Fuse randomly blow::: Please help

[Deepdark]

Hi

I just finished a tuve vari mu compressor. My main transformer is edcor 350v c.t. When start up first, the front panel was open and i tested voltages, all looks good. I shut it off, bolt the front panel and shut it on. Then the gise blow. I found it weird a little. I checked my ground connection and they all have good continuity, my secondary have their c.t. Connected and they have good reference to ground. I change the fuse that was a 100ma for a 400ma. It works. I shut it off and shut in on and then i blew. Is it normal, i mean, maybe the caps haven't fully discharge? 

One thing to remember is that i changed the tube rectifier for a pair of 1n4007. Maybe current build up too fast? Thing to remember is that i don't implement a standby switch (i'm not a standby switch believer).

Another thing is that the pt is rated 350vac but i measured 390vac, but straight after the diodes, i have the right voltages and after my rc filter, i got exactly the right voltage so it gave me the impression that i have the right current through the. Cicuit (30ma) and i got 191v, which is exactly what i need. As i said, voltages are all spot on. Any idea?

Maybe there is a current limiting kinda thing i could implement?

 

[Pip]

a schematic would help.

 

[Deepdark]

 

[pucho812]

what kind of fuse is it? slo-blow?

 

[Deepdark]

sloblo. I dont know if its due to the facture i changed the tube rectifier for a pair of diodes and à rc filter. Maybe current raise too fast?

 

[mjrippe]

First of all, this is the THIRD thread you have started about this same problem.  Not a good practice.  Keep all the information in one place and you won't confuse people trying to help you.

Second, the original schematic calls for a 1 amp slow blow fuse.  Why do you think that 100ma or 400ma should work?  Particularly with a silicon rectifier which can give you a larger/faster inrush current.  None of your other threads mentioned the undersized fuse.

 

[Deepdark]

Ok now that fuse is replaced, i retake all voltages. All is in the ballpark exept 2 things. V1 and v2 pin 6 (screen grid) recall for 118v in the original manual.I got 12v and while readind over the web, some people did get the same result. This is in no signal applied condition. While reading over the bunkerofdoom am864 post, at no signal, v1 and 2 opperate at zero bias condition, so low screen grid voltage. So my 12v should be allright?

Other one, the vu meter. I just strapped the gain reduction side of the meter, so basically across r2. It basically measure the voltage differential change across R2 as gain reduction is applied, isn't it? Mine site at 0v, at the left side, and while turning current control cw, needle move a little to the right. I just use a regular vu meter from hairball audio. Is it possible the internal 3k6 resistor has to do with it? It is also internally rectified but since we are dc, i guess dc will just pass through it without any changes, isn't it?

Maybe both probelm are related to the low screen grid voltage. Is it possible that one bad 6sk7 cause this? Even if the other values are ok?

 

[Deepdark]

Ok so no more blowing fuse. I tested the unit and pass signal, no hum, plenty of gain. Some little things, though. When putting threshold pot at max, i got i kind of rumbling sound. And playing with current ctrl, there is a certain spot wich Is oscillating, but once passed that spot, no more oscillation. I know the 6sk7 need to be really well matched, so maybe mine could be better matching, what do you think?

Secondo, the meter. I do see the gain reduction while compressor is working, but it sit really low. So two choices. I could open it and swap out the diodes and 3k9 resistor OR putting changing the 240ohm resistor (R2) until I get 0db. What do you think?

 

[Deepdark]

I just analysed a little more the schematic. The gain reduction side of the meter is strapped across R2, which mean that if we want to read 0db, the voltage drop across R2 should be around 1.2v, isn't it? So here we talk about around 5ma (1.2v/240ohm). isn't it?

If my meter sit about -10, so it looks like I got a voltage drop of less than 1.2v, isn't it? Since I used this meter (http://www.hairballaudio.com/catalog/parts-store/panel-meters/8020-vu-meter#tab-40), from hairball audio. Is it possible that the internal resistor and rectifier can cause that drop? Since we are DC, i guess it will flow through the rectifier without any modification but maybe the internal 3K9 resistor affect the reading, So maybe I should just swap out the internal rectifier/resistor? Or I should test another pair of 6SK7, maybe they aren't that matched. Am I in the path?

And on another note, I run the unit for testing through my Apogee ensemble. So I don't run into 600ohm load. Don't think it will change anything but we never know

 

[Deepdark]

I did some more test. I found it to lack a lot of lot end and I got the motorboating at certain setting, that last one leading me to think that the 6SK7 are mismatched. But for the lack of low end, is it possibly due to a impedance mismtch (I strap it through my apogee ensemble). To get the full freq, response, is it primordial to be load at 600? Or there is another symptom of that lack of low end? The unit is really mid forwark, which is nice, but I want to hear from you guys. Thanks

 

[radardoug]

So you say you are building this for a client, but you obviously are not good at faultfinding or understanding circuits. I hope you are not charging him for all the time you are spending on it.
I suggest you do a lot more learning about circuits before you proceed. And how to fault find.
If you are losing low end it could be anywhere. Did you use the right transformers? Did you use the right coupling caps?

 

[Deepdark]

Did you use the right transformers? Did you use the right coupling caps?

Coupling caps are 0,1uf, transformer are edcor with center tap (secondary is their PP series).

 

[Deepdark]

Input transformer is XSM600/10K (2-1/2W) which is center tapped. 20hz - 20k
Output trransformer is XPP1-600-10K  (1W, push-pull tube output) with freq. response of about 70hz to 18K / -1 db

A continuity test has been made first to be sure all connections are made and it follows the schematic. Then, I adjusted the current pot to 2.5v, as on the manual to test my voltages. it's all ok. For the lack of low end, It could be the 6sk7, since they are nos tubes. Could it be the culprit?

I remade the psu as attached, I calculated about 150db of ripple red. so I should be ok there, don't think the loss could be from there.

Mine is made with R3/4 at 47K as on the old manual schematic, I coudl try with 12K but this will probably change noting.

Since it's a new build, bad cap is less to be the cause, but we probably never know.

I don't think a mismatching impedance can cause that. Output of the apogee is 100ohm (so I enter the compressor at 100ohm) and is loaded at 10K (the input of the apogee). Pretty sure the trouble isn't there

 

[radardoug]

Have you verified your test hookup? Output of Apogee to input of Apogee, then test.
Tubes of themselves do not affect low frequencies.
Have you calculated the cutoff frequency of the 470K/ .1 uF combination?
Have you measured at every stage what the frequency response is?
For instance, measure right after the input transformer?
Instead of using the Apogee, buy some proper test gear.
Do you have a scope? Oscillator? Noise and distortion meter?
If not, why not?