Fostex 160 4 track rescue attempt!

[rockhorst]

A former guitar teacher of mine asked me if I could take a look at a broken down Fostex 160 4 track from the 80s, that belongs to a producer friend of his. According to its owner this unit adds instant analog mojo to your recordings, and tons of it.

Plugged in the power, LED lights up (nothing else works) and after about half a minute, the unit starts to smoke. So yes, it's definitely broken Inspecting the pcb at the place where the smoke came from, I found a burnt diode in the power section. The pcb around it has been scorched and the goop that covered it has burnt of as well (I guess that's the smoke). The diode had unsoldered itself because of the heat and the traces seem to be gone. I tried to identify the diode, but could only make out the letter 'N'. When I touched it with some pliers, it crumbled. It's construction looked similar to an 1N4001. If anybody owns a unit or schematic, it's the diode right next to the 470 uF capacitor labelled C2.

At the moment, I only have a crap cell phone camera at my disposal, if necessary I'll post better pictures later.


So, the questions I've got:

• Any clue about which diode is in there? Does anybody have a schematic, or even better, a working unit? Some picutres of the power section would be very helpful
• Any educated guesses about a diode substitute?
• What do you reckon are the changes of the unit working properly again by just replacing the diode and maybe fixing some traces?
• Anything else?

BTW: the manual says the unit requires 12-15V DC. The Fostex adapter that came with the unit dishes out a heftier 17.5V. My guess is that this extra bit of juice doesn't really harm the unit, but I thought I'd mention it.

[Skruffyhound]

dishes out a heftier 17.5V

Did you measure that or is that what it says on the adapter?
Edit: Had more questions , did you manage to see orientation of the diode before it crumbled?
Can you be sure the adapter is the correct one and therefore the correct polarity (half ties into my first question).
I'd check the traces to see whats what polarity wise, earth should be obvious. Then work backwards to see what the adapter input wants. That's if it's not marked on the box
Wouldn't be the first time a studio had more than one adapter by the same brand name but with different polarity. If that diode was reverse polarity protection it may well have done it's job very well. Looks a bit deep into the board though for that, but I'd check it first.
Good Luck

[rockhorst]

It's the exact adapter specified in the manual of the original product. Polarity is correct (center positive, sleeve negative). The sticker on the adapter says its output should be 12V, but I measured 17.5 V. I have no clue what the studio owner powered it with when it fried though.

The unit broke down 2 years ago, so I don't think anybody remembers the circumstances. Judging by the guys nature, seeing that the original adapter was also handed to me, I think they used the right power supply.

I haven't had the time to trace the entire power section, but it seems the diode (D6 according to the pcb) is situated parallel to the input of a 78M09A voltage regulator (12V comes in, splits to regulator and diode). After the diode there is the negative side of a 470 uF capacitor. I didn't get the polarity before it crumbled, but this does suggest a direction. I've found another 1N4001 diode in the same section (oriented in the reverse polarity way), so it's reasonable to assume this diode also was a 1N4001.

I've separated the pcb from the rest of the machinery now, I'll hopefully have a detailed picture tomorrow.

[rockhorst]

Update: the 'nearby diode' turns out to actually be a 1N4002 (daylight makes a difference )

Some more detailed pictures

the full power and tape control pcb

The part sticking out on the top right is the power section

power section detail, incl burnt diode


the backside of the power section

if you look at the far right of the burnt diode, it says 'DC in'.

[Skruffyhound]

Congratulations. We still don't know what caused it, or if there is other damage. I reckon someone grabbed the wrong connector in the rats nest and zapped it, maybe without even noticing the error. If it was me, and I am just a hack, I'd check the whole of the board for anything unusual, bulging caps, dirt, discoloured components wire fragments etc, then double check those polarity issues, stick an IN4002 in (just a higher voltage rating than IN4001), cross my fingers and fire it up. That's just me though. Perhaps someone wiser will suggest some other safety checks.
The 17.5 V the adapter is throwing out is without any load when you measure it and as you say it goes straight into a regulator on the board, so I wouldn't think that's the problem.
Sounds like you might get away with an easy fix if the diode burned fast enough. If I was really paranoid I might desolder the regulator and test it before firing up the whole board, and replace the 470uF cap on the other side just for safety's sake. Even if it's not fried the cap may have been compromised.
Good luck with it and let me know the results.
- Aston

[rockhorst]

@Skruffyhound: Thanks for the suggestions, this was exactly along the lines of what I was thinking: replace the cap, fit a new diode and maybe replace the regulator. I haven't found any other parts looking 'suspicious'. I don't think we'll ever know what caused it, but maybe we can fix it nonetheless.

I'm a bit unsure about the diode's orientation though. Is the intact diode (D3) reverse biased, should D2 be oriented the same way, or the other way around? And if anybody has a suggestion for a substitute for the 78M09A, since it seems to be a bit difficult to source for me. Any method by which I could check the cap and regulator 'in situ'?

[rockhorst]

Wow, Fostex sent me the service manual, within 12 hours of mailing them! Not bad for support on a 20 year old product. Relevent part of the power section:

The diode is indeed a 1N4002 and oriented to the positive side of C2. So what's its purpose? Polarity protection?
(sorry, bit nooby at this part, hope somebody has the time to elaborate a bit).

[beans_amps]

I would suspect the downstream cap: C2.  The transistor seems to be working as a pass regulator.  It may be shorted, or the downstream load may be the culprit.  Where does the wire going off the drawing from the Transistor collector go.  Is the trace marked with an 'X' cut on the board?

D1 is present to prevent the transistor emitter to base voltage rating from being exceeded.

In order for the Diode to get hod enough to melt the solder on the legs, it was passing a lot of current.  Likely was not shorted until the very last.  Oddly enough if a diode shorts, it won't always burn up as the power dissipation goes down when the voltage drop across the diode goes nearly to zero.  The 1N4001 is only good for 1A continuous and can carry that for a while.  I would replace all 1N400X types with the 1N4007's.  It is cheap insurance and the price to you is usually the same.  Fostex saved a penny or two by using the 1N4001.

HTH,

Sean

[beans_amps]

Some of the previous post was off topic.  I thought (for some reason) that D2 is the failed diode.  If D6 is indeed the dead diode, the obvious failure mode is application of reverse voltage.  When reverse voltage was applied, the diode conducted all the current it could until it failed. 

The new understanding makes me think that C3 will need to be replaced.  Also Check U2.  The diode held the reverse voltage clamped at 0.6V or so until it finally burned open.  Once this happened, the full reverse voltage was applied to the circuit.  The three terminal regulator does not like to live in a reverse voltage enviroment.

You said that you had checked the polarity of the power supply you used last.  Re-Verify to be sure.  There is a chance that the diode was damaged from a prior use of the wrong adapter.  You want to be sure about the adaptor before using it again.  Check the polarity of the voltage at the disconnected plug that attaches to J1.

Sean

[rockhorst]

Thanks for all the time Sean! I think I'm just going to replace all 4 'big' caps (3x 470 and 1x 1000 uF), though they show no obvious signs of failure it can't hurt after 25 years, like you said. The board looks a little darker around them anyway. New diode and new regulator, quadruple check the power and pray.

I've got a 7809 regulator, the schematic calls for a 78M09. I think it's a substitute, but not sure (found only one manufacturer that stated that they replaced the 78M09 with the 809). Can anyone confirm?

[Skruffyhound]

Check that the pinout isn't reversed by downloading both datasheets.


http://www.ic-on-line.cn/view_download.php?id=1029205&file=0002\ta78m12f_10947.pdf


http://www.fairchildsemi.com/ds/LM/LM7805.pdf

here's two possibilities, but I'm just guessing at the letters before and after.

Good luck

[beans_amps]

I think the M is a lower current rated part.  The M type is rated for 0.500 A while the regular 7809 is rated for 1.00 amp.  Again Fostex saved a few pennies on a part.  As far as I know the pinouts match.  Change the caps and put in the 7809 that you have.  Everything should be ok.

Sean

[rockhorst]

and...it...works...Whahaaaaaaaaa nice!

Thanks guys

[Skruffyhound]

Very Cool. See what you can do with a little care and attention! Instead of landfill it's going back in the studio.
Well done mate.

[beans_amps]

Good News.  I hope it works for 20 more years.

Sean