RH UNIVERSAL AMPLIFIER

[Alex Kitic]

Following the success of the first generation RH amps - RH84 and RH807, mostly, I have developed a 2nd generation of RH amps, for KT88/6550 (RH88) and for 300B (RH300B, obviously).

The RH88 is long in the making, I have not done anything in particular with the basic breadboard of the amp since 2005 when I first posted about it on various forums. As far as I know, very few if any diy-er has built it... mostly due to the fact that KT88 tubes are much costlier than EL84, as well as the output transformers, etc.

Since I was interested in making something truly universal, that would allow me to just listen to music and enjoy tube rolling (since I do not have a lot of free time to devote to hobbies and simila), I have developed further the RH88 schematics to the RH Universal model.

The RH Universal lets you use a wide variety of output tubes, from EL34 and KT66, via 6L6 and 5881, to the more powerful KT88, 6550, 8417 tubes. 7027 can be used as well, but you would need a switch or jumper to disconnect pin 1 from pin 8 (necessary for the relatively common EL34).

The driver tube is, of course, foreseen to be an ECC81/12AT7 (i.e. some tube from this family), while E180CC, 5965, and a few others can be used with success as well.

The double rectifier can be chosen between 5R4 and GZ34, any rectifier capable of 200mA DC output will do.

Since there is a considerable difference in anode dissipation between EL34 (25W) and 6550 (35W), the rectifier choice will let the user vary the B+ for a range of approximately 50V. Thus with a 5R4 anode dissipation with any tube will not exceed 25W (safe for EL34 and KT66), while the GZ34 will allow more power with KT88, 6550, 8417 due to the higher B+.

Any tube inserted in the output tube socket will draw approx. 100mA. Approx. because there will be a difference in g2 current draw...

Well, I hope this one will get people going :D

[Geek]

Image link is bad... try this:
http://www.audiokarma.org/forums/attach ... tid=386045

Cheers!

[EWBrown]

Just keep in mind, that the maximum DC input voltage for the LM317 is 37 VDC, some of these power tubes may exceed that value on the cathode.

/ed B in NC

[Alex Kitic]

Image link is bad... try this:
http://www.audiokarma.org/forums/attach ... tid=386045

Cheers!

Thanks, but unless you are logged in to audiokarma you will not be able to see the attachment (I just tried).

Does anyone have any idea as to where to upload the schematics to make it visible to forum users?

[Alex Kitic]

Just keep in mind, that the maximum DC input voltage for the LM317 is 37 VDC, some of these power tubes may exceed that value on the cathode.

/ed B in NC

I tought about that (I try to think about all possibilities in advance), and it's safe in this case. The "bias" with 6550 is around 20v, and that should be the highest value of all tubes chosen (it's around 10-11V with EL34).

I heatsinked the LM317s (actually, LM217s in my amp) but they are dead cold (at 100mA and 20V it's about 2W dissipation... less with EL34).

[dspth]

Does anyone have any idea as to where to upload the schematics to make it visible to forum users?

http://img21.imageshack.us/img21/1715/rhuniversal.jpg



Uploaded with ImageShack.us

[Shannon Parks]

Does anyone have any idea as to where to upload the schematics to make it visible to forum users?

In addition to the 3rd party image host services, any forum member can email an image (or several images) to me for upload.

Shannon

[Alex Kitic]

Thank you all for the assistance!

Well, now that the schematics is in plain sight, does anyone has anything to ask or comment?

[20to20]

Thank you all for the assistance!

Well, now that the schematics is in plain sight, does anyone has anything to ask or comment?

Not familiar with the voltage regulators in the cathode. How do you determine an accurate cathode current using the LM317?

[EWBrown]

For the LM317 used as a current regulator, the formula to determine the resistorvalue for a given current is: R (ohms) = 1250 / I (mA)

In this design, the target current is 100 mA, so 1250/100 would be 12.5 ohms, the closest standard 1% value is 12.4 ohms.

HTH

/ed B

[20to20]

For the LM317 used as a current regulator, the formula to determine the resistorvalue for a given current is: R (ohms) = 1250 / I (mA)

In this design, the target current is 100 mA, so 1250/100 would be 12.5 ohms, the closest standard 1% value is 12.4 ohms.

HTH

/ed B

YID

[roren]

Hi,


Shouldn't there be a HT supply to ECC81 (V2A) ?

Rolf
Edit: I forgot to thank you Alex. I'm glad that you share your ideas.

[Alex Kitic]

Hi,


Shouldn't there be a HT supply to ECC81 (V2A) ?

Rolf
Edit: I forgot to thank you Alex. I'm glad that you share your ideas.

The ECC81 is directly connected to the anode of the output tube. The difference between RH 1st gen and the 2nd gen designs is in the fact that the driver tube is only connected to the anode of the output tube. In that manner, all the current needed for the driver comes from the connection with the anode, and the circuit is, among other things, further simplified: not only does it simplify the audio part, but the power supply as well...

Thanks for recognizing my work: there is a lot of satisfaction in sharing. I am considering eventually to develop kits of the RH amps for those who find it difficult to build and gather all the parts themselves... but that is a far future, eventually.

[roren]

Okey, I see!
That whas a bit of a surprise to a novice like me. Interesting, I must say.
What if I dont like to have a CCS in the output tubes cathode. And only want
to use lets say 6L6. Couldn't I just use a suitable resistor decoupled with a
cap instead. Or do you have an other reason for the CCS other than make the
amp more flexible for different output tubes?

Rolf
P.S. The 6L6 was only an example. If I remember right a steeper tube (higher gm)
would be better, or...?

[Alex Kitic]

Okey, I see!
What if I dont like to have a CCS in the output tubes cathode. And only want
to use lets say 6L6. Couldn't I just use a suitable resistor decoupled with a
cap instead. Or do you have an other reason for the CCS other than make the amp more flexible for different output tubes?

Rolf (and others),
the issue takes us back to RH88 (or RH6550): I was trying to use it with several tubes and always annoyed by the varying current with one set Rk... eventually I made a combo 270+100 where a switch shorted the 100 ohms when one wants to use 8417, using the full 370 ohms for 6550 and 7270.

While it is not "forbidden" to choose one value Rk for the given tube one wants to use, the CCS under the cathode is useful as it sets one important parameter, the tube current draw, regardless of the single tube used... no need for paired tubes, particularly having in mind that this is an SE amp. Why is it important? For starters, the OPT is gapped and calculated, like any choke, for best performance at one particular current draw. The ones I use definitely sound better at 100mA cathode current draw than at 80-something...

Once we set the current regardless of the tube specimen, there is nothing stopping us to use any given tube, 6L6, EL34, KT77... which fits the socket and the pins designation. The only thing one should keep in mind is anode dissipation - and that is addressed with the ability to change rectifier tubes: the 5R4 has the highest voltage drop, therefore gives the lowest B+ in the circuit, making it suitable for use with EL34 and "weaker" 6L6s -- the 5U4 is a step forward, more suitable for 30W anode dissipation tubes -- finally, 5AR4 is the most efficient, therefore suitable for above 30W anode dissipation (but in this power supply it remains below 35W, therefore safe and conservative for KT88, 6550, etc.).

While one might object to a CCS above the anode, the influence of a CCS under the cathode on the sound is much less pronounced, and if there is any, it can be defined as beneficial. If someone wants to call it a hybrid amp, let him...

Some like to avoid any kind of negative feedback whatsoever, others use it with prejudice in different (more or less) innovative ways... why frown at a CCS under the cathode? The only good reason I see is the fact that the chance of a CCS failing is higher than the chance of a ceramic high power resistor failing... the same is true of the zener diode setting the g2 voltage: why not?

P.S. The 6L6 was only an example. If I remember right a steeper tube (higher gm) would be better, or...?

No, that was the story about the driver tube... you need a driver tube with a relatively high transconductance and relatively high mu... while the ECC81 fits the bill ideally, you can use 5965 and E180CC just as well. ECC83 will not drive to full power, 6SL7 even less suitable... etc.

The output tube needs to be a tetrode/pentode, and those mostly have higher gm :D