Drip 670 v3 HV Heater problem

[Winston OBoogie]

IMHO, very poorly designed  PCB. One 10mF capacitor at input of the LDO and 3  pcs of 10mF at output.

Generally speaking, Drip boards seem to favour form over function.  Surely anyone can see that 3 caps on the output looks better, reducing ripple on the input would spoil the presentation and just look like crap.  I bet the traces surrounding that regulator are beautifully executed too

 

[Pusch3l]

@moamps
Cutting the CT did the trick  ;D
I am feeling a bit stupid right now

@Winston O´Boogie
The Mainboard of the 670 is pretty well designed as far as i can say.
This PSU PCB they send is a new version without primary connections and the lack of any
documentation and a BoM tells me they hammered it together in a short amount of time.
But the most disturbing thing is that you get no absolutely no support if you have problems.
i have contact with Tubemonkey who did build the exact same version and we have both tried
to contact them via email,social media and contact form on their site. but never got any answer....

I am very grateful to know a place like this where you get help from much more experienced people than i am

 

[Winston OBoogie]

Heya Pusch,
I suppose I was feeling badly for you guys who've been having issues with getting things working without documentation and so
I fired off a cheap shot at Drip. 
It isn't like building a 670 is an inexpensive project after you've bought all those transformers and, given the price they charge for the boards themselves, a basic schematic and BOM is the minimum you guys should expect
Anyway, I'm happy to read you and Mo figured it out 

 

[TubeMonkey]

Hey All!

Just chiming in here as my builds have both been sounding great but I am having an annoying issue with the HV line. There are two possibly related symptoms... The first is that when loaded with all 8 6386's and dialed with HV adjust to 240 the high voltage line seems to crash and presents as a "thump" on the meter (crashing and recovering 3-4 times/sec) This seems to stabilize when the HV adjust is dialed down. This has been my temporary solution... but I would ideally like to run this at full voltage and believe this is a symptom of a build or design flaw... it is happening identically on both units i've built. The second possibly related symptom is turning off the power switch and re-engaging it before the HV filter caps have time to discharge seems to cause the 5y3 to self destruct spectacularly when HV starts to come up. sparks galore. The rectifier gets trashed after this happens and needs to be replaced, and i'm unsure if the other tubes in the PS are weakened by this arc. It is possible that triggering this arc could trash the other tubes in the supply. It is also possible that the JJ 5y3 aren't handling the required amperage. on that note the JJ 6386's seem to only last about a year before about 30% of them become noisy or oscillate so heads up on that. Any ideas? I have the .1 wirewound installed.

 

[TubeMonkey]

Forgot to mention pulling 6386's from one side seems to fix the issue so I believe it's a problem presenting from total current draw

 

[scott2000]

Subbing in another supply to check sounds too risky if there is a problem somewhere else....?

 

[Sc]

Forgot to mention pulling 6386's from one side seems to fix the issue so I believe it's a problem presenting from total current draw

Hi
Think you are right. I'm not through with the the 670 yet - the V3 version board I know has a GZ34 (5AR4) and a total of 600uF PSU-Caps on the PSU-board only (!). I'd throw out at least 2/3rd of it, and check the rest of the HV Caps on the mainboard against the original design. It may be a bit more than original, but certainly the elevated values will stress the rectifier valve, even with a first cap of acceptable value. And check DCR of your power transformer against max. values of the valve - two +- 50R resistors in the AC line may be a good idea (fried several mostly new production GZ34 myself). The Sowter transformer is sure strong enough - if you use another one it's relying on old design literature a good idea to have 1.5 to 2X current capability of normal consumption on every winding (peak currents). The pulses in your unit .... As Scott200 said, if you have a strong bench supply you could see an unsusual consumption of the mainboard. But this underdesigned and overcapped PSU issue is a bit familiar (e.g. U73b / STA ....). The sparkling when reengaging before caps are unloaded may happen in less loaded designs ... .
Good luck ...

 

[Winston OBoogie]

What he said

If your PCB version is anything close to what Sc indicated his version is, then there's way too much capacitance for that poor rectifier tube.

 

[TubeMonkey]

Yes this is the V3.2 pcb with 600uf. I had a thought about all the filter caps when build wondering how he was getting away with that much capacitance. I am wondering if I could get away with dropping in a diode octal plugin rectifier before pulling the board and experimenting with filter values.

 

[TubeMonkey]

Subbing in another supply to check sounds too risky if there is a problem somewhere else....?

I actually had the power supply from one of my units powering the other unit and the same problem presented so unless there was the same errata present on both main boards I still believe it's a design flaw

 

[Sc]

It is also possible that the JJ 5y3 aren't handling the required amperage.

I am wondering if I could get away with dropping in a diode octal plugin rectifier before pulling the board and experimenting with filter values.

It seems to me, that GZ34 / 5AR4 (as on mine) would be a better choice - much stronger, voltage drop much lower and less heater current .... (see chart). Pinout should be the same. I can't estimate the outcome of a diode rectifier on this tube regulated PSU except that B+ will be a lot higher compared to 5Y3, and you might loose the "dynamics" of the original PSU. But of course I'm a bit fixed on trying to be close to original designs (including carefully elevated filter cap values) .

 

[radardoug]

Yes this is the V3.2 pcb with 600uf. I had a thought about all the filter caps when build wondering how he was getting away with that much capacitance. I am wondering if I could get away with dropping in a diode octal plugin rectifier before pulling the board and experimenting with filter values.

600uF??? Are they crazy????

 

[Winston OBoogie]

600uF??? Are they crazy????

They know not what they do.