I wouldn't think so.
If there is something, a component, in commercial apparatus,
you can mostly count it is necessary for operation.
.... they will cut any little cost they can ...
not to be able to lower the price, to customer
but to increase profit margins.
Say they sell 10.000 devices and managed to remove for $2 dollars in each device sold.
Will be $20.000 more in their bank account.
I can see two possible reasons for such resistors:
1) As you say, pull-up. To make sure is a digital '1'
if no other output is defining this. Like at power on for example.
2) To give some load to the line - so called termination.
To have a TERMINATION Load makes sure there is a LARGER AMOUNT of current flowing in the rail.
And high frequency signals NEEDS CURRENT to make good signals
and minimize disturbance - distortions.
Compare for example
A) radio / TV-systems,with often 50-75 Ohm coaxials connections.
With our low frequency
B) audio systems, with often 47000 Ohm input impedance!
What this means ... is there is for 1.0 Volt signal
a flow of:
A) 1/50 = 0.020A = 20.0 mA
B) 1/47000 = 0.000021A = 0.02 mA = 20uA
Another way to say it:
Audio below 20 kHz can be transfered with very low current and good results.
1 MHz = 1000 kHz, 1GHz = 1000000 kHz
(mobilephones has got transistors and IC chips for such high freq)
.... Very High Frequency - VHF,
needs more current = lower cable and connectors impedance, to get good results.
My example show a ratio 1:1000 in currents.
Myself often prefer a somewhat lower than normal input impedance
in my audio signal connections.
At least in those cases where have a possible option: Like 10-15 kOhm.
While many, even power amplifiers use 47-100 kOhm inputs.
This is because I wont more currents to travel.
High impedance lines and very low voltage level signals are more sensitive to electro magnetic interferences and surrounding noise sources.
Like microphone cables and turntable cables, as you know
SHOULD BE VERY Well SHIELDED!
To protect the signal wire from being a bare exposed target for such noise and disturbance.
Signal / noise ratio, we often see.
This is how much stronger signal is, than the big sea of noise, which surrounds us everywhere.
( how do you think we can hear radio / mobile telephone,
if there was not something in the air - everywhere - except deep down in the Ground, Ground = Earth Level = 0 Volt )
More currents, make the signal stronger, in power, even at same voltage level!
lineup
If there is something, a component, in commercial apparatus,
you can mostly count it is necessary for operation.
.... they will cut any little cost they can ...
not to be able to lower the price, to customer
but to increase profit margins.
Say they sell 10.000 devices and managed to remove for $2 dollars in each device sold.
Will be $20.000 more in their bank account.
I can see two possible reasons for such resistors:
1) As you say, pull-up. To make sure is a digital '1'
if no other output is defining this. Like at power on for example.
2) To give some load to the line - so called termination.
To have a TERMINATION Load makes sure there is a LARGER AMOUNT of current flowing in the rail.
And high frequency signals NEEDS CURRENT to make good signals
and minimize disturbance - distortions.
Compare for example
A) radio / TV-systems,with often 50-75 Ohm coaxials connections.
With our low frequency
B) audio systems, with often 47000 Ohm input impedance!
What this means ... is there is for 1.0 Volt signal
a flow of:
A) 1/50 = 0.020A = 20.0 mA
B) 1/47000 = 0.000021A = 0.02 mA = 20uA
Another way to say it:
Audio below 20 kHz can be transfered with very low current and good results.
1 MHz = 1000 kHz, 1GHz = 1000000 kHz
(mobilephones has got transistors and IC chips for such high freq)
.... Very High Frequency - VHF,
needs more current = lower cable and connectors impedance, to get good results.
My example show a ratio 1:1000 in currents.
Myself often prefer a somewhat lower than normal input impedance
in my audio signal connections.
At least in those cases where have a possible option: Like 10-15 kOhm.
While many, even power amplifiers use 47-100 kOhm inputs.
This is because I wont more currents to travel.
High impedance lines and very low voltage level signals are more sensitive to electro magnetic interferences and surrounding noise sources.
Like microphone cables and turntable cables, as you know
SHOULD BE VERY Well SHIELDED!
To protect the signal wire from being a bare exposed target for such noise and disturbance.
Signal / noise ratio, we often see.
This is how much stronger signal is, than the big sea of noise, which surrounds us everywhere.
( how do you think we can hear radio / mobile telephone,
if there was not something in the air - everywhere - except deep down in the Ground, Ground = Earth Level = 0 Volt )
More currents, make the signal stronger, in power, even at same voltage level!
lineup
You need these resistors for couple reasons:
1. If they are inputs, the resistor is needed to properly bias the input to get the best noise immunity.
One can just tie the inputs to the +VCC to serve the above purpose. It saves couple of cents of component cost. But the down side is it's very difficult to trouble shoot if one pin is shorted to ground. So normally pull up resistors are used.
2. If they are open collector outputs, you need these resistors to pull the output high.
1. If they are inputs, the resistor is needed to properly bias the input to get the best noise immunity.
One can just tie the inputs to the +VCC to serve the above purpose. It saves couple of cents of component cost. But the down side is it's very difficult to trouble shoot if one pin is shorted to ground. So normally pull up resistors are used.
2. If they are open collector outputs, you need these resistors to pull the output high.
Leolabs said:
Maybe you can try, and see if it works without.
I Don't think anything really bad can happen at a test, like shorts.
But you will see if this set up is working right or not.
you might be right about those resistors
hard for those to say
that hasn't studied this special circuit, like you have, Leolabs
regards
lineup
I assumed that it's a reason for those network resistors to be there. As far I can see, in this sch. the pheripherals (including ATAPI Interface) are multiplexed hard enough to be driven by a Low Pin-Count uController. So... why put some resistor networks there if aren't so important... when the intention was to make this schematic as simple as possible?
From my experience... ATAPI bus is released by Optical Disc Drive when DIOW and DIOR are inactive and if You want to SCAN that bus (especially for Status or Error Registers values) when it's in Hi-Z, the lines will float without pull-ups and You'll read garbage (same for LCDs). The firmware might crash when read unexpected data but of course, that's depends on how good it's firmware in that uController. For reference You may try to look in SFF8020I (ver 2.6 or later), ATA/ATAPI-7 Volume 1 (Register Delivered Command Set, Logical Register Set), ATA/ATAPI-7 Volume 2 (Parallel Transport Protocols and Physical Interconnect), Multi Media Command Set (MMC) and SCSI Primary Commands - 3 (as reference).
Regarding uC internal pull-ups... must be enabled in firmware to be active!
Best regards!
From my experience... ATAPI bus is released by Optical Disc Drive when DIOW and DIOR are inactive and if You want to SCAN that bus (especially for Status or Error Registers values) when it's in Hi-Z, the lines will float without pull-ups and You'll read garbage (same for LCDs). The firmware might crash when read unexpected data but of course, that's depends on how good it's firmware in that uController. For reference You may try to look in SFF8020I (ver 2.6 or later), ATA/ATAPI-7 Volume 1 (Register Delivered Command Set, Logical Register Set), ATA/ATAPI-7 Volume 2 (Parallel Transport Protocols and Physical Interconnect), Multi Media Command Set (MMC) and SCSI Primary Commands - 3 (as reference).
Regarding uC internal pull-ups... must be enabled in firmware to be active!
Best regards!
- Status
- Not open for further replies.