Dan

I have a few question about the Version 1 LED board operation.

I assume that we want board to operate in Bar Graph Mode.  If this is correct the mode select switch is wired incorrectly as pin 9 should be connected to PIN 3 (chip's plus supply voltage) according to Data sheet.  Wiring on board shows it connected to pin 1 on next chip in series which is correct for DOT MODE.

Does the chip pull output low when TPS voltage rises?

I believe I do have the latest Gerber File from above.

Note:  This article also talks about additional wiring to cascade chips and use a common reference voltage range
https://learn.sparkfun.com/tutorials/dotbar-display-driver-hookup-guide/all 

Earl

Yes it pulls low
No i think it will work in dot mode
DD

Good pulling it low is what I thought.

Just starting to test LED replacement board to see what it does.  Just found top and bottom TPS and ECU pads are not connected.  Will jumper them as ground comes from bottom.  I was trying to set upper range and could not get it to move.

I am assuming 5V supply so operating range of LED board is 0-3.5V as it needs to be 1.5V below supply.
Note: If we need a bigger range the card supply can be higher as it does not need to be the same as the signal input.

Installed jumpers between top and bottom pads and was able to set upper limit I choose 3.4V and then tested each of the 32 outputs from 3914 at chip interface and in dot mode each one pulled output down in order last one around 3.3v.   For the Black (A1) I also check that signal reached the 150.  I did not check the white side yet as it is hard to do due to mapping and I had already verified path is correct.

Couple of picture of my test configuration.

ok looking  pretty with those jumpers, I waiting a chip part to arrive here,
did you map the jumpers.

Ronnie had one trace on k11 and 1 on the k8 to showing the voltage rise and the duty width
in his video


Dan

Yes I mapped the jumpers.  Have this all documented in K11 write up.  I have even added the TSP replacement information.

I had already did some testing of K11 and have written up that part.  Found that 150N chip did not appear to be work correctly without having a load on the LED input pins.  This is the reason I built the new TSP LED replacement board.  I now have that working and have done initial testing to setup voltage range and have verified in DOT mode that the output of all 32 pins changed as expected.

Using the board I can now see the data changing on the 150N output pins.

I even can get data all the way to board output but I can not pulse width to change in DOT mode only get a single pulse and pulse is only there when pulse as at a certain point (around 2V).  I put put 5V jumpers on TPS board to put board in BAR mode and was able to sort of adjust pulse.

I say sort because I an see 2 pluses a single pulse then a second and I can only adjust the second one.  Not sure if the 2 pulses are caused by the jumper between pin 9 and 1 on TPS board or the the switch between A2 and A1 on K11.
I am using the test points that Stan used on K11.  There is one pair on pins 5&6 on LS112 that select A2 then A1.  (Switch every 16 clock pulses). And the stay LOW for 16 clock pulses then switch state.  The 150N only puts out data when Pin 9 is low.
The second pair is on pins 4 & 5 on the LS122,  PIn 4 has the time pulse from 112 and it changes every 32 clock pulses.   PIn 5 is the clear pulse from the 150Ns.

By looking at 122 Pins 5 and 6 I can see the data pulses from the 150 change state.  As currently wire the first pulse shows up half way between the  first reference start cycle pulse on 122 Pin 3.   This is not what we what as we get now board output until we start using data from the white input.   NOTE:  This is because the way the board is wired the A2 white is read first as pin 5 goes low before pin 6.

I think I can fix this by just switching location of cables on the connectors on TCP/LED board.  This should work because the interface to K11 Led connectors and 150Ns is exactly the same.  What this should do then is cause the low order LED bits to be read in proper order as I have remapped cable to match order pins are read. (This is all explained in document I wrote which I used to build cables).

I have not done this yet as I want to label cable before I move them so I do not get them mixed up.   I will only move end on TCP/LED board.

Have not yet taken pictures as it is a pain holding scope probes on the chip pins I can see why Stan solder test jumpers to the chip pins.

Dan I have attached Current  version of my document this still in draft form but it does have tables showing what is pins and they should be correct as I built cables from these tables and verified path with ohm meter.  The LED cables mapping are just in front of the test section of Document close to end.

Test section does have scope pictures from my initial testing showing the initial clock pulses and the timing pulse sent to 122.

Dan
Found a problem with mapping of White cables what is currently in draft document is wrong.  During testing I found values for A2 White cables jumping around. Through testing I found that the pin connections between the counter chips 74193 A3 and A4 and DM74LS150N chip A1 and A2 are not what the schematic implies.  The schematic show ABCD connections and pin numbers being different when they are actually both the same. 

In both cases the connections are
LS74193 Pin #   3   2   6   7
DMLS150N Pin #   15   14   13   11
Name for both   A   B   C   D

OK I am reading through this , do you have a picture of your tps connect to k11?
or is it going through new TSP LED replacement board?

reason is we need the input from rotary 3 wire tps sensor

I am just using a pot to simulate the TPS and it is going through TPS LED replacement board

After I switched cables to make them the same, I captures a few screen shots to show the inputs to LS122 Pin 4 is the reduced clock input that starts the output pulse.  Pin 4 in the input from 74LS150N in DOD mode this is a single pulse for active pin and only one pin is active at a time.

CH1 Yellow is from 122 Pin 4 (clock)
CH2 Blue is from 122 Pin 5 (Data input)
 
This is the setting for all the picture in this series

Picture 1 is the output in DOT Mode with low voltage input from TPS

Picture 2 is the output in DOT Mode with mid voltage input from TPS

Picture 3 is the output in DOT Mode with high voltage input from TPS

I switched TPS LED board to Bar Mode and repeated screen shots
Picture 4 is the output in BAR Mode with low voltage input from TPS
Picture 5 is the output in BAR Mode with Max voltage input from TPS
Picture 6 is the output in BAR Mode with Mid voltage input from TPS

I did look at K11 M/M1 output.  In DOT mode only got an output for one of the data inputs the one closes to start of pulse.  In Bar mode I got a pulse that I can change wide but max width of pulse was about 15 per cent.  I think this is due to the break you see in the pulse in picture 6.  At this point most of the system appears to be working.  I think I need to adjust cable to remove data dropouts. I also think I need to switch things around to remove pulse to left of pulse start.  Think this caused by the data delay in the data path as appose to clock path.

ok noted

Dan,

I marked up the TPS Led Gerber to show changes I made to TPS LED Version board to get it to work.
3 sets of changes
Connect top and bottom TPS and ECU pads - Needed to provide all the power paths
Remove connection between Pin 9 and Pin 1 in 3 Places (part of converting to Bar Mode)
Add +5V to all 4 pin 9s (needed to put board in BAR Mode)

I am providing +5V and GND to the board through the TPS +/- Pads.

I am using a 10K Pot to simulate TPS input so Mid lead of it connects to TPS signal pad.  Have set on board Pot to provide 3.4V to set top range of TPS input. (Must be at least 1.5 volts less than supply voltage)

Dan

This is the setup and test points I am using

I am providing Power to the board and TPS (pot) using jumpers to Pot which connect to TPS Pads on board.  In this case +5V and GND.   Output from Pot is 0-5V I used pot on board to setup upper limit to 3.4 volts.

I have mainly been looking at pin 4 of 122N Clock in (This appears once every 32 [G] pulses) as this starts a new output pulse and is the yellow pulse in pictures above).  I also look at pin 4 of 122N (Blue Pulses in pictures above) as it has the data input from the DM150Ns.  In data Mode this will be a variable length pulse.  ( In dot mode a single pulse).  Pin 8 will be inverse of M/M1 but I have been looking at Pin 1 of board as this will be normal output (I just use pin 1 of break out board in my test setup).

I also have been checking the data inputs of DM150N to verify which pulse I am looking at.  Currently I have them wired so  A2 White E0 is from Pin 1 for first LS3914 chip and E1 is from second etc.  This because due to clocking the A2 White is read first.

Note:  The way the 122N is is configure pulse will be extended while the data on Pin 5 is high. At the first low pulse the output on pin 8 will go low and stay low until the start of the next Yellow pulse so any other high inputs will be ignored.  This is way if you only look at 122N pin 5 you will not see the real output from board. 

It is a real pain holding scope probes on these pins and doing anything else.  In Picture of Stan's K11 he as soldered test leads 122N pins 4 and 5 and also to 112 pins 5 and 6.

Earl

 
ok Thanks Earl , I am  working on this
did it trigger smooth via tps to k11 ?
any video
Dan

With the current setup I am getting a variable pulse width pulse signal into pin 5 of the 122 and it is smooth in that it changes by the pulse width of the 150N pulses.  I did have a problem with a little noise in the system.  I think that was caused by bad solder joints on the LED board.  It is hard to get good connection with my solder iron with the small pads on the LED TPS board.  After fixing the noise this is now very repeatable. But I am still only getting a pulse out of 122 on K11 that is only about 10-15 % of pulse width.  It does vary from zero to this limit.  It does jump at this limit but I can see a 0.2V variation on TPS sensor input voltage that I think is causing that.

I have been looking at the 122 tech documentation about the variable pulse length function.  It can be done 2 different ways. 
1) Extend the pulse with another pulse in on A or B inputs in a narrow time widow.  (Pin 4 is one of the B inputs)
2) Cut off the pulse width when input on Pin 5 goes Low.

I had gotten it in my head that we are using method 1 but that is not the case as input from 150 and LED board is not a series of input pulses . it is instead the clear pulse (which may mean the dot mode should work).

The other thing that has been bothering me is I have not been able to change the pulse width of the start pulse using the on board POT.  I am begin to wonder it should be used to step a maximum length pulse then cut it short using the clear pulse input on pin 5.

Both methods rely on timing which is setup by the capacitor and resistors on K11 122 on pins 11 and 13.  If they are set correctly non of this works right.

Other thing I just notice is normal output of 122 is on pin 8 which is what I have been checking.  But yesterday I also checked Pin 6 which should be the inverse of pin 8 but instead where I go one long pulse on pin 8 I got a series of several pulses on pin 6 so something is wrong.

I will check with a different chip to see if I get same results.

Earl

seems we are getting a good signal but it is not varying from 0 its only varies in a range so we are  working on a solution for that and hopefully you will see it working properly very soon

I have the LED TPS vary from 0 to 3.4V and that works with the 5 Volt input from TPS POT.  Range must be 1.5V below the board supply and I am using the same supply to power both.  I can see that input on pin 5 of the 122 and it is changing over the full range.

If need the full 5V range we can use separate supply for card with higher voltage.   Some article recommend doing this to provide finer control.  I am seeing at times a 0.2V fluxuation. 

just making note    that G must have signal in to have card start right?

  i think we need higher control voltage to get full variation
 i try to make a small breakout that can provide higher voltage variations before merging
if you have tps driving k11 please post video here Dan

Yes [G] must be present as it provides the timing for everything.  Without nothing works.  I also found that the 150N chips do not work correctly with out load on data input pins.

As this point I am getting full range out LED TPS board and can see each of the input pins on the 150's go low.  Normally high from pull up resistors.

I can see this signals from the 150 on Pin 5 of 122.  This is the clear signal.  Pin 4 is the data Pulse.    With that I can see a variable pulse out of the 122 on Pin 8 and out of K11.    At this point I believe the issue is the configuration of the 122.  I should be able to change the pulse width with the Pot on pin 13 of 122 but it does not appear to do anything.

Going to look at the Math on the timing of the pulse.

Earl

so tps into k11 is  scaling  ok ?

I try get some picture  of my side please try take pic of video of your board arrangement
Dan

Yes as I am looking at the input to 122 which is the output from the 150 chips.  The center 150 is connected to lower input and both sets of cables are exactly the same.   The schematic on where pins goes was miss leading as it turns out addressing of 119 counter chips is the same.

You just have to adjust range with POT on board.  You use it to set the top voltage.   The chips have a wide range but I am only using 5V supply.  So range is only 3.4 V volts as it can not be more than 3.5 volts. As top range MUST be 1.5 volts less than supply voltage.  I am also providing 5V to a pot simulating TPS.  If you want or need a wider range you just have to increase the card supply voltage and it can be different from TPS voltage though you would need to separate the supply and TPS voltages on card.  For my testing I was doing I was not to concerned about this.   NOTE. Using separate supplies is recommended in for more accurate results.

I did have a problem with some solder connections on some resistors I think it was due to small solder pads.  I could see these when I looked at inputs to 150 pins.  I also saw smaller pulse on some pins until I cut the traces be pin 9 and 1 that where there for dot mode.

Picture of my board set is above I will get my wife to help take a video with my phone.

I have update document to show both cables are the same and the tables that show pin connections It also has a some for screen shots from my testing.  I use those tables to check that I have good connections and that cables where wired properly.

ok thanks  I,  any pics or videos will help show it running from tps,  input
like ronnie did just using tps in replace,   
I will keep trying mine here more and  start work to merge the board into one

I will increase the solder taps more to assist people,
if any tips for notes or  things for new board let me know and i include it on there
of coarse all those wire will be now hard traced into pcb.
so if any further note son those welcomed as i do it

Dan

The drill holes for the large capacitor need to larger at least the one I purchased I and to use leads from a resistor as the then solder capacitor to them as capacitor leads were to big to fit.  Also had a problem with ribbon headers had to purchase small round headers as ones with square pins did not fit.

I would recommend adding separate power supply input for board and TPS.  If TPS input is 0-3V current version works but it TPS is 0-5v then you can only use 0-3.4v of the TPS range.

ok noted
I start on that while we  try get pic videos DD

Dan,

After looking at LS122 pulse timing I found I had set time too short by using too low a frequency.  I had been using around 13hz at that was as high as I could get using one switch setting on K2.  I switch to next higher K2 output and then lowered it to 41.32hz (which is what I have been using in my other testing as a default).  Now this is the K11 M/M1 output frequency not the [G] from K2 which is higher.  This made a huge difference in the duty cycle output.  I still could not get full range but could get it to change from about 96% down to about 46%.   Note sure we need to go lower as I have K3 set to be 50%.

While I did make a video with my iPhone it is huge about 360MB.  I will put several pictures in a table and post that as it will also let me show data at three points.  122 Clock reference pulse (Pin 4 of 122), 150 Data output from LED -TPS board (Pin 5 of 122) and K11 board output M/M1.

Note:  M/M1 output is inverse of Pin 5 input.  So at low voltage M/M1 is high duty cycle and Pin 5 low duty cycle.

Another case of Stan's boards being designed to operate in a specific range.

I had tried to set frequency to this range and just did not keep at it long enough and did not think change from 13hz to 41hz would make that big of deal.

It will take a while to collect the pictures and write up what they are showing.   I will also put them test document along with information of what I changed.

Earl