Printable View
email sent, if I had realised how small the file was I would have sent it to you before anyway.
I also mentioned that we checked +FL but not -FL which I think are something to do with the filaments in the display.
C24 is worth checking as it connected between these to supplies and if faulty could cause the display problem.
Well, I’ve had a chaotic few days, but have found some peace and quiet this morning. However, I cannot report progress, in fact it looks like a step backwards.
First, I installed a new 470 10v cap at C21, instead of the temporary. Foolishly, I didn’t do to the resistance check before fixing it, but seems to be OK. I also had to take a couple of minutes out to do a temporary fix to the mains input as the socket had come off the board.
The result was where we were a few weeks ago – no +5V or +3VS on the output cable. So had to restore the link between R95 and R98, and then I had those voltages. No SATA power though, but that may be because my test HDD has failed.
Voltages on FL+ are 22.v and FL- 16.5v. I could check those at both the PSU and front panel ends.
I’m still getting 5.1v DC at the input to the front panel but no voltage on the two output pins. Curiously, and I discovered it by accident, if I put a probe on one pin only, I’d get a steady 3.0v AC, but as soon as I got the other probe on the other pin, the voltage ran down to zero.
Thanks, Stroker, for the flash tools. I’ve installed them, and then tried connecting the receiver and PC with the USB A-B cable. Of course, I’m working blind, but holding down standby and the right arrow for several seconds produced no reaction on the PC – so I never get the opportunity to load the drivers. Nothing shows up in Device Manager.
I am beginning to think we made the wrong decision when we tried flashing because we seem to have lost any signs of life we had previously. If we had done it with flash-tools the result may have been different! Anyway onwards and upwards there's a few things we can check while I come up with a new plan of attack!
1. On the display board I would like you to measure the voltage (to GND) on the first and last pin of the display itself, I expect this to be FL- and FL+ but I could be wrong.
2. Remove the battery/cell for a while and if you can remove the tuners - some people have mentioned problems with these. Later refit the battery but not the tuners power on and see what happens - no change I expect but always worth a try.
3. Check CE1, CE2 and ZD1 on the display board, you may be able to check the caps on-board otherwise remove them. ZD1 is a zener diode but can be checked in the same way as a normal diode - only 2 pins need testing, I think the 2 nearest the transformer but not sure. If you cannot test ZD1 try melting the solder to freshen it up but be careful that all 3 are not melted at the same time or it may slide of the pads.
4. There is a 20 pin header(connector with nothing connected) towards top left of the motherboard is there anything to identify it's purpose? I think it may be a JTAG connector (used for low level programming). Perhaps you can take pic and perhaps a close up pic of the 64 pin chip you posted recently.
From your post FL voltages are probably OK.
The 3V ac reading is probably just sent to try us. You could check continuity between the 2 pins on the transformer
It reminds me of a mains powered unit I was fixing a while ago there appeared to be believable voltages everywhere in the unit but it turned out the mains transformer primary winding was open circuit - fitted a new transformer and it was fixed - wasted a few hours on that one!
Just thought - when checking out the suspected JTAG connector please measure the voltage on each corner pin - this may give a clue what it's about. I expect one corner to be 5V or 3.3V another should be GND and the others two may be a bit random.
Sorry – only just got this posting. As to previous posting, if by the pins on the display itself you mean the long line of pins on the underside below the two sets of connectors, then there is no continuity between neither +FL (Pin9 on the long connector), nor-FL (Pin 10)and the first couple or last couple of the long line. Voltage to the long pins was 25.2Vor so. I tested several of them at random, and they came in the same, or a bit lower. At the same time +FL was reading 22.2V.
I had come across the possibility of faulty tuners, so I had removed both for most of the checks I’ve reported.
There is continuity across the two pins of the output of the front panel transformer.
ZD1 OK – one probe on the single pin and the other on the pin next to the transformer – 0.6V and infinity.
CE1 and CE2 left in situ, and both show increasing resistance and with probes reversed.
I should have mentioned yesterday that I tried both C24 and C25 on the PSU board while in situ, and they both passed the test.
Can’t make anything out of the twenty pin header – not helped by my magnifying glass going AWOL. Will try and upload some pics tomorrow I’ve come out without camera.
Thanks for the info - will digest later.
I have sent you an email with slightly different (USB) flashing instructions.
We still have the possibility of flashing using RS232 but I think this only allows flashing the boot loader.
It is usual for RS232 to work when nothing else does but you will need RS232 on your PC/laptop.
I am interested in this possible JTAG connector there would be a reset pin on it which could be useful to us.
I am having trouble uploading pic of what I am expecting so will send by email.
If we briefly short reset and GND we may get the box into a better state.
Looking at the 20-pin header from the front, the top left pin is GND. To the pin just below it there is 1.8v. There is no voltage to the pins on the right hand side.
Just thought I’d check voltages at the standby switch. On the middle pin of the bottom row, I get 5.1v, which is the +5VS voltage. However, I don’t see continuity from that pin to any of the other five, whether or not the switch is held down. There’s something like 1.7v on the other two bottom pins, and nothing on the top.
I also thought I ought to re-check the voltages at the output end of the PSU in case something had changed. In the course of that I found I did have SATA power. But, with front panel disconnected, reconnecting the PSU to the main board caused SATA power to be lost.
Attachment 66372Attachment 66372
Sorry second pic hereAttachment 66374
Need to think about your header reading - not quite what I was expecting.
Don't understand what's happening with the SATA supply - will look at that later perhaps.
With the standby switch I think the middle pins will be the LED and checking your post from 09 Jan (I think) the switch contacts are top and bottom corner pins.
I think you will have continuity between top and bottom corner pins with switch pressed.
I believe the flash is stored in the socketed 44 pin device to the right of the battery.
Do not try to remove this device yet as it can be easily damaged - there is a special tool to remove it but you can do it without the tool although you can damage the pins of the device and the socket if not done properly. In fact it's not much easier with the tool. Anyway you could try pushing the device down into the socket - sometimes they work loose and lose contact on one or more pins.
Assuming the switch is working I think we need to try blind flashing the boot loader I don't think we can make it any worse!
You need to rename the file to usb_uboot.img, I am thinking that the software for the display etc must be in the boot loader.
We should be able to use RS232 to flash as a last resort later anyway or at least I hope so. Flashtools can be used with RS232 but of course we have to sort out a cable and suitable PC/laptop to go down this route
Apologies for a longish silence – poor internet access, and, as you’ll see, a lot of confusion on my part.
I’ve had another look at the 20-pin header, and found the following:
Top row, even nos 2-20 – all GND.
Bottom row odd nos:
1 1.75v
3 unstable, hovering around 1.0v
5 as 3
7 1.75v
9 nil
11 nil
13 nil
15 as 3
17 1.75
19 nil, but once got 0.88v.
I also noticed one or two changes. I had previously measured voltages from the hotside heatsink – getting 333v to TH81, 21v to R98 or 11.6 with the link to R95in. Now, I got unstable readings from it. From the chassis screw, I get 325 and 7.6v to R98 and to R95 with the link in.
Well that’s what I found an hour ago.
Now, I repeat the check and get something different. I get stable readings from the heatsink, but not from the chassis. And I get no voltages at all on the odd number pins of the JTAG header.
Furthermore, I get a different picture near the heatsink. I’ve removed the link between R95 and R98. From the heatsink to R95 I get 1.26v, same to both ends. To R98 I used to get 21.6v, which would be present through to the cold side of R96. Now I get 1.27v to the hot side, 0.61v to the cold side and nothing to the cold side of R96.
Earlier today, I got SATA power with the main board disconnected; now I don’t.
As far as I can tell, I’ve done nothing to produce this change within the hour.
Now I have made a change. I removed the coin cell, removed both tuner cards and put back the cell. So check, No 3:
R98 comes at 21.1v – same through to the cold side of R96. R95 comes in at 0.4v on the hot side 0.9v on the cold side –yes, treble checked, that way round. Still no SATA and no voltages on JTAG. After replacing the link, there is 11.6v at R96 but still nothing at the JTAG.
Stroker, if your patience is beginning to get stretched, I fully understand.
If you’re prepared to stay with it one more day, I remember your first prime suspect was IC91. I do have a replacement, so could try that. Also replacements for Q3 and Q4 if that helps us get an AC voltage on the front panel.
Just an added point about the standby switch and its LED. When the system is powered up, the LED flashes for a minute or so and then stops flashing. During the flashing phase, the voltage at the lower middle pin alternates between 5.10v and 1.98v. When the flashing stops, it’s a steady 5.1v
Another noticeable feature now is that there is no voltage at the fan header. Previously, I’ve always had 11.9v there, and two days ago the fan even came on for the duration of one powering up. Also at the output connector (CN41) +8V comes in at 7.2v, whereas previously it was 8.3v.
I've only had a quick look at your post - I won't have much time today.
I would not rush into changing IC91 as long as you have 3.3V and 5V with the link fitted. As far as I can see the SATA 12v is only used for the hard drive - things to check are D41, C41 and IC41 - I like the way they have numbered the devices, trying to keep the numbers the same for everything in the same circuit. The SATA 12v supply comes off the same transformer as the 3.3V and 5V (but not 5VS). Looking at the PIC of you board the soldering on IC41 does not look too good so you could remove some solder and re-solder or just use flux to tidy it up.
I am not sure what's happening with your voltage measurements I will re-read your post when I get the chance.
The connector does look like a JTAG so pin 15 should be reset. You could check what happens when you short it to ground for a short time - you could use a jumper as used on PC motherboards shorting pin 15 and 16. I would expect the box to reset and go through the flashing LED sequence again.
I have to admit this has become a bit of an obsession and progress at the moment has ground to a halt but I still think we can fix this so I am not giving up trying at least for the moment.
Thanks, now out until Sunday p.m.
I have been net surfing again and found someone who recovered one of these boxes with similar problem - no USB access and unusable display.
It involves using the RS232/serial interface and a terminal program to gain command line level access to the box, enable the USB interface and then use flashtools over USB to flash the box as normal. If you do not have RS232 on your PC you can get a USB to serial interface cable - choose one with female 9 pin connector so it will plug directly into the box. I have seen some advertised as for use with Dreambox etc.
Connect box to PC with serial and USB at the same time.
Open terminal program such as "tera term", putty or hyperterminal.
Select the corresponding comm port and set the speed to 115200.
Power on IPBox and repeatedly press ctrl+z on the PC.
When connected some messages should appear on the PC, you can type help for a list of commands.
You should see "startusb" type this and the PC should detect a new connection.
You should now be able to use flashtools.
Flash uboot first and then your required image.
Turn the IPBox off and on and.....
Small update.
Before closing down on Friday, I did another voltage check and found I dis have SATA and I did have 11.9v at the fan header. Presumably something intermittent again. My suspect is CP22 which I refixed about three weeks ago. One of the solders looks poor as if a bit broke off after I did it. So I will re-do that and IC41 at the same time. Now in this house, if there’s a soldering job to be done, the kit has to be got out and then put away again. So, is there anything else I should try. Is there any harm in trying to replace IC91? Any other caps I should remove before testing?
Also I’m wondering whether the time has come to screw everything down again and try connecting to a TV. Or is that likely to be a waste of time?
I do have the capability of doing the test using Serial and USB cables, so that will be the second job after the re-soldering.
I am not sure what you mean by CP22 did you mean C22 or CN82?
Anyway going by your pics apart from IC41 (SATA 12V suppy) as already mentioned you could check soldering on
D42 - in the 3.3V supply looks like someone has been soldering this could do with tidying.
D43 - in the 5V supply looks like someone has been soldering this could do with tidying.
Transistor above R98 - used to turn on IC91 (5V and 3.3V supplies) - looks like someone has been soldering this could do with tidying.
CN82 - could do with tidying - I think you already had trouble with this.
IC91 I would not recommend changing this - there is probably heatsink compound (like cream or grease) between the metal part of the IC and the heatsink which is used to improve heat transfer. Unless you have some heatsink compound I would not bother as long as we have 5V and 3V3 with or withoutt the link on R95/98. If we can flash properly we should be able to remove this link.
It may be an idea to screw the PSU in place so that the filters work properly - we don't want any interference while flashing. I would put the lid on but maybe only fix with one screw - yes I am a pessimist!
As to strange meter readings I have had trouble with intermittent meter leads in the past causing confusion.
Thanks – good to know that multimeter probes may not always be reliable, because I’ve often wondered and then thought is it me? So today, when I got an unexpected result, I didn’t panic. I did have another multimeter in tow, but in fact, after powering down and back up, everything was normal.
OK – today’s homework – have only managed a couple of hours.
First sorry, I meant C22. While I was about it, I replaced it with a new 22µF 50v. Actually, I bought the box new, so no-one has soldered it except me. So I’ve tidied up CN82 and IC41 (I think I touched that with an iron by mistake while doing C21). However Q91, D42 and D43 all look fine and I thought if I do something, I could only do damage. So I left well alone. As regards IC91, I do have some thermal paste, but it’s left over from the last desktop computer I put together, which must be at least three years ago, therefore it may have dried out so I left that for another day. I tried a “repair” to the track to the transformer on the front panel but didn’t really manage a small enough bit of solder. However, I never severed the track and the transformer still gets 5.1vDC – but still produces no AC voltage.
So there are three stages to I went for:
First, I connected the PSU board to the mains, but not to the motherboard. With, and without the link between R95 and R98 I had the same voltages as before i.e. with the link I had all voltages and SATA, but without it no SATA and no +5V and no +3VS.
Secondly, I connect the motherboard with the link, and again, same results as before. – SATA working. Now I do get voltages in the JTAG – same as before with two exceptions.
Finally, I connected the front panel, and again no surprises. At this point I screwed it down, and also screwed down the PSU board. No changes except now the JTAG voltages are as I reported a few days ago.
I’ve closed down now, but tomorrow will try connecting to a TV. And then tackle the flashing issues.
Now there’s a point. The TV and the desktop are on different floors. I do have a serial to VGA cable so could connect to a laptop and a desktop at the same time. Is there any future in that?
My comments on the soldering are based on the pics you posted so anything I saw may have just been a trick of the light at that particular point in time.
I would not worry too much about connecting a TV I think you should be able to follow the flashing progress in PCTools. You will only need the TV after flashing has been successful.
I had a thought - what if you cannot enable the USB?
Well you should still be able to flash UBoot with PCTools using serial assuming you can get this to work - just select serial instead of USB.
If I read the manual correctly you cannot flash the main image over serial on this box only UBoot.
If you can flash UBoot you may be able to get the USB to work - may need to restart the box - and then flash the image with USB.
Good luck and fingers crossed!
No luck with the serial plus USB connection – probably the result of my having an extension cable, so that input. So I will try swapping Pins 2 and 3. I take it the USB connector at the box should be Type B, marked client, with Type A at the computer. If something does work, would I expect to see any messages in the PuTTY box? And would I expect to see the box in Device Manager? Come to think of it, shouldn’t I see it anyway with the USB cable connected – just like my printer.
You will not be able to use USB until we get it enabled.
The first objective is to get some form of communication going which we are hoping to do over serial.
At each end of the cable ie the sat box and PC there is an input and an output, the output of one end has to be connected to the input the other end and vice versa.
Perhaps this may sound obvious but this is an important point.
A suggested method of working out which is the output of an RS232 port but does not always work :-
Unplug the cable, since each port in your case has a male connector this will be easier.
At the PC end measure the voltage to chassis of pin2 and 3, expect one to be near 0V (input) and the other negative (output, up to -12V).
Repeat the measurements at the box end and note the results. If you get good results check that you cable connects input one end to output at the other and vice versa.
You could of course just swap pin 2/3 at one end but I hope doing this test would be a more certain way of proving the connections are correct.
We are also working on the assumption that the box is alive to the extent that the RS232 port is working (which will usually be the case).
It may be working while the LED is flashing and not when it stops or the other way around. It may also work when the box is in standby.
Some but not all boxes output a data string over the serial port while booting giving a sort of running commentary on what is happening.
I am not sure whether you need to be pressing anything on the box to get a response from the serial.
You may need a lot of patience and a cool head with this.
Sorry I did not fully answer your questions.
USB yes PC to client connector on box as you would normally do using PCTools.
Initially we are looking for messages in Putty, USB will not say connected until the USB is enabled in the box, at which point you may be prompted for the USB driver n the PC. If you get serial to work but not USB try flashing UBoot with PCTools over serial.
It might be worth running PCTools at the same time as Putty - a response from anywhere would be good, although I am not sure if PCTools will say connected when serial connection is detected.
Make sure in Putty that handshaking or flow control is disabled or set to none in the port settings. If ctrl+z does not get a reaction try pressing buttons on the box eg up, down etc
I expect my last post about the serial cable may have confused you perhaps this may help.
Example of RS232 connection.
Voltages measured without cable connected.
-12v pin2 <--------------------------> pin3 0V
0V pin3 <--------------------------> pin2 -12V
GND pin5 <--------------------------> pin5 GND