• 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.

Help with these Russian GU50 tube amplifier

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Someone plotted out the curves for the GU-50 with different loads.

10K gave very linear performance at the cost of power,

5K gave a balance of distortion and power,

3.5K gave the most power but at expense of high distortion.

Guess I chose the middle road. :)

Besides: The suggested value for the OPT was from the guru of GU-50 amps... Wavebourn. He has vast experience with these tubes and can design any circuit with the GU-50 that will run circles around any of the other commercial offerings using other NOS tubes.
 
Last edited:
Hello quikie22

How much power you get now with your set up.
I ask because I want to buy the OP transformers. I think to go with 2 tubes I want to buy 3.2K that transformer also I can use for one tube.
To me 7-8W it would be plenty of power in triode mode (of course if the distortion not to high) for example with classical music. You know that sometime very powerful sometimes you hardly hear anything
I need at lease 7-8W with out high distortion

Greetings gabor
 
Translating GU50 specs

One of the hardest things I find with using the GU50 and working out circuits is my lack of ability to read russian :(.

I've attached a couple of pages of GU50 specs - could some kind soul please translate the words under the curves traces, and tell me how the tubes are connected up in each instance?

This would make planning bias circuits etc vastly easier and would help I think a lot of people take advantage of these awesome tubes!

Thanks in advance!
 

Attachments

  • GU50-datasheet-150.jpg
    GU50-datasheet-150.jpg
    86.3 KB · Views: 687
  • GU50-datasheet-152.jpg
    GU50-datasheet-152.jpg
    92.6 KB · Views: 900
Thanks for the info and the links.

My russian is rusty however, and google translates the first graph text into:

Removed the CVC PG-50 for different stresses safety net. It turns out interesting picture.
When 0V:

From the russian I can see it's the RY50/GU50, but is the CVC/BAX the G3 pin?
So that would make the graphs for cathode at 0V, G2 at 250V and G3 = 0V at the top (conventional), then the 2nd pic is G3 = 0V, then G3 = 10V etc.
I'm assuming that Ie (backwards e) is the G2 screen current (Ig2)? And Ee or Ug3 are the voltage at grid 3 (Vg3)?

I read a page recently where they connected a pentode's G3 to the cathode only via a film capacitor and said it sounded much smoother and softer - have you come across that trick? (DECWARE - The Hazen Mod)

Interesting to see if it works.. I guess that G3 will float to a voltage it feels happy at and for all AC signals will stay at they voltage...
 
No ВАХ means Volt-Ampere-Characteristic. :)

Yes, graphs show pentode mode for different G3 values. The last graph is for triode mode when 2 top grids are connected together to the anode. But he did not have possibility to test on positive grid voltages for complete picture, unfortunately. A1 mode for this tube is waste of power, in triode it should work in A2.

Also, if to connect all 3 grids together you may get a nice right-handed triode (similar to high voltage BJT transistor).

Yes, it would be nice also to experiment with positive voltage on G3, while G2 is connected to anode. A2 may not be needed at all.
 
Last edited:
This is the latest sketch of my amp design, it's all rather subject to change and tuning but seems to be the most interesting I can come up with currently:

The front is an 6AR8 tube, which then splits into both sides of the PP amp, of which one side is shown here.

There is a neat feedback point at the 6AR8 beam deflector 2 (BD2) but as the 6AR8 is common for both push and pull that could only be utilised from a transformer winding, which I do not want to have inside the feedback loop as I consider them too slow and phasey.

So here the idea is that a combination of DC Schade feedback and DC feedback to the SRPP(ish) stage create a stage where all feedback avoids capacitors. the SRPP(ish) stage may use 6n6p or 6n2p tubes, the 6n6p allow more grid voltage (less gain though) and therefore more effective cathode feedback as the cathode resistor is bigger.

The transformer shown is half of the OPT primary - i.e. the 'push' of a push-pull. Also shown are a couple of B+ correction feedback paths. The issue with these common cathode output stages is that when the GU50 draws current the rail will sag, which may result in non-linear compression. Therefore the power rails are fed into the amp at just the right levels to correct the output.

The 500V B+ corrector feedback works like this: falling B+ reduces grid voltage of the driver, which allows the driver anode to rise switching on the GU50 a bit more, causing the correct voltage over the primary to be maintained.

The 400V driver stage corrector feedback works by allowing a dip in that rail to cause the driver grid to fall, making the anode rise to compensate. These can simply be tuned for hum nulling for optimal value.

The 6AR8 also has a deflector tuning pot - like vertical position on a CRT - to null out any static magnetic fields - it is adjusted until A1 - A2 measures 0V on a multimeter.


Any thoughts? (I know about the G1 gu50 bias error on the diagram!)
 

Attachments

  • GU50-halfPP.jpg
    GU50-halfPP.jpg
    57.9 KB · Views: 647
Last edited:
Hello
May I ask why do you want a new circuit?
What I mean why U do not take a proven good working schematic and build that.
I would do that, all do it depend on your experience.
I like these schematic a lot, specially with the two tube.
All do having problem to find the right power transformer. I think I need at least 400mA rating plate voltage for the two tube each side.
I do search many month for a power transformer on EPay and nothing came up to get around 420V.
It is annoying to drop half power (go with one tube each side) just because I do not get the right transformer.
To order something from Plitron it is horrible expensive. Pay more for power transformer than the OP Transformer I think waste of money.
I will order a pair Edcor 25W output transformer after I get a power transformer.
Also It would be stupid to use 1 power tube to get 7-8W and I'll buy 25W OP trannies.
Îäíîòàêòíûé óñèëèòåëü íà ÃÓ-50
I'd like these amp a lot. I have most of the parts, I just need the transformers.

Greetings Gabor
 
Here it is - a proper viewable version :)
I've been meaning to invert the colours for ages!

Hi Gabor,

I want to create my own circuit because it's more DIY that way - I don't want to just assemble stuff, I want to be listening to something I have designed too!
Plus just copying stuff is boring


For this one I'm just re-using some mains and OP transformers + chassis from a cheap KT88 amplifier (an MA-800MkIII shipped without tubes), so as far as the iron is concerned it's just an upgrade from a cheap type of tube to some decent ones ;).

Then there is the chance to stamp out some global feedback, which has no place in my world. If I had my way I'd wind a decent interstage step-up transformer and use cathode followers on the output, hence eliminating coupling capacitors AND lowering output impedance in one simple stroke.

Without those transformers I have to be use a more conventional approach and use a few tricks to get what I'm after. Oh - and to see if any of my ideas work with a scope, sig-gen and variac ;)
 

Attachments

  • PP-GU50-WB.png
    PP-GU50-WB.png
    121.8 KB · Views: 657
Just some calcs added so I can remember them:

The 'SRPP' will be unbalanced by the DC feedback so the resistors need to be chosen to something that works first of all. Assuming 6n2p:

Assuming we get 1mA from each feedback resistor, we have 1mA from the top triode, 1mA from the direct GU50 anode FB, and 1mA from the Schade FB.
So allowing the bottom tube to sink 2mA @ 200V needs a grid of 1.15V, which sets the bottom Rk = 1.15 / 3e-3 = 380ohms.

So the top triode's Rk needs to provide 1.6V for 1mA in the top tube. Actually this SRPP doesn't need to be balanced as we are only driving +ve going pulses as this is half of a push-pull but we'll balance it for now: Rk = 1.6/1e-3 = 1.6k.

Then the direct anode resistor (dropping 500V at 1mA) = 500k, and the indirect Schade one dropping 300V needs to be 300k.

This gives something that should at least balance the 'SRPP'(ish) bit in the middle, whether it gives enough feedback is debatable though ;)
 

Attachments

  • PP-GU50-WB2.png
    PP-GU50-WB2.png
    111.3 KB · Views: 676
Hello Globulator

Thanks for the answer!
OK if you have the knowledge to design your own in that case go for it.
May be you can give me some advice to.
I plan to use 2 GU50 paralleled, each biased around 90mA. So for the two chanel I need at least 400mA transformer to the plate.
That is correct? Last few month I didn't found power transformer to get 420DCV and min 400mA on Ebay.
Otherwise I have to go with two power transformer. After if I have the two power transformer why not build a mono block.. Need two choke so on....
On the end all cost twice as much I planed to spend on these amp.
Honestly I build these amp to get back into the tube world. All do I read these is a great tube to in a good circuit so if I build something I want to have some value.
Last 15 years I just upgrade tube amp but I didn't built not one.
Just Solid S amps.
After I want to build a GM70.
Actually I wanted only that amp, I collected the parts for that to(but that is to dangerous to start) 1KV to get at least 30W.
My friends told me start with something smaller (lower voltage, not like 400V is so safe)
to have some success and experience.
Greetings Gabor
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.