• 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.

#26 pre amp

Hi All,

I am in the process of completing Phaeton RC 26 preamp published on http://www.vt52.com/ site (Jim de Kort) as my latest DIY project. But still I am little puzzled with few things with my current project. Hope someone could provide an expert advice to me to figure out those.

One thing I am confused with is the hum coming out from B+. I figure out it is something to do with the grounding however. Lately I found out that if I ground the filament negative with B+ negative with 470uf capacitor, the hum is almost disappearing. But not sure whether this method is right or safe because what I did was sheer coincidental . Need advice on this. I used separate power rail using a SMPS for filament and both filaments were connected serially with 3V supply. Only connection for B+ negative from this rail is 470uf negative to negative (capacitor + to filament's negative and capacitor - to B+ negative).

Second thing I am puzzle with is the bias supply. Should I use separate 12 V supply for bias for 26 tube or should I use voltage divider to get the bias supply? Where exactly should I connect the bias voltage in this schematic?

Greatly appreciate if someone could clarify these for me.

An externally hosted image should be here but it was not working when we last tested it.




Added as requested by djn -

Jim de Kort's site
Kevin Kennedy's PF article
Keven Kennedy's Schematic PDF
Where to begin? At the beginning of this thread
Coolzero 1
revintage 1
revintage 2
mach1 sneaks in a 12B4A schematic
Richard's double-fisted 156C choke load
coolzero's hum avoidance diagram
coolzero 2
revintage 3
andyjevans's 10Y preamp
regal's 26 driving 71A
Rod Coleman's DHT linestage
Rod Coleman's DHT linestage with power supply
andyjevans 26 with AZ1 rectifier
coolzero 4
mogliaa's 26 testbed
Original Burnedfingers
euro21
mogliaa's 01A
mogliaa's 26 source follower
dhtrob's 26
andyjevans still can't find the rotate 90° command
Rod Coleman's 26 with a filament-on-filament bias with IDHT output
euro21
mogliaa version 2
Nightpuma's 26
dhtrob 2 (now with grilled pencils)
H.Michi
andyjevans 26 with VR tubes.
andyjevans 26 with VR tubes and CCS
 
Last edited:
One thing I am confused with is the hum coming out from B+
I think the hum is more likely to come from your filament supply and or grounding issues.

The 26 is notorious for humming if the filament supply is not up to scratch.

The cap from negative filament to ground is safe.

What bias voltage do you mean? The design you posted is auto-bias. Therefore the bias you refer to is probably the voltage across the cathode resistor.
 
The hum is ripple in your filament supply. What you did "by mistake" was twofold. You "star grounded" and added a "C" filter.

You can actually incorporate this into your filament supply circuit.
Maybe increase the value of that capacitor also. 1,000uF??

Make sure the B+ capacitors and chokes are of good quality, but I doubt the hum is coming from there.

By the looks of it the "bias" of that tube is actually more around -15V (~6.2ma thru a 2K2 cathode resistor.

It appears he took the operating point for the 26 straight out of the RCA manual. Vp 180 Grid Bias -14.5V Current 6.2mA

What you might consider is placing the 2K2 Cathode resistor and the 470uf on the center tap of a 50 Ohm 1W potentiometer. The two outer tabs of the potentiometer then go to the filament pins.

This is a "hum balance" pot.

Check your voltages proper operation should be at idle (short the input)

~194 - 195V at plate connection to ground
and pretty close to -14.5V grid to ground.

He is running the 26 kind of "hot" the other designs I have seen show it running more at Vp135 current 5.5mA
Find a copy of RCA RC-10 and look at specs.

Good luck, I am working on a 27 preamp right now.
 
Thanks BH. I am suspecting grounding because as soon as I put the cap, hum is disappearing almost to zero. I ordered two of IXCY CCS from Kevin@K&K audio (http://www.kandkaudio.com/accessories.html) to replace the 33K anode resistor which I change to increase the gain a little bit. I tried simple CCS using a IRFBC840 and it gave good result. 150V B+ gave the best sound out of this schematic. Now I can enjoy it without hum since you help me to clear my doubt. Thanks again.

I think the hum is more likely to come from your filament supply and or grounding issues.

The 26 is notorious for humming if the filament supply is not up to scratch.

The cap from negative filament to ground is safe.

What bias voltage do you mean? The design you posted is auto-bias. Therefore the bias you refer to is probably the voltage across the cathode resistor.
 
Thanks coldcathose for all the suggestions. Sorry for my limited knowledge of electronics first of all. I haven't had much of an idea of what I was doing by connecting cap from filament negative to B+ negative, that is the reason I told it was by mistake. BTW, it get rid of all my hum issues and now it is zero hum :). Also I used simple CCS and 10000uf filter cap after the SMPS output to feed the filament which I assume sufficient to get rid of the filament hum issue. I use JJ caps and Hammond chokes (14H-->4H) for my B+.

BTW, I like the idea you gave of this "Hum pot" which I will try soon.

About the OP. points. So far what I found was 150V (with 33K anode resister) is best sounding. Smooth treble, tight and tuneful base and I am in heaven:). I had opportunity to listen to many of 1st class Audiophile equipments including audio gears like LAMM mono blocks with Wilson speakers connected to; but this little project is a serious contender from the sound and musicality standpoint for those, even it is running with my NAD C 372 SS amp and DALI Evidence 470 speakers (at least for my ears, this preamp change the sound of my audio gear drastically. Tone, image, PRAT whatever I could think of). Really glad that I did this project and best I did so far.

Good luck for with your 27's. Isn't this tube a "IDHT" btw? I tried few IDHT pre-amps schematics as well but DHT's are winners hands down if compares.


The hum is ripple in your filament supply. What you did "by mistake" was twofold. You "star grounded" and added a "C" filter.

You can actually incorporate this into your filament supply circuit.
Maybe increase the value of that capacitor also. 1,000uF??

Make sure the B+ capacitors and chokes are of good quality, but I doubt the hum is coming from there.

By the looks of it the "bias" of that tube is actually more around -15V (~6.2ma thru a 2K2 cathode resistor.

It appears he took the operating point for the 26 straight out of the RCA manual. Vp 180 Grid Bias -14.5V Current 6.2mA

What you might consider is placing the 2K2 Cathode resistor and the 470uf on the center tap of a 50 Ohm 1W potentiometer. The two outer tabs of the potentiometer then go to the filament pins.

This is a "hum balance" pot.

Check your voltages proper operation should be at idle (short the input)

~194 - 195V at plate connection to ground
and pretty close to -14.5V grid to ground.

He is running the 26 kind of "hot" the other designs I have seen show it running more at Vp135 current 5.5mA
Find a copy of RCA RC-10 and look at specs.

Good luck, I am working on a 27 preamp right now.
 
Result

Finally I have completed the pre-amp. Attached the schematic (sort of Kevin Kennedy, Jim D'Cort mixture). Thankful to Kevin for his great advices to come up with this final circuit which work nicely and worth all the efforts. Used a PSU with Hammond 372 X (which I had already otherwise an overkill I suppose) +5U4GB+10K(5W)+OA2/OB3 tube regulators+IXCY CCS (kit module K&K Audio -http://www.kandkaudio.com/accessories.html) to produce 250V hum free B+ which I built out of randomly picked up items I had already. Otherwise, any simple PS schematic would work I assume. For me, use of regulated B+ helped towards reducing the hum drastically. Hum with unregulated B+ was unacceptable to me (even though it sound great). For filament, I used a cheap China made external DVD SMPS which worked greatly with no hum at all. :)

Attached a picture of the final product (pardon for my poor craftsmanship:no:).
 

Attachments

  • 26 mypre.jpg
    26 mypre.jpg
    19.6 KB · Views: 29,345
  • PIC_00008.jpg
    PIC_00008.jpg
    275.3 KB · Views: 27,574
Last edited:
Looking at the 220nF coupling cap at the output; isn’t that generally too small?
In normal (not DHT) linestages this is usually 2-4uF. Are there special considerations with DHTs?

A very nice build btw., coolzero.:cool:

I am confident that a well constructed 26 linestage is one of the best linestages money can buy... regardless of price.
I have not jumped on the DHT linestage wagon myself yet, but I have a lot of 26 tubes.
Think I will try a 27 linestage first. I have collected a small truckload of Arcturus Blue 27 (tipped).:cool:
 
Hi there! A 26 preamp is indeed the highest of high end. Hard to better!

I have one built which sounds pretty good - I use a glow tube to get 150v B+ and then a plate choke - the cheap Hammond 156C (150H). A resistor on the cathode right now, but will be replaced by something better.

However, a better solution is a cascode IXYS active plate load. I believe you have one already. Then I would try some red LEDs as cathode bias. Each is 1.7v, so say 130v plate, -8.5 (5 LEDs) for 6mA. I don't know what that would be as a sense resistor on the IXYS cascode - must try it and see.

For filaments I use a 22,000uf cap, a LM1084 chip as current source with a 1.25 sense resistor (min. 6w but better 12w). That's the "end bit" of a Ronan Reg. You can add a voltage reg before it but you may not need it. The current source is the important bit - better sound. You'll need a heat sink about 3" by 3" for the LM1084 - gets hot depending on the voltage going in.

andy
 
Last edited:
Hi there! A 26 preamp is indeed the highest of high end. Hard to better!

I have one built which sounds pretty good - I use a glow tube to get 150v B+ and then a plate choke - the cheap Hammond 156C (150H). A resistor on the cathode right now, but will be replaced by something better.

However, a better solution is a cascode IXYS active plate load. I believe you have one already. Then I would try some red LEDs as cathode bias. Each is 1.7v, so say 130v plate, -8.5 (5 LEDs) for 6mA. I don't know what that would be as a sense resistor on the IXYS cascode - must try it and see.

For filaments I use a 22,000uf cap, a LM1084 chip as current source with a 1.25 sense resistor (min. 6w but better 12w). That's the "end bit" of a Ronan Reg. You can add a voltage reg before it but you may not need it. The current source is the important bit - better sound. You'll need a heat sink about 3" by 3" for the LM1084 - gets hot depending on the voltage going in.

andy

Andy,
What value is your coupling cap?
 
Administrator
Joined 2004
Paid Member
Looking at the 220nF coupling cap at the output; isn’t that generally too small?
In normal (not DHT) linestages this is usually 2-4uF. Are there special considerations with DHTs?

A very nice build btw., coolzero.:cool:

I am confident that a well constructed 26 linestage is one of the best linestages money can buy... regardless of price.
I have not jumped on the DHT linestage wagon myself yet, but I have a lot of 26 tubes.
Think I will try a 27 linestage first. I have collected a small truckload of Arcturus Blue 27 (tipped).:cool:

While the 27 is cool the 26 seems to be much more reliable and somewhat less microphonic if you can believe that! (Frankly they are both horribly microphonic) I would go straight to the 26 although the 27 is a lot of fun. Sadly almost all of the 27 line stages I have encountered are now "retired" - mouldering in closets somewhere because as good as they sound the tubes are just unreliable. My prototype 26 dht pre just keeps soldiering on year after year with minor mods here and there. I'm only on my second pair of 26's in 8 years of relatively steady use.

On coupling caps choose the smallest that places the -3dB corner well below the LF cutoff of your amplifier/speaker combo.. Generally I think 5Hz is low enough in most instances.. A 1uF output capacitor into 100K has a -3dB corner in the vicinity of 2Hz which IMO is more than low enough, perhaps a bit too low if you are playing vinyl.

You should also consider plate to line transformers, Electra-Print can make these and old UTCs work well too. Refer to my old 26 construction project on Positive Feedback Online for more details if interested in this approach. Providing that the transformers are well shielded from magnetic fields you get lower noise, and a much lower output impedance at the expense of gain.


Those Arcturus 27s are cool looking aren't they.. :D

Edit: WRT to the 27 I recommend using fixed bias with the cathode grounded as this reduces the likelihood that cathode insulation leakage will result in excessive hum particularly if AC is employed for heating. Unlike later types it is not a good idea to elevate the filament supplies more than a few volts above ground as the cathode to filament insulation may fail. Unfortunately this seems to be one of the most common failure mechanisms in these tubes. (Well it was the first idht - so what do you expect? :p )
 
Last edited:
OK guys, in the interests of science I just did a blind test of a Holco cathode resistor bypassed with 47uF polypropylene and 4 x red LEDs. Changed one side but didn't know which until I checked after, then played a mono CD, listening to one side then the other.

The LEDs were much clearer, though ?? maybe a little more etched. The Holco combination was smoother, but a bit more muffled and not as exciting. So on the basis of this, I'm changing both channels to LEDs. I used four, for a bias voltage of 6.8v and I got 130v on the plate, corresponding to a current of 7.5mA. Should be just OK though if I use 5 LEDs the current should come down to 6mA which would be more comfortable. OK, so soldering iron out! Will report!

This is with a 156C Hammond plate choke coming off a 150v glow tube.

andy
 
Looks like that last 150uF cap in the power supply is in parallel with the gas regulators. They don't really like that, maybe put a 100 ohm resistor in series with them; added bonus that you can measure current through them.

Zigzag,
Thanks I forgot about them not liking a cap at the input, I think I will put a 5 watt rheostat there then I can adjust the current taken up by the VR tubes to keep them happy.

I picked up an old Motorola radio chassis the other day in exchange for some work on a guys old Philco. Has LOTS of 5 pin sockets I think I might just bust out the soldering Iron and try this thing.
 
Looking at the 220nF coupling cap at the output; isn’t that generally too small?
In normal (not DHT) linestages this is usually 2-4uF. Are there special considerations with DHTs?

A very nice build btw., coolzero.:cool:

I am confident that a well constructed 26 linestage is one of the best linestages money can buy... regardless of price.
I have not jumped on the DHT linestage wagon myself yet, but I have a lot of 26 tubes.
Think I will try a 27 linestage first. I have collected a small truckload of Arcturus Blue 27 (tipped).:cool:

Cap value was taken from the Jim D'Cort orginal design without a change. He had mentioned something in his site with regards to the maximum value of coupling cap. Pretty sure it was not that high as you mentioned. Didn't give much of an attention for that since I was hurried to listen to it :)
 
Hi there! A 26 preamp is indeed the highest of high end. Hard to better!

I have one built which sounds pretty good - I use a glow tube to get 150v B+ and then a plate choke - the cheap Hammond 156C (150H). A resistor on the cathode right now, but will be replaced by something better.

However, a better solution is a cascode IXYS active plate load. I believe you have one already. Then I would try some red LEDs as cathode bias. Each is 1.7v, so say 130v plate, -8.5 (5 LEDs) for 6mA. I don't know what that would be as a sense resistor on the IXYS cascode - must try it and see.

For filaments I use a 22,000uf cap, a LM1084 chip as current source with a 1.25 sense resistor (min. 6w but better 12w). That's the "end bit" of a Ronan Reg. You can add a voltage reg before it but you may not need it. The current source is the important bit - better sound. You'll need a heat sink about 3" by 3" for the LM1084 - gets hot depending on the voltage going in.

andy

Hi Andy,

How does the plate choke (H156C) sounds compared to tha normal plate resistor in place? Kevin recommended me to use MagneQuest but it is way above my budget. I have seen few articles about using 156C as plate chokes with good results.
I treid CCS as active plate loads but sound become uneven with unmatched tubes and didn't know how to fix it. So went back with plate resistors. I tried 3 stage filament regulator with LM317 (somethign similer to Ronan reg.) but the hum was too high. Best I had with the SMPS which I a musing now with no audible hum at all.
Alkiline battery bias worked for me very well and thinking of repalcing the batteries with small 9V PSU later.