Parallel KT88 SE by Kegger

I

ianm0

Member
In another thread: http://www.audiokarma.org/forums/showthread.php?p=6394369&highlight=kegger+parallel+se#post6394369, #260, Kegger posted a schematic for parallel KT88 SE which got me interested since I need a bit more power for my speakers. Kegger later suggested to replace the 1.66V LED + 100R with 1.4-1.5V LED + 160-180R to all those usable tubes.

1. Are there any other updates to the schematic?
2. Should the primary impedance of the OT be twice of that for a single-tube SE?
3. Recommendation for OT?
4. What is a safe B+ to use EL34 in triode?
5. Can 6L6 be used?
 
I'd suggest 2.5 - 4k output tranny. (something like this)
https://www.edcorusa.com/cxse25-8-3_2k

For 2 tubes you actually "halve" the primary impedance.
(but going 3.2 - 3.75k you can still run just 1 tube as well if you want)

425v is fine for EL34 and 6L6GC, may want to adjust cathode resistor for EL34.

Minor update/changes to the schem yur refering to, 6DJ8-ECC85 to use in here.
(I suggest ECC85 as the best option, but 6DJ8-6BK7/6BQ7/6BZ7 work fine to)

attachment.php
 

Attachments

  • KT88-SE_UL_fbk_LED_6DJ8_ECC85_425V.gif
    KT88-SE_UL_fbk_LED_6DJ8_ECC85_425V.gif
    8.3 KB · Views: 740
The resistor and LED should get swapped for either one LED or two. The standard intensity/low impedance red LED's tend to be slightly less than 1.5V so there is some wiggle room there. The lower gain tubes like 6DJ8 can do this comfortably.
cheers,
Douglas
 
PakProtector said:
The resistor and LED should get swapped for either one LED or two. The standard intensity/low impedance red LED's tend to be slightly less than 1.5V so there is some wiggle room there. The lower gain tubes like 6DJ8 can do this comfortably.
cheers,
Douglas
Click to expand...

Actually Doug it's done like that on purpose.
(for some feedback and higher output impedance)

From testing with a distortion analyzer it's the way to go in this circuit.
 
Kegger said:
Actually Doug it's done like that on purpose.
(for some feedback and higher output impedance)

From testing with a distortion analyzer it's the way to go in this circuit.
Click to expand...

Brilliant...should have assumed you'd been that thorough...:)

While you are out gathering data/reading, I must recommend Pete Millets signal amplifier testing. Pentodes, triode rigged same, with some neat data like output swing at n %, and THD at y V output with loading given, sometimes a few different ones. The 12BY7 is a most useful tube per his data...not like you'd be surprised.
cheers,
Douglas
 
Last edited:
Kegger said:
KT88

UL roughly 24 watts.

Triode roughly 16 watts.
Click to expand...

Thanks. Sounds good with a 3.5K OT which allows the option to use only one output tube to maintain true SE. If one tube delivers enough power for the speakers, so much the better (soundwise, I presume).
 
I've seen some schematics that place resistors between the plates and the opt for tubes in parallel,for load sharing. Would this amp need matched tubes because there are no resistors?
 
When paralleling tubes it is quite possible they can oscillate. A small stopper on their g2 is usually sufficient, and a couple hundred ohms on the g1 is not going to adversely effect the frequency response within the OPT's bandwidth.

With the TV sweeps, a stopper on every input is useful when going parallel. In the case of the cathode, the un-bypassed 10R current sensing resistor is usually sufficient. A few tens of Ohms on the plate should suffice, and then go and stopper the g1 and g2 nodes. All of them if there is more than one pin for the same bit( for example 6DQ5 has two pins each for k/g3, g1 and g2 ).
cheers,
Douglas
 
I wish I understood the schematic better, What is the purpose of the 220k 1W resistor from the plate of the KT-88 to the plate of the 6DJ8? it looks like feedback from the plate to the grid of the KT-88 , but I'm not sure.
 
You got it,
That's exactly what is, a bit of local feedback instead of global feedback.
 
As I said that is roughly/approximate output.
It's derived from taking a single KT88 output version, measuring with a scope till clipping, then doubling it.

I have not built this exact amp (using 2 KT88 per channel), but have other parallel single ended amps though.
And pulling an output tube the only real difference I can discern is the output power when pushed/driven hard.

So you'll have to build it, test it, and listen to it for yourself to get any more real insight into it's performance.
 
HI Kegger
I got every thing together to build a pair of these mono blocks.
My power supply will be 380-0-380 5AR4GRAETZ 10uf 10H 40uf 2H 500uf
I am thinking that my B+ will be around 460v using the 5AR4 . I bought some extra resistors 2.7K, 3.7K, 4.7K and 5.7K for the
RC stage to the input tube. I am guessing that the 4.7K will be the right one to keep the voltage at 120v. I am hopping to be able to use the GZ37 to reduce the B+ and be able to also run EL34.

Do you have any thoughts about a feedback switch with like 180K, 220K and 270K ?

Thanks Brian
 
brians7 said:
HI Kegger
I got every thing together to build a pair of these mono blocks.
My power supply will be 380-0-380 5AR4GRAETZ 10uf 10H 40uf 2H 500uf
I am thinking that my B+ will be around 460v using the 5AR4 . I bought some extra resistors 2.7K, 3.7K, 4.7K and 5.7K for the
RC stage to the input tube. I am guessing that the 4.7K will be the right one to keep the voltage at 120v. I am hopping to be able to use the GZ37 to reduce the B+ and be able to also run EL34.

Do you have any thoughts about a feedback switch with like 180K, 220K and 270K ?

Thanks Brian
Click to expand...
I built a similar SE amp called the RH Universal which has local feed back or plate to plate feedback from the power tube to the driver tube. In my amp i used the 6sn7 driver tube and 100k fb resistor. Then i made a switch for adding in series another 100k resistor which would be connected so i could use the 6sl7 as a driver tube. Switching the fb resistors allowed better operating conditions for the different driver tubes. I accidently put the wrong driver tubes in with the incorrect fb resistors and that didn't seem to cause any issues. The sound was slightly different but not very much. So, if you want to experiment with different levels of fb a switch would not be a problem. I would caution you about switching the resistor while the amp is on as it may make a loud thump when it switches over.
 
Kegger said:
I'd suggest 2.5 - 4k output tranny. (something like this)
https://www.edcorusa.com/cxse25-8-3_2k

For 2 tubes you actually "halve" the primary impedance.
(but going 3.2 - 3.75k you can still run just 1 tube as well if you want)

425v is fine for EL34 and 6L6GC, may want to adjust cathode resistor for EL34.

Minor update/changes to the schem yur refering to, 6DJ8-ECC85 to use in here.
(I suggest ECC85 as the best option, but 6DJ8-6BK7/6BQ7/6BZ7 work fine to)

attachment.php
Click to expand...
Hi Kegger, i have not read of your assessment of the sound quality of this amp. Do you have a write up that i can check? Also did you find any benefits to have 2 x kt88 vs just a single kt88 , other than the increased power? Was there less information in the reproduced sound as some people report about paralleled output tubes?
 
You must log in or register to reply here.

Similar threads

eedork
Kegger SE UL build questions
Replies
1
Views
267
Dave451
Dave451
kward
KT120 Amp build
25 26 27
Replies
534
Views
55K
kward
kward
Back
Top Bottom