
Some ELLA technical information:
The circuit, at least for the older version (=Conconance M100), is also based on the Mullard 5-20, but uses a triode SRPP input stage. This is DC coupled to LTP phase splitter with resistor tail, and EL34/KT88 output in UL at 430V.

The Ella signature with remote volume control may be somewhat wasted if it is to be run as a pure power amp?

Svein.
 
Thank you for all your input on this sorenj07, SY, Eli Duttman and Svein B.!

There exist a schematic for the Ella on the net, but it is of an older version. The difference to the one offered today I believe is "just" a dual-mono psu. Here

I thougt I had chosen wisely, when asking for your advice of one of these three designs ... I guess I have to add at least one other construction that originally was "skipped" because of the 20W output. I thought I would need more, but this is not so, since I'm going to use it from 1000Hz and up. The loudspeaker-drivers are configured in MTM with ribbon mids and air motion transducer tweeter. These will give a stable impedance of approx. 7 to 8 ohms from 1KHz to approx. 3KHz and falling to 4 ohms from this point and staying there further up ... Sensitivity is 92dB/1m. and the energy needed to drive mids/higs are less than what will be needed below 1000Hz (where I have 50W).
Since I see that phase-splitting, sufficient drive for the output and the need for NFB with pentode-mode KT88's is issues with the three designs, I thought this design might be better:
Andrea Cuiffoli's PP KT88

And, guys, - thank you very much for your input and sharing of thoughts and knowhow! It is very much appreciated!

Aril
 
hi,
I've had an ella amp for a number of years and did a lot of modifications to it for my own enjoyment. Its a good platform to work with.

I'd agree with using a SRPP at the input... not really necessary. I think it was just fashionable to use that topology wherever you could put it a few years back. Still, it peforms reasonably in that role. I have changed to an active loaded 6fq7 for the input as I don't need much gain. I also modded the 6sn7 LTP to swing 200Vpp for driving DHT outputs, which I use.

not sure what the considerations for your amp choice are now. If its fully DIY, there are many design choices. The Cuiffoli design looks good if you get the lundahl IT. Another option is using a center-tapped choke for splitting. The ella is easily modified for triode, no-feedback operation if thats of interest.

-z
 
SY said:
I hope I didn't imply anything different- the 5-20 is a classic for a reason.

I haven't used the ECC99, but it sure looks good on paper. For an input tube, I might consider the 6AB4 (it's half a 12AT7 if memory serves). A 6AQ4 might be close to optimum; the higher mu will result in higher feedback, and the linearity is excellent. Trioded D3a would be overkill, but if cost were no object...


The 6AQ4/EC91 sure looks interesting. The 6AB4 is 1/2 a 'T7. It seems RP for the 'AB4 is a tad lower than that of the 'AQ4. How that works out in the total gain structure, I will not guess.

My penchant for using the 'T7 in PP amps is well known. The 12AT7's distortion spectrum mates WELL with the internal cancellation of even order HD products in PP "finals". The net HD spectrum is an ear pleasing "waterfall", with 2nd > 3rd > 4th. As a consequence, NFB level need be large enough only to ensure an adequate damping factor.
 
Thank you for all your suggestions and comments :)

Through this, I have found that I might not go for any of my previously mentioned three designs, but rather for either the Andrea Cuiffoli or the zafira1981-modified Claus Byrith design as reenberg suggested. The Cuiffoli giving more than sufficient power (and getting hold of the correct IT from Lundahl is no problem). The Modified Claus Byrinth has much more than sufficient power output for my need. In the descriptions given in the string reeberg linked to, I lack technical info on OPT used.

Does anyone have any comments on these two designs?

Looking very much forward to all input.
I'm very grateful to be able to have such knowledgeable diy-friends helping me out with this process!

Best Regards, Aril
 
As an owner of a heavily modified Bill, i can say it is a really awesome amp, that alows for vigorous mods.

Most of my mods were implemented in Dick's later designs, allthough we had no contact on that.

The service and quality of AE is better than outstanding. I emailed them I had a slightly buzzing input transformer I bought second hand, and they promptly made me a brand new one!

I ended up using even other output tubes (gu50), but that is a matter of taste and wallet.

The two thing that made this amp bloom was 6n6p instead of ecc99 and the use of an exotic input tube. The ecc99 is brilliant on paper, but makes for a cold sound used here.

Keeping the three stage topology as in the original bill gives you ample gain to experiment with feedback and triode strapping.
 
Eli, wait until you see the 6AQ4 spectrum. Wow. Just... wow.
 

Attachments

  • 6aq4 spectrum 80vpp sample 2 ir red.jpg
    6aq4 spectrum 80vpp sample 2 ir red.jpg
    43.9 KB · Views: 673
- So Eli D. and SY, what you are talking about is modifying the WAD KaT88, that actually is a version of the Mullard 5-20 circuit, by using the 6AQ4 or 6AB4 for the input and a ECC99 as phase splitter ...? This means a total re-design of the KaT88 circuit ...
As of today, I am really not yet competent enough to pull that off in any successful manner. Do you guys mean this to be a better design than the zafira1981-modified Claus Byrith?
The Ella is allready a modified 5-20 with a driver stage that really swings. Compared to the modified version of the KaT88, wouldn't this be easier for me as a relative "rookie" to start out with? - Would modifying the 5-20 in the way you suggest be very difficult, and would it be that much better than the Mono Bill's? Many people seem to have acheived good results with this design, perhaps somewhat modified, like beamnet describes.

I introduced the Andrea Cuiffoli PP KT88, since it is a quite different design, using IT for phase-splitting and KT88's in PP triode-mode together with a SE 5842 input-stage. This should give good swing and drive for the KT88's and a better phase-splitting? (ref. Lynn Olson and Sakuma). Cuiffoli also suggest 1/2 7119 for the input/driver stage. What about a russian 6H30 instead (being equivalent to 7119, only with different socket/pin-out).

Today's New Years eve ... Happy New Year to all!

Cheers, Aril!
 
SY said:
Eli, wait until you see the 6AQ4 spectrum. Wow. Just... wow.
I'm impressed.
WE91 said:
I introduced the Andrea Cuiffoli PP KT88, since it is a quite different design, using IT for phase-splitting and KT88's in PP triode-mode together with a SE 5842 input-stage. This should give good swing and drive for the KT88's and a better phase-splitting? (ref. Lynn Olson and Sakuma). Cuiffoli also suggest 1/2 7119 for the input/driver stage. What about a russian 6H30 instead (being equivalent to 7119, only with different socket/pin-out).
I've built lots of variations around this and the 6H30 will work if you can deal with the lower gain. A 6C45 is great here.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.