JVC DR-M10 Loading and Repair Question

[oyster]

oyster,
I would love to know if your M10 now has R5327 and R5328 on the regulator board changed to 1,000 ohms either with 1/4 or 1/2 watt parts. Yours being a late model should have the "loading" fix in it already,

By the way, mine has s/n: 099B6403. Maybe DR-M10 owners could post their serial numbers here..

but I would not go so far as to have you void your warranty. The warranty is 90 days I believe.

Here in the EEC the warranty is 24 MONTHS!!

[emlsnws]

I've looked at the schematic as I can't power the board while probing the two resistors I found earlier this evening. And I don't have any thin wire to attach and then lead out of the digital can. It seems that -29V does not even appear on the digital board (correct me if wrong) so I think that's a dead end.

I started thinking about R5327 and R5328, why would they be touched during a factory mod? Initially they look innocuous. I checked the value of DV12V (supply to dvd drive, 12V). It occurred to me what would happen if the transistor Q5313 was not driven hard enough? Answer: DV12V would drop. which would be "a bad thing" if the drive were in writing mode, and (say) seeking at the time, when current demand on 12V peaks during that time.
So with my unit in dubbing mode, when any drain on 12V would be highest, I saw 12.4-12.2V on Q5313 emitter, and 12.1-12.0V when seeks occurred. Drops were definitely noticeable when the drive seeked (sought?). So I closed in and measure the C-E voltage, result 40-50mV when not seeking, and temporarily read 60mV when seeking. This was on a multimeter, which would not respond to fast transient peaks at all (perhaps why some of the numbers don't add up). The B-E voltage was -0.77V to -0.78V even during seeks. (DV5V is ok at 5.04V, and 5.07V when seeking. Yes 5.07V)

My unit had not been on long and was running with top cover off. Given heat buildup over time, and a helping of chance (perhaps a bad area on the disk requiring repeated seek operations) perhaps the DV12V rail would drop to the point where the drive had an error, in writing or in data transfer? Then the firmware has a panic and crashes.

Perhaps a hare-brained idea but there you are. If it were my design I would use a power transistor for Q5313 to minimise voltage drop and thus heat dissipation in it, but we see a small device in place there (2SA1585). The same as used for the fan. The net shows the transistor to have a 1A peak Icc. I always thought DVD writing drives had high current demands.

trhouse, what do you think? I think maybe I will look at these parts (they're on the concealed side of the regulator board) and check for evidence of heat there. Then perhaps try a higher powered transistor for Q5313, or adjust the resistor values to drive it harder. Perhaps that's the factory mod we are seeking? Though with b-e at 0.77V that seems ok to me, perhaps it's Vcesat at high Icc is the issue.

[trhouse]

The schematics for the M10 and MV1 must be slightly different. The M10 has three sources of +12V. There is an unregulated +12V labelled AL12V. This is the line that supplies the fan plus some other things but does not go to the dvd-ram drive. There is the unregulated +12V to the dvd-ram drive labelled DC12V. Finally, there is a regulated +12V labelled SW12V.

In the M10, the transistor supplying the unregulated +12V to the dvd-ram drive is Q5316 ( 2SA1585S ). This device acts only as a switch to turn-on the +12V. Even with 60mV for Vce and one ampere current, that is only 60 mW dissipation. I would not be concerned.

R5327 and R5328 are on the power supply board in the M10 not the digital board. They along with zener D5302 provide the only -12V in the M10. It is labelled AL-12V. It is derived from the -29V. These resistors dissipate power as long as the M10 is plugged in because the -29V and AL-12V are not turned off in any mode. These parts may have different part numbers in the MV1. The M10 has no 1/4 watt resistors on its digital board.

I suspected these parts because CEF posted that JVC fixed his "loading" problem by putting in two 1,000 ohm, 1/2 watt resistors. The nearest values in 1/4 or 1/2 watt resistors to 1,000 ohm were R5327 ( 681 ohm ) and R5328 ( 820 ohm )in the M10. These resistors should not cause a problem if they are not defective but several sources have indicated the problem is related to bad parts.

[emlsnws]

trhouse, check your pm's for a link.

Yes I think last night I latched onto your M10 resistor part numbers, and ran with them, but these part id's are used for a different job in the MV1S. Still, I think the idea has possibilities and will try to prove it out to the end.

The MV1S has a -29V which is generated by R5326 and a zener+rectifier in series on the junction pcb. It goes to the flourescent display and generates the -7V rail, see below.
I noticed that to Mig-45 earlier in the thread, you named -29V as CN5301 pin 3. On the MV1S this pin is named "BT2" and goes to the tuner board to the 2 tuners (+30V is needed by each if the diagram is correct). The voltage is obtained from a (positive) diode reg from the main smps transformer. It is fed thru R5308 and R5309 (1k2 + 1k, which I need to check for signs of heat damage) and a switching transistor Q5303. Certainly this part differs from the M10 which uses -29V for the tuners, am I correct?

There is no AL-12V in MV1S, but there is a SW-7V which is used by some audio op-amps on the audio/da-ad board and the main(audio i/o) board. Interestingly there is no zener on the SW-7V line, but there are 2 series resistors back to -29V. Presumably regulation occurs by virtue of the voltage drop (and is not precise at all). These are R5501 & R5502 which are 1/2W (560R each fitted on the board). There is browning around these 2 on the pcb, I have pictures if necessary.
Interestingly the schematic shows these as 820R & 680R 1/2W which matches what you were thinking for the factory mod. I wonder what SW-7V would be at with 2 * 1K resistors? Lower. This needs measurements making on the board.

The 12V sources are: DV12V which is switched via a transistor (q.v.) from its own winding on the smps transformer. There is also M12V which operates the fan and other things. And SW12V which is M12V after a transistor/zener regulator, switched by P.CTL[H].

[emlsnws]

Quick update: SW-7V is at -8.3V when DVD playing.

[trhouse]

The cable company is working on the Internet connection at the moment. There are some differences between the MV1 and M10 as follows,

R5501 ( 820 ohm ) and R5502 ( 620 ohm ) are equivalent to the R5327 ( 680 ohm ), R5328 ( 820 ohm ) in the M10 except these are 1/4 watt parts in the M10 and 1/2 watt parts in the MV1. These resistors with 12V zener D5302 provide a regulated -12V labelled AL-12V in the M10. In the MV1 there is no zener ( at least on this schematic ) so they just lower the -29V to about -7V.

The power dissipation dropping -29V to -7V is 146 mW and 176 mW for the R5502 and R5501, respectively. I suspect that in the M10, these resistors were changed to 1K, 1/2 watt units to solve the "loading" problem. It is a little different in the MV1 because the -7V is switched off in power save mode where the -12V is not in the M10. The MV1 has no power in these parts in power save mode where the M10 does.

Here are a few other interesting differences. In the M10, AL5.3V is designed to be +5.75V but the MV1 has no AL5.3V, instead it has AL5.8V which is set to +5.9V. The M10 has capacitors in the system controller section only rated for 6.3VDC which has the +5.75 applied to them. Wonder if those capacitors are on the MV1 controller board? They would be very close to their limit if the +5.9V is applied the same way.

The design values for the D1.8V in the M10 set the output of the PQ5EV3 regulator to +2.03V. In the MV1, the design values set D1.8V to +2.48V. This is the voltage that goes to the DMN8652 in the M10 media processor section. The MV1 must have a way to drop that +2.48V down to about +1.8V otherwise the DMN8652 will be damaged.

In the M10, -29V goes to CN5101 pin 3, BT2 goes to pin 18. On the MV1, -29V goes to CN5304, pin 13 and BT2 goes to CN5301, pin 3. They are not on the same connector.

P.S. If the R5501 and 5502 have been changed to 560 ohms and the voltage at SW-7V is -8.3V, the current flowing is about 18.5 mA. That means the values on the schematic ( 680, 820 ) would lower the SW-7V to -1.27V.

P.P.S. In the M10, the maximum current from the AL-12V line that can be drawn is only 11 mA with the 680 and 820 ohm values. For some reason, the MV1 draws a lot more from the SW-7V. I was able to download the photos today.

[emlsnws]

Many thanks trhouse, your dedication is exemplary. I will think over the points and make measurements, but I have other priorities for the next day or so. I'll be back!

Recording and dubbing with DVD-ram went ok today.

[MiG-45]

Trying to catch reason of comparition for this two models. M1V have VHS module, and I guess it also use some voltages so their values are a bit higher then in M10. So and topology. Most interesting to measure voltages near the end of line.

As for me is now only one question - where is the problem of such a strange behaviour (green lines) - power supply or digi board. And now its seems more like problem is on diginboard, because voltages outgoing from powersupply in common are used in different parts of scheme. Except maybe 1.8V. So if there were problem with power supply it affects then not only video but other parts too.

[trhouse]

Mig-45,

Have you solved the problem with the thermal pad which is too thin? At this time, it seems to be the most likely source of problems by allowing the DMN8562 to get to hot.

Manufacturers typically introduce a recorder first, then the vcr/recorder combo, and last, the hdd/recorder models. Studying the units that go to production later can sometimes reveal changes in the designers thinking. For example, is it possible that in the MV1, D1.8V is +2.48V ( or is it a mistake in the schematic ) because they felt it necessary to filter that voltage more which then drops it down to +1.8V.

We never resolved how your +2.03V managed to drop to +1.86V. I think you already proved it cannot be the resistance of the pc board. There must be a part not accounted for in the schematic.

In your unit can you tell if R5327 and R5328 are what the schematic says they are ( 680 ohm, 1/4 watt and 820 ohm, 1/4 watt )? Also, what is the manufacture date on the back?

Oyster,

A two year warranty is too good to void. I would not open the unit much as I would like to know about those two parts.

[emlsnws]

Back again. I've been observing the voltages from the sw.reg board as the unit warms up. I have also made a change to the STRG6653 heatsinking, by removing the black vaned TO220 heatsink which I added to the 1.8V regulator way back (it didn't really need it), and screwing it on top of the STRG6653 heatsink. The unit still has the top cover off. The results:

Unit just turned on at the wall (cold). Exhibited usual "loading" for 1-2mins.
SW-29V @ -29.4V SW-7V @-7.7V then slowly rises to -8.0V (over 30s)
SW12V @ 10.5V (this one is interesting, I hadn't realised it was so low)
Ambient temp 25oC

Unit been at idle for 1 hour
SW-29V @-29.5V SW-7V @ -8.24V
SW12V @10.5V AL5.8V @ 5.87V (first time this was measured)
STRG6653 heatsink temp 43oC (used to be 49oC with oem heatsink)

After dubbing for 2 hours
SW-29V @-29.8V SW-7V @ -8.31V
SW12V @ 10.5V AL5.8V @ 5.88V
STRG6653 heatsink temp 42oC (used to be 51oC with oem heatsink)

Unit on standby overnight (power save on)
STRG6653 heatsink temp 56oC (this used to be >70oC with oem heatsink)

The SW-7V is interesting. If there is no zener on this rail, how is the voltage determined and why does it change? It may be internal protection diodes operating, their breakdown voltage varying with die temperature. After 30s to 1 hour the chip die warms up and the -7.7v increases to -8.24v.

The temperature readings are much more reasonable now. Combined with the vent holes in the top cover, when I refit the cover it should be cooler.
I still have to investigate the digital can, ie. does the thermal pad contact the shiny metal.
Good to see you back Mig-45.

[MiG-45]

Yeah, I was travel far from home - now I`m back )
So was unable to check if new thermal pad works - now M10 working and I will wait for next few days for results! )

[emlsnws]

Great.
I have now opened the digital can, and like yours, the thermal pad is not contacting the LSI chip. I might modify the limit stops on the can so that it can be made to contact.

Also regarding the SW-7V changing over the warm-up time, I see now that SW-29V changes also in approximately the same way. I will measure the drop across the resistor chain when I get the chance, as it could be that SW-29V is changing and the reason SW-7V changes is because the voltage drop across the resistors is constant. ie. it's not a problem.

[MiG-45]

Are we sure that designers idea was to use metalic shield as a heatsink ? Maybe they put thermal pad to avoid in any case contacting metalic top of chip ?

[emlsnws]

Definitely. Large ASICS dissipate lots of power, and the metal slug visible on the LSI one tells me that this is a "warm" one!
Thermal pads are usually/always electrically non-conductive so I have no fears about making the thermal pad contact the ASIC.
I will check this one with my meter to make sure!

[MiG-45]

Of caurse pad are non conductive thats why I make guess that it can be used to avoid electric contact and not to provide thermal contact.
Anybody can check if metal plate on top of chip is grounded ?

[trhouse]

emlsnws,

The SW12V is designed to be about +10.3V. The regulator uses an 11 volt zener minus the pass transistor Vbe drop.

There may be an error in the schematic for the -29V. It shows D5503 ( 27 volt zener ) in series with D5502. This would indicate -29V should be about -27.7V. Can you measure your D1.8V on the regulator board to see if the value calculated from the schematic is correct ( +2.48V ). That is quite a bit higher than the M10 value.

mig-45,

It would serve no purpose for the designer to dimple the cover to bring it close to the DMN8652 then use a pad to prevent it from touching. You may not have seen this, but some IC's have the thermal pad built-in to the package. This has been done for many years.

[MiG-45]

ok ok ...just thought "what if..." )

[Yngwie38]

I own a drm10 (made in germany) for almost a year now and never had problems. On day 1 i've turn the brightness of the display 1 step lower. A friend also has a drm10, after 3 months the problems began: the green lines and he had the loading problem. After turning the brightness of the display one step lower the problems are gone for more than 6 months now.

Greetings from the netherlands.

[MiG-45]

I used to have brightness lower for even 2 steps - Brightness 2 from the beginning and after 5 months problem with green lines appers! Also tryed to set brightness 1 (as your friend) but it doesn`t help.

But thanks anyway for information

[emlsnws]

I also am using brightness low (only 2 steps on DR-MV1) as it causes the annoying blue lighting to be off. With brightness on high, the blue lighting is on -- poor sense of style!

trhouse,
So there's nothing to worry about the SW12V being so low, good. I have rebuilt my unit with the digital can definitely contacting the LSI Logic ASIC.
Surprisingly I found the temperatures roughly the same when this was done, perhaps it was just that the heat was conducted into the can instead of warming the air inside it. The only data I have is that the top of the can was measured at 53oC (in a 29oC ambient room ... it was hot here on Friday). This is 24 degrees above ambient, and I was seeing 19-22 degrees above ambient with the can not contacting the LSI.

Furthermore I've now added a small heatsink on the dimple area. (A low profile green finned affair - the kind of heatsink you see on northbridge devices on motherboards). I now find the temperature to be 35oC on the can, with the top cover on, after hours of recording etc.
Also I have the DVD drive support bracket open on 4 sides to permit ariflow and don't forget the extra holes in the top cover (I can provide a picture of this later), so there is airflow in the side at the front, then over the digital can & exhaust via the fan. Fan is still at top speed (12.2V.... I measured it). This is a good outcome, as I know the heat from the LSI is going into the metal can and then into the air.
Don't forget, my D1.8V was reduced to 1.77V early on, it was my first post on page 5 of this thread. The original values for that reg. were 3.0k and 4.7k, giving 2.03V (according to formula, I don't think I ever measured it in practice) and I now have 2.2k and 4.7k (giving 1.77V measured).
I think the 2.48V (if that's what 47k & 4.7K give) is erroneous and units couldn't have been built like that. The 3.0k I see on my pcb is more likely, giving 2.03V.

I will check the voltage drop across the 27V zener, as it seems slightly out. What causes a zener to be inaccurate? Excessive current? Possible, as I see brown marks on the board around it....

I have not seen any more "loading" events, but I have a new problem with the vcr deck, it is reluctant to load tapes. Upon insertion, It doesn't lace the tape around the head, but ejects them immediately. Something new to work on! But recording from tuner has worked very well, and I have been playing tapes at the same time (=equivalent load on power supply?) using sometimes DVD-RAM and sometimes DVD-RW.

Soon I think I will slow the fan down to an acceptable noise level and see if the reliability is still there.

[trhouse]

That was my mistake. I looked at the MV1 schematic and thought I saw 4.7K and 4.7K when it actually says 47K and 4.7K. Yes, no unit could have been built with those values and as you mentioned the real values were 3K and 4.7K.

The DMN8652 will be much happier with those temperatures. The good thermal contact insures that the IC will not be too much warmer than the can temperature. Let's hope it solves mig-45's green lines too.

You are correct about zener current increasing the zener voltage. A perfect zener would not do that but they all have internal resistance.

[emlsnws]

The zener is dropping quite a lot more than the part's value of 27V. I also have made an error in the state of the -29V line, it is not turned off until the MV1 is unplugged from the wall. The SW-7V, as you'd expect from its name, is off in standby mode (power save=on). SW-7V is obtained by switching (via a transistor), some series resistors and the load consisting of op-amps & analogue switches, onto the -29V line.

The zener itself drops 27.4V when the unit is in standby (-29V line voltage -28.3V), rising to 28.2V immediately after switch-on (-29V line voltage -29.2V), rising to 28.8V after 2 minutes or so (-29V line voltage -29.6V).
It looks like it is passing a lot of current and the dissipation is thus high, then temperature affects its zener voltage.
The behaviour after switch-off is interesting too. Just as the fan keeps running for some 5-8 seconds after pressing the power button (at corresponding time to a "please wait" on the screen), the zener's voltage stays around 28.8V then after 5-8seconds it drops back to 27.4V. Perhaps this is related to the number of VFD segments alight.
Perhaps a power zener would be less subject to voltage changes with current drawn (as this is what I think is happening as unit powers up).

I have also slowed the fan down to its design speed & noise level, by reinstating the fan series resistor at 39R. The fan voltage is 6.4V, and temperatures with the (vented) top cover on, are as follows:
After 1 hour at idle, digital can temp was 40oC. After recording on DVD for 1/2 hour, digital can temp was 41.5oC. Back where we started.
I'm not best pleased with the amount of airflow from the fan running this slow, so I am going to try adding 39R across the existing 39R to increase fan voltage & speed, hopefully the noise level will still be acceptable.

[trhouse]

emlsnws,

We have a bit of a mystery. The way a zener works in this type of circuit, the zener voltage will be maximum when there is no load. The load in this case is the SW-7V and the displays. The zener current ( no load ) is set by R5326. This current must be greater than the required current when both SW-7V and the display are both active. We know that when SW-7V is active, the current draw is about 18.5 mA.

Let us make it a simple case that SW-7V is the only load. If the zener current is set to 20 mA. then this entire current will flow through the zener when SW-7V is not active. This will make the zener voltage maximum. When SW-7V is active, it removes 18.5 mA from the zener so only 1.5 mA is flowing through the zener. This low current should cause the zener voltage to be minimum.

You are measuring the reverse. The zener voltage is minimum in standby which means little current is flowing in the zener. If that is so, where is all this current going since SW-7V is not active? With the unit on, the zener voltage is higher means that less current is being drawn from the zener than in standby.

[emlsnws]

I hadn't got around to thinking about that yet - but yes, it seems odd. It's not different (ie. opposite) because here we're talking about negative voltages?? My detailed power supply theory is now a long time in the past.

I was thinking that the green lines Mig-45 and others have seen could be a result of the tuning voltage or high voltage supply to the tuner going away from its correct value, de-tuning the received frequency. On an analogue system we would see black & white noise in some lines of the tv picture, but in the digital domain, if it can't encode the noise (or detects it and substitutes an alternative) perhaps it settles on a green pixel. Some equipment uses a blue screen when no channel is tuned, I'm thinking along those lines.

[MiG-45]

Nah! Green lines has specific view! when I just started to analyzie sympthoms I noticed that they exists not only on output cascades, but also can be recroded to disc, and also appers when I`m whatching DVD. Also, they have a bit other structure then noise snow - left edge of the screen is there base when they achive their critical mass they bahves like graphic equalizer = pickes rising from right edge. And quantitiy of them depends on "CPU usage" on computing terms. Keep in mind that they don`t apper s soon and their quantity grows. So it more looks like over heating or capasitor leak.
Already five days of work and still no green lines .

[trhouse]

emlsnws,

It does not depend on polarity. The voltages are behaving as if there is a heavier load on the -29V in standby and less load when on.

MiG-45,

Have the green lines ever appeared when you were playing back a commercial dvd?

[MiG-45]

emlsnws,

Have the green lines ever appeared when you were playing back a commercial dvd?

Yep, any kind of them ! Even when playing audio CD the splashes on screen and when I whatching signal from outter source passed through M10

Btw, speaking about tuner, do you also have LG tuner in your M1V ?

[emlsnws]

The schematic shows LG/SANYO/ALPS are possible, and also Matsushita. I will have to check in my unit to see what it was built with.

So, have you re-assembled your digital can with the limit stops removed? I pressed mine out with a vice and a 1/4" square metal bar. LSI IC now is the first thing that touches the metal can, and the soldering is arranged to exert gentle pressure.

trhouse,
It is confusing indeed about the zener. It may be something, or it may be nothing and just a result of 'lax' design. It would be good to find a correlation with the problems. Perhaps running without VFD may eliminate something (there's a 4 pin cable that can be disconnected I believe).

I have now increased the fan voltage to 9.5V by putting 39R in parallel with oem 39R. The fan speed is about as fast as possible to preserve acceptable noise level. Digital can temp is 37-38oC after an hour of operation, with vented top cover on.
Last night I saw "loading" again some 6 minutes into a recording with vhs section playing a tape at the same time. The cause may have been media again, as it was one of the first lot of discs I have.
An odd symptom is that the unit seems to recover and commence recording again, ie. if I look at the navigation menu for a disc, after a 'loading' crash, I see 2 (or more) separate recordings with a small gap between. Bless it, it's trying to do its best and pick up the pieces.
So I changed to DVD-RAM and then it recorded 8 hours without a problem. I need to record the disc on a PC to check it out, then use it again in the MV1. Still thinking about that DV12V pass transistor......

[MiG-45]

Yeah I remove limit stops on can those are on sides and slightly change two of them in the middle to make solidthermal contact between LSI and top cover.
But its not enough - I have to change a bit the way it screwing to the main board. There is a trap - when you connecting digiboard to mainboard it fited tight on connector, but side legs of can limits and dont let it fit completly when digiboard are sitting deeper in can (as it is now) , so you have to work with can`s legs to make this contact good.

[emlsnws]

Oh yes, the middle ones. I had to cut those down with sidecutters to allow the board to sit deeper in the can.

I didn't notice a problem with the connector when re-mounting into the unit, but as they say, your mileage may vary...

And no green lines yet? That's good news so far.