Help Troubleshooting Millenia Media HV-3B PLEASE

[moamps]

Ok, turn off the power, take out all the ICs, and test the input transistors as diodes. Do you have scope at hand?

 

[sonolink]

I have this.

 

[sonolink]

But I'm not sure I know how to use it to measure transistors though...

 

[moamps]

Could you test diodes (power should be off) on both channel according the picture, also test all diodes (zenners too), compare readings on both channels.

 

[sonolink]

I can use a normal multimeter with the diode position right?

 

[moamps]

Yes, you can use DMM in diode measurement position. You need to use a continuity meter to check the connections between the points. It could also happen that when removing the original ICs and installing the sockets for them, PCB damage occurred and that there are no thru-hole joints in some places. It would be great if you could send me to PM high res focused images of the top and bottom so I can better analyze where each track is going.
During the lunch break you can turn on your OWON scope and measure the voltage waveforms of the correct channel at points U3,4 pin 6 and U8 pins 1 and 7 when you are speaking into mike. That way you will know what and where you need to get on the channel you are repairing.

 

[sonolink]

From left to right on the pic:

First diode reads 0.68v and OL the other way around.

2nd reads 0.7 either way.

3rd reads 0.66 and OL

4th reads 0.68 and OL

This is the faulty channel of course.

Channel 2 (working one) readings:

First diode reads 0.68v and OL the other way around.

2nd reads 0.7 either way.

3rd reads 0.7 either way

4th reads 0.68 and OL

 

[sonolink]

I have checked the other diodes on the board (grey ones and orange ones):

D2, D3: 0.75/1.9 both channels same reading
D4, D5: 0v both ways. Both channels same reading
D6, D7: Ch1(faulty) 0.75/OL, Ch2 0.75/1.8
D8, D9: 0.63/OL

Cheers!
Sono

 

[sonolink]

I've used the scope and here's the results for U8 pins 1 and 7 of the working channel:

 

[sonolink]

I tried doing the same for U3 and U4 on pin 6 but the scope doesn't detect anything at all...

 

[Ike Zimbel]

I tried doing the same for U3 and U4 on pin 6 but the scope doesn't detect anything at all...

Its says "Osc Option" on your scope. Does it have an oscillator output? If so I recommend using that as a test source instead of a mic as it will give you a more consistent test signal. If not, anything that you have that will generate a tone would still be a better test source, IMHO.
Lot's of good advice in this thread, so far.

 

[ccaudle]

I'm not sure I know how to use it to measure transistors though...

The base-emitter junction of a transistor is a diode, so in the forward direction (base more positive than emitter for NPN, emitter more positive than base for PNP) it should measure like a normal diode, in the 0.6V to 0.7V forward, and no conduction in the reverse direction (although don't get fooled by resistive paths that may exist on the board if you are trying to measure in circuit).
The base-collector junction is also a PN junction, so acts like a diode, it should conduct if the base is more positive than collector for NPN (not the usual condition in operation), but more importantly should not conduct from collector to base if the collector is more positive, and should not conduct from collector to emitter if there is no bias on the base. You basically want to go all around the pins on the transistor and make sure there are no shorts across what should be a reverse biased PN junction. Again can be difficult in circuit because of resistors on board confusing things, but when the circuit is powered is relatively straight forward to check the voltages to make sure collector is significantly more positive than emitter and base, base should be around 0.6V to 0.7V more positive than emitter. Of course you have to trace out the connections to figure out which pin is which on the transistor pack to be able to do that.

the element mentions 16 pins on the top but there are only 14 pins underneath

Just a guess, but that might indicate that the original part used was 16 pins, but the board had to be redesigned later due to part obsolescence to use a 14 pin part. Analog Devices has a matched quad in a 14 pin package (although currently only SMD it may have been available in 14 DIP in the past), and That Corp has a matched quad also in 14 DIP. I saw past quoted talking about a matched octet of transistors for the front end, but I can't find any current 8-pack available, only quads, which is what makes me suspect the input circuit probably had to be redesigned at some point.

 

[ccaudle]

This is the basic topology of the input stage. On the Millenia there may be multiple transistors in parallel to reduce noise voltage, but the general idea is the same. The signals labeled as + output and - output then go to a pair of diff amps (op-amps with four resistor configuration to convert differential input to single ended output). The front end gets exposed to more abuse, so seems the more likely place for failure as compared to the diff amps.
Not shown are the protection diodes, which as the manual pointed out are a possible source of failure.



Note that this picture is for a single supply operation, with bipolar supplies the non-inverting inputs of the front end op-amps may be tied directly to the 0V ground net instead of getting a bias voltage from a voltage divider as shown in this picture.

 

[Ike Zimbel]

I just remembered something that might be relevant: Does this unit have the HV switches on it? Those are a well known point of failure on Millennia pre-amps, especially if you have a unit that does not have the HV option fitted. If it does, try cleaning and then exercising the HV switches and see if your results change.
"Beware the switch you never use!" Me.

 

[JohnRoberts]

I will check if it's a multilayer but I think at most it'll be a double layer.

I have checked the opamps. And they are all filed!! One of the heatsinks has been GLUED to the opamp....this unit has been clearly manipulated in the past...

The thing is that on the good channel opamps are also filed except one: an NE5534. I don't have any spare ones. I wonder if I could use a NE5532 or a TL72 just to check if it's the faulty one...

the 5534 is a single op amp so different pin out than 5532/072 duals.

JR

 

[sonolink]

Its says "Osc Option" on your scope. Does it have an oscillator output? If so I recommend using that as a test source instead of a mic as it will give you a more consistent test signal. If not, anything that you have that will generate a tone would still be a better test source, IMHO.
Lot's of good advice in this thread, so far.

Hi Ike,

That label stands for "Oscilloscope Options" really.
Thanks for your input

 

[sonolink]

I just remembered something that might be relevant: Does this unit have the HV switches on it? Those are a well known point of failure on Millennia pre-amps, especially if you have a unit that does not have the HV option fitted. If it does, try cleaning and then exercising the HV switches and see if your results change.
"Beware the switch you never use!" Me.

This one hasn't got the 130v bit. The board is prepared for it but the components aren't present for 130v operation

Cheers
Sono

 

[sonolink]

This is the basic topology of the input stage. On the Millenia there may be multiple transistors in parallel to reduce noise voltage, but the general idea is the same. The signals labeled as + output and - output then go to a pair of diff amps (op-amps with four resistor configuration to convert differential input to single ended output). The front end gets exposed to more abuse, so seems the more likely place for failure as compared to the diff amps.
Not shown are the protection diodes, which as the manual pointed out are a possible source of failure.

View attachment 93610

Note that this picture is for a single supply operation, with bipolar supplies the non-inverting inputs of the front end op-amps may be tied directly to the 0V ground net instead of getting a bias voltage from a voltage divider as shown in this picture.

Thanks for all that info ccaudle

 

[moamps]

Check that the pins are connected according to the figure, then that pins 2 and 3 of the XLR are connected to the pcb.



Find the zenner diodes on the PCB and take off one leg each, then try the preamp. As Ike suggested, try to find a signal source that you will connect to the XLR input as a test signal. It can also be a signal from a sound card, in some software you turn on the generator at about 440Hz (phantom shold me off).

 

[sonolink]

Check that the pins are connected according to the figure, then that pins 2 and 3 of the XLR are connected to the pcb.

View attachment 93620

You mean check continuity between the pins you marked on the pic, right?