DX-4D repair, adjustments ??? (those three trimmers...)

[Blue Jinn]

Hi,

I am selling one of my DX-4D's and made Ethan's mod for standalone operation. Ran through a quick function check* (with a tone, not with program material) and noted the level drop on ch 1,2,4 (which I'd expect is the 2:1 compression kicking in.) Channel 3 goes straight to zero. offered the guy to drop the deal or try and fix. He opted for try and fix. OK.

I came across a thread on audiokarma (which for whatever reason won't let me register....)

http://www.audiokarma.org/forums/printthread.php?t=206575&pp=40

Re: the type II on a Tascam 244 and a dbx 224. A lot of similar parts AFAIK. Both found that a TL072 opamp in the output had failed. That is a $2 part, so I'd figure it can't hurt to replace it. I'm also ordering the dbx chips, NTE makes replacements:

The uPC-1252H2 crosses to NTE1794 and the uPC-1253H2 crosses to NTE1795.

My test equipment consists of a cheap multimeter right now. There were no *obvious* signs of failure (puffy caps, burn marks etc.) If the TL072 proves to be the culprit I'm a lucky bastard, (and the complete lack of output signal (based on the meter, not on the ear) seems to suggest so.) If not, (and I'm rarely that lucky,) (and here is my ignorance at work) how do you make the adjustments with those 3 trimmers assuming I replace one the dbx chips? One is "Gbias" for the 1794 the other is "Is" and the third is ????? That is where I am totally lost. Looking at the block diagram Ethan supplies on his website (which is diff from the one I have) two of them are adjustments between +/- 15v and the other is on the "Is" pin and between the pin and ground. How do you measure those three? And what is the Is. what value are you looking for on Pin 2 ("Is")


(And that is assuming one of those is bad...) There aren't that many electrolytics, so I suppose I could replace all them as well...

I'm posting here instead of DIY, but if you think I should move the thread over there (or the Tascam forum) let me know as well. Seems the DIY path and the analog path cross quite frequently....

As an aside, it looks like transistor on the output of the 072 is just part of teh muting circuitry.

*0vu tone on Teac 5, to via dbx to 22-4. Adjsut 0vu on 22-4, engage dbx. Did not actually put a pair of headphones on to see if any noise on channel 3.0

 

[Blue Jinn]

Well,

I was wrong about the output IC being bad. Replace the TO072, (and wrecked one of the PCB pads, but enough copper to connect it anyway. However, still no signal. So, I'm waiting for hte dbx chips to come 3/15. WIll try that. Have the service manual now, but I don't have the time or test equip to trace down where I lose the signal, so I'm basically replacing parts willy nilly at this point....

 

[Blue Jinn]

Transistor theory help

OK.

Here goes my understanding: transistor: acts like a switch, voltage on the base (like the grid of a tube) allows the transistor to conduct. Basically voltage at base switch closed, remove voltage, switch open. Can a transistor fail in such a way that it is always closed? In troubleshooting this unit, when the dbx function is engaged there is no signal on channel three. There is a transistor in the output that looks like it effectively shunts the signal to ground when 12v (labeled "mute") I'm wondering if this could have failed which permanently sends the signal to ground. I don't have a sensitive enough volt meter to determine if there is any signal at the output of the opamp, so my next step is to desolder that transistor...

 

[jviss]

OK.

Here goes my understanding: transistor: acts like a switch, voltage on the base (like the grid of a tube) allows the transistor to conduct. Basically voltage at base switch closed, remove voltage, switch open.

Well, that's so if the transistor is being used as a switch; it can also be operated in its active region, and used as a buffer or an amp or a filter, or part of a larger circuit that fulfills a similar function.

 

[Blue Jinn]

It looks that way from the schem as if it were "on" it shunts the output to ground. So I'm wondering if it coudl have failed such that it is always "on?"

 

[jviss]

Yes, assuming it's a bipolar transistor, the common failure mode is shorted, collector to emitter, sometimes the base is shorted with them.

 

[Blue Jinn]

Yes it is bipolar. I'm going to test for shorts.

 

[Blue Jinn]

Help!

OK, if anyone recalls, channel 3 has no signal. Did some rsch, got a schemo, and replaced U101 (opamp) based on assumption from another thread where two folks had this part fail. Didn't fix. So I replaced the dbx VCA chip with a NTE equiv. This is U102. The NTE part is 1794. I verified the correct orientation. INstalled. Optest unsat.

Now the NTE chip gets extremely hot, hotter than it should, (actually hot enough to stink....) and none of the other (working) dbx chips are warm to the touch. I'm assuming that:

1) something is making it overheat, (obviously)

2) whatever that something is, probably whacked the orig dbx chip (the vca) too

so, any one out there have suggestions on what would cause this chip to overheat? I am going to double check my solder work to see if I shorted something I don't think I did. I'm really at a loss as these circuits aren't too complex...

There is one other dbx type chip, the RMS detector, this I am assuming is the "side chain" and provides the control voltage for the dbx vca. I have a replacement for this one, but I don't think that should have an impact, or could it??????

I don't have a data sheet for the orig part, but the none H2 part and this one both have a 15v operating limit, the circuit is designed for 12, and I'm pretty sure that aspect is OK, and the chip takes it Vcc directly from 12.6 v source off the muting board, and I checked the voltages before.

I am in the process of setting up a PC based O'scope & distortion analyzer (found some appropriate freeware for that and an old laptop that will do nicely) for the calibration, but I don't have that yet.

 

[jpmorris]

This is a completely stupid suggestion, but - you're quite sure the new chip is in the right way around, yes?

 

[Blue Jinn]

This is a completely stupid suggestion, but - you're quite sure the new chip is in the right way around, yes?

Not stupid at all, and I didn't check that before I installed just assumed that the "N" would be 1 and the writing would be on the same side. After a good self I am assuming that the I character on the board is "1" and that the notch on the orig chip is for pin 1. (They are 8 pin in line) the NTE chip has a beveled side which is the "front" and "1" is to the left, so I am certain the orientation is correct. I'm going to get a better dvm and see if the control voltage coming off the RMS detecter is in the 100mV range next.

 

[evm1024]

Not stupid at all, and I didn't check that before I installed just assumed that the "N" would be 1 and the writing would be on the same side. After a good self I am assuming that the I character on the board is "1" and that the notch on the orig chip is for pin 1. (They are 8 pin in line) the NTE chip has a beveled side which is the "front" and "1" is to the left, so I am certain the orientation is correct. I'm going to get a better dvm and see if the control voltage coming off the RMS detecter is in the 100mV range next.

post photo....

 

[Blue Jinn]

post photo....

I will post a photo tonight, but I am pretty sure the orientation is correct. I looked at the datasheet for the NTE part and the way the chip is bevel cut indicates the front. http://www.nteinc.com/specs/1700to1799/pdf/nte1794.pdf

The bevel side is facing the same as the orig chip which has a notch on the left as you're looking at it. ( I don't have any of this in front of me at teh moment.) ALso, the "N" character is in the same spot as the "N" character in the datasheet for the upc1252h, which is IIRC the previous version of the chip.

 

[jpmorris]

If the notch is in the same orientation as the old chip, it's probably right. However, the two times I've managed to overheat a chip for no obvious reason were both because I reversed the pinout. I could just imagine them printing the text upside-down relative to the original chip and you forgetting to check the pinout.

The most memorable time I screwed this up was when I installed a new memory chip on my brother's video card in the late 1990s. I didn't check, assuming that the socket was non-reversible. The display went wild and the chip was scorching hot when we removed the board. Remarkably it still worked afterwards.
I seem to recall doing something similar to some kind of melody generator chip. Again, it worked when reoriented, somehow.

 

[Blue Jinn]

If the notch is in the same orientation as the old chip, it's probably right. However, the two times I've managed to overheat a chip for no obvious reason were both because I reversed the pinout. I could just imagine them printing the text upside-down relative to the original chip and you forgetting to check the pinout.

The most memorable time I screwed this up was when I installed a new memory chip on my brother's video card in the late 1990s. I didn't check, assuming that the socket was non-reversible. The display went wild and the chip was scorching hot when we removed the board. Remarkably it still worked afterwards.
I seem to recall doing something similar to some kind of melody generator chip. Again, it worked when reoriented, somehow.

I looked at the datasheet for this one and it will work up to 75 degrees celsius and it probably wasn't quite that hot when I cut the power, so I'm going to guess it is still OK. The chips themselves are pretty cheap anyway, but the PCB is not very forgiving of all teh desoldering.

 

[Blue Jinn]

Ethan:

Couldn't get the camera to focus close enough last night, will try again when borrow better camera. Did double check the orientation, and positive the orientation is correct. I will double chekc the solder joints too.