diyAudio (
-   Tubes / Valves (
-   -   "Mullard 5-20 KT88 PP blocks! (

TubeMack 13th December 2009 04:48 PM

"Mullard 5-20 KT88 PP blocks!
I'm deparately seeking comment from you experts so this project can begin NOW! After alot of research, and reading, and input here. I have decided once and for all on topology, and output power.

I should first state that i'm quite green, but have decent build skills, and some general idea how tube amps work. I am not able to design, and trouble shooting without hand holding would be very limited. I have one ST70 New build under my belt.

I have decided on what Eli had suggested. A modified mullard 5-20 LTP type driver with some type of CCS in the tail. Outputs will be 2 KT88's run Ultralinear, and output goal is about 60W. Builds will be constucted on custom mono block chassis. All parts will be new, and of good to better quality. The build should be as simple as possible while still offering quality equal you better than commercial offerings of the same class. Cost is not a concern within reason.

I'm looking at 3 options to get started, and would like the disscussion limited to those option. I'll just confuse myself otherwise.

Option 1

Was created by forum member Zafir 1983 by modifing an existing schematic with input from all the Gurus on this forum. IT meets all the requirement except complexity. I have no idea how to design a first class PS, or pick the transformers. A fully P2P project could be over my head. However, I know you guy's like the Schematic. Wavebourn said it was on of the best for PP he has ever seen! Another issue is that it has not been put on a scope and tested. Also there are no other proven builds for refference. This one sounds very risky for a newby to me Here it is.
( Anyone willing to help with this one? that would mean a full PS schematic, and recommended Transformer model numbers!)

Option 2

would use a Mark III upgrade board from Diytube, and very loosely follow the rest of the MarkII schematic. . I used this build option with the ST 70, The results were great, and this may be more within my ability. Shannon has created a Modified 5-20 circuit that includes the CCS. The key here is I can use the Comlpete MarkII schematic as a great guide. Repro Trannies are availible, and the power supply could be upgrade, or replaced with a SDS upgrade board. Some here didn't like the driver tube choices (12ua7), but some limited options are availble (12BHt), and even some 6 volt options. This one is very doable for me.
Here's the schematic.

Option 3 would be the same as option 2 except using the Triode Mark III replacement board. This appears to also be a 5-20 type driver? I don't see where a CCS is used on this one but it does seem to have move driver tube options. I can't find a schematic, but here's the info.

Driver Boards for Dynaco Mark II and Mark III

Any comments keeping in mind that my design / electronic skills are limited?

SY 13th December 2009 05:42 PM

On the surface, Option 1 appears the best. I could quibble with some design elements, but it's basically solid. Option 2 has an input stage that is loaded poorly, compromising gain and linearity. It then has a driver stage using a tube with relatively high distortion. Option 3 is unknown, no schematic.

Eli Duttman 13th December 2009 10:28 PM


I agree with SY. Go option 1.

AnTek's model AN-4T550 torroidal power trafo seems suitable for B+ and heater power. Parallel the 2X 550 VAC windings and bridge rectify with 2X snubbed 1000 PIV PN diodes (on the ground side) and 2X 600 PIV Schottkys (on the "hot" side). Pseudo-choke I/P filter the "raw" B+. Not much capacitance will be needed in the "fudge factor" position to bring the rail in at the spec'd 520 V.

We'll work out the 2 negative voltages necessary later.

Edcor's model CXPP100-MS-3.3K will do nicely for the O/P trafo. Super premium "iron" in a Mullard circuit is nice, but it's not essential.

I suspect you don't have an o'scope. Getting CF (phase compensation) right without a scope is next to impossible. Instead, I suggest you use the brute force method for phase compensation. Use 50 KOhm inductive wirewound resistors as part of the 'T7 loads. Wire 1% tolerance 100 Ω resistors in series with the 50 KOhm parts. Connect 1 end of the assemblies to the tube and the other end of the assemblies to the ends of a linear taper 10 KOhm potentiometer. Wire the wiper of the pot. to B+. Either 100 Ω part serves as a test point for the DC balance adjustment the pot. provides. A 0.3 V. drop across either resistor means both 'T7 sections are "idling" at 3 mA. Adding inductance to the load peaks HF gain. All that remains to keep from biting off more than you can chew is to short the NFB loop out at approx. 2 octaves above the audio pass band. Making RF 560 Ω will feed back about 15% of the O/P voltage. 2.7 nF. of capacitance between the "hot" speaker terminal and ground takes care of shorting the GNFB loop out away from the audio band.

pjanda1 14th December 2009 12:02 AM


Originally Posted by TubeMack (
I have no idea how to design a first class PS, or pick the transformers. A fully P2P project could be over my head.

Looks like you've already got help picking the iron, and a good start on the PS. Given that you have some experience already, I don't think you should worry about diving into a P2P project. The entire schematic might look daunting, but work through bit by bit and you'll be fine.

Most tube amp projects consume considerable time and financial resources. Why choose compromised designs when you can build something great? If SY and Eli recommend "option 1", dive in. You'll figure it out and be happier in the end, even if you burn a couple of parts.

Wild Burro Audio Labs - DIY Full Range Speakers

TubeMack 14th December 2009 12:15 AM

Now were talking! I'm going with Option 1 I want the best. I

Eli, I have puchsed a new 30MHZ scope and low Dist. audio generator so I do have that.

Could I request another Ps trannie suggestion as these are going in the main room, and must look nice. I'd love standard verticle mount trannies in black if possible. Somethiong that "LOOKS" simular to the Mark III trannies would be nice. Cost is ok within reason, and if nessasary. I'm still reading but iit appears the B+ voltages were thought to be high, and Zafira1981 has lowered them, but the revised schematic is now missing. I think he finally went with 450V to the OP transformer. Not sure if he dropped any other voltages, as i'm still reading the thread. Here's the thread were additional changed were made. What I want to ask is if the required voltages should be reexamined first?

He also changed the CCS a final time, and mentions going with a 12UA7 for some strange reason. I don't know whats up with that?

Eli, if you or anybody cares to help out with this, i'm ready to gather a BOM, and put in several orders as soon as everything could be determined.

Thanks everyone.

TubeMack 14th December 2009 01:37 AM

BTW, i'm open to the idea of ordering custom Electroprint OP & PS trans. Would these be a step up? I'd just need the specs to put the order in.

Ultima Thule 14th December 2009 03:51 AM

Hi everybody, I'm jumping on this thread too, I have been thinking for some time to build my first tube amp and decided to go with KT88 PP, maybe the reason is they look good but they seem to be popular and the power output is quite on spot too for what I'm looking for.

I noticed that an Edcor transformer was suggested with 3,5k input, I have skimmed through a some KT88 PP schematics and most seem to be 5k transformers, can anyone tell what makes the difference, why some are higher input impedance while other lower? What does it mean in terms of power output and distortion and what else it could mean?

A second thing I noticed on the schematic by zafir1981 is the UL w typical 40% tap, but in his schematic its 43% and yesterday I went through Lundahls homepage and found a suitable transformer (think it was 1679) but it has a 37% tap and 4,5k input impedance instead, what's the difference and why?

Cheers Michael

TubeMack 14th December 2009 04:06 AM

Wow, I had all those same questions!

Please do join me in this build. It will make things easier I believe... I've spend a year researching, and believe this is the topology to go with for a "standard" KT88 PP amp. I also thought some good arguments were made for the driver tube choices. As for the schematic, i'm still unsure of the various stage voltages, (there was some debate over 520v), and if the CCS needs altering (cascode mosfet). I think all this was later changed by Zanfir, but the later schemetic is missing? If the 520v is needed for 60W output, i'd be inclined to stay with it.

SY 14th December 2009 08:50 AM


what's the difference and why?

Really, these differences are pretty negligible. Don't sweat it.

Bandersnatch 14th December 2009 10:21 AM

I'd suggest the Hammond 1650R 5k a-a output TX for starters. Very nice output, and in comparison to *MANY* custom and vintage units it came out ahead...let alone it isn't very expensive...:)

I'd also suggest dropping the power a few Watts to 45-50 and cut B+ to ~400V. It is much better to run the KT88 at 80-85 mA/tube without getting so close to maximum ratings the higher B+ would expose them to.

All times are GMT. The time now is 05:27 AM.

Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 17.65%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Copyright ©1999-2018 diyAudio