DIY rtp5 help please?

[jazz]

Hi,

I'm trying to build a rtp5 linestage according to the design by mr Allen Wright(third schematic from the top):

Schematics.

I had most of the parts present so this seemed like a nice and easy build but it is giving me quite a headache.

The way I've build it now it seems to have two functions but not the intended one. First it acts as a very effective fet destroyer. Every time i power down and then power up again the bootstrap fets in the output stage go to fet heaven and i don't understand why.

the anode voltage on the output cathode followers is about 150V, cathode voltage at about 70V, grid at about 66V and running at about 10mA each. The only difference from the original schematic is higher grid/gate resistors and 4,7K in stead of 47K in the bootstrap bias circuit. Why do i keep blowing these fets? I used 5687's for the output.

Second the unit is a very good and stable noise generator. I've scoped back the noise and it's first appearance is the anode of the input tube. Thus i figured i needed some anode stoppers and while it helps the noise doesn't go away even after going as high as 100R and there is very little difference between 22, 40 and 100R. The noise is present all around the output stage slcf except for the power supplies which are all on target and pure dc.

The input tubes are 6dj8's running at about 5.5 mA each, so a bit higher then specced but lowering the running current hasn't helped. Grid at 0V, cathode at 1.15V, anode at 66V. CCS as per the cookbook, p4 of mod's and maybe's best.

What could i do for the fet destruction and for the noise?

Thanks,
Joris

[SY]

I'm sure Allen will have some comments. My first thought was, "Easy build?" Not at all- without using a well-laid-out board, the chances of oscillation are near 100%. And I think that may be the root of your problems. Cascodes are tricky and insanely power supply sensitive, and using high transconductance tubes makes things even trickier.

This is a very advanced and challenging project. Very high performance, but not something to just throw together.

[jazz]

Hi,

for additional info here is the pcb layout, all jumpers are resistors.

grtz
Joris

[Allen Wright]

Those circuits are not intended as "paint by numbers" DIY projects, they are there to show that our designs are very different to most current products that are just 50's and 60's designs re-issued in modern boxes.

That map, and you, don't show the protection zeners the fets need. If you put a 10V zener from gate (cathode) to source (anode) on those fets, they will stop blowing.

But If you've sent the time and effort to make up a FVP-5 line stage, is there any way you can make the far better sounding all tube version, the FVP-5A?

It takes another tube per channel, but is really on another level as far as sound quality goes. The difference between these two designs had me rethink the sonic value (or otherwise) of MOSFETs in an audio signal path.

Noise? As SY has suggested, it's most likely oscillation. Are the gate stopper R's RIGHT on the tube socket pins? I mean RIGHT on the socket, not 1/4".5mm away! 6DJ8s are VHF RF tubes after all.

Regards, Allen

[Allen Wright]

I just looked over your PCB layout and I'm confused.

I cannot see where Pins 1 & 9 are, as they all seem to be equallu spaced, so cannot orientate.

And in one tube you seem to have one pin unconnected, which if it's pin 9 that's ok, but on the second tube it appears that you have TWO pins unconnected.

???????????????????????????????????????

Regards, Allen

[Andrewbee]

I cloned an FVP4 or 5 several years ago, cannot remember the model at the moment but it was from Allen's preamp book which I bought. It used 6922 or
6dj8's and IRF830's I believe and was quite impressed and wrote Allen about it, singing its praises.

He responded that if I thought it was that good I should try the the FVP5A, which I then cloned, superreg and all. While I would not say it was a night and day difference I thought the FVP5A was "smoother".

I still have it after many years, the only amp I have never cannibalized which knowing myself is unheard of.

In other words, its well worth the effort.


Andrew

[jazz]

Hi,

Just looked at the posted pcb and realized it is the wrong layout. My sincere apologies for this error. The mistakes around the output tube have been corrected but i can't find the correct layout now. I'll keep searching.

Thanks for the hint on the zeners mr Wright. My FVP5 is due for service anyhow so I'll try to change it to the FVP5A version when i'm at it to

regards,
Joris

[takiskatsikas]

[ATTACH][ATTACH]100_1135.JPG[/ATTACH][/ATTACH]hi all .i work on this preamplifier

[jazz]

Hi,

Well i didn't have the layout anymore so I made a new one. The dimensions may vary slightly from the actual unit but not by much and should give a good impression. This is bottom view for the tubes only. I've marked pin number one and supply voltage inputs as well as the values for resistors shown as jumpers.

I did try to get those gridstoppers rather close but maybe not close enough...

regards,
Joris

[Andrewbee]

Joris,

the grid stoppers should be soldered on the tube socket pins as close as possible, in other words the resistor body should be right against the socket connection, touching it. The idea is to have the grid stopper as close to the tube grid as possible. Perhaps you may want to consider soldering one end of the grid stopper directly to the tube socket and the other end to the PCB.

While it may not be pleasing to the eye, it will be pleasing to the ear and once its installed you will not see it anyway


Andrew