TiVo Community
TiVo Community
TiVo Community
Go Back   TiVo Community > Main TiVo Forums > TiVo Roamio DVRs
TiVo Community
Reply
Forum Jump
 
Thread Tools Rate Thread
Old 08-07-2014, 04:57 PM   #1
telemark
Registered User
 
Join Date: Nov 2013
Posts: 629
base Roamio power calculations / tests

I'm starting to have suspicions about this and I'm hoping an inquisition will put it to rest definitively or find a work around that will improve the picture.

I'm ignoring the 6 tuner Roamio because it should have a completely different, beefier supply.

I'm not 100% certain how to go about everything, but it should become clearer as it gets further along.

To gather some initial information:
Specs:
the included Power brick
the included HD (idle, operating, spin up)
typical (AV-GP?) HD upgrade (idle, operating, spin up)
worst case (10k rpm?) HD (idle, operating, spin up)

Measurements:
confirm above specs by measurement
board consumption (boot, typical, peak)

Subcomponents (hopefully they can just be aggregated):
fan consumption (low, high, spin up)
cable card?
Wifi card
RF card?
USB?
telemark is offline   Reply With Quote
Old 08-07-2014, 05:16 PM   #2
telemark
Registered User
 
Join Date: Nov 2013
Posts: 629
I had 4 mysteries in mind, that might be explained by the inadequate power scenario.

nooneuknow's unreadable bits, suspected weak writes
my flash corruption
stuck blue LED on boot
failed drive upgrades

If these all symptoms only inflicted the Base Roamio and not the other Roamio, it would lend some credence.

As an initial calculation, the Power Supply is 12V, 2A?
Some Seagate specs says spin-up is 2A at a nominal 12V ??

Two types of tests that might simplify everything:
One of the SMART values is average spin up time. If the spinup time by a Roamio is greater (slower) than when using a known good supply, it would imply the drive was being power starved.
On the RaspberryPi wiki, it gives the example of monitoring the voltage (ideally with a scope) on the board, and when additional loads come online, if the voltage sags below an acceptable value, that would be conclusive of a problem.

This is an interesting paper on HD power usage, a little out of date.
https://www.cl.cam.ac.uk/~acr31/pubs...harddrive2.pdf

Last edited by telemark : 08-08-2014 at 02:08 PM.
telemark is offline   Reply With Quote
Old 08-07-2014, 06:14 PM   #3
aaronwt
HD Addict
 
aaronwt's Avatar
 
Join Date: Jan 2002
Location: Northern VA(Woodbridge)
Posts: 13,923
My Roamio BAsic draw around 16 watts. My Roamio Pro draws around 22 watts.

I wouldn't even consider one of the larger size Seagate drives for the Roamio Basic. They draw a lot more power than the WD drives do.
__________________
Roamio Pro
TiVo Mini x4
Roamio Basic OTA
39TB unRAID1--53TB unRAID2--36TB unRAID3
XBL/PSN: WormholeXtreme
aaronwt is offline   Reply With Quote
Old 08-08-2014, 06:37 PM   #4
nooneuknow
TiVo User Since 2007
 
Join Date: Feb 2011
Location: Cox Cable Market, NV
Posts: 2,788
The wall-warts for my base Roamios are 12V 2.0A.
The BRICKS for my tuning adapters are 12V 2.5A.

I verified the plugs make proper contact when swapped, and the polarity is the same. I have not tried swapping the two around, live (yet).

Any comments or questions? It would seem like a TA should require less than a whole TiVo.

They are the Cisco STA1520 TAs. Could it be a solution is right there for those with these TAs (possibly even other TAs)?

If I wasn't such an honest guy, I'd have three extra of those Cisco bricks (made by LiteOn). When I went to return three of them, the rep told me all they needed was the TA, and not the rest. I explained that they were wrong, and gave them the bricks, power cord, and coax that each came with. But, I kept the self-install kit(s) contents: splitter, high-end coax, MoCA PoE filter, etc.

If I'd just gave them the TA's alone, I'd have three 2.5A bricks to use, without using the 2.0A TiVo wall-warts for the TAs.
__________________
Cisco tuning adapters should never be used inline (using the TA coax OUT port) to connect a TiVo, if MoCA is in use. Use a splitter w/PoE filter on leg to TA, use other leg for the TiVo. Enjoy!
nooneuknow is offline   Reply With Quote
Old 08-08-2014, 10:02 PM   #5
nooneuknow
TiVo User Since 2007
 
Join Date: Feb 2011
Location: Cox Cable Market, NV
Posts: 2,788
Found my temp gun.

Roamio base wall-wart (with open-air ventilation) TiVo not busy: 47.5C, 117.6F
Cisco STA1520 LiteOn brick: 31.3C, 88.4F

Random sampling of other warts and bricks shows idle TiVo wart to be the highest temp.

The temp of the TA brick is always the same (going by touch). I'll measure at other times, using temp gun, to be sure

I'll try to get the temp of the Roamio wart when recording 4 programs and processing data, which I expect to be substantially higher.

So far, I'm thinking the TA will do fine with the TiVo wall wart, and the brick from the TA might remove any borderline-power issues with the TiVo.
__________________
Cisco tuning adapters should never be used inline (using the TA coax OUT port) to connect a TiVo, if MoCA is in use. Use a splitter w/PoE filter on leg to TA, use other leg for the TiVo. Enjoy!
nooneuknow is offline   Reply With Quote
Old 08-08-2014, 10:02 PM   #6
telemark
Registered User
 
Join Date: Nov 2013
Posts: 629
Ya a lot of CableCo hardware are power hogs.

I don't think it's necessary harmful to upgrade the supply, unless it's unregulated or noisy.

What would be really useful is if we could get voltage measurements off the board during drive spin-up. Maybe that has to be done twice, once each for 5v and 12v.

Then if you upgrade the supply, you should expect to see a smaller dip.

Regarding temperature, a higher capacity charger is expected to be lower temperature for the same load because you're no longer operating at the edge of it's range.
telemark is offline   Reply With Quote
Old 08-08-2014, 10:29 PM   #7
nooneuknow
TiVo User Since 2007
 
Join Date: Feb 2011
Location: Cox Cable Market, NV
Posts: 2,788
Quote:
Originally Posted by telemark View Post
Ya a lot of CableCo hardware are power hogs.

I don't think it's necessary harmful to upgrade the supply, unless it's unregulated or noisy.

What would be really useful is if we could get voltage measurements off the board during drive spin-up. Maybe that has to be done twice, once each for 5v and 12v.

Then if you upgrade the supply, you should expect to see a smaller dip.

Regarding temperature, a higher capacity charger is expected to be lower temperature for the same load because you're no longer operating at the edge of it's range.
I seem to recall being corrected by somebody, like CoxinPhx, when I had a fit about the TA power brick being rated higher than a TiVo.

I recall a spec sheet being brought-up showing that the TA used far less than the brick was rated at, bringing it to lower consumption, with excess power supply capacity.

I think making up an in-line test fixture where I could test amps, under varying conditions would be the most telling. Since the TA & Roamio both use slightly odd DC connectors, it would be easier for me to just measure AC amps going into the power supplies. Something I could leave, and not have to hold probes on test points... But, I think setting something up inline on the DC side would be much better.

I'll have to see if I can find all my "universal" power bricks/warts, with the assorted tips, and enough clip-on leads and banana jack jumpers to get a running reading, then move onto other things.

Since the spin-up time for my drives, in-TiVo, didn't vary enough to mean anything, and were better than I'd even expect, I don't think the weak writes/weak sectors cause is going to expose itself there.

It doesn't seem like we're getting much interest here. I found the thread about the internal LEDs, and that thread's kind of slow too...

Red NAS drives have a supercapacitor that is supposed to complete operations with loss of power. They are also designed to spin-up without drawing much load. I wonder how long that capacitor holds a charge, and if it might be more hurtful, than helpful, used in conjunction with a device designed to be resistant to drive corruption on power loss (maybe my weak writes were made when I pulled-power at some time?)

I can't help but wonder what was wrong with that power-loss mechanism, that it was "improved" in the newer drives that just came out...
__________________
Cisco tuning adapters should never be used inline (using the TA coax OUT port) to connect a TiVo, if MoCA is in use. Use a splitter w/PoE filter on leg to TA, use other leg for the TiVo. Enjoy!
nooneuknow is offline   Reply With Quote
Old 08-08-2014, 11:01 PM   #8
nooneuknow
TiVo User Since 2007
 
Join Date: Feb 2011
Location: Cox Cable Market, NV
Posts: 2,788
More thoughts:

I have always had a "redundancy TiVo" setup, where I have more than I need for if (when) one has a problem, or if (when) a drive has a problem.

Since bedroom #2 was the one with the drive issues, It's been offline, while I do my three-drive testing I described in the hdd thread.

Upon it's first startup, it's due a drive format, a TiVo software update, and a cablecard update. That would seem like an interesting time to be measuring, but also the most likely period where one slip, or bad connection while measuring, could be catastrophic. But, those would seem to be opportune times where the load on the power supply would be greatest. If the wall-wart is iffy/borderline, I'd be worried if I should use the TiVo wart, with all that being due (since it's a suspect).

I'm pretty much talking to myself here, unless I hear from you, telemark, in case you hadn't noticed.

If you feel measuring loaded voltages, rather than going for the amperages is best, I'll go with that. I never ask anybody who is working with older models to give me amperages (due to the complexity involved). So, maybe it's a longshot just hoping that we can even get a decent number of people to probe voltages, giving a large enough sample size.

I'm also still hoping you'll give me your opinion on what drive testing/benchmarking suite(s) are worth buying. I was just about to buy HD Tune Pro. Then, you gave me the weak write/weak sector page from the HD Sentinel page...

If there are any weak and/or slow sectors, I want to know about them. It's more important to me than benchmarking, unless the software can benchmark the drive in a TiVo-like mode of operation.
__________________
Cisco tuning adapters should never be used inline (using the TA coax OUT port) to connect a TiVo, if MoCA is in use. Use a splitter w/PoE filter on leg to TA, use other leg for the TiVo. Enjoy!
nooneuknow is offline   Reply With Quote
Old 08-08-2014, 11:27 PM   #9
telemark
Registered User
 
Join Date: Nov 2013
Posts: 629
Volts vs Amps, they're both useful. Do whichever is easier. I personally find Volts easier cause it doesn't require building the inline cabling.

Measuring on the DC side is better than on the AC side because it's isolated what's going on in the Power Supply.

I don't expect anything to be more demanding than the HD spinup, having a good handle on those few seconds is probably enough as the likely worst case scenario. Unless a drive is spinning down later in operation in which case that will be the worst case as the processor will also be busy.

I don't know Windows software. I only do manual benchmarks on Linux. I linked to the Sentinel page cause the description was thorough, but not cause I can vouch for the software.

One write through of the pending sectors is suppose to be enough to make it go away. Is that not happening as expected?

Maybe if you set the TLER to the lowest level, dunno what that would be, you'll force it to give you error messages instead of slower responses maybe due to retries.

Could be unlikely, but just in case, most UPS's don't switch instantly. I wonder if that can affect a drive's write performances if the UPS is tripping and switching over..

Since one of the failure modes is a bricked box, I don't suggest flat out testing unsafe combinations. It's safer to extrapolate when it would fail, from safe operating combinations.
telemark is offline   Reply With Quote
Old 08-17-2014, 12:41 AM   #10
nooneuknow
TiVo User Since 2007
 
Join Date: Feb 2011
Location: Cox Cable Market, NV
Posts: 2,788
Quote:
Originally Posted by nooneuknow View Post
Found my temp gun.

Roamio base wall-wart (with open-air ventilation) TiVo not busy: 47.5C, 117.6F
Cisco STA1520 LiteOn brick: 31.3C, 88.4F

Random sampling of other warts and bricks shows idle TiVo wart to be the highest temp.

The temp of the TA brick is always the same (going by touch). I'll measure at other times, using temp gun, to be sure

I'll try to get the temp of the Roamio wart when recording 4 programs and processing data, which I expect to be substantially higher.

So far, I'm thinking the TA will do fine with the TiVo wall wart, and the brick from the TA might remove any borderline-power issues with the TiVo.
So, I made the swap (and waited for the highest temps after many hours) and here's the results:

2.0A base-Roamio wall-wart running Cisco TA: 34.8C/94.8F
2.5A LiteOn (Cisco) brick running base Roamio:35.2C/95.5F

I trimmed less than 2mm of rubber away from the plug on the LiteOn bricks, so they went into the Roamios further (full depth). As they were, they went in far enough to catch the groove that the Roamio's plug didn't even have (makes accidental unplugging more difficult). I simply made the exposed outer shell identical length, without any detriment to structural integrity.

I say the temps spell things out. I'm not over-loading the Roamio wall-wart, using it for the TA, and the Roamio now has 500mA more power available to it.

I dug and dug and dug for every wart and brick I have for external drives, or that came with my USB adapters/docks. The MINIMUM rating is 1.5A, with the majority being 2.0A, and some as high as 4.0A (dual drive docks, and enclosures that came with 7200RPM drives).

That a whole TiVo gets a 2.0A wall-wart, while every external low-power drive gets 1.5A, just to run the drive, makes it seem like 1.5A is for the drive, and all the rest of the TiVo, plus what else the TiVo has to power (like the cablecard) gets 500mA (as a gross oversimplification).

I've also noticed that the Roamio wart cable is unshielded, and without ferrites. The bricks for the TAs, and for most of my enclosures are shielded, plus have ferrites.

Call it placebo-effect, if you must. I couldn't help but notice it seemed like my Roamios each booted faster with the bricks, vs warts.

The only potential issue I see is if the Roamio wart's lack of shielding and ferrites causes TA issues, due to noise from the wart, and ingressing of RF from the air.

It used to be hard for me to unplug/plug the TAs, due to the bulk of the plug. Now, it's easier/faster to do it without fumbling around (with either).

I'll try to get around to making the actual measurements the thread was created for.

As usual, I had to do a search for "find more posts by telemark", to even find this thread (and I'm subbed to it). It's a pity how quickly telemark's threads wind up on page 25-50 of subbed threads...
__________________
Cisco tuning adapters should never be used inline (using the TA coax OUT port) to connect a TiVo, if MoCA is in use. Use a splitter w/PoE filter on leg to TA, use other leg for the TiVo. Enjoy!

Last edited by nooneuknow : 08-17-2014 at 12:45 AM. Reason: typo
nooneuknow is offline   Reply With Quote
Old 08-17-2014, 08:40 AM   #11
slowbiscuit
FUBAR
 
Join Date: Sep 2006
Location: In the ATL
Posts: 2,596
Quote:
Originally Posted by nooneuknow View Post
Call it placebo-effect, if you must. I couldn't help but notice it seemed like my Roamios each booted faster with the bricks, vs warts.
LOL.
slowbiscuit is offline   Reply With Quote
Old 08-17-2014, 09:56 AM   #12
telemark
Registered User
 
Join Date: Nov 2013
Posts: 629
I also did an audit of my External Hard Drive power adapters as well and they're all:
12V 2A
or
12V 2A & 5V 2A (dual rail)
Not really the best way to do a comparison, but concerning still.

Surface temperature is power consumed x inefficiency and something about cooling and ambient temperature. A few more unknown variables there than ideal.

Oh right, you also have the prior temperature measurements. Yeah, that looks like a good trade off.

A number of people have those LiteOn power supplies because reps say they don't need them back. (this may be true if they don't refurbish them because of liability, in which case you're saving them from a worse fate)

Quote:
I couldn't help but notice it seemed like my Roamios each booted faster with the bricks, vs warts.
A Hard Drive could spin up faster, but the Roamio board is well into root filesystem verification before the Hard Drive is ready. There might be some solid state differences as well, but no need to speculate, if it's detectable, would be enough to settle that question. Stopwatch is the simplest tool for this.

I have a triggered stopwatch that would be ideal, but I don't have a Roamio on bench because I'm working on Premiere for a while, which is also why I haven't added datapoints back to this thread.

Among the Tivo power adapters, the one that came with a Stream had a Choke but the one that came with the Base Roamio did not.

Last edited by telemark : 08-17-2014 at 04:46 PM.
telemark is offline   Reply With Quote
Old 08-17-2014, 09:20 PM   #13
nooneuknow
TiVo User Since 2007
 
Join Date: Feb 2011
Location: Cox Cable Market, NV
Posts: 2,788
Quote:
Originally Posted by telemark View Post
I also did an audit of my External Hard Drive power adapters as well and they're all:
12V 2A
or
12V 2A & 5V 2A (dual rail)
Not really the best way to do a comparison, but concerning still.

Surface temperature is power consumed x inefficiency and something about cooling and ambient temperature. A few more unknown variables there than ideal.

Oh right, you also have the prior temperature measurements. Yeah, that looks like a good trade off.

A number of people have those LiteOn power supplies because reps say they don't need them back. (this may be true if they don't refurbish them because of liability, in which case you're saving them from a worse fate)



A Hard Drive could spin up faster, but the Roamio board is well into root filesystem verification before the Hard Drive is ready. There might be some solid state differences as well, but no need to speculate, if it's detectable, would be enough to settle that question. Stopwatch is the simplest tool for this.

I have a triggered stopwatch that would be ideal, but I don't have a Roamio on bench because I'm working on Premiere for a while, which is also why I haven't added datapoints back to this thread.

Among the Tivo power adapters, the one that came with a Stream had a Choke but the one that came with the Base Roamio did not.
I checked the SMART value for spin-up time. While the 2.0A power wart was consistently a 180 (ish) value (out of 200, with higher being better), The LiteOn brick has it at perfection, with a 200 attribute value, sometimes higher (meaning better than would be expected).

I did a stopwatch (old-school) check (great minds think alike, again), and found up to a 2 minute improvement between power applied, and TiVo being able to get far enough to be able to start a recording that was scheduled, and interrupted, by removal of power (which will start recording again, as soon as everything is far enough along).

The temps of the bricks/warts are staying solid. The MBT value of the TiVos has decreased/improved from mid to high 50s, to 47 to 49C. I can knock this down by nearly greater than 5 degrees (up to 9), by simply leaving the cablecard cover off, and using something else to take the place of the rubber foot that is on the cover. This was noted before I changed the power supplies around.

I have also noted the pitch of the fan noise no longer shifts lower when load is increased by becoming "busy". This implies that the voltage isn't dropping, like it seemed to be doing. This effect has to be caught before the fan kicks up its speed after the load increases temps for long enough. I've also noticed the fan kicking into higher speeds less often, less quickly, and for lesser time. Since a voltage drop, due to load strain, causes higher temps, this makes sense.

Still, there is more work to be done, as in actual test-point voltage measurements with the original wart, at different stages, modes, and loads, of operation.
__________________
Cisco tuning adapters should never be used inline (using the TA coax OUT port) to connect a TiVo, if MoCA is in use. Use a splitter w/PoE filter on leg to TA, use other leg for the TiVo. Enjoy!
nooneuknow is offline   Reply With Quote
Reply
Forum Jump




Thread Tools
Rate This Thread
Rate This Thread:


Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Advertisements

TiVo Community
Powered by vBulletin® Version 3.6.8
Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
vBulletin Skins by: Relivo Media

(C) 2013 Magenium Solutions - All Rights Reserved. No information may be posted elsewhere without written permission.
TiVoŽ is a registered trademark of TiVo Inc. This site is not owned or operated by TiVo Inc.
All times are GMT -5. The time now is 05:27 AM.
OUR NETWORK: MyOpenRouter | TechLore | SansaCommunity | RoboCommunity | MediaSmart Home | Explore3DTV | Dijit Community | DVR Playground |