No Erase/Record on Chan 3 (Tascam 38)

[explodingPSYCH]

I recently had some issues with various channels on my 38. After sorting out most of the issues via card swaps, etc I blew a fuse to the PSU, specifically Fuses F503 and F504. After replacing diodes D814 and D815, the fuses stopped blowing.

Everything seemed to be functional, but upon testing repro I noticed no erasing or recording on channel3 . Input and Sync both work on the channel. It's almost as though the function switch isn't arming properly, although it's LED does light up correctly.

I'm having a difficult time tracing this to components that could be an issue. I've also read other posts about similar issues that point to the oscillator or PSU (may something else got damaged besides the diodes, but I was unable to identify any additional damage). I've reflowed all the solder on channel 3 and cleaned the boards with alcohol. I also reseated the pins that plug in above the erase head. I also tested flipping the tape and playing it backwards, confirming no recording/erasing on channel 3 (channel 6 playback when tape reversed)

(The issue doesn't move with swapped cards, and the card installed works fine in other channels.)

 

[SkywaveTDR]

Wow, Normally that problem can be in the slave oscillator block. The way you can test is the oscillator is going on is with a scope hooked to the feed side of the Bias adjust variable capacitor. They are called C134 and so on. You can also ohm out P163 in comparison to other coils on the erase heads if it measures short or Infinite then the erase head need to be replaced. You need a good service manual with clear schematics to work on this not like you get at Hi Fi Engine with no schematic.

 

[explodingPSYCH]

Good call on checking C134. I'm getting voltages around 4 volts (on one of the pins) on all cards except for the card in slot #3. However, removing and moving that card to other slots, the voltage is still not present on that card. Normally I would think "something wrong with the card" but when I exchange it for one of my 8 backup cards, they all have the same result. No matter what slot I plug them into they don't get voltage on C134!! (well, about 100mVs and most).

I swapped the oscillator card with an extra I have from my back up machine, just to see what would happen. No change.

The other 7 installed cards get voltage on C134 on slot #3. This makes me think something is up with the master oscillator and how it's feeding the cards. It's as though it is not sending it's signal to my extra 8 cards (and the original card #3) based on something about those cards... but I can't tell determine what the variable would be on those cards

Next I will check the erase head and mobo pins you recommend.

 

[SkywaveTDR]

If a card has the same symptoms as it does in another slot where other cards work then it is on that card. If the problem stays in that slow no matter what card then a motherboard solder through problem. A 32 I was working on the other day I found a lot of cracked pin on the motherboard when inspected with a bright light and magnifiers. I typically remove the RCA panel and go in and resolder every pin that looks bad- most all the pins I have access to. They are shiny and well bonded to the pins as this is what I look for when soldering. This particular 32 has a lot of pots that just were open. They said 202 which means 2000 ohms but when you ohm them out the resistance is infinite from outer leg to outer leg. I have not seen this before with so many pots on both cards- I put Cermet trimmers in. They were 502 and they worked great. In the slave bias block there has been some caps that have shorted. I open these up when needed and put better 1 amp transistors in and check all the caps. I bought some 250 V caps at the value that was shorted and have some to replace the bad ones. I have only found a couple of shorted caps in all the many decks I have worked on. I would ohm out the erase head to see what resistance is on those coils. I did have a gold colored erase head or two go bad on me but these were the older 1/4 track erase heads as could be on the A1200 model. I don't ever remember the erase head going bad but surface contaminants can keep the tape off the gap and many people use Isopropyl alcohol still even though I have told them Denatured (Ethanol) is what should be used. I don't know what way you measure voltage but with bias at 160KHz a meter can not be used and a scope is most often the tool of choice for RF like signals.

 

[explodingPSYCH]

The issue now appears to follow the cards. 7 of the original cards work perfectly, but the original card for channel 3 does not get votlage on C134 (as measured by an oscilloscope). Furthermore, all eight of my back up channels behave as the bad channel (original #3).

I uninstalled all working cards and only put in non-working cards, and they still didn't work in any channel slot.

At this point I have to assume that the same component(s) are bad on all my back up channels. Relays seem to be a reasonable starting point, but if they are bad on all of my back up cards, then I will have to order replacement parts. Any suggestions on parts to grab first?

 

[SkywaveTDR]

You seem to be jumping to conclusion not supported by any testing and this usually results in lost money and more time to fix the unit. Relays only send the signal from the card to the head. If the slave oscillator is not working as tested by the C134 then I would test the transistors that turn this on and also the pins that are soldered to the connector- most of then are poorly soldered and once in a great while a tin will be broken internally which can be detected by ohming the inside pin location against the pad of the card. These are Molex connectors not easy to find and I researched and put in a stock of them. The Bias block can also be bad but the way you tell is if voltage goes into them and nothing comes out. Sometimes they are turned on by a transistor in ground circuit.
So 15 Vdc comes from J107 pin 3 is there all the time at osc lead that goes to C133. Pin 2 is the ground transistor that goes to Q119. Tp 3 and Tp2 allow you to measure current going through the Osc transistors. Pin 7 of J107 is the 24Vdc that comes in and turns on Q118 that then turns on Q119. If one does not have a extender card, you can tack on wires to parts and insert the card and make measurements this way. I have done this many times and I do have the extender card the very one I used at Teac.

 

[explodingPSYCH]

Right on! The transistors (Q118, Q119) were giving me weird voltages. I noticed Q119 was not the exact same type (had been replaced from one of my backup boards) so I swapped that first. That made the voltages correct on the transistors.

However, still no voltage into C134. I started looking for any other differences between the working boards and non-recording boards and discovered two things.

First, there was a discolored resistor (R198) so I swapped that with one that looked normal. This actually seemed to solve the issue and now I have voltage back on C134 and reproduction when recording!

The second thing I discovered after inspecting the boards for a while was that all of my backup cards are slightly different. The main differences are
1.) They are numbered with a "-02" instead of a "-01" at the end of the TEAC number on the trace side of the board.
2.) There is an extra resistor (R208 under jumper W125) on the "-02" boards.
3.) Some of the transistors are slightly different model numbers (Transistors marked C1685 have additional labelling "R81" instead of an "R38" on my original cards). Additionally the backup cards all use a transistor marked "C1318, S.7N" for Q119 where the cards that work in my machine use a taller transistor marked "D667, A C, 1 B 2".

I'm unsure why none of my backup cards are interfacing properly with the oscillator from my 38, but based on what got this card working, the Q119 transistor needs to be the taller D667, and perhaps something about that extra jumper is causing issues.

Any other suggestions on what would cause these other backup cards to not work would also be welcome. Next I will attempt to fully calibrate the card. I will also order some D667 transistors for the backup boards, unless there are better suggestions.

Thanks!

 

[SkywaveTDR]

The R 198 is the 8.2 fuse resistor used for current limiting. If the circuit shorts then this will burn out. I have no idea of the different cards but I do know that guys that buy cards willy nilly that come from 34B and other decks do not have the right circuits on them and so are not going to work. They must come from a 38. The 2SC1318 is a 60V, .5 amp transistor and in general the 2SD667 transistor are more powerful that the 2SC types. The D667 is a 120V and 1 amp transistor with 900mW dissipation. Other model of transistors can be used in circuit as long as they have equal specs to what was designed in there. The original transistor is a 2SC1318S type so the circuit make use of less than .5 amp. The 2SD667 is a good transistor and is ECB so whatever you sub in should have the same pin out. You should learn to look up transistor specs and know how to interpret the numbers plus this web site give lots of subs- https://alltransistors.com/crsearch...=80&uce=80&ic=1&tj=150&ft=70&hfe=60&caps=TO92
I stock a transistor called KSD161A which would work as well. Anything .5A and above with ECB leads.

 

[SkywaveTDR]

R198 should be a fuse .5W and this one would work. https://www.mouser.com/ProductDetai...=sGAEpiMZZMtlubZbdhIBINaGfyemSiGisY80V6qUUvc=