Denon AVR-1912 E2/EA

[Michael Studio1]

Hang in there Michael, you will succeed with the help of the hugely knowledgable guys helping you. Just be methodical and keep copious notes of the results of your actions.

Regarding the track lifting, maybe up the temperature of your soldering iron and do put some rosin and extra solder on the solder point you are working on

Thanks Bushtech. I'm very grateful indeed for the prospect of assistance in my quest to repair this seemingly un-repairable Denon AVR-1912 E2/EA
Please note there seems to be a discrepancy in the Denon Service Manual for this AVR. Page 168 pinout of IC801 does NOT agree with 801 pinout on page 114 AUDIO_VIDEO UNIT (1/3)
Soldering: I have:
1) an 'HV/K badged 858D which I haven't used much, Three soldering irons:
2). an 'old' 25W.soldering iron with no temp controls. I need a new one as the tip is getting filed down to its point of connection - and won't remove.
3). 20W. 3/16" tip which doesn't seem to hold solder and is reluctant to heat to melting point.
4). Weller 12W.'Micro-Point' which I haven't really experimented with much - it never seems to be large enough.
What is the best iron wattage - presume it should be temperature controllable. I use a bit of sponge in a plastic plant-pot saucer (which I keep wet and covered when not in use.)
5) a 40W. pump-action desoldering iron.
A small non-heated pump-action desoldering tool.
Solder: Bobbin of Lead-Free Silver Solder - it doesn't say whether it has a flux core.
Bobbin of 22swg flux-cored 60/40 Sn/Pb
Flux: A small screw-top tub of 'Powerflow Flux (idela, it says, for lead-free solder. I apply it with wooden tooth-picks. Also Liquid Flux - which I use on re-flowing Computer Mo-Bo's.
Test Equipment:
Meters:
1). The usual old pre digital Analog type
2). an old Thandar TM351 Digital Multimeter - this was 'state of the art' originally.
3). UNI-T UT61E 22000 counts True RMS Multimeter dvm - this has connection for read-out on the Lab Computer
4). HoldPeak HP-9804 this does not automatically switch to a range of Volts/Amps/Ohms/AC/DC but the dial lights!
5). Digital Temperature Meter. I bought this to check the 858D hot air soldering unit.
6). Peak ESR70 Cap. tester.
7). Peak DCA55 Transistor Tester.
8). Oscilloscope: battery operated pocket-sized DS201 pro. Open Source 200Khz bandwidth 1Ms/s sample rate.
9). FeelTech FY3200S Dual Channel Arbitrary Function Signal Generator/Counter with BN5 out/input sockets.

So that's the Lab's equipment. I've listed it just in case it needs printing to help anyone advise on my best use of it!

Right nor the Denon AVR-1912 E2/EA has two horrid problems -
a). reversed polarity - the A+8v. rail reads A-8V. and the A-8V. rail reads A+8V. !!! Why?
b). the A+8v. is arriving at IC801 pin 80 (AVCC) measured at the Cathode of D800 as -7.41V - well, that is, in a way, an improvement for it had been reading much lower! So maybe things are improving if I can get the polarity corrected!
Michael Studio1 UK 11:25GMT

 

[bushtech]

That's certainly more equipment than I have. I use a 50W temp controlled soldering iron. Runs from 150 to 450. I find 350ish good for soldering and dial it up to about 450 for problematic desoldering.

I try to help where I can but I'm still a long way from knowing whats going on. For that you need the gurus.
Over to the experts again.

 

[Michael Studio1]

That's certainly more equipment than I have. I use a 50W temp controlled soldering iron. Runs from 150 to 450. I find 350ish good for soldering and dial it up to about 450 for problematic desoldering.

I try to help where I can but I'm still a long way from knowing whats going on. For that you need the gurus.
Over to the experts again.

Thanks for your interest and your reply, Bushtech. The Guru who occasionally pops up in my thread is 73's de Edd - and I always look forward to his help. But I must find the reason for this polarity issue otherwise the whole project is 'toast'. It has been suggested that maybe even the IC801 has gone up in smoke - not those exact words, but that same meaning. So I must get back to the basic Power Supply. BTW I am using what I call 'Tagged Test Tails' - jumper wire which I solder to the solder side of Connector Block pins in order to test voltages and tones at points otherwise inaccessible - due to the AVR not powering up when any of the PCB's are removed. If you look at page 43 in the Service Manual, they refer to a 'Jig for Servicing' It comprises just a few lengths of ribbon cable with connector blocks to fit the connectors on the PCB's. First Denon said "We do not sell this Extension Unit Kit to other than bona-fide electronic engineers. Subsequently they offered me a Kit at the astronomic price tag of over £300. So I didn't answer that and continue with my Tagged Test Tails. I bought a bunch of electrical connector strips and mount the ends in them, duly tagged and identified with Connector and Pin No. This works very well - particularly around IC801 which is on the ground floor and virtually inaccessible. However I also have an Endoscope which, with its ring of powerful LED's around the camera lens, allows me to see the sharp end of my probe on the Lab Monitor. In use it's rather like riding a bicycle by sitting on the handle-bars - like Paul Newman in 'Butch Cassidy and the Sundance Kid' - nothing goes in the expected direction - and then it goes too fast. But with applied skill one can manage! It's called 'Android and PC Endoscope' and I can thoroughly recommend it. I originally bought it to investigate the repairs done to my very old German Lions' Head 'Cello. Repairers marks were inside the body. Very interesting. Now I shall continue to unravel the mysteries of this AVR - and, hopefully, discover the reason for the reverse polarity! Fingers X'ed! Michael Studio1 UK 12:33GMT

 

[Michael Studio1]

Readings from Tagged Test Tails all reference REG_cnt 1/4 AGND CN101 pin 5 page 110:
SIDE CONNECTOR page 111 CP13A pin 7 should read A+8V. Actual reading: -8.158V
" " " CP13A pin 6 should read A-8V. Actual reading +7.829V.
" " " CN14 pin 15 should read A+8V. Actual reading -9.049V.
REG_CNT 1/4 page 110 CN101 pin 1 should read A+8V. Actual reading -14.175V.

I really do not know where to go from here. Please? Michael Studio1 UK 15:55GMT

 

[Michael Studio1]

A propos Solder. N.B. from page 93 to 103 of the Service Manual, boxed at the bottom of each page is a note about Soldering. It says: Lead-Free Solder. When soldering, use the Lead-free Solder (Sn-Ag-Cu) My bobbin from Maplin Electronics S63 5DL UK specifies this as Code N32AR with a composition: Ag.4%, Sn 95.5%, Cu: 0.5%. I am currently finding IC 106 - it's on page 97, REG (COMPONENT SIDE) and with my magnifying head-band on I can just make out which one is IC106 - odd isn't it, that the last two IC's are not in numeric order! I shall now reassemble the AVR PCB's and test IC106 with a DVM - my long pointy probe can get at these IC pinouts quite readily. First I must see the Spec Sheet to find out what I should expect to see. The Parts List (page 180) gives it as an NJM7808FA. Here I find I have pencilled a margin note: A+8V. CN13A pin 7. But we already know this shows -8.158V. (see above) so the problem must emanate from further back in the power supply circuitry. Nothing as yet tells me that IC801 is fried. Michael Studio1 UK 18:29GMT

 

[Tha fios agaibh]

page 110 CN101 pin 1 should read A+8V. Actual reading -14.175V.

I really do not know where to go from here. Please? Michael Studio1 UK 15:55GMT

Sounds like a rogue ground fault on the positive side. In other words, when you read from Gnd to pin 1, you are reading the positive voltage on Gnd in addition to the negative side back feeding through the components.
Because you have both a pos and neg rail, your Gnd reference will be giving you all kinds of confusing voltages as long as the rogue ground exists.

To find something like this without first isolating portions of the circuit will be a challenge to say the least.

 

[bushtech]

Have a look at 73's de Edd's VCC block diagram in post #21. There you can see the 2 rails going to the two IC's giving + and - 8V and you can also see where you can find a ground point which should not give you confusing answers

 

[Michael Studio1]

Sounds like a rogue ground fault on the positive side. In other words, when you read from Gnd to pin 1, you are reading the positive voltage on Gnd in addition to the negative side back feeding through the components.
Because you have both a pos and neg rail, your Gnd reference will be giving you all kinds of confusing voltages as long as the rogue ground exists.

To find something like this without first isolating portions of the circuit will be a challenge to say the least.

Hello John - thanks for your suggestions. I happily take up the challenge! However there was a note from 73's de Edd (see this conversation page 2) that this was 'the initial RAW DC supply voltage from which the + 8 VDC is to be derived AFTER it comes out from the IC106 . . . a 7808 linear regulator. At the time 73's de Edd made this comment, the voltage I was seeing at CN101 pin 1 was +10.78V. The really confusing issue is that of the Polarity change. 73's de Edd was suggesting that, due to the low Voltage reading I was getting at CN14 pin 14, (then it was +4.589V.) there was the possibility of a faulty IC106. Maybe now I have not one, but two faulty IC's - 106 and 107 - these being the regulators for the (+)8V. and (-)8V. rails. Unfortunately my Peak DCA Transistor Tester does not test a complex IC like these NJM7808's. BTW I do believe the suffix 'FA' refers to the long-tailed form these particular IC's take - for these are chassis mounted and not SM types. Mouser have them if these are indeed faulty. Michael Studio1. UK 10:46GMT

 

[73's de Edd]

Sir Michael . . . . .

I just didn't initially include all of the +8 and -8VDC supply connections terminations.
In this one, the top center is where the + and -8VDC supplies have their origins at their respective regulators.
I just can't perceive of the misread of polarities . . . . start there and move across to the connectors to the left and then down and across.
Notice at the last top right connector, that you have to transport your ground, since there is no Agnd connection on that connector.

REFERENCE . . . . .

73's de Edd
.....

 

[Tha fios agaibh]

..this was 'the initial RAW DC supply voltage .
The really confusing issue is that of the Polarity change.
Maybe now I have not one, but two faulty IC's - 106 and 107

Even if the raw supply voltage was coming through, it would still be the correct polarity.
The voltage you read can't be higher than the source. Therefore your reading has to be the combination of + and - voltage.
Don't think of it as the + rail suddenly switched to negative, think of it as the negative rail feedback back around like a loop and your meter is adding the positive voltage that's now on Gnd.

 

[Michael Studio1]

Even if the raw supply voltage was coming through, it would still be the correct polarity.
The voltage you read can't be higher than the source. Therefore your reading has to be the combination of + and - voltage.
Don't think of it as the + rail suddenly switched to negative, think of it as the negative rail feedback back around like a loop and your meter is adding the positive voltage that's now on Gnd.

Thanks, John, for your reply. I agree 'the voltage can't be higher than the source' - this is what threw me. There is something odd here - and I'm trying to find out what it is. I have a message from 73's de Edd which you should be able to see above. I wonder how I 'transport my GND' as he says. This is a term I am not familiar with. But he does go on to say: 'this has to be done as there is no AGND on that connector' - this is, of course what I have been doing - using the AGND of any convenient connector, assuming all AGND's are at the same potential. - or am I wrong? These reverse polarities are taken at the very same points as before, from the same pins in the same connector blocks as before - 'before' being when +7V was +7V. and +8V. was +8V. and the -8V. and -7V. were tested likewise. Is there a way the Power Supplies can be powered up without that nest of PCB's being in place? At this very moment the HDMI, AV, SIDE_CNT, FRONT_CNT pcb's are all out of the chassis. I no longer remove CX4 FFC cable - but drape it and the HDMI board over the Front UNIT when inspecting the pcb's or adding 'tagged test tails' to them. I do hope 73's de Edd isn't tearing his hair out at my inept procedures! MIchael Studio1 UK19:03 (and a good year that was. King Edward VII was just on the throne)

 

[Michael Studio1]

Sir Michael . . . . .

I just didn't initially include all of the +8 and -8VDC supply connections terminations.
In this one, the top center is where the + and -8VDC supplies have their origins at their respective regulators.
I just can't perceive of the misread of polarities . . . . start there and move across to the connectors to the left and then down and across.
Notice at the last top right connector, that you have to transport your ground, since there is no Agnd connection on that connector.

REFERENCE . . . . .
View attachment 39756
73's de Edd
.....

I am extremely delighted to read your reply, 73's de Edd. Thank you! My readings - the ones which showed the 'reverse polarity syndrome' were all taken from the same reference points as before:
All voltages Reference to AGND from CN101 pin 5
A+8v. at CP13A pin 7 SIDE_CON B'D page 111 reads -8.158V.
A-8V. at CP13A pin 6 " " " reads +7.829V.
A+8V. at CN14 pin 15 " " " reads -9.049V.
A+8V. at CN101 pin 1 REG_CNT 1/4 page 110 reads -14.176V.
To check this all again I shall re-assemble the PCB cage and power up the AVR.
I have done this and the readings are much the same as above. i.e. where it should be + it is - and where it should be - it is + Neither can I perceive of mis-read polarities, but the readings are reverse polarities.
As you can see, I have had to 'transport' AGND from CN101 pin 5. for reference.
Now I shall measure the Voltages of the 5 NJM7800 series on the REG_B'D in their
order: IC103, IC104, IC105, IC107, IC106. Where of the three pins, pin 1 is 'IN' and pin 3 is 'OUT'. pin 2 is GND.
IC103 pin 1 -13.968V pin 2 Video_GND pin 3 -8,709V,
IC104 pin 1 -3.953V pin 2 Video_GND pin 3 +1.41V.
IC105 pin 1 -8.168V pin 2 AGND pin 3 -5V.
IC107 pin 1 -0.002V pin 2 AGND pin 3 +7.8V.
IC106 pin 1 -14.358V pin 2 AGND pin 3 -8V.
I hope this is making sense? Michael Studio1 UK21:43GMT

 

[Michael Studio1]

Even if the raw supply voltage was coming through, it would still be the correct polarity.
The voltage you read can't be higher than the source. Therefore your reading has to be the combination of + and - voltage.
Don't think of it as the + rail suddenly switched to negative, think of it as the negative rail feedback back around like a loop and your meter is adding the positive voltage that's now on Gnd.

Thanks, John, for your reply. I agree there is something odd here - and I'm trying to find out what it is. I have a message from 73's de Edd (which you should be able to see above). I wonder how I 'transport my GND' as he says. This is a term I am not familiar with but since I am referencing to an AGND in a different connector - this must be what he is referring to.

Have a look at 73's de Edd's VCC block diagram in post #21. There you can see the 2 rails going to the two IC's giving + and - 8V and you can also see where you can find a ground point which should not give you confusing answers

Thank you, Bushtech, I have just posted my current findings at +8V. -8V. and the IN and OUT at IC's 103/4/5/7/6.
I am showing what I find and then I shall see what the points 73's de Edd made actually show as. Right now it's late in UK and I am being told to finish . . . . Michael Studio1 UK 21:54GMT

 

[Tha fios agaibh]

I wonder how I 'transport my GND' as he says. This is a term I am not familiar with. But he does go on to say: 'this has to be done as there is no AGND on that connector' - this is, of course what I have been doing - using the AGND of any convenient connector, assuming all AGND's are at the same potential. - or am I wrong?

I just take that to mean; move your black test lead to the proper AGnd reference. (As you are doing). It's a good idea to try different testpoints of (Agnd) to verify their all reading the same. Perhaps another dodgy ribbon cable is throwing a monkey wrench into the mix?

I believe your best approach may be to forgo the idea of interconnecting the boards at this point, until you can pinpoint the location of your ps fault.

If you have a regulated power supply, now would be an ideal time to take each board on its own and inject your own +7vdc and Gnd to the appropriate Cn and check that its path is on the straight and narrow.

This method should steer you to the board that is giving you trouble in a more straightforward fashion, as opposed to playing wack a mole with strange + and - voltages popping up all over the place.

 

[73's de Edd]

If you read improper polarities right at the 7000 series voltage regulators, and if you are using an autoranging meter, will you now look down and check the leads plugging into the meter to see if they might be reversed there ? ( ¿¿¿ BLUSH-BLUSH ??? )
Then . . . pull out a standard 9V rectangular battery and inspect the battery contacts vewy-vewy cawefully . . . .a la Elmer Fudd .
Out-ee is being the battery POSITIVE
and
In-ee is being the battery NEGATIVE
And see if that then reads right on your DC metering.

And yes . . . .the connector of the:
RC5/M-X PORT
REG-CNT 4/4 B3
does not have an Agnd pin connection on it, so you would rely upon a negative metering connection being from one of the other modules A gnd connections.

73's de Edd
.....

 

[Michael Studio1]

I just take that to mean; move your black test lead to the proper AGnd reference. (As you are doing). It's a good idea to try different testpoints of (Agnd) to verify their all reading the same. Perhaps another dodgy ribbon cable is throwing a monkey wrench into the mix?

I believe your best approach may be to forgo the idea of interconnecting the boards at this point, until you can pinpoint the location of your ps fault.

If you have a regulated power supply, now would be an ideal time to take each board on its own and inject your own +7vdc and Gnd to the appropriate Cn and check that its path is on the straight and narrow.

This method should steer you to the board that is giving you trouble in a more straightforward fashion, as opposed to playing wack a mole with strange + and - voltages popping up all over the place.

Thank you, John, for this sage advice. I am truly glad to have your ideas and methods of approach. Now, the main thing seems to be an adjustable DC supply. I can make one up from the following: Two series connected 110V. Variacs (brought back with me from Vancouver BC.) and one heavy duty home-made Full Wave rectifier. This is a small, very simple metal box containing four heavy-duty screw-in diodes and a 16μF 450V.dc Electrolytic Capacitor, together with a pair of high amperage input cables and another of output cables.
I have just made an AWFUL DISCOVERY. - there's a fault with my dvm. This is where the problem lies. I checked a 9V. re-chargeable battery which was just off charge - only to find this meter showed it as -9.6V. SO there you have it. I need a new dvm - unless I can find out what is wrong with it and repair it. Apologies to all!! Michael Studio1 UK08:21GMT

 

[73's de Edd]

What meter / brand / model is that ?

 

[Michael Studio1]

What meter / brand / model is that ?

My HUGE apologies about this! Thank you for your question: The dvm is: (or was?) a UNI-T, model UT61E where one feature is its ability to be sensed and its read-out shown on the Lab Computer screen. Michael Studio1 UK

 

[bushtech]

Meter's battery OK?

 

[Tha fios agaibh]

My apologies for not suggesting that you verify your oddball voltages with another meter.
Congrats on discovering it.