Any tivo hardware experts around???

[Hubmaster]

Here is a copy and paste that I put in the general forum section a few days ago,no reply.
I know you guys like dealing with the software end of tivo but I fell I have a hardware issue (can do the work just not sure where to start
.....................................
My Hughes 6000 (2.5 Ext) has been up for the last 3 months without a hitch!

Well all good things come to an end and now time for my first post:

I turned on the TV and saw "searching for sat" and no picture. I went to setup to look at signal strength and saw 80% on both inputs. Then a few seconds later it bounced to 0. Then it went back up to 80% again a few seconds later! Etc...

I shut it down and went to bed disgusted. I tried again in the morning and this time there was no signal strength at all.

Took RG6 cable and swapped into a normal Direct box and all was well with signal strength and no bouncing etc...

Took TIVO to friend's house, unplugged his working TIVO, plugged mine in and had 0 signal strength. Checked voltage on inputs, 13 and 18 volts, on both with different transponders. Then tried the delete everything on box hoping this would fix it, but still no signal...

I tried wiping hard drives and reinstalling 2.5, but still nothing... AARRRGGGG!@#%@%

I have some reasonable curcuit board experience. Any ideas would be greatly appreciated! I live in the Great White North, so returning the box is a definate no go!

Thanks in advance

[BubbaJ]

you have good voltage but it's not tuning yes?

on both tuners?

That's whacked. Have you been playing with any funky software? The tuners are controlled by a serial interface on the 5505s.. if the soft that runs the 5505 were partially hosed, you couldmaybe get that result. or if the serial connection were broken you should get an error. in short, what'd ya do to it?

[Hubmaster]

I know, this all makes no sense to me either! To answer your question I am using

standard software 2.5 hack with showcase
additional 60 gig maxtor
no other trick stuff or files

It's just that one time I went to setup and saw my signal strength for both tuners bounce from 80% to 0%, then back again (both tuners doing this in unisen) and they are now stuck at 0. If I had not seen it myself I would not beleive it!

Does anyone have any thoughts on possibly some voltage checks on the curcuit board that I can perform or is it something like this serial interface (software issue?)

I know that TIVO is not the most stable machine, but I refuse to go back to my old non-TIVO ways!I really would love to fix this problem and this may help other members in the future if this is a possible glitch or weakness somewhere. I have heard of other legit people returning their boxes under warrantee for tuner type issues.

Thanks again!

[BubbaJ]

The whacko part is that it's affecting both tuners.

1) test the tuner control
hook a voltmeter up to the sat 1 input
scroll through transponders, the voltage should alternate between 13v and 18v
if it does, well the tuner is talking to the 5505

2) hmm.. if the tuner is being properly controlled, perhaps one of the data lines has a bad solder joint? you could try swapping all of the /dev devices that point to the 5505s and see if it is tuner specific. or physically swap the tuners, though I understand that's a bit of effort..

[Hubmaster]

O-K ,,so now my ignorance is going to really show,I have tested the voltages and they go between 13-18 .I guess what I am trying to figure out is ,if this sigle test is enough to prove my tuners are working(I kind of think not??)

Also as in previous post I saw the signal strength on both..inputs jump up and down @ the same time then zippo.so I am still thinking control prob(.or is this the 5500 thing you are talking about?)

Ever night I come home I stare blankly into the TIVO with the lid off and look at the tuners tying to figure out what the h@#l controles these things.(sorry my rant for the night)

If anyone has any ideas on stuff I can start unsoldering let me know and I will start testing stuff .

[BubbaJ]

if the voltage on the tuner changes when you change transponder, then the 5505 is talking to the tuner.

on some transponders, the signal strength will be 0 thanks to spot beam. on other transponders, it will be higher. If the only transponder you get signal on is 31, then you are pointed at the wrong sat. if it jumps between 0 and something higher than 70, you probably have vegetation in the way in just the wrong spot, though that wouldn't explain other receivers getting signal from the same feed.

Tuner one may be defective and not tuning properly, if nothing has been screwed with, tuner 2 should NOT also be defective, unless you have REALLY bad luck, and even then, it would probably not show the same behavior. Are you doing something STUPID like using a regular splitter? that could cause this behaviour.

[Hubmaster]

BubbaJ,, thanks for hanging in,let me give you an update as to whats been done these last few days

-TIVO is at friends house, (he has 2 working TIVOS)
-reloaded 2.5 and tried aquireing sat and still no sig strength (using dual lnb seperate lines no splitter!)
-this weekend we will try putting virgin image on factory drive and not use 60 gig at all,
-if that does not work try useing my 60 gig as primary drive with virgin image
-useing the single drive will alow me to use the factory ide cable to eliminate 1 more thing that might be a problem

One question ,, if the voltages are correct comming out ,you say the 5505 is talking to the tuner ,,does that meen the tuner is o-k ,or does that mean the path out is o-k but in could still be dead coming into the tuner,,and both atthe same time ,,NAAAA??
thanks again

[BubbaJ]

Correct voltages means that the serial control connection (I2C) is carrying traffic in both directions. The fact that you have no signal on either tuner is disturbing. Was there a lightning strike in the area? is your cable suitably grounded? do you get feedback voltage AC or DC on either the shield or the center?

You'll notice that the tivo does not plug in to a grounded outlet, it is very susceptable to interference or damage from poor installation. You may have burned the tuners out slowly. .. But even then it would be a surprise for them both to partially fail simultaneously.

[Hubmaster]

BubbaJ,,, no luck reinsalling virgin software did not help..

It is an interesing point you make about the non grounded electrical plug,and in all honesty the unit did not get grounded at all during use ...but for both tuners to go up and down in sig strength at same time then wink out at same time ???? I will take box to a electronics tech. guy for some basic checks and see if he finds somthing.

you wrote about feedback voltages,,I am assuming rg-6 shield to ground or rg6 center to ground with voltmeter,this is measuring at ird end of cable going up to connected lnb??(never heard of trying this before ,, how would voltage get there just curious)

will let you know if my chiphead finds somthing in or around tuners
thanks

[BubbaJ]

with the cable connected to the receiver and again without, measure AC and DC voltage on the shield. (Voltage on the center is supposed to be relative to the shield)

If you get AC then you are providing too much load for your wiring. Probably an inadequately small ground wire was used. Remember, voltage can feed back through any device sharing a circuit, and there are at least 3 circuits in play.

1) center to shield circuit

2) shield to ground circuit

3) AC to ground circuit

and those are repeated and shared on any receiver sharing the switch.

[Hubmaster]

at the risk of sounding stupid here I go,,,

I am amazed at how much emphisis you put on the ground,,

<If you get AC then you are providing too much load for your wiring. Probably an inadequately small ground wire was used.>

I guess I have been horse sh#t lucky with all my outher systems because me and the ground wire have never met before

I had to look for myself last night,,all my outher box's wallplugs do have a ground pins so that may be why I have gotten away without in the past????

Anyway Tivo is going to the hospital in the morning will let you know what we find,I suppose we will look inside the tuner and around that area unless you have outher thoughts as well??

thanks again

[BubbaJ]

I HAVE had this problem in the past, and it's affected more than just DTiVos.. Here's an easy experiment for you. Buy a grounding adapter, clip the metal tab Plug it in and plug in a computer. Get a cable where the shield is connected to earth somewhere (real ground) measure the voltage between the chassis and the real ground, you'll probably get around 60VAC because current is feeding through the chassis to ground, and you're effectively using the chassis and some wire inside the power supply as a large resistor. (You're providing a truer ground than the ground wire is providing.)

You wouldn't BELIEVE how many improperly installed electrical circuits I've come across in my life time. And a circuit breaker doesn't work too well on ground loops.

Have you ever seen a wire with its casing burned completely off before the copper oxidized too much to conduct electricity? I have.. in a house I was renting. The house had even been inspected by the building inspector for code violations.

Proper wiring, and a guarantee of suitable ground is VERY important for electronics even if it seems like everything is fine.

[Hubmaster]

informative post ,, I hope outher people heed this because if this is what killed my TIVO I will be very pis#ed at myself.

Anyhow buddy at the shop could not find anything obviously wrong with box ,,,surprise,surprise!!

but had an interesting thought ,,,if or does TIVO see tuner # 1 as primary tuner ( may use it for initial aquire of sig to get things rolling) he is going to unsolder tuner 2 and install into # 1 position and retry the reboot???hey nothing ventured nothing gained I say!!will let you know how that goes.

Is there a source for TIVO tuners or is this the possible end of the road ???

Please??

[khmann]

yeah, so this is completely irrelevent, but some folks might think it interesting...

Because I'm insane, I run non-standard power to all my AV equipment, a technique commonly used in AV production studios called 'Balanced Power'.

In the US (and most parts of the world I believe) the two wires responsible for delivering electricity to your appliances are 'Line' (120V in the US) and 'Neutral' (tied to ground). This creates an unbalanced potential inside the equipment with respect to ground- the 60 volts BubbaJ was referring to.

In balanced power, the Line and Neutral are replaced with two out-of-phase 'Hot' connectors of equal voltage- ie +60,-60. The trick is that the capacitive forces inside the power supply and the device itself then cancel out- leaving the circuit reference at 0v (ground).

Using this technique (in my experience) results in a drastic reduction of 60Hz hum and other noise in both audio and video. I was pretty unhappy with my $1500 Sony WEGA until I switched it to balanced power- (my previous Panasonic TV had been connected via a small balanced power transformer, which didn't have nearly enough juice to run the Sony...)

a number of companies sell these types of systems commercially, for large ammounts of money. The reality however, is that this type of power can be easily achieved with a $250 transformer from your local Commercial/Industrial electronics supply company.

In my application (because I'm inside), 90% of my entire house is fed with a 7.5KVA (120volt, 60amp) transformer. The remainder (HVAC, kitchen appliances) is fed with normal power.

This has all been in place for about 18 months, and nobody in my house has been shocked and/or killed, so I guess that's a good sign.

Yeah, and I think the transformer isolates my equipment from the power company somewhat - I don't have any decent surge protectors on my equipment and nothing has exploded...

equitech dot com sells commercial balanced power transformers for thousands of dollars - and has some good white papers on the subject.

yeah, a couple safety related notes:
Since both wires coming from the wall are now live... it's not safe to rely on single-pole power switches to disconnect power from a device. Also if you're gonna wire up the whole house, you gotta use double pole (240V) breakers for everything, to interrupt both conductors.

Yeah, and lightbulbs... normally the tip is 120V and the shell/thread is ground (neutral). With balanced power, the thread is hot - 60V at all times, so tell the wife and kids to be careful.

It is an absolute requirement (both from a safety standpoint, and according to the US National Electric Code) that _all_ receptices be ground fault protected. Note that with Balanced Power, nuisance tripping (eg- during a storm) of GFI outlets/breakers occurs much less frequently.

Yeah, so, yeah, as I said - irrelevent. but it _really_ can improve the performance of AV gear.

take care.

[BubbaJ]

:D

It's an Isolation Transformer. Its real purpose is to create a completely seperate second circuit fed off the first circuit. The reduction in hum comes from the fact that there are no capacitive or inductive loads sharing the circuit. (An AC to DC power supply is a mostly resistive load)

The isolation transformer itself is a capacative load.

khmann: I hate electronics. some circuits that I was imagining running based on your setup made me consider using paper and pencil.

[khmann]

Originally posted by BubbaJ
:D

It's an Isolation Transformer.

khmann: I hate electronics.

yeah man :) I personally dislike computers quite a bit (though I am paid to dislike them). I never liked TV either until I got TiVo...

Anyway, yes I am talking about Isolation transformers. And they really _do_ make a difference in AV systems. Most people would only need a plug-in 'power conditioner' to run their home theaters. Jameco has a 2000VA (watt) BussMan unit on sale for about $70 which seems like an excellent deal...

My situation is a little more complicated in that I've got my AV and computer gear all interconnected and spread out over 4 or 5 locations (I don't like to look at it :). In this situation, plug-in isolation transformers can be problematic/dangerous because (even with a transformer at each location) you end up with different ground potentials and induce a bunch of ground loops by interconnecting the stuff.

Note that while isolation transformers in which the secondary is floating (ungrounded) may provide slightly cleaner power, they can be extremely dangerous. With no ground reference (other then the week capacitive coupling of the wire and equipment), if a ground fault (short of one hot conductor to ground) occurs, the protective devices (GFI/breaker) will not operate because there will be very minimal current flow- only the system leakage current. In actuality, the entire system will usually 'tip over' one conductor grounded through the fault, one at 120v.

This is not, by itself, dangerous. It's only when a second ground fault occurs simultaneously with the first (which never got detected) and to the alternate hot conductor, that it gets really dangerous. Since the transformer itself is not connected to ground, all fault current flow occurs through the two faults in series (and any interconnect wires), yielding a much higher impedance fault that may fail to trip protective devices.

Also without a solid ground on the electrical system, the entire thing gets grounded through your interconnect lines - telephone, satellite, and then you get current flow... again, bad.

and with no solid ground reference, you run the risk that the entire system will 'float' way up above normal ground potential. This can occur through power surges, or intermitant ground faults - ie a faultly motor or surge protector.

By providing a 'center-tap' to ground on the isolation transformer, all these problems are eliminated. Circuit breakers and GFI devices are still able to operate normally, because while the individual phase conductor voltage has been halved, the total system voltage and current remain the same.

One interesting side effect of all this- normally you get three wires from the power company. Two hots (240v apart) and a neutral, which is used to power 120v loads, and is tied to ground outside at your meter/disconnect. Since I've used a transformer with a 240v primary, I draw almost no current on this neutral connector, which makes it a much cleaner 'ground'.

The only downsides are really the requirement for double-pole (240volt style) circuit breakers and all the GFCIs. For me that wasn't too much of an issue because I had to re-wire much of my house anyway (1960's construction). The additional cost was minimal- $5 additional per breaker, and I had to get the really large panel 'cause double pole breakers take two spaces...

yeah, and I guess I should mention that as of the US 1999 National Electric Code (I haven't checked 2002), this type of power is only allowed to be installed as 'technical power' for electronics. And you're very specifically _not_ allowed to use it for circuits which supply lighting loads. Primarily because of the issue of the light bulb screw thread being live I think. And the single-pole switches that are built into everything - because normally there's only one the hot conductor...

hey, it works for me :)

[BubbaJ]

That post answered a lot of my concerns about how you had it installed.. :D Primarily the center tap to ground.. (which I guess you did mention before and I just overlooked..)

You could get around the lighting issue by using fluorescents everywhere.. :)

[Hubmaster]

FYI:

Took the box to a friend's house and unsoldered tuners and switched them around. Still no go! Zero signal strength...

I know I'm asking for a flame, but does anyone have parts for these. Am I wasting time with this?

A friend still has one new Tivo in a box, which he at this point will not let me play around with or swap tuners. So finding parts would be awesome.

Thanks and I would REALLY appreciate anyone's input at this point.

[BubbleLamp]

Here's (http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=1363140623) a DSR6K parts machine on eBay.

[Mjolinor]

Tried following this thread a few days ago gave up because it seemed to wander off onto the subject of earthing (grounding for those on the wrong side of the pond;) ).

I think the guy still has a non worker and maybe this will help.

I have some experiance with mending satellite decoders in the UK and I assume the systems over there are similar

It is unclear from the thread whether you are talking about the tuner voltage or the LNB voltage.

The driver voltage for the LNB is normally derived directly from it's own power supply rail. As you have 2 tuners in this box they will undoubtably be driven from the same power supply and as both tuners seem to have problems then maybe this power supply is up the spout. I have seen lots of these that measure 14 /18 volts DC on the coax but they still have a fault and this voltage needs to be measured with the LNB connected because the voltage goes away under load.

Problems arise in this circuit because there is normally a facility in the microprocessor to detect short circuit on the coax and shut down the LNB drive voltage and from the fact that switching is required to go between horizontal and vertical polarisation (left circular and right circular if that is what is used over there).

The actual circuit is normally fairly straightforward and can be traced out in a short time.

If I were you I would find, on the power supply connector, the voltage that is used to make the LNB drive voltage. This will normally be higher than the other supply rails. In my experiance it will be between 20 and 36 volts.

Once found see if it stays stable as you connect and disconnect the coax. If it does then check it at the tuner end.

It is fairly easy, but does require soldering abilities, using an external power supply to drive the LNB. You need to isolate the internal LNB supply from the coax normally by removing the drive transistor and just connect a DC supply to the point on the LNB side.

The LNB supplies need checking before you embark down the road of checking the tuners, be they I2C or analogue tuning. If they are analogue then a ramping voltage will be present on 1 of the tuner pins as the decoder searches the transponders, this will be probably 0 to 5 or 0 to 12. If the tuner is I2C then you either have to build a Philips I2C interface to catch the data or go into the tuner, Major work I think and probably you have excluded this by changing the tuners.

All this assumes that these direct tv boxes are satellite decoders with TiVos in. Never having seen one I don't really know what htey are.:D

[Hubmaster]

BubbleLamp,,thanks for the e-bay link but I still have a bad tast in my mouth from getting this Tivo in my house with 2 bad ebay sellers (took 3 purchases till i finally saw 1 at my door ,,,ouch!)

Mjolinor,,I agree with you that something that controls the tuners sould be at fault???cause for both to go at same time ,must be a control or power supply prob. or somthing like that
I will see if I can check the voltage with the lnb connected and see wasss up.

thanks

[Mjolinor]

The LNB drive voltages are never part of the tuner (at least I've never seen them that way) they are made/altered/set directly on the motherboard and are fed directly into one of the tuner pins that only goes straight to the coax.

If you pull one of these tuners apart immediately behind the coax connector the path will split. One path will have a small capacitor in to prevent the LNB drive voltages from entering the tuner and the other will have a 'curly' copper track on the PCB and will immediately go to one of the external pins of the tuner. This is the DC LNB supply line. The curly track provides some inductance to prevent the signal from passing down the track, this may very occasionally be done with a surface mount inductor though this is not common because of the current it needs to carry in order to drive the LNB.

Once on the main PCB, through the tuner pin, it will undoubtably go to a transistor, this is not often a surface mount device because it has to switch a lot of current and it can normally be spotted fairly easily around the tuner area. In your case I suspect there would be 4 of them, 2 for each tuner, one for short circuit protection and one for voltage switching.

Unfortunately I would also guess thatif I were designing a borad and I needed 4 similar transistors in a particular area I would use a chip of transistors which in your case would be a ball ache because you can't spot them without a data sheet for the chip :)

Anyway happy hunting.

[Hubmaster]

Mjolinir,, o-k I think at least a little of the fog is clearing,so tell me if I got this right

2 voltages ,,one seems like it is just passing through tuner ,(sould be able to do a test probe for voltage at tuner pin into pcb, and check that bit of funtionality of lnb voltage ) and if with lnb connected voltage is good then at least that is good if voltage bad then look further up river.???

the second voltage i guess is the tuner voltage ???so there must

A- be a way of checking

B- should some power source not be controlling both,,, and maybe it's n.g.

I take it you have never seen the tuners or inside one of these box's if it might help you identify (or if you are nosey)I can post some pics of tuner and pcb stuff around the area of concern??

thanks

I hope that is allowed in the forum??

[Mjolinor]

From the beginning.

LNB (bit on the satellite dish with the coax connected to it)

Inside there is a frequency convereter and an amplifier. These need powering. On modern satellite stuff this is provided up the coax that the signal goes down. In order to perform switching within the LNB this supply voltage can be varied in certain ways as follows. (may have voltages wrong way round)

If 14 V DC (approximately) is supplied then the LNB only looks at vertically polarised signals.

If 18 V DC is applied then the LNB only looks at horizontally polarised signals.

If there is a 22kHz tone present on the LNB supply voltage then the LNB switches to a higher frequency band for its source signals.

More modern designs may use disecq switching but I don't think this is relevant in your box.

Because not all LNBs work this way the Tuner in the set top box is not designed to handle this requirement but it is normally done on the motherboard of the IRD.

Tuner
The tuner is designed to accept, via the coax connection, a signal containing all channels from the transponders on a satellite converted into the frequency range 1 to 2.5 (?) GHz. It is the tuners job to scan that frequency band and lock onto any particular frequency that contains the signal for the TV station.

In oder to scan the frequencies it will either be an I2C interface or an analogue one. The analogue interface will accept, probably, 0 to 5 volts and this supply is converted to a frequency using a voltage to frequency converter inside the tuner such that 0 v in would correspond to the lowest frequency and 5 v in would correspond to the highest frequency (may be 0 to 12). If the tuner is I2C then what happens is that the microprocessor on the main board will supply tuner specific instructions via the I2C bus (2 wire bi-directional master slave communication protocol designed by Philips) to the tuner telling it what to do. This may be "go to frequency X" or "scan till input voltage = X" or some thing else. These instructions are totally dependant on the tuner design engineer and without a data sheet all you can do is see if the data and clock lines are behaving correctly i.e. going up and down from 0 to 5 volts :).

So thus far connections to the tuner will be

Ground (earth to me in the UK) this will be a pin connected to the silver tin of the box. There will probably be more than one of the pins connected to the ground because sgnals around here are very small and noise ingress needs to be minimised. This gound will also connect to the coax outer and, depending on the power supply configuration and the internal American wiring methods, may connect to the ground on your house wiring system.However it may not so don't go touching both at the same time. Reference all voltage measurments to this pin / tuner case.

Supply for tuner probably 12 volts. It is possible to have 5 and 12 volt supplies to the tuner the 5 being for any digital goings on and the 12 for the analogue type bits inside.

Supply for LNB probably 14 or 18 volts and may have a signal of 22kHz superimposed on it.

2 connections for I2C or
1 connection with a ramping voltage on when the decoder is scanning for channels

1 connection for signal off the board into the decoder

There may be a feedback volltage for an automatic frequency control i.e. if the signal level drops the decoder may tell the tuner to slide frequency a bit and try to increase signal strength but this may be done external to the tuner and change the I2C or ramping control voltage.

There will probably also be an automatic gain control to increase signal level when a cloud gets between you and the satellite.

So a minimum of 7 ish pins with perhaps multiple earth pins

Each of the signals is usually reasonably easy to identify with a voltmeter or preferably an oscilloscope

You will find the LNB supply by removing the motherboard from the decoder and disconnecting the power supply using a resistance meter connected to the centre of the coax and run down the tuner motherboard pins. One of these will be a short circuit. It is at this point that you need to check the LNB supply voltage with the decoder powered back up, connected to the LNB and not putting fingers anywhere near the mains power supply. They do seriously hurt and can easily prove fatal. You can check this by measuring, going into menus and switching to horizontal, you should see it change. If you have an oscilloscope then you can see the 22kHz tone when you switch to the top band.

I will stop here for now. If you have some pictures then I will look at them but it may not provide much info.

If you understand this and try it and need more info please post again.

[Hubmaster]

Mjolinor..

Thanks for the post ,,,my electrical abilitys are limited to things like building avr boards etc.but I do have an electronics repair guy that is good at trobleshooting most things (but has never seen a sat.reciever before)and has repaired electronic tuners before and the like .(this is a 12c)

anyway he helped me swap the tuners and said that if I can find some info on this stuff he is game.

so I guess this stuff you gave is a starting point ,,and yes we will be sending a pic or two of the tuners and surrounding pcb area,

I will be giving him an lnb with 2, 50 feet pieces of coax to do the test with as well ..

will send a couple of pics soon

thanks again

[volkovoi]

Has anyone get a solution to this problem? I just got a Dtivo for Xmas from my sister. I know its early but I COULDNT WAIT! :)

I Plugged it in and noticed that the signal was jumping from 90% to zero every few seconds. And it stuck when recieving the guide
at 26%.

I had my dad bring over his old RCA, we plugged it in and it picked every transponder and all signals were high 80's, no problems.

Please, if anyone can help I would GREATLY APPRECIATE it. :)


Thanks,
Volkovoi

[Hubmaster]

should have posted my status sorry,,I sent my box to truett electronics and poof it was fixed.80.00 and next day service awsome people to deal with.

hubb

[cali]

Did they give any info as to what was wrong?

[Hubmaster]

not sure but I saw a few new solder spots on the lnb power regs.???all I know is it is up and running 100%

hub.

[cali]

U mean the LNBP15's?