Fostex 3070 Power Supply Considerations and issue With Schem

Fostex3070 compressor.  I installed machined pin sockets for u1 and u2 ics. I added .1uf ceramic bypass caps from + to - and - and ground. In one unit  On both units c3 and c4 popped/vented. I'm not experienced at all but have had success upgrading a few pieces of my gear and am familiar with bypassing the opamp supplies and feedback resistors if need be.  I've attached a schematic. I plan to get to the print store and have them scanned so I can upload the manual but for now this picture is all I have. If anyone can maybe tell me what is happening, I'd appreciate it.

scott2000 said:

I have 2  Fostex3070 compressors. I wanted to start upgrading them with some better op amps. I went ahead and installed machined pin sockets for u1 and u2 ics. I added .1uf ceramic bypass caps from + to - and - and ground. In one unit I installed op2604 and the other a lm4562. On both units c3 and c4 popped.

Click to expand...

Popped? You mean they literally blew off the rubber bung that seals them?  The changes you list should not stress those caps.

I managed to add some additional 47uf caps from pins 8to5 and 5to4 and the units work and sound good.

Click to expand...

sounding good is good.

I'm not experienced at all but have had success upgrading a few pieces of my gear and am familiar with bypassing the opamp supplies and feedback resistors if need be. To be honest, I don't even know where I got the idea to add the 47uf caps to keep the power supply caps from blowing.

Click to expand...

So adding caps elsewhere kept C3 and C4 from blowing?  Are the VR U1 and U2  putting out clean DC?

I'm suspect that the power supply may not be up to par but, I only added 2 additional amps per unit so, I'm guessing that should be only around an additional 10ma draw?(tl072 vs 2604 and lm4562) I've attached a schematic. I plan to get to the print store and have them scanned so I can upload the manual but for now this picture is all I have. If anyone can maybe tell me what is happening, I'd appreciate it. I was hoping to at least swap u1 and u2 as I believe they will provide the most help? I notice there aren't too many caps around. Could lack of decoupling have anything to do with it? Can I upgrade the supply with a newer transformer? I appreciate any help and I'll get the manual uploaded within the next week or so. Let me know if you need additional info...Thanks!!! I should add that this happened as soon as I replaced ic u1...I never got to installing u2.......

Click to expand...

Unclear what is happening from that symptom.. to blow up C3 and C4 would require multiple component faults (shorted rectifier and shorted VR, which is low probability and would not be repaired by adding caps elsewhere. 

Sorry but I have no good advice.

JR

Hey
Did you replace c3 and c4 at the same time as adding the two extra 47uf caps?  If so then the original c3 and c4 might just have been old and ready to blow... Otherwise you would have also blown c3 and c4 replacements before adding the two extra 47uf caps and replacing c3 and 4 A third time...

I have one of these things and like it too, I've always been meaning to mess with it to get it cleaner so im real interested in your work!

Your ceramic caps look screwy.
One cap from pin 4 to ground side of r3 is ok. Bthe other cap you have from pin 8 to pin 4.  This should be from pin 8 to the ground side of r3.  Or instead of using r3 you can use pin 5 like you've done for the 47uf caps.

Both these caps might already be there I think as c25 - c30, c32.  0.01if caps.  Check on the board.  They need to be close to the op amp and probably increases to 0.1uf.

Well, if it works now and sounds good... Then there's no problem.  If that's what the op amp you're using likes then so be it, I haven't used that one before.

From what I understand it's usually one or the other.  Ie one ceramic between the V+ and the V- or two between the V+/- and ground. 

But if it works then great!

I checked the layout the c30 bypass cap is pretty far away from the op amp so it's not really useful.

No they are too far away to do any good.  Leave them as is, the caps on the pins is the way to do it.  If it ain't broke...

Off hand in not sure why it would blow the filter caps.  To me that's very strange.
Did you use the same bypassing setup for the 2604?  I wouldn't use the cap between the + and - pins for that one.  Does adding the 47uf caps stop c3 and 4 from exploding in the new unit?

I'll dig mine out and play with it next week.

IC voltage supply bypass connection: CAPACITOR BETWEEN THE SUPPLY PIN AND GROUND

Ok so either our schematics don't line up or maybe you are misreporting your cap numbers?
For me
C1 and 2:  1000uf filter caps
C3 and 4:  10 uf filter caps post regulators (these are the ones destroying themselves? Check their orientation).

C7 I have as a 4.7uf cap in the power supply mute section.  Definetly shouldn't be a increased to 47uf imho (no benefit?)

C2... If you've change this to 47uf from 1000uf, that's not good.

One thing off the bat that I see is that the regulators are case mounted on leads.  They look long to me and that affects the psu.  Read the datasheets to see what they say specifically about distance (wire distance!  My leads are tucked under the psu pcb so are quite long) if regulators from filter caps.  It May need an extra cap or two to make up for that.

scott2000 said:

Well.....It was the bypass from + to - I guess.........I changed everything -to ground and + to  ground..at least on the unit with the 2604..... Seems great so far... Thanks!!!!


is there any truth to the lm4562 liking bypass from +to - and -to ground?????


cuz lucy gonna have sum splainin to do.....

Click to expand...

I read on another forum the 4562 likes having V- to ground best.  Some things were also said about V+ to V- as being helpful if option 1 wasn't enough.

Maybe it has something to do with the power mute circuitry after the regulators on the power board?  I'm not sure but I suspect the 4562 decoupling might have caused a conflict.  After reading up on it the method you used has been suggested and implemented by lots of people for that particular op amp (meaning your other gear is probably fine!)

C2 and 7 in main board totally fine.  Very confusing to have one schematic with multiple c1s 2s 3s etc.

For phase compensation I have no idea.  I'm still figuring that stuff out.

If you're worried at all you can always pop in some 5532s in there I doubt it would sound much different.

Things I would do with infinite time:
Change c1 on main board to film or increase its value for an el cap.
Change the 78m15 regulators to either lm7815/7915 or 317/337 adjustable ones (would need to add a couple of parts but there is probably room since there case mounted.  Just be careful to insulate the negative regulator.  This psu uses dual positive regulators I guess so they could use the case as a heat sink without needing mica waffers?  Its a little screwy to my eye... Transformer has dual secondaries rather than center tap.  The 2nd regulator is wired out to ground so that it's ground leg is now V-.... I'm sure there are betters ways to get +/- !

To be honest I have zero clue about how to alter attack and release times for compressors... Any ideas?

I was looking though at the input stage and thinking of changing c1 to a 4.7 polypropylene and r3 to 10k (to keep the hp pole freq the same as it is now... It would be an experiment not necessarily the right thing to do)

Inverted formula?  Don't know what you're getting at there.  Looking at it again its probably better with the original resistor and a 3.3 uf cap.  Sets the frequency to about 10Hz.  Don't think it needs to be lower than that?

My c9 is listed as 1uf/50v and is also that part in the unit.  I wouldn't change anything in that little muting section because I don't know what would happen and don't see it as a place for big gains .

The impedance , if I'm not mistaken, is more to do with r1 ... It's listed as being 30k ohm input impedance so I'm missing something.  Hopefully someone will step up and explain it to both of us.

Changing c1 is to get "better" sounding cap instead of the el cap that's there.  I have some 3.3 Panasonic film caps that will fit easily so I'm going to try that.  The frequecy bit is a high pass filter formed by c1 and r3 (rc hp filter) and should be set very low to avoid low end phase shift in the audable range.  Right now it's like 3hz I think.  Changing c1 brings it up to 10Hz which I think is ok, but it could always be lowered again by increasing r3 to 10k (i happen to have some 10k resistors handy too).

Let me know if you get any results from changing c11 for the attach and release times!

Why not just rebuild the power supply?  Or bypass the muting circuit, I have a feeling that there is something funny going on in there.

Did you use these units often before modding them?

Is it just c3 that's swelling or c4 as well?  C3 is on the positive rail side that's connected to the muting circuit (it mutes the outputs when you turn it on or off... I'm sure it's there for a reason but maybe the new op amps aren't playing nice with it).

I'll get you a pic.  Everything is populated on mine.  All the little caps look the same size but have different values.

C9 defintely 1uf/50v

C3 and 4 probably blew because they are supposed to be 10uf for the regulators stability... You should check if the + and - rails are balance ie equal to each other (not +15v and -14v) or just did up the psu to something more standard (imho).

Why did you change the 330pf to 22pf?

Neat rule for phase caps.
Glad you didn't blind yourself.  Ditch that psu and get something regulated (and reliable).  Maybe you've Madera solder bridge somewhere?  That could blow caps.  Still blowing them?

C37 doesn't exist from what I can see.
C19 is between pin 1 and 2 of u7.

Play with c11 yet?

C19 is actually empty on my unit physically but it's well labeled.

C37 doesn't appear in the parts list neither can I find it on the schematic.  I don't find it on my board anywhere.  Where is it on yours?

I do have a 220uf 16v cap as well as a diode on the backside.

Well that's right were I have (o r rather don't have) c19!  Weird.  Do you have a cap there on the other side?  Of so what's its value?  My pcb markings are on the other side (component side not solder side).  Must be different revisions?  Mine's labeled 1194v-0 and 8251095 100