|
|
So Todd, there you have it.
Did you read the 768mV P-P with or without your circuit attached?
What's the counter chip you're using?
I would imagine that you damaged it in your 20 volt tests, and it now clamps
Tracy's output to 0.7V (typical diaode clamp voltage). Or you had it attached
but unpowered.
I thought the absolute max (power supply) ratings for CMOS chips (like 4000
and 74C.. series) were 18V.
Certainly TTL chips (74..) are limited to 5V supply (5.25V max).
Either that or you didn't read the scope right.
I, for one, appreciate the 12 pulses per rev. Gives you a precise 5 RPM resolution
at 1 second update rate.
Finn
Tracy Crook wrote:
Everything you ever wanted to know about
the EC2 tach output.
The tach output from the EC2 is an open
collector output with a 2K pullup to battery voltage (+12). It is a completely
independant driver so builders have the freedom to hook it up to any random
place they want to and blow it out without stoping the EC2 from functioning.
Several builders have exercised this freedom. Unfortunately, some of them have
used this same freedom on virtually every other EC2 connection with less
than good results.
Due to the very few tachs that use 12
pulses /rev, I am dropping this feature from future versions of the controller.
Even Grand Rapids seems to be dropping this option. A new Tach output may
take its place with a more common output pulse rate.
The 24 tooth crank sensor (running at
1/2 crank speed) is the source of the signal timing so it is 12 pulses per
rev. This happened to be a pulse rate option on the Grand Rapids EIS I
was using so it worked out.
To answer the questions about EC2 software
updates, I will be posting details about when & how much (it's cheap)
on the website.
Tracy Crook, RWS
-----
Original Message -----
Sent:
Friday, August 22, 2003 12:56 AM
Subject:
[FlyRotary] Re: tach divider
Might be an open collector output, in which case you'll need a pull-up
resistor. Just make sure you don't pull up to a higher voltage than what
Tracy's output chip is rated for.
In my design I use a chip with SchmittTrigger input (for example 74C14).
Still would recommend a series resistor before the input to the chip,
and the input to the chip protected with a Zener diode (voltage rated
at less than max input rating of the chip and supply voltage).
>Input - resistor - chip input ->
|
Z
|
>Ground--------------------------
For example, if you're supplying the chip with 12 Volts, use a 12V or
less Zener diode.
Size of resistor depends on type of chip, but for a 74C14, 10K or more
would be just fine.
A capacitor in parallel with zener to filter out noise would also be a
good idea.
10,000 RPM x 12 / 60 = 2 KHz would be the maximum you'd ever need to pass.
A .047 uF capacitor and 10K resistor would be fine (3 KHz).
Finn
Haywire wrote:
Hi Finn;
You got me thinking that I could come up with a
better solution, than a voltage divider. I built the circuit on a breadboard
again, but using a capacitor across VDD & VSS this time. I tried
various capacitors and was unable to improve on the stability at higher
voltages without a voltage divider. I was considering ordering a couple
of regulators (still the best idea) when I started considering that
I really don't know anything about the output pulse from the EC2 other
than it is a 12pulse/rev signal. So I borrowed a Fluke scope from our
shop and brought it to the airport to have a look at the wave. It is
a square wave 768mV peak to peak, base voltage 8mV (I guess I could
have just asked Tracy :-). This surprised me as I was expecting 5V.
In experiments on the breadboard I found that the counter performed
well at voltages up to 20V as long as the clock input was not greater than
5 volts below supply voltage. Now it's becoming a little more clear
(remember, I'm just an apprentice). When I showed the waveform to a
journeyman at work, he also was so surprised, he suspects it's not correct.
Tracy can you confirm this P/P voltage?
The voltage divider, while being a little crude,
is effective, so I may just stick with that, unless I can find a couple
of 5V regulators in the spare parts. Since I expect that once the EM2
becomes available, most builders will chose to use that since it really
is a great tool, so there won't be any reason to get too carried away
with this thing. Besides I'm just giving the darned things away because
I had 24 extra counter chips, so I shouldn't start buying parts -- no
wonder I can never save any money :-)
John and Dave;
I'll try to get a couple sent off before next week.
John, I'll need an address.
S. Todd Bartrim
Turbo 13B RV-9Endurance
C-FSTB
http://www3.telus.net/haywire/RV-9/C-FSTB.htm
"Whatever you vividly imagine, Ardently desire, Sincerely believe
in, Enthusiastically act upon, Must inevitably come to pass".
A capacitor accross the supply pins should be fine there. I was thinking
about the input. Of course if your input comes from Tracy's controller,
that's probably not an issue. But from anywhere else, like coil pick
up, some input filtering/clamping would be wise.
Finn
Haywire wrote:
Hi Finn;
It is a CMOS device capable of handling 20vdc on
its VDD pin. So I didn't feel it was necessary. In reality I found
it was unstable when VDD was less than 5vdc or greater than 11vdc,
so since I had a bag of 470ohm resistors handy, I just put a voltage
divider on the VDD. With a regulated 13.8vdc it gives me just under
7 volts and a stable reading.
a side note; I originally had an internally
regulated alternator but the regulator failed (have since converted
to an external reg as per "electric Bob"). When running with the
alternator tripped and less than 12vdc, I found the tach would become
erratic at high RPM's, due to a low VDD. When alt is charging, everything
is fine.
S. Todd Bartrim
Turbo 13B RV-9Endurance
C-FSTB
http://www3.telus.net/haywire/RV-9/C-FSTB.htm
"Whatever you vividly imagine, Ardently desire, Sincerely believe
in, Enthusiastically act upon, Must inevitably come to pass".
I don't see a Zener diode to clamp any spikes.
Finn
Haywire wrote:
Attached are a couple of pics of my tach
divider. It can be set to divide by any number necessary. I can't
guarantee it would work with any tach but it works very well
with my cheap off-the-shelf-at-any-autoparts-store, Hastings
tach. I have a few more if anybody wants one.
S. Todd Bartrim
Turbo 13B RV-9Endurance
C-FSTB
http://www3.telus.net/haywire/RV-9/C-FSTB.htm
"Whatever you vividly imagine, Ardently desire, Sincerely
believe in, Enthusiastically act upon, Must inevitably come
to pass".
The EC-2 tach output has 12 pulses per revolution,
and won't directly work with any normal tach. Todd has made
a divider to bring the pulse rate down to normal, and it sounds
like it's working well. The TT that you have expects 1 pulse
per rev, so if you don't mind dividing the rpm reading by
12... :-)
|
|