LED , Meanwell , Typhon and a circuit question

[oreo57]

This is going to be a bit long but my curiosity has gotten the best of me..(NOTE: Really green amateur circuit builder here)
first a parts list:
Meanwell ELN-30-48 (both volt and amp out adj.. 10V analog dimming )
LEDS... (won't be too important as to type ect for this discussion)
Typhon controller from StevesLEDS w/ both 5v PWM out and 10V PWM out..
Home made circuit using a voltage doubler (I will state the reason in a bit and yes I've looked at OpAmp circuits that do what I want.. .also will clear this up in a bit)
FORGIVE the crudeness but it should be understandable. The 2V(f) is the added voltage drop of the 2 diodes.. I threw in the other "losses"..assuming some in the Meanwell dimmer circuit


"no load output" from VOM of circuit using the Typhon 5v PWM outputs:
New diodes in.. dimming down to zero w the Typhon..
0%=0
10%=0.51v
20%=1.08v
30%=3.05v
40%=4.06v
50%=4.99v
60%=5.93v
70%=6.78v
80%=7.66v
90%=8.46v
100%=9.32v

Reasoning.. The 10V PWM out is not filterable to analog for, as I understand it, they invert the pulse so as to pulse the ground (sink) not the source (v) unlike the 5V PWM out (see lower diagram) . I needed to interface w/ the Meanwell "D" driver and use the controllers dimming function..
It would have been real simple IF I could have just used the R/C filter to smooth the 10v pulse but that didn't work (works great on the 5V "normal" PWM circuit)..It also would be easy to build an op-amp based circuit (found a schematic for that but it involves another power tap point) but I was trying for a completely passive circuit and the voltage doubler (and/or inverter) interested me for this..
As you can see (and I verified by testing) it does smooth the PWM and double the v-out in a simple form..

Problem: Even though the v-out is 9.32V (10 would be preferred but there are losses in the diodes, already substituted Schottkys to minimize this) it doesn't seem to appear to the Meanwell as 9.3V (in regards to light output).. as an example, it will ramp up the light output but at MAX (9.32V) the intensity is lower than if I just plug a used (7.6v) 9V battery.
THAT confused me and I was wondering if anyone had an answer (my guess is that the Meanwell dimming circuit is lossy at low mA out that the PWM was converted to, unlike the still capable 9v.. BUT THAT is a real uneducated guess..This is my FIRST attempt at anything like this..
OH, one more note: w/ the Meanwell being output adj. I can "probably" make up the difference from 9.3 to a real 10 as to brightness.......and according to Meanwell the dimming circuit does not like anything over 10 by more than a 0.1V or 2..
(the voltage doubler chip was $1)
One other odd question is the spec sheet says it runs at 1.5 v to 10 (something to do w/ powering the logic circuit in it) but, as you can see, and if I understand correctly works well below this V ...........though it loses linearity..
Bit more background.. and THANKS for any help insights.. BTW I bought new drivers to avoid this mess but I'm still curious as to this circuit and application
http://www.reefcentral.com/forums/showt ... ?t=2318572

[pebe]

I’ve looked through the links and the data sheets for the various components you mentioned but I am having difficulty sorting the wheat from the chaff in your posting.

Am I right in assuming that the Typhon controller has nothing to do with the problem and what you are actually trying to do is use the PWM output from an Arduino to control the output voltage of the Meanwell SMPS by means of a varying analogue voltage of 10V max?
Is the Arduino fed from a 5V regulator, and if so what is the input voltage to the regulator?

[oreo57]

I’ve looked through the links and the data sheets for the various components you mentioned but I am having difficulty sorting the wheat from the chaff in your posting.

Am I right in assuming that the Typhon controller has nothing to do with the problem and what you are actually trying to do is use the PWM output from an Arduino to control the output voltage of the Meanwell SMPS by means of a varying analogue voltage of 10V max?
Is the Arduino fed from a 5V regulator, and if so what is the input voltage to the regulator?

Hmmm..yes and no... See the Typhon has 2 banks of PWM 1 5V, 1 10V.. The 5V bank is what I'd consider normal i.e it pulses 5v+.
The 10V PWM "pulses" the ground (sorry best way I can explain that)
IF you hook a R/C filter circuit to the 5V PWM you can smooth it to DC w/ V being proportional to the Pulse width.. Easy peasy.
On the 10 V side it doesn't work that way.. Any simple RC filter goes right to 10V w/ no proportionality to pulse cycle.

So the Typhon IS a problem in a sense.. It is a commercial board so I'm not too interested in redesigning it as it is..

Now to get to the meat.. Since it is EASY to create a analog dimming circuit from the PWM 5V dimming circuit IF I had drivers w/ a 5V (analog controlled) dimming circuit I'd also have no problem...

THE problem is getting a clean variable by PW 10V analog signal out of the unit as it stands..

Sooo to make a long story longer, My only useable output(short of hacking the board) that can be converted to analog is the 5V PWM out. I need a variable 10V out.
Now as I said a simple opamp circuit to double the voltage, using the 5v pulse as a gate (correct term?) is fairly straightforward:
Simple opamp controlled voltage double analog out circuit


Getting a supply voltage to the opamp is really not much of a problem since the Typhon uses a 12v wall ( I assume input V of regulator) wart and I could just tap off that.. .
BUT I like my circuit for its passivity. since in theory I need 4 and can build a plug and play on the fly without tapping a power branch
At this point it is more academic. Opamp circuit is almost "perfect" but needs power, my circuit is almost "functional" w/ limitations as noted...
I'm pretending I'm stuck on an island w/ no more power available .. how would I convert to analog and double the voltage of a PWM 5V signal....

This Circuit of mine was an attempt to answer the problem in a "different" manner than usual (using a passive double IC).. Oh and the orig. board I assume has a 10 V plus "bus" somewhere that feeds the triac (or trans. circuit) to be able to even do "backwards" (in my mind) 10V PWM..

I doubt if I made it much clearer.. I'm willing to try again.......The Typhon is an Adruno w/ LCD, RTC and a custom circuit board.. so to speak..


whe I found out you couldn't filter the 10v PWM to analog I asked them why.. This was the answer:

Thanks for sharing that. Regarding the 10v side is definitely different looking on the oscilloscope, (not a clean square wave), but oddly, it works just fine on the 10V PWM drivers, so we never really looked into it.

Thanks!
Jeff

A SIDE QUESTION (not really needing an answer) is can a "normal" filterable to 10V analog PWM channel be built from the adruno core ? In a sense this 10v issue is an "engineering defect".. to me..

[oreo57]

Am I right in assuming that the Typhon controller has nothing to do with the problem and what you are actually trying to do is use the PWM output from an Arduino to control the output voltage of the Meanwell SMPS by means of a varying analogue voltage of 10V max?
Is the Arduino fed from a 5V regulator, and if so what is the input voltage to the regulator?

You make it sound so simple....
One side note.. My circuit works.. It converts a 5V PWM to a 10-ish V (10V minus the 2 diode V drops) analog out.. in a relatively linear factor (see data points above)
My current REAL question is .. when presented w/ a real load..it dims fine BUT the LED top end out is not what is expected for a 9.5V output.. somewhere I'm losing "signal" as compared to the passive V-out measurement of the VOM.. Another way to put is that the board puts out 9.5V but the dimmer circuit in the Meanwell is responding as if it was 9.5V minus some V factor.. I'm clueless as to why this would be..

One thing I didn't measure was the V-out between the circuit and the Meanwell dimming circuit.. when they were hooked together..

another side note ; "rumor" had it that the Meanwell "d" drivers worked fine w/ PWM but I found that, w/ the Steves's "board" that is not the case.. at least as the 10v PWM circuit is designed.. A "normal" (pulse source voltage, not sink) 10V pwm circuit MAY work.. but who knows..
Most have just gone the 5V PWM route..

[oreo57]

I’ve looked through the links and the data sheets for the various components you mentioned but I am having difficulty sorting the wheat from the chaff in your posting.

Am I right in assuming that the Typhon controller has nothing to do with the problem and what you are actually trying to do is use the PWM output from an Arduino to control the output voltage of the Meanwell SMPS by means of a varying analogue voltage of 10V max?
Is the Arduino fed from a 5V regulator, and if so what is the input voltage to the regulator?

note.. I have another post coming which sould be the FIRST response to your question. The above is an addendum... .. Seems that one is held for moderation (has images) so they may be out of order... sorry..

[oreo57]

ICL-7660A data sheet:
http://www.intersil.com/content/dam/Int ... fn3072.pdf

[pebe]

It’s all a bit confusing, but tell me if I understand the position correctly.

1. The Arduino shown in your first posting is actually a part of the Typhon controller.
2. The Typhon outputs 5V and 10V PWM signals.
3. The Meanwell can be controlled by either 0V-10V analogue signals, or 10Vpp PWM signals.

So why are you trying to use an analogue signal: why don’t you use the 10V PWM signals from the Typhon? Or have I missed something?

[oreo57]

1. The Arduino shown in your first posting is actually a part of the Typhon controller

Yes, as I understand it.. the underlying platform is an "adruno" like microcontroller ect..

2. The Typhon outputs 5V and 10V PWM signals.

Yes BUT the 5 V PWM is like "native output" to the Adruno platform..
The 10V PWM is an add-on circuit.. and though it "pulses" it is not what I'd call "Standard" or an R/C circuit would smooth it to analog just fine.. It doesn't

3. The Meanwell can be controlled by either 0V-10V analogue signals, or 10Vpp PWM signals.

No, that was the "rumor" and I believe it would work w/ a "normal" 10v PWM signal where the Voltage is pulsed.. not the ground is "pulsed".

either way it is irreverent to this discussion since it doesn't work. Neither 10V PWM (tested w/ my "d" led driver) or trying to rectify the 10V PWM to analog (tried it w./ the same R/C filter that I used for the 5V PWM) works.
I need 10v analog out (well 0-10V dimming actually)... (or possibly "normal" 10V PWM.. but it is best to just leave that out of the discussion)

Or have I missed something?

Yes, see above...

[pebe]

No, that was the "rumor" and I believe it would work w/ a "normal" 10v PWM signal where the Voltage is pulsed.. not the ground is "pulsed".

either way it is irreverent to this discussion since it doesn't work.

It’s not a ‘rumor’. The Mean Well data sheet from ‘alldatasheet.com’ (which site address won’t reproduce here) states clearly on page 3 that an analogue (or 100% duty cycle PWM) control voltage of +10V on DIM+ relative to DIM- will turn the supply output on full, and the graphs show that any lesser value will turn on the output proportionately.

You cannot ‘pulse the ground’. Ground (see DIM- and the negative output in the block diagram on page 2) is the reference level for the pulses applied to DIM+. In a circuit like this where the DIM- is connected to ground, and the other DIM input is clearly shown as positive (+), then DIM+ must pulse from 0V to 10V. If the ELN unit doesn’t work as specified, then it is faulty. No doubt about it!

If it is OK, the fault must be with the Typhon unit. The spec for the Typhon published by Steves Leds states that it can deliver 10V PWM signals and that it is suitable for working with ELN units. At 100% duty cycle you can check its output with a digital meter; it should read 10V. At a lesser ratio you can measure the average voltage with an analogue meter, because its needle inertia will average out the voltage. If you don’t get those results the unit is faulty!

I hope that will enable you to find out what is wrong. You should have no need to resort to trying to smooth a PWM signal.

[oreo57]

Err not quite.. There is the ELN60 (P) and the ELN60 (D)..The data sheet doesn't clearly differentiate the 2.
Each works differently.. the P series uses PWM and does not work w/ analog out..
D is supposed to be controlled by a 10V analog signal and a 10K pot.. Which I do use currently.. some have reported the "D" works w/ 10V PWM
but my evidence says no and I cannot find any other solid evidence that this is true... Mostly hearsay and I believe confusion from the "mixed" data sheet.

MeanWell LPF-60D-48 is better as it offers dimming via 1~10Vdc or PWM signal or resistance unlike the ELN-60-48D/P .

I've tried to find PROOF that the D does behave well w/ a PWM signal but have not, to date, found this... Even priv. conversations have just given me "I hear the D series works fine w/ PWM's' which is worth nothing really...Well there is this:

http://scapeclub.org/forum/showthread.p ... post153419
http://www.mouser.com/ds/2/260/ELN-60-spec-45820.pdf

easiest thing for me to do would be to find a dimming driver that uses 10V PWM and test my unit.. OR an Oscilloscope to see if the 10V PWM is working correctly.. the "weird" comment from Steves LED leads me more to the side of there is a difference.. somewhere. Be it a defective board or a changed circuit in the Meanwell..
I realize that the difference between opening ground or opening supply seems to not be anything that would cause this problem........

This, to me, has become the "do I reinstall windows or find out what is wrong w/ it" syndrome.. if you get my drift...
bit more background..........

Your situation is that the 5V PWM output of the Typhon works as a "source" and the 10V output as a "sink". Briefly explained, the 5V PWM switches on/off the +5V and the ground is fixed. With the 10V PWM signal, the +10V is fixed and the ground is switched on/off.
When you connect the low pass RC filter to the 5V PWM to smooth it out it works since the switching 5V charges the capacitor to a certain point depending on the PMW ON duration.
On the other hand with the 10V PWM, since the +10V is fixed the RC low pass filter have no effect. There is no variable voltage to smooth out.
I guess that with a PNP transistor, the 10V PWM "sink" output can be converted to a "source" output. I have not done this, it is just a pausible solution to try.

http://www.reefcentral.com/forums/showt ... ?t=2314368

PNP or NPN or my circuit (oddly enough I understood mine.. )

[pebe]

I wish I had known that there were two different ELN versions at the outset, because there is nothing in the block diagram of their data sheet that would imply there is a difference. In fact, the data sheets for both the ELN and Typhon are very vague; OK if you are only wiring them up, parrot fashion, but pretty useless if you are trying to do what you are doing.

However, all is not lost. I think you need to confirm that the ELN is working correctly by connecting a 9V battery to the DIM+ and DIM- pins and check that it gives 90% output. If that is OK then we need to establish if those pins draw current from the battery. Most complex ICs these days use MOS transistors internally which have a very high input resistance, so it is only necessary to provide voltage – not current.

To prove the point, fit a 100K resistor between the +ve of the 9V battery and the DIM+ input pin and see if the output is still the same.

Try that, and depending on the results we can proceed from there.

[oreo57]

I wish I had known that there were two different ELN versions at the outset, because there is nothing in the block diagram of their data sheet that would imply there is a difference. In fact, the data sheets for both the ELN and Typhon are very vague; OK if you are only wiring them up, parrot fashion, but pretty useless if you are trying to do what you are doing.

However, all is not lost. I think you need to confirm that the ELN is working correctly by connecting a 9V battery to the DIM+ and DIM- pins and check that it gives 90% output. If that is OK then we need to establish if those pins draw current from the battery. Most complex ICs these days use MOS transistors internally which have a very high input resistance, so it is only necessary to provide voltage – not current.

The driver won't even work without a "load" on the dimmer circuit. I used a 9V battery just to use it when I first bought it..
Currently I use a 12v wall wart hooked to one of those ebay Chinese voltage buck regulators to drop it to 9.97V.. Then a 10K pot is hooked to this circuit to manually dim.. Works perfectly fine..

To prove the point, fit a 100K resistor between the +ve of the 9V battery and the DIM+ input pin and see if the output is still the same.

Try that, and depending on the results we can proceed from there.

See above.. dimming circuit works fine in full manual mode.

high input resistance, so it is only necessary to provide voltage – not current.

That is the part that has me baffled since my circuit outputs enough voltage to be perfectly useable.. Like the manual circuit..Unfortunately, to me , the Meanwell seems to have a current need.. maybe.... IF using a 9v battery (@7.4v) gives me higher light output than my circuit @9.4v.
BOTTOM line: for now I've given up trying to use the built in 10V PWM and converting to analog.
Tried to convert the 5v PWM to 5V analog and double the voltage.. Worked but has the problem of not setting the driver to close to fulloutput w/ a 9.35V analog signal..
Bit more background on how the Typhon is "probably" built......and why, for me, it is a PIA..
http://forum.arduino.cc/index.php/topic,20632.0.html

[oreo57]

Hmmm.. thinking about the above diagram it dawned on me that I could possibly just reinvert the 10V line using a PNP transistor in a cascade fashion.. hooking the base to the 10V PWM out..
of course it gets back to the use of the 10v tap again.. just like using an opamp to smooth double the voltage on the 5V PWM..

[oreo57]

ANOTHER thought using my "chip".. Invert the 10V PWM signal.. then filter w/ an R/C circuit.. no need to double the voltage then..

The ICL7660 and ICL7660A contain all the necessary
circuitry to complete a negative voltage converter, with the
exception of 2 external capacitors which may be inexpensive
10μF polarized electrolytic types. The mode of operation of
the device may be best understood by considering Figure
12, which shows an idealized negative voltage converter.
Capacitor C1 is charged to a voltage, V+, for the half cycle
when switches S1 and S3 are closed. (Note: Switches S2
and S4 are open during this half cycle.) During the second
half cycle of operation, switches S2 and S4 are closed, with
S1 and S3 open, thereby shifting capacitor C1 negatively by
V+ volts. Charge is then transferred from C1 to C2 such that
the voltage on C2 is exactly V+, assuming ideal switches and
no load on C2. The ICL7660 approaches this ideal situation
more closely than existing non-mechanical circuits.

Yes I know I'm getting in way over my head here as well as ignoring the simple solutions.. but what fun would that be.........

[pebe]

I'll ask again!

Put a 100K resistor in series with the DIM+ pin, ie. between 9V battery+ and DIM+. Do you still get same output as without the resistor?

If not, what is the voltage a) across the 9V battery, and b) between DIM- and DIM+?